Syonyk's Project Blog

Deck Part 4: Stairs, Edges, Umbrellas, and Bats

2024-02-10T16:00:00+00:00

I left last time with the bulk of the deck surface done - but the edges weren’t finished out, and I didn’t have stairs (at least, not written about). That’s what this post covers!

Building the Stairs

I thought I’d save time by getting pre-formed stair stringers and cutting them down to the exact size I needed. What I didn’t really consider is that I wasn’t able to actually find enough of the two stair stringers locally to meet the Trex requirements (it’s soft, so you need about 12” on center for stairs, too). I was able to find three. Not the four or five I should have gone with. And I probably should just have purchased a big hunk of 2x14 or something to build my own. But, this is what I built with what I had laying around.

The concept for stairs is pretty simple: I have a door, and the deck is a good bit below it. So, add stairs up to it. The deck is low to make building it easier, and also to allow access to the crawlspace vents and some other things I need occasional access to, and the end result is that I need a pair of stairs at the 6” and 12” mark to make a nice even rise up into the house. Of course, the stringers aren’t exactly that size, so I had to cut them down.

When cutting stair stringers down to size, recall the height of the treads you’re about to put on them - the Trex I’m using is an inch thick. Take that into account when trimming things off! I have to say, I’m quite happy with this metal speed square - my orange plastic one shattered some while back, and this one just keeps going. Caution: Hot in the sun!

I like my compound miter saw. It can make some impressively long cuts when asked, and doesn’t complain a bit about it. My saw has been out and hooked up quite a lot this year, and it’s worth every penny I spent on it.

Of course, we have to be careful with praying mantis cocoons! This went right back where we found it so it can hatch and chew on other bugs.

I initially planned something fairly complex - a step on top with a big step around it on all sides below. I even started building it, and you can see the general concept here. What I didn’t work out was the details of how to actually secure everything to the stringers and to generate a structure that would actually work. If I want to build something like this, I think the right answer is actually to start with regular lumber instead of stair stringers, and just build it entirely from scratch. So, why didn’t I? I was building this a few days before a rather large “deck warming” gathering at our place for the 4th of July, so I needed something working, and was running out of both time and tolerance for working in the heat. I may go back to this design at some point in the future, when it’s cooler, and I’m not under time pressure - it’s easy enough to rework the stairs should I want to.

What I ended up with instead is something that I’m going to claim is more than strong enough, and spreads the load out decently - but also looks more than a tiny bit Kerbal in design. It does provide the needed support for the stairs, though. The whole thing sits on the deck - it’s not anchored to the house at all. Find the joists, drill some pilot holes, and some very long structural screws go through the base at the front to hold the entire contraption in place. Should I want to redo things down the road, I can easily remove it.

But, it supports steps, and allows people to go up and down into the house. I just don’t really like it. I’m considering ways to do a fully enclosed set of stair treads in the future, with some overhang of the treads. I’ll just need to pull out a table saw and do a bunch of Trex ripping to do it - which is a project for future me.

Finishing the Surface: Square Sided Trex

As of the last update, I had the bulk of the Trex in place with hidden fasteners - though I’m suffering some tennis elbow as a result of all the hammering and need to get that resolved. The last step in a “full deck” is to finish it out with the square edged Trex (the stuff I built the stairs out of). There are two ways of doing this - either secure the outside of the previous “hidden fastener” board with screws and fully screw the final board in entirely, or do what I did, get a Trex Hidden Fastener router bit, and make one of the square-edge board edges compatible with the hidden fasteners - and then screw the outside of that in. Benefits include “Cleaner, better looking, and only marginally more work.” Of course, you have to borrow a router, if you don’t have one…

I start by cutting the boards to the proper length - matching the previous hidden fastener boards. Yet more fun with a compound miter saw, and yet more wrangling a few small children to help hold stuff, where I want it, without wiggling it around.

Step one is to set the depth of the router bit properly. You mill (or, if you’re Jeremy Clarkson, you probably call it “routering!”) from the backside, so the depth needs to be set from there. You can guess and check, or you can just use a piece of the hidden fastener board to get things close. I found I had to extend the bit a hair further than “aligned in the groove” to get it to cut the right depth - just a smidge further out than what you find when aligned in the groove.

Route some test pieces to verify the depth. It’s easy enough to get set - just be sure it’s accurate, otherwise you’ll have issues with the hidden fasteners. They’re somewhat tolerant of misalignment, but don’t make them work harder than they have to.

Finally, with your router depth adjusted, it’s time to start making a square edge board less square! What I found worked best, since you’re working from the back of the board, is to set the board in place and make sure it fits, then just flip it up and over the previous deck board, with an inch or so hanging off the end. Apply a few clamps, and mill it out in sections.

Start with the end, then once you’ve finished that part, move a clamp over and route the next section. It doesn’t have to be continuous - the bit, if you’ve secured it properly, won’t go anywhere. Just start the router in the previously cleared groove and start working on over! Repeat until you have a full groove completed.

It leaves a wonderful mess. But, before securing the final board in place, I still had one thing left - trimming the perpendicular “blocker boards” down to length. Instead of trying to estimate exactly how long to cut them, I just left them long, with a plan to cut them down once everything else was worked out. The trick is, you really don’t want to rip up your new end board - so they have to be cut before you secure the new board fully into place.

Get the final row of hidden fasteners installed, slide the end board into place, and then make a nice mark to cut along. Remove the end board, and apply a circular saw (one of the only times in this project I had to use one - I really don’t like them much) to trim to length! Just square the marks first.

Once done with that, secure the boards! I’m using the hidden fasteners, and then one screw per joist. You might, rightly, consider the boards to be “a bit more cantilevered out than reasonable,” and you’d be right. Once the railing is up, I’m going to put a fascia board in under them to finish out the side - but I don’t want to do that until I’ve got the railings and some shade cloth structures finished out.

The result? One finished deck surface, complete with stairs!

Shade Umbrellas

You’ll notice in some of the photos above some shade umbrellas, and they’re worth talking about. I’m using them more as a short term solution, because I have plans for something better down the road - but they provide some useful shade from the sun and rain should you care to have something more than a full sun exposure. They’re not cheap at $500 or so each, but they are well built and should work for some useful while. I’ll see - we have a stupidly high UV environment here that just destroys stuff in short order, so I’ll be interested in how the more expensive ones hold up.

They all look generally similar, though a lot of them don’t come with a base (they’re designed for hard mounting to something). Not having had any idea of where I want ours, they’ve got a plastic base that I filled with sand (or, rather, got my kids to fill with sand and paid some other kids to fill with sand, because stuffing wet play sand into them is just tedious). The whole thing folds up nicely for storage and doesn’t take too much space - but they’re still large and heavy when stowed.

You can fill each of the four base pieces with whatever you want - sand, water, antifreeze, etc. If they’re going to be out all winter long, you probably don’t want to fill them entirely full of water, because it will certainly crack them when it freezes. I went with some play sand, though it’s of the wetter variety and a pain to pack in (you can’t just pour stuff out of the bags in). They’re not entirely full, but there’s enough weight to keep the umbrellas mostly planted when deployed. Wind can still push them around a bit, so be careful - Trex provides a slippery enough surface for other plastics that these can be slid around by either a person or the wind.

The crank at the bottom serves to rotate the umbrella around 360 degrees, though it will bind up if there are strong loads on it. Rotate it before you deploy it in the wind.

Squeeze the handle, and the bottom slides up and down to cantilever the umbrella out. I’ve found that to deploy the umbrella, you need to slide it out somewhat - it just won’t deploy properly when stowed.

There’s a crank on the handle that pulls on a rope. Crank it a bit, flip all the umbrella ribs out (at least one or two tend to like to be over-center and the umbrella won’t deploy until you fix it), and the umbrella pops out. Presto! Instant shade! Well, at least, during the middle of the day. Towards the evening, when we like it use our deck, the sun is coming from lower in the sky and this doesn’t offer much useful shade. You can see the shaded section across the door.

At the cost of far greater wind area, you can then tip the umbrella up and add a lot more shade. I’ve found that if the wind is coming at it from the peaked side, it’s fine, but if the wind is coming into the “cup” of the umbrella, it’s likely to slide around more than I’d prefer. Deck umbrella sailing will probably never be an Olympic sport, but it’s certainly a thing.

Overall, I think these are a decent enough short term solution - but they’re not a great long term solution. You have to stow them when not in use, or the wind (at least out here) will come up and do something nasty to them. They’re stout, but I just can’t leave them deployed regularly.

Batty Batty Bat!

Apparently, though, a stowed umbrella is a compelling seeming place to sleep the day away, if you’re a bat. I went out one afternoon to deploy them, and as I started cranking it up, I saw a large black spot. I was initially annoyed, thinking it was mold or mildew, but… nope! This is what I found!

This would be the “What on earth are you doing waking me when that thing is still in the sky?” sort of death glare you get from a bat, mid-day, that has gone from a nicely shaded sleeping spot into full sun. It was not happy with me.

And, having shaken off some of the grogginess, proceeded to crawl down a bit and fly off (erratically - I’ve never seen one fly in the day before, but the Hotel Transylvania movies did get it right). Anyway, I suppose I should find a good spot to put some bat boxes. I’m just concerned about the heat - most of our land is sunny in the summer, and high temperature isn’t good for bats. So I need to find some shaded areas for them. But I’ve never seen a bat out here before! Sorry, little guy! I didn’t know you were in there!

That’s the state of the deck - until I get the next parts for it in!

And Lent…

Lent 2024 starts this Wednesday - Feb 14, and goes through the end of March. As is becoming my tradition, I substantially reduce internet use (and computer use) during Lent, and focus on the world in front of me. I’ve written about it before, but I’m finding other people working out much the same (my end of the bell curve seems to be getting crowded lately) - so if you’re curious, I’d encourage taking a look at a recent post from Ruth and Peco over at School of the Unconformed (and subscribe to their substack, if you’re that sort). They’ve laid out a good plan of action for how to take about six weeks off computers, and find ways to do something else with the time. After Easter, you can add things back in, but it’s amazing how much less I find myself caring about stuff after that time. The forum threads I’ve argued in have gone somewhere else or died out, the draw of the doomscroll is dramatically reduced, etc. Try it out, and report back!

Posting will resume in April, along with the Nampa Farmer’s Market. I’ll be out there this year in a booth with a bunch of other people I know. We’re going for a bit of an “old west trading post” sort of feel, and I’m providing what should be some pretty solid quantities of basalt rock dust and biochar, along with ferro rods and blow pokers that weren’t at the market last year. It’s a very much analog, in person, aggressively local (everything must be made or grown within 100 miles of the market) sort of place, and if you’re in the Treasure Valley and haven’t stopped by, you really should check it out this year!

Deck Part 3: Trexing

2024-01-27T16:00:00+00:00

After the last post, I had the deck framing done - and so this post, logically enough, covers the process of installing all the Trex. Hopefully, if you’re considering a deck, there’s something useful in here for you, too!

I had hoped I’d be able to get through this process a lot faster than I actually did… it took a month worth of weekends to get it done, having started with most of a week off to get it all started. But it’s done, usable, and still waiting on the rest of the project work for it!

Trex Decking and Hidden Fasteners

I’ve waffled back and forth on decking over the years, but I’ve been convinced by various people that the right answer for a deck, especially if you’re doing it yourself, is Trex. The older stuff had some problems with high UV environments (which I absolutely am), but the newer stuff is better encapsulated and, at least in theory, holds up far better. You can see the hard plastic shell around the core inner recycled plastic layer. Time will tell…

There are a few variants of Trex. Some are thinner and use less material, some are thicker. I got the high end thick stuff, under the logic that I’m saving enough money doing the work myself that I may as well put the money into materials. Also, I plan to put a hot tub on here, and I plan to have reasonably large gatherings. The deck is a third the size of our interior space, and it’s very much intended for “expanding in person gatherings outside.” Once I get the railings up, I’ll add some misters and some shade canopies and generally do as much as I can to make the space useful all year round.

I’m putting the deck together with the hidden fasteners. These are buried down between the boards, and serve to hold the deck down without a lot of visible screws poking through the top. There are other ways to do it as well, involving screws and plugs, but this seemed worth trying - and, it allows the deck to expand and contract a bit with the heat. Trex is rather more thermally expansive than wood.

From the end, this is what one of the hidden fasteners looks like. You can see that it serves to hold the boards down, but it also has the tabs on the end to handle the spacing between them. This is one of the perks of them - they automatically space the boards if you’re using them, consistently and evenly down the length. Be careful with them near the ends of the boards, though - the decking core is a bit closer to particleboard than you might assume, and it’s not too hard to rip part of the bottom lip off.

Leveling the Joists

One of the first things you’ll learn when you get a trailer full of Trex is that it’s very bendy compared to wood. For wooden decks, the joists can be reasonably far out of level with each other and the stiffness of the decking wood will just pull them into shape over time. Trex isn’t that stiff, so before any decking goes down, the joists need to be planed pretty darn level. I’ve seen suggestions of 1/4” variance, and I’ll suggest that you probably want less than that, if you can at all manage it. A low joist wlil certainly pull the decking down a bit, and it’s visible if you look from the right angles. The better you get this, the better the surface will look from a “Tour of my errors” perspective.

My tool of choice for this was a DeWalt surface planer. No need for a cordless one, plug in is fine. I bought it for this project, though if you could rent one, that might work too.

The process I used is pretty simple. I used a 6’ hunk of box tube and just went back and forth on the deck, finding the high spots. Rock it back and forth - it’ll show the high regions. I marked those with some chalk to indicate where I need to grind things down. You could also use a stringline across the whole deck, but it doesn’t matter quite as much if the entire thing is flat - just that relative sections are lined up with each other. You won’t notice a half inch variance over 10 feet, but you will notice half an inch between two joists.

Set the planer fairly shallow. If it’s running too deep, it’s a lot harder to control the cut, and it doesn’t do as good a job. Start with it shallow and adjust as needed. Once the areas are marked with chalk, spin it up, start sliding it along the joist, and drop it down. You want to feather the ends so there’s no step - just a smooth transition from the bare wood to the stuff you’ve cut down. Do this repeatedly as needed, letting your box tube guide you to something nearly level. Again, the closer you can get it, the better. Trex won’t level the joists - the joists will form the Trex.

For the blocker board framing, the key is to ensure that the two side boards on each side are level, and that the center board isn’t any higher. It can be lower - that’s no problem. But if it’s higher, you’ll have trouble getting everything to line up properly. So take your time on these. Again, it’s more important that they’re smooth across the transition from one side to the other than that everything is 100% level across the deck. You should probably hit the trimmed edges with some wood protectant after you’ve trimmed the edges, though - the pressure treating doesn’t go that deep.

Also, use metal framing. Seriously. Most of these problems go away with metal deck framing, and this project has made me wish I’d figured something like that out early on. I try to make 1/32” tolerances with pressure treated, and it just doesn’t do it.

Also, replace any joists that are just really trashed. This is one upside or downside of the process taking a while - the joists can dry and weather and you find out which hunks of pressure treated that looked perfectly fine when you installed them didn’t age right! Fortunately, they’re quick and easy to replace, and it’s well worth doing this before you put the surface on.

There exists a downside to running a planer over random pressure treated wood, though. I clearly hit some staples or something else embedded in the wood. The blade is a bit chewed up now, and probably needs replacement before any sort of quality finishing work. However, it’s totally fine for this sort of thing, so I’ll keep it around and replace the blades when I have cause to do so.

Joist Tape

One of the common questions I’ve gotten working on the deck is, “What the heck is that nasty black tape you’re applying to the joists?” It’s joist tape, and it serves an important purpose if you want a long lived deck (or so they claim). Apparently the problem people ran into with the Trex (and other synthetic) decking material is that the frame would disintegrate from rot long before the surface was having problems. This is mostly a result of water intrusion into the wood, along the nail/screw paths. The tape is designed to seal the top of the wood such that water won’t get into the grain and cause rot, improving your frame longevity. Sounds good to me!

I’m using TrexProtect, since it’s designed to work with the decking I use. It’s a thick rubbery butyl tape, and if that makes your fingers cringe from the sensation you expect, oh yeah. It’s that, and worse. It’s nasty, sticky stuff. But, that’s what it’s supposed to be.

Deploying it is far easier with two people. Have one person working in the deck laying the tape down, and the other person at the far end, simply holding a tape spool, letting it come off, and making sure it doesn’t touch anything. It will stick to anything when it touches - deck, person, pants, weeds. So try to avoid that. If you’re laying it down, just guide it into place and stick it every foot or two - it’ll get pressed into the wood as you lay the surface down, so you don’t have to do anything but make sure it’s not going to come loose before you’re finished with the decking.

Cut the end with your handy pocket knife. You’ll rapidly discover that the stuff leaves a snail snot sort of residue on your knife that builds up. When you’re done, a bit of kerosene on a paper towel does a grand job of cleaning all the nasty stuff off. Water and soap don’t really do much against it, since it’s designed to be waterproof. If you don’t have kerosene around, try something else similar, but… come on, it’s 2023, why don’t you have kerosene for your low-blue evening lanterns?

Snake, it’s a snake, ohhhhh, it’s a snake!

Badger badger badger… erm. While working on the deck, we disrupted a pretty good sized bull snake that was lurking around the house. It really didn’t want to be disrupted, but neither did I want it to get stepped on, so we chased it off. You’re welcome back to eat mice, just not while I’m working on things! I do like our snakes - we typically have some bull snakes and a few blue racers around the hill.

Blocker Boards

Another common question I get is, “Why do you have those perpendicular boards from the main run? Why not just run the boards end to end like most people do?” The reason for this involves the thermal expansion of Trex - which is a good bit higher than for wood. It doesn’t really swell with humidity, but it does get longer with heat. If you have butt joints on the boards with enough gap for the summer, they’ll end up very far apart in the winter as things shrink. So, the recommended design is to have these “blocker boards” for every section. It cuts the effective gap in half for each set of boards.

When installing them, the key point is having the two for each section absolutely parallel. Measure it in a bunch of spots and ensure it’s parallel. There’s more than enough flex in the boards to either align things or throw them off, so feel free to use plenty of screws to hold it in place once you’ve lined things up.

Good metal tape measures are your friend! You’ll use one quite a bit. I’ve got a nice long fiberglass one, but it’s also a tiny bit stretchy, and that’s just a good way to get erratic measurements.

Once the blockers are in for a section, cut the rest of your boards to length as accurately as you can. Once again, the big compound miter saw shows up. A friend, years ago, suggested that if I was going to get something like this, buy the biggest, baddest one I could find, put it on a nice work stand, and don’t look back. Once again, he was right. This is easily one of the most common tools I use around the hill. I did put a Trex-specific blade in, and while it makes nice cuts, I can’t actually tell the difference between those cuts and the cuts of a reasonable quality wood blade. Except that the wood blade cuts an awful lot faster. So you can probably skip the composite-decking specific blade if you want to save a few bucks. There are some differences, but I can’t tell the difference in the result.

You need to trim both ends of the board. The end from the factory is a sort of rough cut end that’s jagged, ugly, and not always exactly straight. So trim off something like a half inch from the first end to get you a good clean edge to work from. Then measure and cut. For the 16’ boards, this is a two person job - they’re too long and floppy for just the table supports, though you could probably rig something if you had to do it solo. But I borrowed some friends, and we were able to get through the boards quickly. Load, trim, slide, measure, trim, send off to be stacked.

These are not my kids working! Perks of having regular gatherings at your place (which the deck is designed to improve): You can borrow other people’s kids as hauling labor when needed! They’re just stacking the boards up on a finished section for easy access when I’m slinging them out. Two people cutting and two people hauling made quick work of each section - and, yes, the boards are a lot easier to move with two people. They’re quite heavy, and very, very floppy.

Finally: Trexing the Decking!

With all that done, one can finally start actually attaching the decking!

Start out by throwing a bunch of the boards across the joists. You get half the work of deploying joists done, and, more importantly, you get a lot of nice walking and sitting area. Also, a great spot for your kids to run back and forth and complain about you taking their jungle gym away. My kids, while nominally approving of the deck, took a while to get over the disappointment of their latticework of wood being covered up. It was quite the popular thing to run across!

In terms of tools for this part, I ended up mostly using an impact driver (with a different bit in it for the hidden fasteners), a drill (for toenailing the boards in - I’ll cover that later), a square for ensuring things were aligned (after learning the hard way that they aren’t always going to be perfectly aligned), a screwdriver for various alignment tweaks, and, of substantial use, a good dead blow mallet for ramming the boards into place. Cordless tools are properly nice here… you don’t end up using that much battery, but the portability is nice. I’ve torn apart a bunch of battery packs and I object to the DeWalt 20V Max stuff the least of that which I’ve pulled apart. I do like my DeWalt tools…

The first step is to get the edge clips in. These are aligned with the starting edge of your deck and are somewhat critical to get lined up - or, at least, to get close enough that you can get the first board straight. If it’s curved or crooked, everything else will be. The good news is that these bend a good bit and you can use a hammer on the board to align things. The bad news is that you shouldn’t just beat the carp out of the board, because it’ll end up somewhat wrong. Try to get the spacing between the wall and the clips even.

Also, you did caulk the gap between your ledger board and the foundation, right? It helps keep water from down between the ledger board and foundation.

Depending on the spacing, this may be a good place to use your handy right angle drive. I still can’t believe how well this thing holds up to an impact driver.

If you’re doing long board runs lined up through the blocker boards, it is really important to ensure things are lined up. Trust me, you’ll see it if they aren’t. A good metal square here solves a lot. Every board, ensure it’s dead on with the neighbor across the blocker - there’s a lot of adjustment room, but you need to check it, because if you don’t, alignment will drift.

To put the hidden fasteners in, you’ll either need the Trex tool to do it, or a scrap chunk of decking. I tried both, and settled on the scrap piece - it just worked better for how I was installing things, and seemed easier to use to align things properly (given the persuasion of the dead blow hammer). When installing the hidden fasteners, you’ll just use the scrap piece to hold them in place and then drive the screws most of the way in - but not all the way. Consult the Trex Installation Guide for details, of course…

You want the screw head below the top surface of the board, but not all the way in or it’s a pain to get the next board in. Leave it too far out, and it’ll chew up the edge of the two boards as you install the next one. You’ll get a feel for it quickly enough.

Your hammer is used to ensure the boards seat tightly against each other. The hidden fasteners have those little tabs to handle spacing - so all you need to do is ensure that the boards are tight up against the previous one - assuming you got the first one in straight. It just chains out from that one, so take the extra time on the first one.

Run down a board, pounding it into place, and installing the hidden fasteners. I found that I could do every third one and then the two previous - get the third one pounded in firmly and the new fastener screwed in place, then do the previous two and there’s no need to hammer things. Just get the fasteners in tightly enough. I also found that it worked best if I got the far end secured with one fastener, and then came back to the other side to work across - it was easier to keep everything lined up and tight. You can feel when the board is firmly seated - once it’s firmly pressed into those tabs, you don’t need to keep pounding it.

According to the installation guide, if your area has high day/night temperature swings, you should secure the edges of the boards in place with a toenail screw. This means putting a bare screw in at a 45 degree angle, down through the Trex and into the joist. I used a handy cordless drill to do it, and my suggestion is to use as high a speed as your drill supports. The Trex is slippery to a drill bit - so I found that the best approach was to get the bit spinning, touch it lightly to the Trex at the desired location, and then add some pressure to chew through. If I tried to place it first and then drill, it would reliably walk across the groove and make a mess of things. Drive a screw in, and you’ve reduced the amount of walking around the boards can do. I definitely have high temperature swings - 105F during the day to 65F at night isn’t unheard of, and as I’ll show later, the deck gets more than a tiny bit hotter in the sun than the air is…

After each row is in place, go through and tighten the previous fasteners. You will snap some screws off - the amount of torque they can take seem to vary wildly between batches. Some were constant, some were infrequent. I found that while I could move the connector over slightly and drive another screw in, it also worked just fine to simply put another screw down the same hole - it would work in next to the existing screw and fasten things tightly. Yes, you should deal with the snapped ones, though. Every joist should have a fastener to keep things tight.

Do that… uh… a lot… I didn’t finish out the edge, because that’s going to be done with a square edged board routed to fit, and I’m coming back around later to do that.

Tricks and Tips

My deck involves a lot of sitting on the south side of the house, facing north. I rapidly worked out that my neck was going to be charcoal if I didn’t do anything - so I came up with a handy contraption to keep my neck in the shade. It worked well - so well, in fact, that my wife and kids got me a hat with a built in version of this thing for Father’s Day. Mostly, I think they were tired of seeing me wandering around with this contraption…

There were some days that were just hot and calm, with no breeze. Well, when all else fails, “making your own breeze” is totally an option. This is one of my “I have no idea how many CFM it is but a lot” air movers that moves air, sand, grit, and small rocks. However, it does a nice job of keeping me cooler while working. It was quite loud, though… so pick between that and music. A smaller misting sort of fan may have been a wise purchase, but DeWalt doesn’t seem to have any options there.

The boards are very flexible - in both directions. If you can, and you notice some boards are curved from how they’ve been sitting, having them bend away from the previous row helps make things faster - once both ends are in place, the center will be held a lot tighter. I also played with caches of hidden fasteners across the deck, but without someone to regularly keep moving them, ti’s easier to slide the big bucket up and down the deck. Plastic slides on this really easily.

Don’t miss the forest for the trees. Every few rows, get up, step back, and sight down the deck. Make sure everything still looks straight - if you have long sight lines down the deck like this, it’s easy to get things “not 100% right,” and it will show. A good deck rug is always an option when you find this out too late. Someone asked why a lot of decks have 45 degree decking, and after having done this deck, it’s because having everything at an angle like that will hide an awful lot more mistakes and misalignments than this sort of fully square arrangement. You can do it, just be aware that you’ll have to pay a lot of attention to the long form alignment while you’re installing the deck.

I’ve still no idea how to make this task go faster with more people, though. With a skilled team, the right answer would be to do the sections in parallel - just making sure that they remain aligned. With myself and some extra hands, I’m not sure how to speed it up. I was able to get through one section in two solid mornings of work - though I was faster towards the end. By mid afternoon, it was just too hot to continue working out here.

Deck Thermal Images

Do you have an appreciation for just how much radiant heat off a house and deck impacts comfort on them? I didn’t - but, after having installed a deck in summer, I do now! One of the problems I found rapidly was that the decking gets very hot to sit on by early afternoon. I was curious as to what that meant, and discovered several interesting things. First, my blocker boards, with no airflow, get really hot - quite hazardously so, at times. Yeah, that’s 164 coming off one of them. The rest of the surface gets hot too, and while I’d not thought about it before, the section over the joists is hotter than the rest - because there’s no airflow around them. The rest of the surface has some cooling airflow, but not the section over the joists. I’d have assumed the thermal sinking of them would mean they were cooler, but, not in the summer!

Also, the side of a house gets properly warm. It’s like standing in a shipping container in the heat, and even though you don’t notice it for short periods of time, sitting out there for hours in the heat, the radiant heat makes a big difference. I was going through almost two gallons of saltwater a day while working on this deck in the heat - and still ended up somewhat dehydrated.

That’s it for this week, because I’m trying (mostly failing…) to keep my posts at least a bit shorter. Next time, I’ll talk about the edges, as well as my stairs!

Deck Part 2: Framing

2024-01-13T16:00:00+00:00

I ended the last deck post with the full row of concrete pilings completed, and the ledger board on the house. That still leaves a few holes left, and, of course, the rest of the wood framing! But after that, I’ve got something ready to Trex!

Figuring Out the Mid-Beam

Like most of my projects, I tend to buy ahead of time. This pile of framing lumber spent a while on the trailer until I needed it for something else, at which point I offloaded it (while rather sick - it wasn’t much fun) over to the side of the house.

Pressure treated is always annoying to stack and dry, because it will warp given half a chance, so take the time to make sure it’s going to dry at least mostly straight. I probably should have added a few more spacers in the middle between boards for air circulation here, but it was more or less fine.

You’ve probably seen regular joist hangers - but did you know there are “hidden” style ones for the corners? This neatly avoids having bits of metal hanging off to the side, or the joist somewhat inboard on the framing. I’ll eventually have another joist just inside this for deck surface reasons, so I’m not too worried about the strength of this hanger (it’s a lot of screws into the very edge of the ledger board).

The reason for getting this joist mounted is so that I can start working out the beam heights - both for the far beam, and then for the midpoint beam. I know I have to dig down some for the mid-beam - but the question is, “How far?” And to answer that, I need some good references. A five gallon bucket and some rocks makes a fine temporary support for deck joists! If you’re wondering why I’m putting in a beam at what looks like the 5 foot mark for something that doesn’t need it, the intent is to eventually put a hot tub on the deck, and this is just extra support in the section that’s likely to be heavy. If it’s worth DIYing, it’s worth over-DIYing!

Cut some posts, temporarily align things, add a few more joists to get the spacing, and I can place the mid-beams to figure out exactly what has to happen to make everything fit properly.

The general concept here is simple enough: Pour concrete, add supports for the beam. I dug down so the beam would clear the holes, worked out heights for the supports (which you’ll see a bit later), and got ready to do yet more concrete pouring.

For these holes, since the height of the top mattered a bit more, I did anchor the form tubes in place. A 4” long screw or three, driven through the side of the tube, straight into the dirt, will work very nicely indeed to anchor a tube for concrete pouring! I didn’t have any problems with the tubes shifting during the pour. Because they had to be low to the ground, I wasn’t able to use my normal elevated form holders (squares of 2x4s).

Once the concrete had cured for a week, I brought out the usual set of tools: A rented hammer drill, and my air compressor. Bore a hole, blow it out. Repeat until done. At this point, I really should probably just buy a hammer drill…

The post bases I found for this support beam are almost perfect for the sort of thing I’m doing here. They have the post cup at the top, then a 3/4” threaded rod welded on, with a nut and washer to set the height. As long as the top of the concrete is close, these can be adjusted up to fill the gap and spread the load! Again, avoiding precision concrete work is ideal for me, though I’ve recently learned some tricks I want to try next time I have to bore precision holes in concrete.

I’m using two beams, and the center point has two of the post bases - one for each side. This gives me some adjustment, and doesn’t involve having a beam only halfway supported at the end where they butt together.

The beams even fit! Once the joists are in place, I can lift these beams with the nuts and secure them in place. If there’s a bit of tweaking that has to happen later, someone small can slide under the deck and add another turn or so to them - but I’ll set them on the tight side initially to compensate for some sag.

Moving More Rocks

Like any project on our hill, “putting holes in the ground” means “removing rocks from the ground.” This is no exception, and the tractor, complete with new tires, does a wonderful job of moving rocks around the hill. With the pallet jacks on the back and a random pallet (this one is from one of the loads of concrete), the rocks can be rolled on easily, even if they’re too large to lift - though nothing here was quite that large. I collect all the rocks I remove over on a corner of the property for later use - and the pile is getting rather impressive!

Finishing the Beam

Once I finally had some tolerable weather, I finished out the beam. The first few segments were done, but that’s hardly a full deck worth. Keep working down, leveling things, cutting to fit, and eventually the beam runs all the way down the house! Here, it’s just sitting in place for test fitting, because I had it in and out constantly to adjust things - but ready to secure.

Before anchoring everything together, I actually followed the recommendations and put some wood preservative stuff on the cut ends of the posts and beams. I don’t think it really matters out here, but you’re supposed to do it, and I want the deck to last, so I got a bottle and applied it. The stuff I found is insanely thin, and wicks into the wood like nothing I’ve ever used before - which, fortunately, is exactly how it’s supposed to work. You can dip the wood in it, or apply multiple heavy coats with a paintbrush, which is the method I used. On the shorter posts, it would sometimes run all the way down through a crack in the wood and pool on the underside - exceedingly low viscosity for something you apply with a brush.

I ran into an interesting problem once or twice as well: My post bases wouldn’t fit on top of the beams. Like any lumber project, dimensions of lumber are variable (moreso for pressure treated), and some of my posts were cut from something that varied wide. It wasn’t going to go on without cutting the wood down - I tried.

Nothing like a good miter saw for shaving a tiny bit off the edge! One clamp to hold the wood in place later (because I’m not going to put my hand quite that close to a blade!) and I was able to shave just a bit off the edges.

After shaving a bit, the wood still has the pressure treatment slits, and can see sort of how the stuff soaked in - the top stretch is harder wood, and the preservative didn’t soak in as deeply. I’ve never shaved the edge off pressure treated before - it leads to some visually interesting results!

The posts are held down, and the beam brackets are held on, with a large number of screws that are sold by the same company (Simpson), and are designed for this sort of thing (they’re structural screws). Often, stuff like this is held together with nails, and I just hate driving that many nails. I also lack a good nail gun. So, screws it is. You can see the wood preservative on this piece of wood, having soaked in and through it. These will be well hidden under the deck, so I don’t care that they look a bit weird.

Before putting the beam in, it’s important to tighten the post base to the concrete. Once it’s tight, it’s not going to lift - but you can also adjust it slightly side to side if needed by using precision taps from a sledgehammer. Try to get it right before you tighten it, though!

The beam slips in, and an awful lot more screws go in the side to secure everything in place. Even without any joists in place, the beam is solid.

And one beam, very well secured in place, with a midway beam also ready to help support the hot tub.

Adding the Joists

The process for adding joists is as simple as it is repetitive: Add a joist hanger to the ledger board, add a joist. Secure at the far end with a hurricane tie to the beam (we do, on occasion, get a lot of wind out here). Then add the blocker boards in between them to prevent the joists from “racking” and collapsing.

It eventually ends up looking something like this - after a lot of joists and blocker boards. I’ve used the wood protectant on all the various cut ends here, which involved a lot of letting things sit, standing upright, in paint trays. If some boards are being obstinately twisted and need pulled back into shape, a selection of clamps and ratchet straps goes a long ways - apply as needed. You can get a lot of force on big ratchet straps, and once the boards are secured at a few points, they’ll relax over time into the desired shape.

One thing that’s been very, very useful for this project is a long fiberglass tape measure. I spent a lot of time making sure the end joist was straight, so I can use it as a reference point for all the other joists. I spent a lot of time with this tape laid out over the joists for reference - with everything being 12 inches on center, it didn’t take much math to figure out the alignment!

Preparing for Railings

My local railing supplier gave me a handy template for pre-drilling the joists. I’ll be bolting the railing supports to this after the deck is finished - but it’s easier to drill them now, before all the decking is on top.

How does one fit a normal drill between 12 inch on center joists? You don’t! You just get a “stubby” bit for your impact driver and use that instead!

Right Angle Drives and Final Framing

When I did the initial framing, I knew I was going to have to replace a few sections for the deck surface design I planned - but I’d not yet finalized what I wanted to do (I eventually went with “the cheaper option” instead of “everything even”). Working between joists when you need to add a range of “blocker boards” can get tight - so I had an excuse to pick up a 90 degree impact driver adapter! Somehow, this manages to survive the abuse of an impact driver and turn the impact torque 90 degrees. I’m interested in what the gears look like, and will probably pull it apart after it fails just to see what’s in there - because “impact driver” and “gears” aren’t two things I normally consider compatible.

But you can fit it in some nice tight spaces! The only thing I wish it had were some way to anchor it to the gun - it turns most things into a “Well, I could use three hands…” sort of challenge.

And, while I won’t claim this style of “Two boards, a roughly 1.5 inch gap that doesn’t matter, and two more boards” is the most efficient way of accomplishing the task, it does work! How this interacts with the final design, you’ll see in the next post.

With the framing done, all I have to do is plane the boards so the top is level (it matters with Trex, far less so with wood decking), apply joist tape, and lay down about 660 square feet of Trex…

More Flip Phones: The Sonim XP3+

2023-12-30T16:00:00+00:00

Alright. Let’s end the year on something nifty! It’s been a while since my last post about flip phones (the AT&T Flip IV), and… well, I spent a while back on the iPhone, because iOS 16 and Lockdown mode really solve a lot of my problems, and because the Flip IV basically stopped working. But I’ve been back on a flip phone again for quite a few months now, and it’s that one I want to talk about in rather more detail: The Sonim XP3Plus. It’s a glorious, rugged, overkill, brick of a modern flip phone that does almost nothing useful. And I love it!

I’m going to talk about the actual phone first, and then some of the surrounding ecosystem issues later on in this post.

The Sonim XP3Plus

Have you ever thought that modern phones were a bit on the delicate side? The XP3+ is the opposite of that. The thing is an absolute brick. It’s half a pound of brick-like flip phone goodness with the ergonomics of a Glock, and “casually tossing it to people to check it out” is the sort of wonderfully entertaining thing you can do - without worrying about it, should they miss!

Fundamentally, it’s a flip phone. It clamshells up, you open it, make calls, receive calls, send texts, and slam it shut on the scammers, spammers, and fundraisers that make up the bulk of modern calls (though I am slowly training people to call me again - it’s kind of nice).

But I cannot reiterate enough: This phone is massive. It’s thick. It’s heavy. It feels amazing in the hand. I’ve got a stack here of some phones, with the XP3+ on the top, crushing the Flip IV in the middle, and an iPhone 2020 SE (so iPhone 6, 7, 8, basically anything of the last few generations with a home button) on the bottom. But, on the plus side, there’s absolutely no reason to bother putting this phone in a case!

The keyboard is a pretty standard sort of flip phone layout - your normal dial pad, some arrow buttons for navigation, call, hangup, and a few extra buttons along the top that various apps use for menus and settings. The keyboard feels robust - one of the reasons I stopped using the Flip is that the keyboard literally failed on me, and was double and triple pressing, missing presses, etc. I’ve not had this one long enough to give a confident answer about longevity, but so far, it feels far better - and with the phone being milspec rugged, I have some greater confidence in the ability to handle actual use for more than a year.

If you’re the sort of person who is frustrated that your smart phone is utterly useless in the cold with gloves on, there’s no problem here - you can use it with gloves, and it works fine.

Below the keyboard is a very big speaker, and if your problem has been that your speakerphone just isn’t loud enough, this should solve your problems.

There are two ports on the phone - one USB-C charge/interface port (defaulting to charge only - I’m glad to see this change over time), and a headphone jack. If you have headphones in, the phone also will work as a FM radio in the way that most of the modern flip phones do (which I appreciate). The red button is, by default, an emergency button, but you can reprogram them in the phone for a range of features.

The other side has another programmable button (I use it for the flashlight), and volume control buttons. Again, big push buttons, glove friendly, satisfying feeling.

The back comes off and exposes a removable battery, which I find rather handy in a range of situations. I like being able to physically remove the battery from a device - a lot. And I can do that with this phone. You may also notice that the battery is massive in terms of capacity (at least for a flip phone) - 2300mAh, or 8.85Wh. This phone will casually do a week between charges, or closer to 14 days if you turn it off at night. I’ve no idea what active use time is, because there’s not exactly much to “actively use” this phone for, but I can say that if you’re the sort who forgets to charge your phone often, this one is no problem at all. Just charge it weekly and you’ll be totally fine. When it gets to the “low battery” warning, you still have a few days left before it dies.

Under the battery lives your SIM card slot and a SD card slot for the storage of whatever you feel like. About the only thing that could take up much space on this phone is music - and while the speaker isn’t amazing for it, it’s still quite a bit better than most smartphones. The media player supports a range of formats, including FLAC and Vorbis (!). It also supports video formats, and… sure. It will play a video if you must, on the tiny screen. It even supports H.265 and VP9 playback. I’ve no idea why you’d do that, but if you want to, it will. It’s actually a fun little party trick…

The back cover has a nice stout gasket around the battery compartment to keep the water out of the parts that matter. You’ll also notice a screw in there - unlike a lot of the flip phones and removable battery devices out there, the back won’t pop off when the phone is dropped. It just stays firmly in place, keeping everything protected. It’s a nicely thought out device for rugged use!

Environmental Ranges

Have you ever had a problem with your phone shutting down from either heat or cold? That shouldn’t be a problem with this phone. It’s rated to operate from -4F to 131F. It’s also rated for total immersion in water (the IP-68 rating), explosive environments, and it passes the MIL-STD-810H tests, which are a wide range of “Abuse the device” tests that mean it should continue to work even in the stresses of a warzone. I mean, is your phone rated to withstand typical gunfire vibration and explosive decompression?

For comparison, an iPhone 15 is rated for operating temperatures of 32F to 95F. Kind of explains why they have trouble in the sun and ski slopes, doesn’t it?

Does it matter? That probably depends on how hard on your phone you are. If you can keep a naked or near naked smart device intact for 5+ years, it’s of no particular importance. If your phone is held together with packing tape and screen protecters after six months, this might hold up a good bit longer! But, practically speaking, the phone is literally military-tough, and you’re not going to have any problems with it from the environment. Water, dust, rain, ice, sand, temperatures, shock, abuse, dropping it (it has a 1.5m drop rating - onto concrete)… it’ll just keep on going. Toss it in your pocket or bag and don’t worry about it. The phone will be fine.

Using the Phone

How is it to actually use? As long as you don’t expect it to do anything other than calls and texting, and you remember your T9 predictive texting, it works just fine! The homescreen has a simple set of options - contacts, messages, and camera. Notifications live on another screen (and with the volume of the ringer, you’re not going to miss them in a noisy environment easily), and to the left is a settings menu.

Settings are simple enough - and for a device like this, you can safely leave WiFi and Bluetooth off constantly, because there’s just no need for them. A tiny bit of cell data is enough for receiving MMS messages (group texts or photo messages, if your social circles use them), and… I guess there’s a browser, if you’re bored. It doesn’t do much. I consider this a massive feature. It does support being a hotspot, though - so you can at least do something with the couple gig of data that most modern cell plans come with.

The menu opens up a grid of applications. Several of them only matter for enterprise-managed stuff, and I’ve no idea how they work because my plan doesn’t support EPTT. Everything else, for reasons that make no real sense to me, are in the “Applications” folder.

Buried in Applications, you’ve got the usual stuff for a flip phone - downloads, backup, calendar, clock, image gallery, calculator, a music player, etc. If you sideload stuff on it (see below), you’ll find a few more things in here, though there’s not a lot that really supports this sort of limited interface. It really mostly is what it is, and you’ll find extending it somewhat frustrating.

The Front Screen

On the other hand, it has a useful front screen. Some phones (the Flip IV, for instance) have a front screen that is “A useless decoration.” The XP3+, on the other hand, has a wonderfully useful front screen for a range of things. By default, it shows the phone status - time, date, mode (“silent” here), cell signal, battery… which, by the way, I charged 5 days ago in these screenshots. But the buttons below allow you to unlock the screen (right front button, then volume down on the side) and change things around. The most useful, that I use regularly, is changing the sound settings - though the label indicates what you’re changing to - not what mode you’re in. Here, I’m in the silent mode, and if I hit the button, I’m going to be in Meeting mode (vibrate, mostly).

You can also see recent calls - nothing useful, really, since it’s just scammers and spammers, unfortunately. But you can see what call you missed without bothering to open the phone, and that’s quite handy to call back (on speakerphone, even!) when you miss a call.

Even nicer, when a text comes in, you’ll see the sender pop up on the front screen, and you can view it if you want (the volume keys will scroll through it if it’s more than can fit on the screen at one time). This is what a front screen should be - a micro-interface to the useful features of the phone. The screen is also rather blindingly bright, and there’s no reason to bother with a flashlight at night. I’ve not found a way to dim this screen…

But, again, big chunky buttons you can use with gloves. There is literally no capacitive touch anything on this phone.

And it does support MMS - group SMS messages. That’s important to me. I know it’s weird, but this is how a lot of my social circle communicates, even in 2024, and one of my standing policies with phone experiments is that I won’t cause excessive pain to other people. Not supporting MMS group texts is the sort of thing that does actively annoy other people in my life.

How’s the Camera?

Another problem with some of the flip phones out there is that their camera has an awful, 2003-grade sensor - in the mid-2020s. The XP3+ comes with an 8MP sensor that, while not amazing, isn’t half bad either. The Flip IV had a 2MP camera that was… well, present and accounted for. And that’s the best I’ll say about it. The XP3+’s camera is utterly usable, though you won’t ever mistake it for the latest iPhone with 72 computationally augmented lenses. I’ve taken a few photos to demonstrate that it’s absolutely workable, though it gets quite grainy in low light.

Standard indoor lighting of some stuff my kids have been doing this winter:

Afternoon lighting of lanterns, and you can certainly see that it’s not the crispest image. There’s some (ok, quite a bit…) of noise up in the ceiling, and the fine detail on far objects isn’t great. I’m not going to be replacing any of my standalone cameras with this, and it’s not up to modern smartphone performance, but that’s fine in my book. If you’re reading along this far, you probably aren’t too happy with a smartphone anyway.

Outside, it’s… fine. Mostly. Again, it’s not the crispest, sharpest images you’d ever find, but neither is it awful, either. It’s an adequate camera for the sort of stuff you’re likely to use a flip phone camera for, and if you want to get better images, well, carry a standalone camera with you. Don’t expect fine detail in a crop. And, yes, I have gotten back in the habit of carrying a standalone pocket point-and-shoot when we go places. There’s no shortage of “rugged” cameras that can take a drop, get wet, get pounded around, and still be entirely fine. They pair rather well with a rugged phone.

Installing Apps: Yes, but…

If you really insist, you can sideload Android APKs onto this device. Don’t expect them to work well, if at all. There’s no Google Play services (yay!), the screen is tiny, and there’s no touch input at all. This violates “literally all the assumptions of a modern app developer.”

Do the standard “Android Settings/version” thing to enable an exceedingly verbose Developer Mode menu, enable USB debugging, plug into a computer, and then “adb install [whatever].apk” should manage the install process for you, and you should find the newly installed application over in the Applications list.

But it probably won’t work. Most modern phone applications do not support “arrow keys only” style keyboard navigation, which means you’re limited in what you can do - although if you really insist, you can pair a Bluetooth keyboard and mouse to get a mouse pointer (and better text input for long text messages). Which probably won’t help, because all the UI elements will be on top of each other or off the screen with no ability to scroll to find them. Nobody seems to test their stuff on a keyboard-only 240x320 interface - which is sad, because quite a few things could work quite well on this screen, with a flip phone friendly UI option.

And some stuff is pretty crash prone. I was able to get Element (the Matrix client) to install and it was marginally usable with a keyboard, but it crashed every time it tried to do anything with encryption. I don’t know why, and at this point, I simply don’t care. So, don’t expect to sideload a range of interesting things onto this. In practice, it just doesn’t seem to work. On the other hand, if you work for a company with an Android app, I’m happy to help you debug it for flip phones, because there’s no reason stuff can’t work in this environment - and I’ll suggest that supporting modern flip phone users is a good distinguishing trait at this point! There are dozens of us!

What Happened to the Flip IV?

Why this phone? Why not the Flip IV I was using? Well, let me tell you…

The fundamental problem with the Flip started with the hardware: It was not good. I had problems with moisture on cool nights - if it got foggy, odds were good that the phone (in my pocket) would start malfunctioning. More than once, the exterior display simply stopped working in the moist air. It would work again once things dried out, but it wasn’t exactly confidence inspiring.

The bigger problem was the keypad, though. You have one job, keypad: Work. And after a year of decently heavy use, it simply didn’t. I would be getting double and triple presses, which makes even dialing a number hard. Let’s not discuss trying to T9 text with that behavior - it’s almost impossible.

But, worse is the long term behavior of KaiOS. I’d known from very early on that it couldn’t handle having more than few hundred text messages (total - not conversations, just the number of received and sent messages) on the device before slowing down dramatically, so I had to prune threads regularly. But after about 10 months, the phone stopped notifying me about incoming text messages or calls while closed. I’d have it closed on the counter, silent. Open it up, and, boom, about 5 text messages and a missed call notification all come in at once. Unfortunately, I consider that entirely unacceptable behavior in a phone - if I have you turned on, and you’re connected to a tower, you’d better notify me when something comes in!

It’s possible that I could have factory reset things and resolved some of it, but that’s unlikely to fix the hardware issues with the keyboard. I went back to the iPhone 2020 SE, with Lockdown mode turned on, unhappy with the state of it, but needing a working phone. But I kept looking around, and eventually decided to try this device out.

Why Not a Nerfed Smartphone?

I do pay attention to some of the online discussions about smartphones versus flip phones (or candy bar phones, any form of “not-smartphone”), and one thing I see regularly is, “Well, why bother doing that when you can just disable all the stuff you don’t like on your existing smartphone? Why buy something different?”

It’s a fair question, and I’ve certainly run a more-nerfed smartphone than most, but I still think carrying an actual flip phone is superior, for several reasons (and, yes, I regularly make use of all of these - this isn’t just theory).

First, if you’re not going to make use of most of the features of a smartphone, why deal with the compromises of a smartphone? Find me a smartphone with an extended temperature operating range, a week+ battery life, drop proof, etc, and… well, it’s probably not particularly cheap, or particularly good at being a smartphone. And by the time you put a case on a smartphone that makes it many of those things, it’s just a larger version of this device that I still can’t use with gloves on!

But the second (and more significant to me) reason is best described as “style and conversation starter.” Pulling out a chonk of a flip phone, and literally tossing it to people, makes an impression that no nerfed smartphone can ever do. People (of a certain age, at least…) remember flip phones, but most people have no idea they still exist. And it opens doors to talk about how I just use it for calls and texts, nothing else, no social media, yes, I could put music on it, but I generally don’t. I’ve talked to more than a few people who are very interested in them, having recognized that smart phones are violently human-toxic, but not realizing the state of the current alternatives.

It shouldn’t surprise anyone here, but I’m quite opposed to smartphones at this point. I think that they’re a massive net negative on society, and that we’d be better off without them. Yes, I know about all the fancy things they can do, and having rather predated them, there are plenty of ways to do those things without them (some of which no longer exist, but we can bring them back), and I think the negatives for society far outweigh the positives at this point. So being able to literally toss an alternative to people (with the confidence that if they miss, oh well, it’ll bounce) is worth a lot to me.

Why Not an Old 3G Flip Phone?

What about all the older 3G flip phones and candybar phones you can get for a dime a dozen? Why not one of those?

Mostly because the carriers all collectively “fixed the glitch” and have obsoleted the entire stock of them. In the US, at least, just about every carrier has recently (a) turned down the 3G voice service that would have allowed them to work, and (b) gone to VoLTE - Voice over LTE. This, of course, requires carrier certification to run on their networks, and isn’t well supported in random older devices that were 3G only. Even the ones that support 4G connections for data typically don’t have VoLTE support, and no carrier I’m aware of is doing 4G voice. Gotta sell those upgrades!

So, unfortunately, none of those are working options anymore.

But Work Needs Me To Install This App…

I also hear objections about “But work needs me to install…” - and unless you’re very young, I suggest the following hard principle: “If work requires me to have a smartphone, work will provide me a smartphone.” It’s amazing how well that works - but this is also another argument for “not just nerfing a smartphone.” If I have a modern, updated iPhone that I refuse to install apps on, well, I’m just being cantankerous (in a common, if not entirely reasonable view of things). If I have a phone that’s my daily carry phone that literally cannot install apps, well, now that’s their problem. Further, at this point, I’m just not merging work and personal phones anymore. Been there, done that, learned about just how tempting “checking work email” can be at odd hours of the weekend. Nope. Never again.

Updates, Security, Threat Modeling

One major downside of a device like this: It doesn’t get updates. Not in any meaningful sense. You may get a bugfix or two early in the device’s life, and then… that’s it. It is what it is. And what it is, most likely, has plenty of interesting security vulnerabilities down the road as they’re discovered in things like media processing libraries, the cell baseband, video decoders, etc. So, why would I suggest something like this instead of a device that’s being frantically updated to patch the latest and greatest 0day, 0click, hardware-enabled attack?

Because there’s nothing on this device that hasn’t gone (unencrypted) through the public cell networks, and because I can remove the battery. And, that said, there are relatively few interfaces to attack. Apple devices seem to get remotely compromised through 0click, 0day attacks that come through iMessage and some weird old image format - none of which this device supports. Or, via Safari 0days. I don’t use the browser on this as the browser is only barely usable. I also don’t install apps, so there’s nothing that can get around and pop the kernel that way. It’s just a limited set of attack interfaces compared to a smartphone.

But, beyond that, I don’t use this as “root of identity” like a smartphone. If you compromise this, you can get my phone logs, and my SMS/MMS messages, and some idea of my taste in music. That’s about it - and the phone logs and SMS messages have gone through the carrier already, subject to various ways of getting them. So there’s not really much on that front you can get from the device you can’t already get. You don’t get my Google credentials, second factor keys, or anything of the sort, because it’s not on here. And, on the flip side, an older smartphone used for those things doesn’t have the SMS/iMessage/etc interfaces open for compromise because there’s no SIM card in it (and again, try not to use the browser).

As far as room audio and location goes, well… I just treat the device as already compromised. I regularly turn it off around conversations, I regularly remove the battery, and I regularly leave it behind. Or carry it, turned off, with the battery out of it, in case I need it for something (I generally don’t). I’m trying to bring back the model of “Remove phones from the presence of activities of any interest,” and modeling that myself seems a reasonable thing to do. And I’m also more likely to carry an older laptop with me these days for the functionality I once used a phone for. My Thinkpad X250, running Qubes, tucks nicely in my backpack.

Fundamentally, I’m not sure that any modern mobile device can be said to be secure against any moderately well resourced attacker - so having one I can simply “treat as compromised” makes life a lot easier on that front. Though I’m not honestly sure what the best model anymore is. An iOS device with Lockdown enabled may still be a harder target. Hard to say, unfortunately. The rabbit hole is deep, and the computers are crappy through and through.

Final Thoughts

In some ways, this device is even less capable than the Flip IV I used to run - I don’t even have maps on it! But in other ways (the ones that matter to me), I think it’s about the perfect device for our modern age of “computers can’t be trusted with anything we use them for, and also everything is a computer.” It’s a sort of “minimally viable phone” for a world in which phone calls and SMS messages are still required. It’s not going to come apart from use and abuse, which is nice. I expect it to last me for many years (until VoLTE goes away for MegaWideband5GUltraVoiceBlasterMaxxPro, probably). And it has a passable enough camera for those times when I don’t carry a separate camera with me.

Should you get one? That depends on what you’re looking for in a phone these days. You’ll need to rework your life so that your smartphone isn’t the center of your world - which is quite healthy anyway, and I’ll be talking about that more in some other posts. But it’s a great conversation piece for exactly how and why you’ve done that sort of thing - and as many people have commented as they’ve seen mine, “I didn’t know they even still made flip phones!”

But, so far, I’m quite happy with mine!

Solar Shed Part 19: Going Lithium with EG4-LLs

2023-12-16T15:00:00+00:00

It’s been a while since I’ve posted anything about my office - and that’s simply because there’s been nothing to post about. It’s been absolutely wonderful as a daily use sort of office, doing what I need, powering what I need, and not causing any drama. I really can’t say enough good things about just how well this space has been behaving for me!

But, coming up on 8 years of use, I decided it was time to preemptively replace my lead acid bank out back with something… a tiny bit more modern. That’s right. My office energy storage is now based around lithium batteries!

Eight Years of Deep Cycle Lead Acid

First, I want to review nearly eight years of daily cycling on a 48V, 225Ah deep cycle lead acid bank. They’re a bank of eight Trojan T105-REs, which have long since been replaced by various other (similar) models, and… well, they’re a bit messy by now. Water use is going up, and despite the water miser caps, misting has been increasing too, which is a big part of why I’m replacing them. That’s a good indication of end of life on lead acid batteries, and while I’ve not abused this set, they’ve certainly been cycled - typically down to 80% SoC, sometimes deeper, and for a good 2500+ cycles. It’s not at all bad for a $1400 bank of batteries! The white powder is just a layer of baking soda I tend to sprinkle after watering to help absorb some of the acid - it looks gross, but does neutralize a lot of the acid mist!

General wisdom is that “You’ll kill your first bank of off-grid lead acid batteries in a year” - so having gotten to nearly eight years on this bank means I’m doing something right!

But I’ve got some corrosion going on, and… while I’m sure I could get another few years out of this bank, “run it to failure” no longer has the appeal it once did. I’ve found out just how much lead time there is on lead acid (see what I did there? lead, lead…) this past year or two with solar trailers, and I can’t just bounce down to the corner battery store anymore and find a compatible bank. Yes, I can run off one of the solar trailers or a generator for a while, but this bank will remain in use for quite some time to come as a big UPS bank, and it seemed the time was right to swap stuff out.

Nothing I’ve said in the past has changed - my 11k word epic of “how to treat a lead acid bank” still holds just as true as when I wrote it, and treating the batteries that way has certainly worked out well for my bank.

I put my batteries out back of my office in a plastic deck box for reasons consisting largely of “I didn’t have space inside, I didn’t want to deal with the hydrogen inside, so a box seems a good idea.” Would I do it again that way? Maybe, but with a few modifications to seal the box up better, and I wouldn’t just leave the lid cracked open to vent the hydrogen. It works - but it also leads to a lot of dirt getting in. The electronics down there were behaving, but you really should seal up your battery box better and put in deliberate hydrogen vents. I’m not saying my solution doesn’t work - and if you’re on a budget, it’s certainly a good enough option - I got eight years out of it. But find a way to seal it up better.

The dirt also makes it harder to see what’s going on down there. I discovered, to my utter surprise and concern, that one end of my main battery fuse holder had started getting rather toasty - and I had no idea. The insulation on the wire is a bit blackened, and there’s that cavity melted in my foamboard insulation. I probably should have taken a thermal imager to my battery bank when it was under high load… but, in any case, things have been working.

You can also see quite a bit of corrosion on the battery terminal. If you’re doing a lead acid bank, it’s worth finding some sort of coating for the terminals that stands up to acid and won’t corrode the copper (I think a layer of vasoline works fine). The bank still works - but it’s messy, and it will eventually cause problems for you.

I think the core of my problem was using this sort of screw terminal on a fine stranded welding cable wire. Looking into it, I should have crimped some sort of terminal on the wire, and then tightened that down in here - not just tried to crush the fine stranded wire in here. It definitely got toasty. And, no, I don’t think that’s funny, cute, or any other “O-face” YouTube reaction. It’s quite concerning, and I’d have fixed it immediately had I known it was a problem. When people say “Don’t use a screw down terminal on fine stranded wire,” they’re not kidding - this is exactly why. Now you know…

LiFePO4 For Stationary Storage!

So, yes. I’m going lithium. But I’m going with lithium iron phosphate - LiFePO4, also commonly seen as LFP. This is a “less potentially exciting” lithium chemistry. Energy density is lower than the standard 3.7V nominal/4.2V fully charged chemistries, though power density is quite good. Importantly, though, LiFePO4 is a much more stable chemistry - it’s not nearly as runaway prone as the other chemistries can be, which means it’s safer for something like “the corner of a shed.” Also, the longevity on them is just excellent - these packs are rated for a whopping 7000 cycles at 80% of pack capacity (I use far less), and come with a 10 year warranty. They’re also (now) UL listed - so you could potentially use them for a grid tied storage system legally in many areas.

I don’t mind the use of high energy lithium chemistries where it makes sense. I just don’t like seeing them used in places where a more stable, long lived, less exciting chemistry is the better option. And for stationary storage, energy density just doesn’t matter that much.

The EG4-LL LiFePO4 Rack

The EG4 batteries come in a 4U rack mount form factor - and if you’re stacking up a bunch of batteries, EG4 has two rack options too - a 12U chassis that will fit 3 batteries, and a 24U chassis that will fit 6! These come complete with busbars on the side, which is handy enough that I think the rack is worth picking up if you’re using the batteries. They also make it a lot easier to roll the bank around while you’re installing it. Again, if you’re on a tight budget, you can skip it (or save even more with lead), but I think it’s a handy little gizmo.

The rack shows up in a box, and everything comes on a pallet, so the box isn’t likely to be very dinged up. But if it does show up damaged, make a note, because there is no padding - it’s literally just a cardboard wrapper around the steel rack chassis. Trust me, you’ll notice the weight when you try to get it out!

It is a 2’ cube of steel. And it’s absolutely every bit as unfriendly to wrangle around as that sounds. Fortunately, it comes with wheels. Unfortunately, you get to install them.

The casters are properly stout little units. The batteries are 100lb each, and the rack weighs most of that too - so it’s 400 pounds rolling around. These are absolutely up to that sort of weight, and even fully loaded, the rack rolls around easily in my office.

They bolt in. Make sure they’re snug, because once this rack is loaded, you’re not going to want to be under it again. On the bottom, there’s one opening for wiring. On the top, there are two. This rack is really designed to have cabling coming in the top - you can do it from the bottom, as I have, but it’s quite a bit more work than just coming in from the top. Consider that if you’re designing a system around this rack design from scratch. Of course, you don’t need to use the rack either - but “having everything in a nice, custom fit case, with busbars on the side, and good coasters” is a nice little feature to have.

With the wheels on, I promptly ran into a problem. See, I measured to see if the rack would fit back here. I didn’t measure to see if I could actually wrangle the rack back in there - and the answer is, “No, really, I can’t get it in there.” It goes in diagonally, but wouldn’t clear everything enough to slide how I wanted it. It hung up on the shelves to the left, as well as my storage bins over to the right. No, trimming the shelves didn’t help - I tried it.

Long ago, I had a whiteboard in here - which, for my use of my space, was “mostly useless.” It got removed at some point so it could be used for something else, and I’ve never put it back up. So this open space, and my need to move things around, got me thinking.

This time, I was smart. I didn’t tighten down the screws after I mounted the stuff-holder-boxes on them, so I can remove them far more easily in the future if I have to! Each of these hangs on four screws, but I’d gotten the bright idea to tighten them down in the past. Anyway, this solves my problem, gives me more storage space, and, most importantly, let me wheel the box of batteries where I wanted it!

The EG4-LL-V2 Batteries

You know how a 4U server is awkward to move around, but not terrible? Imagine one that weighs 100 pounds, with absolutely zero handholds on the back. None of that usual lip you hold onto when moving it. That’s what these batteries are. They’re absolute little bricks that suck to move. There’s no good spot to lift them, and no good way to hold them.

I found that “standing them on end, then grabbing under them” worked well enough. I’ve got the fancy ones with a built in BMS display that shows me some useful stats like per-cell voltage and temperature, as well as overall SoC. The basic interface on the left is “Four state of charge LEDs, a run light that flashes when it’s working, and an alarm LED.” Over on the right, a 100A breaker is the disconnect for the output terminals - you can turn the battery electronics on and off, and then the breaker has to be on in order to let the electrons flow.

There’s really not much to set up on the batteries - just set the ID (lower left DIP switches) to unique values for each battery in the chain (the manual comes with a table).

There are shelves in the rack already installed, so the batteries just slide right in! You’ll want to secure them in place later, but they can just slide in and sit there while you ponder the rest of the install. It’s a good looking bank of batteries, really!

Wiring it Up

One of the perks of this rack case is that it comes with a pair of very nice busbars - one positive, one negative. There are three screws for the batteries to connect (the “extra” small ones, top and bottom, are mounting bolts - and while you probably could use them, you shouldn’t, because they’re not threaded directly into the busbar). Plus a set of three terminals up at the top for aggregate current extraction. This is why it’s easier to have your big wires coming in the top - the main takeoff points are at the top. I don’t think the busbar can be flipped over and still clear the case, either - I pondered it briefly, though I didn’t actually try after looking at the asymmetry of the busbars.

Two pieces of advice here: Fold the handles behind the busbar clips before you start, and don’t worry about connecting each battery to the nearest terminal. Play around with the angles, and you may find that it works better to swap two batteries (lower battery connects to the mid terminal, in my case). Theoretically, these wires are rated for the full 100A output current of the battery, though if I were planning to use anywhere near that much current, I’d double them up. I know what wires are rated for, and I know what temperature they’re allowed to get at those ratings, and I’d rather be conservative. You might also consider building your own wires for the interconnect, if you’re bored - the provided ones are really quite a bit too long.

Some of my wiring locations here are just so I can reuse the existing wiring from the lead bank. I’ve put my main measurement shunt (the WhizzBang Junior) down at the bottom, since it’s a good spot for it. This interfaces with the Midnite Classic and provides “net current” values to and from the battery bank, which allows me some ground truth about how the battery banks behave. It matters less with lithium than with lead, but it’s still a very, very useful tool, and allows state of charge guesstimates on the Classic.

All of the other connections are on the far side of this shunt. Power out to the inverter, charging current from my various solar arrays, all comes in on the far side so it can be measured. This shunt doesn’t see current flowing from the charge controllers into the inverter directly, though I used to have a load shunt that showed me “inverter DC draw.” I couldn’t fit it in here, and I’ve got some ideas for other places to mount it down the road closer to the inverter. But, for now, it’s missing in action.

Crimp. Your. Terminals. Seriously. The one problem I had with my lead acid wiring was a spot in which I didn’t crimp my terminals, and relied on screw terminals on the fuse! A good crimped connection won’t ever give you any trouble. Bonus points for heat shrinking them - though my heat shrink setup is offline while I wired this up, so I wasn’t able to make use of it.

My positive terminal area is a bit more crammed than I’d like. There’s just no good way to get these wires run in and hooked up - the busbar area is pretty tight. But, I did get it all hooked up (no shunts on the positive side to worry about), and after some testing, everything’s online!

Using the EG4-LLs

Setting these batteries up, once they’re plugged in, is just a matter of setting the charge controller to match what they want. It’s rather different, and far simpler than setting up lead acid batteries. If you’re setting up a Midnite Classic with a WhizzBang Junior (WbJr) shunt, you’ll want to set things up roughly like this:

Absorb voltage 56.2V. Disable temperature compensation in the T-COMP menu! There’s no need to jack voltages in the cold like there is with lead acid, and you really don’t want to as you can easily exceed the BMS upper limit in the cold.
Float voltage 54.0V, though talking with their tech support, the batteries seem to like 55.0V somewhat better (and I think I agree). Play with both and see.
End Amps (the transition from absorb to float) should be around 1.5A per 4U battery - so 4.5A for my bank, plus or minus. You’ll see them drop to well under an amp per battery if you force things to keep charging, until they (temporarily) disconnect and enter standby (see below). There’s no point in continuing to charge them at this point.
Set your WbJr shunt SoC estimation for (in my case with 3x 100Ah batteries) 300Ah, no temperature compensation, and about 98% charge efficiency. You can get better SoC off the batteries themselves, but this seems to be within a few percent.
Disable any automatic equalization! You really don’t want to suddenly zap these batteries with 65+V, and if you try, they’ll just shut down on you. I have no idea what this does to loads, but I assume nothing good. They’re not lead acid.

Do that, and… then just ignore them. However, if you’ve limited your charge current to something lead-acid-ish (because they just don’t sink a lot of current for long), you may want to upgrade stuff. I initially put a 50A breaker in for my charging, and limited my charge current to about 45A. This wasn’t a problem on lead, because they almost never rode the limiter. This bank? It’ll take all the current you throw at it until it’s fully charged, so I upgraded my charging circuit to 80A in order to better charge this bank. It doesn’t really matter in practical use, but it makes somewhat better use of the “random sunny times” in the middle of cloudy winter days.

Otherwise… they’re batteries. Just a very stiff, well behaved battery that can soak and source immense amounts of current on short notice. Be careful with them - this bank, as installed, is good for over 15kW. Sustained!

If you’re familiar with lead acid charging behaviors, though, you’re in for a surprise! Lead acid takes a long time to charge. This is what a lead acid charge curve looks like - voltage goes up, and stays there, as the battery charges and current tapers off over time. It took 5+ hours to go from “hit absorb voltage” to “switching to float.” This is a big part of why I like my east-west facing panels - to give me a longer solar day for this sort of charging behavior.

The EG4s? Nothing of the sort. They charge, and charge, and then when they’re full, they just hit a brick wall. Here, a reasonably depleted bank charged as the sun came up, and when it was full, it was full. Wham. Down to float, and hold it there. It’s a very different style charging, though it still benefits from a long solar day in terms of “not needing to use batteries because you can directly power loads off solar.”

Random Other Operating Notes

For lack of a better place to put my random notes and observations in a month or so of operation (and I’ll extend these as I learn new things):

When the batteries are charged and floating along happily, if there’s no strong demand on them, they’ll enter standby and internally disconnect from the external battery bus. The “run” light will stop flashing, they’ll show “standby” on the display, and you may notice the BMS reported voltage is drifting down from the external battery bus voltage (an easy way to force this is to hold the voltage up at absorb after they’re full with no net discharge on the battery bus). It seems perfectly fine, and they’ll connect right back in as the bus voltage falls from loads.
Related, for very light net loads, I’ve noticed that one battery will remain connected, while the others are in standby. For me, it’s the top battery, which is the lowest ID - though I don’t know what the algorithm here is. Again, they sort things out as needed, but don’t be surprised if it doesn’t look like all your batteries are running when you’re purring along in float. As I write this, I’ve got 400W flowing from my panels to my inverter, and all my batteries are in standby.
The SoC meter on them does qualify, as always, as a Guess-O-Meter. After a week or two of PSoC (partial state of charge) cycling, don’t be surprised if one or more is wandering from both the rest and anything resembling what you’d expect. They seem to reset to 100% when charge amps up at the absorb voltage (56.2V or so) drops to almost nothing. They don’t seem to reset after a while sitting at 55V as a float voltage. If you’ve got a Midnite Classic, you can set the “skip days” parameter in advanced charge settings to start most days at float, and only go up to absorb every N days. This works wonderfully with this bank, though they go from 55V to 56.2V so quickly (and so decisively if you’re using end amps) that I’m not sure it makes a real difference. They’re not lingering up at absorb for any length of time.

Bonus: Morning Panel Conduit

While I was rewiring stuff, I routed my morning panel wiring through conduit. I’d originally run the wiring down low along the base of my shed, and that ran into a certain problem this year that I believe is called “bunnies.” They chew on everything, and I’m fairly certain, by the shape of the cuts, that this included my wiring, somehow. But the wiring was getting some UV damage on the insulation and the time was right to replace it anyway. May as well run conduit!

Something new and rather recent in non-metallic junction boxes is the arrival of “universal 1/2 and 3/4” conduit boxes. They come with a bunch of 3/4 blockoff plates, and several 3/4 to 1/2 bushings you can use to reduce an opening down. Should you care, you don’t even have to use all the same opening sizes - which may be useful for some other projects of mine!

If you recall my office build, you’ll recall that my walls are rather thick. I’ve got standard 2x4 studs for the walls (filled with rock wool), then 2” of foamboard on the inside, then some plywood. If you’ve ever had a bunch of 1/4” drivers laying around from “free drivers in boxes of screws” (say, from a deck), and you want to build an extender - superglue is your friend, and I’ll suggest not using a magnetic (but shallow) bit on the end. The hole saw wiggles around a bit too much.

On the back of my office, I sealed up a junction box tight against the wood (silicone is your friend here), and ran it around to the other side. Nothing fancy.

On the side behind the morning panels, the conduit zips along, shoots up a bit to better match where the wires come off the swinging frame (at the top, because solar panels have the junction box up there), and ends with a gland facing down to allow the PV wire in.

Always have a drip loop. I did have to add some splices to the PV wires, and then join them to the rest of the wires running back into my office, and I used some heat shrink crimp connectors. As far as I can tell, these are entirely fine to use inside a junction box - so I did. I assume they’re fine without electrical tape, but i decided to add more insulation, just because I can.

Closed up, this is even rated for marine environments! It’s a well gasketed little junction box. And should last a long while out here, shaded behind the panels at almost all hours.

Even though I don’t swing them around anymore as I initially designed them for, it’s nice having the morning panels back. They’re a solid 6-8A of charge current when the sun just starts coming over the hill, and that makes a real difference in how quickly the battery bank comes back up. It doesn’t matter nearly as much now as it did with the lead (when i wanted all day long in the winter to top them off), but it’s still nice to see the “whomp” of power as the sun shows up. They’re not critical anymore - I ran all summer without them - but they’re definitely a nice couple hundred watts to have!

So, Should You LFP?

I’ve been using this new bank for a bit, and… I have to say, it’s properly nice. However, it’s also significantly more expensive than my previous lead acid bank was. So, if you’re considering an off-grid system in 2024, what should you do? If you have the money, I think at this point, going with a good LFP bank is well worth it. They (supposedly…) are long lived, and the charge/discharge behavior is a lot nicer to deal with than lead acid. People in the quiet backwaters off-grid power IRC channel who have been living with off-grid power systems for a lot longer than I have also say the same thing - if you can afford it, go LFP and don’t look back.

But if you’re on a really tight budget, there’s nothing wrong with a lead acid bank. They’re a bit more work to deal with (watering and such), but as far as “doing the job they’re rated to do,” they do a very good job. A set of deep cycle lead acid batteries, ideally focused around off-grid use (the “solar” or “renewable energy” lines), still works just fine. And they’re not nearly as bad as assorted “influencers” would like you to believe. They’re long lived, well behaved, and lack the energy to get too terribly exciting. Just vent the hydrogen properly.

I have to say, though. I’m really liking this bank so far!