Project Conclusion

It’s a good feeling when your ideas are validated by companies taking a chance and investing in them. It’s probably a better feeling when the company in question doesn’t come up with the idea on their own, without you, but I’ll take what I can get.

An unexpected conclusion to the handheld Steam Box project has come in the form of Valve’s very own Steam Deck.

Previous posts on the project:

SteamDeck

The project had been on pause since I’d moved to Detroit, nearly two years ago. My friend had solved the problem of powering the screen off of the computer’s USB power/charging port, which was a huge step forwards.

We’d yet to fashion a case, but we were planning on leaving this iteration corded for power, and tackle a battery in version two. Along with portability, we were planning to add a modern processor and motherboard, as well as a touch screen.

Valve has tackled all of these issues (custom AMD processor!), as well as adding one touch pad on each side of the unit, underneath the analog sticks. They say that it will enable thumb typing, as well as precise mouse pointer control for games that need it. (Stellaris?) I wasn’t so keen on the Steam Controller‘s implementation, but I have high hopes for this device.

Am I upset that Valve beat me to it? Not in the least. I wanted portable PC gaming on Linux to be An Actual Thing in the Real World, and that’s what’s happening. They’ve moved SteamOS from Debian to Arch, but it’s still Linux, baby. Valve’s Proton project is actually commercially viable. Take that, haters!

I’m confident enough that I’ve dropped $5 USD to reserve a unit, and I’m smart enough to have waited to push that button until my unit would ship months after release, ensuring some bug fixes will have already rolled out. Now I just need to wait.

Giving Up and Making Progress

I’m a little frustrated, and a little irritated. MyMateVince is using Anker power bricks to supply power to many of his mobile projects. Anker themselves said that the power brick I ordered would supply the juice I needed. But, dang it, I couldn’t get it to work. I couldn’t get more power out of the USB ports than the standard amperage, and I couldn’t figure out how to trigger their PowerIQ to open up the floodgates. So, finally, after all of the tries in the previous post, and a few more, I gave up and contacted support

Assembled and updating

Support immediately told me that it was not possible with their current generation of power bricks. There may or may not have been some slamming of my head into my desk. I will neither confirm nor deny. The battery is now being returned for a full refund.

Ahem

This led me to table the idea of powering the first version by battery, and instead focus on a portable, if plugged in, version. Which led to re-assembling the machine, updating the software, and confirming that all the things still worked with all the other things. And they did! The games run even better now that Steam is pushing so hard for more Linux compatibility, some visual artifacts have disappeared, and my available library has more than doubled.

Mocking up the case

On top of that, my friend noticed that the mount plate (that piece of metal that the motherboard screws into) was attached to the case with several spot welds. I was confused by how this mattered, until he pulled out his drill and started drilling out the weld points. A butter knife and some elbow grease later, and the plate was free of the case, and available for us to use. We can now mount the motherboard on one side, and affix it to the case on the other.

Speaking of cases, my friend and I began considering. And, when he considers, it’s more like he starts mocking it up on the spot. He just happened to have this small sheet of steel lying around, and got right to bending it. So far, the only thing that we disagree on is the speaker placement.

I’m right, of course. :)

Nerding in parallel

Did I promise to give an update if and when I worked on the portable Steambox project again? I did? Well, here you go! We worked on it! And ALSO on something else.

Again, things came down to power. This project needs a battery pack that can power both the screen AND the motherboard, and at different voltages and amperage. We’ve got some bits and bobs that should have done the job, but none quite did All The Things. So, I went back to my thinking at the start of this project, and looked for a consumer battery pack that would do what was needed. And, after an email chat with a representative, I found the Anker PowerCore AC. I picked it up and… didn’t test it. I’d given the power cords to my partner in crime, so to speak, and he was out of the country for work. Well, last week, he came back.

Whatever tells the Anker power pack to increase the amperage to support faster charging isn’t happening with the screen. It’s getting the 5 volts and 100 milliamps, but isn’t getting the quick charge IQ goodness. So, the screen tries to turn on, fails to get enough, and turns off. So, we needed to find a way to check if the battery pack could put out what was advertised.

Tech mess

My friend set up the following power chain: Anker A/C port, router plug, power-over-ethernet injector, RJ45 cable, step-down converter, D/C cord into screen. And it worked! The screen turned on and stayed on! So we know for a fact that the power supply can do the needful, at least out of the A/C port. Now we need to figure out a way to either trigger the battery pack to engage its fast charging, or combine the power from two USB ports to support the screen. We’ll see which way it goes.

Tech mess

The other project is re-purposing an old Dell PowerEdge server into a moderately useful gaming box. It’s currently got one Xeon quad core processor running at about 2.5 GHz, and has a slot for another. It’s got 8 gigs of RAM, and can go up to a whole lot more. My friend had thrown in a 5 year old video card with 1 gig of VRAM, that had sat in its box since he picked it up. It’s got a 500 gig hard drive, but it’s one that came with the server and is a bit slow. Windows 10 runs on it, and runs pretty well, which wasn’t a surprise. It’s done a really good job of scaling down to old hardware since its release, and hasn’t reversed that trend, which I’m thankful for.

Our first test was Minecraft, cranked to full everything, and it ran smooth as butter. Next, we installed Everquest II, a game that the end user (my friend’s wife) plays, and I hadn’t played in about seven years. My login still worked! My characters were still there! I had a succession of “what the crap” moments, and tooled around the winter area in which I spawned, and figured out that anything over “Balanced” graphics quality wasn’t going to be smooth. But! The game worked! For old time’s sake, I let off a PENITENT KICK! and took a photo.

EQ2

Pieces and parts

Pieces and parts

Power can be a problem. I don’t mean the kind where you have too much power, and you get corrupted six ways from Sunday. No, I mean the kind where you need electricity to flow to the pieces and parts in exactly the ways that they expect. And since we’re using consumer parts to build the project, these things aren’t easily modified. But, you know what? This is a problem that can be solved.

We just haven’t been able to solve it yet. The PC takes the majority of the power, which we expected, and the screen takes a comparatively minimal amount. I bought this specific screen model because I saw it powered by USB in this video. Unfortunately, it won’t pull enough power either from a power brick (same brand as the video, but different model) or from a USB cable plugged into a wall wart rated for more than enough juice. I’ve reached out to the maker of the video, but haven’t gotten a reply.

It’s worth noting that the difficulties we’ve run into have had me considering whether or not I should continue the project. And then, at one point, we had the screen on and connected to my laptop, mirroring the display. And, still, that little 10″ screen is beautiful. It’s so pretty. And with that, I was determined all over again to see this thing through to its conclusion.

If you have any ideas or expertise that you’d like to lend to the situation, please contact me with the details you’d need to help with the problem, and I’ll send them your way.

Long-awaited update

A physical mock-up of the hand held steam box project.

There’s this thing called a mock-up. My good friend had to explain this to me, because I was entirely unfamiliar with this stage of the prototyping process. A mock-up is where you arrange the parts, or bits that represent the parts, close to how they’ll be on the final product.

Why? Well, this way you can make an effort to predict how the details will come together. You can see where there will be room for air flow, test different orientations of the motherboard, and where the ports will be exposed, things like that. In fact, we did that very thing, and made an unexpected decision in the process.

We decided to make it wider.

With the weight of the batteries, we were a little bit worried about holding the device up by the joy-con rails. We didn’t want the mounting screws pulling out, or damaging the joy-cons themselves. Also, they’re small, especially in relation to the rest of the device. If we slightly angle the battery packs underneath them, we provide a better grip and more leverage to hold the device. (If you look at the photo, the mock-up is looking at the back, with the screen down.) We tried to make the angle similar to the angles used in the charging grip. Why waste Nintendo’s R&D?

This made the device a bit wider, but the angle of my arms now feels far more comfortable. I did think about another drawback, though. During charging, the battery packs may warm up to create sweaty palms on the grips. We’ll have to test that once we actually form the grips, and use real battery packs, rather than dead ones.

Full disclosure – due to travel on both of our parts, as well as home repairs, the project has stalled a bit. BY NO MEANS has it stopped. I’m still excited about this and still want to make this a reality.

Off by THIS much!

With everything else tested and taken care of, it was time to look at housing. What about the battery, you ask? My good friend (remember him?) is working on the battery pack, I reply, and I’m not intending on getting in his way. So, housing it is.

Getting the screen apart was pretty easy. Five or six screws, one behind a Quality Control / Void Your Warranty sticker, and it just slid apart. There was plenty of room in there, and everything was placed in a logical way with standard screws and such. Easy to move around, easy to play with.

Getting the PC apart was a little less easy. I had to take out the wireless card and antennas, the hard drive, the fan and air tunnel unit, and the CPU heat sink. And that was just to get to the screws attaching the board to the case.Once the screws were removed, the motherboard was… reluctant to come out of the case. Once I figured out the proper angles, I was able to extract the board without damaging it. It was touch and go there for a minute, so there were many sighs of relief once it was out.

Then came the moment of truth! I lay the motherboard on top of the screen back to see if it would fit, and… well, you can see the photo above. The board is too big (or the backing is too small) by a matter of millimeters. It wasn’t terribly disappointing, because now we get to make a custom housing for the unit, and make sure both the screws for the screen will fit, but that the screws for the motherboard will ALSO fit. We’ll also get to account for internal wiring, cooling vents, and which ports we want to be accessible.

Originally, the plan was for vacuum-molded plastic or acrylic. My friend sent me a video for a fiberglass tape that he thinks will do a better job. If we make the housing out of tape, I’m never letting anyone forget it. Or, for that matter, living it down.

Small is better than large.

System v3You may notice a few changes to the PC in the photo, there. The project has been humming right along, and a lot has changed. The most obvious change, I think, is the hard drive.

The S-ATA SSD that I had did the job, but it was under the system required space, and took up quite a bit of room. Skippy, you might ask, how can a laptop hard drive take up a lot of room? Okay, that’s fair, I respond, space is definitely relative in this case. As I mentioned in the last post, this motherboard has M.2 slots, one of which I used for the wifi card. I spent a little money and ordered a M.2 hard drive, and it’s ridiculously tiny, despite having 250 gigs on it. It’s the Crucial branded number on the left, there. It’s incredibly fast, was immediately compatible, and generates very little heat. With all of these benefits, it was the way to go when space will be at a premium.

The wifi card looks a little different, too. That’s because I ordered antennas for it! The first antenna I tried to attach was a bust. Apparently, when the standard changed from Mini PCI-E to M.2, they took the opportunity to make the antenna connectors smaller. Incredibly small. I was expecting a struggle after my last attempt to attach antennas, but these clipped right on. AND! They have adhesives on the end for when I run them around the edges of the screen. Bonus!

Also! My Dualshock 4 controller seems to have mysteriously morphed into Nintendo Switch Joy-cons, attached to a Charging Grip. With the wires attached to the wireless card, bluetooth works under SteamOS, and can see the Joy-cons. Unfortunately, it sees each of them as a separate controller, rather than seeing both as one. If you look closely in the back, you may see the Magic-NS adapter sticking out the back. (Thanks to My Mate Vince for this!) When using that, the Joy-cons show up as a single controller that can be easily mapped through Steam’s interface. I played some Bastion and some Talos Principle with them, and was NOT disappointed.

The next step is to take apart the screen and get the motherboard out of its case, determine if we need a custom housing, and move forward from there. SPOOOOOOON!

Peripherals are exciting!

To see all the posts for this project, click on the “Handheld Gaming” category below.

So, we have SteamOS working on a Super Small Form Factor box. We have a couple of games downloaded and tested. What’s next?

Well, I’ll be mounting controllers to the side of the unit, so making sure that SteamOS handles non-Steam Controllers well is an important step. Before I drop $70 on the Switch Joy-cons that will be in the final product, I hooked up my trusty DualShock 4 controller with a MicroUSB cable. SteamOS automatically recognized it, and accepted it as an input device. I tweaked the controller layout in the OS a little bit, because I’m a picky person, and the config changes worked AND saved between games.

The next step was adding wireless and bluetooth connectivity. The motherboard has two M.2 slots on it, one designed for a hard drive, and the other designed for just what I was looking to add. Once the card came in, it was incredibly easy to install and use. SteamOS recognized it right away. This was the exact opposite experience I’d had when trying to replace the internal wifi/bluetooth card on my laptop, so I was overjoyed. OVER. JOYED. Unfortunately, I couldn’t successfully test it for connecting to the local network or to the DualShock 4, because I was (and am) waiting for the laptop antenna I’m going to mount on the inside of the screen. Hey, when your friend is hooking you up with free parts from his collection, you can wait. But, it works! In the OS! And can see things!

Figuring out which screen to use was difficult. I needed something that wasn’t going to suck a lot of power, was able to handle full 1080p, and wasn’t too large to use comfortably on an airplane between my substantial belly and the seatback in front of me. That’s the whole point of this project, right? Thanks to this sweet video, I ended up going with this screen, originally meant for videographers and photographers. It’s incredibly light, and has a lot of room under the back panel. It’s full 1080p, and is incredibly crisp. It honestly looks better than the test monitor I had it hooked up to, which has just added to my excitement. Its only drawback is that it can’t be powered by a conventional USB port.

The project is coming along so smoothly that I’m terrified of jinxing it. BUT! Next post will be about swapping out the hard drive, and adding the Nintendo Switch Joy-Con controllers!