Building a Handheld PC | Hacker News Reader

Building a Handheld PC

bytewelder.com

202 pointslo2y27 comments

This is a fantastic demonstration of how accessible all this stuff is these days. Polish used to be the let down on DIY projects like this but 3D printers (for the case) have really moved the needle when it comes to high quality prototyping. I know this has been the state of things for nearly a decade now, but it still amazes me. Raspberry Pi, Arduino, 3D Printers, and the open source hardware community, will mark a paradigm shift in prototyping accessibility in the history books.

It's super cool, but at the same time for many folks an iPad Mini with/without external keyboard would be a more adequate solution considering processing power.

I don't think so. It could be very tricky to modify an iPad mini sufficiently to become a piece of custom hardware that you're prototyping. Except in the case you were prototyping an iPad mini itself, which of course would be child's play.

This sort of comment is actually a rather obscure compliment.

If you make something that's obviously shoddy, nobody confuses it for a serious product. Nobody is telling grandma that her knitting isn't as practical as a waterproof jacket from J. Random Outdoorswear Company. Nobody is telling the electronics newbie that a blinking LED isn't actually of any use to anyone. Everybody has nothing but positive encouragement.

But if you start to make actually really good stuff, requiring an enormous amount of skill and experience, you instead get "cool, but actually not quite as good as the nearest commercial competitor".

If you're so good at your hobby that people are comparing you to the literal state of the art, you're doing something right.

I love this comment! As someone who works in hardware in a demanding industry, although not in consumer electronics, this example obviously has a long way to go to compete with a real product, because you have big hurdles in moving to integrated designs instead of integrating devkits, and in moving to manufacturability at scale. I don't think that's even the intent but I can see it at least as a niche product. I only say this to set up the next statement.

This thing is great to show vision and usability, and I'm more excited by this than a lot of dead-end things I've seen presented as a bunch of requirements and if you're lucky, renderings made by a lot more than one person and 3 weeks.

And think about the software world, it's understood that framework X makes a great example but needs to be replaced by microservices Y, and also understood that people exist to make that happen.

It's not like grandma's knitting, but it could be something like MVP software to get a cofounder and an angel investor.

I love this comment for similar reasons as the sibling comment to this one :)

In the late 2000s (pre iPhone), it just so happens I was lucky enough to work for OQO -- the "Rolls-Royce of Handheld Computing". This project is a DIY attempt at building an OQO, and I'm IMPRESSED.

You must making pretty good stuff for someone like me, who did this professionally, to think "Oh, our product is obviously still better, but wow, this comes closer than I thought!"

Things have come a long way in 15 years!

I think you're missing my point. Sure, there are better handheld computers! My comment is about the ability for home hardware hackers to churn out something like this (relatively) easily. At least comparing to my experiences 20 years ago trying to produce similar.

Going for the iPad mini of all devices is an interesting take.

Ignoring the fun DIY aspect, this project's main advantage would be:

- extensibility and modularity: hardware wise it could fit a PS2 port if it needed to, software wise you can put any OS that runs on arm.

- repairability and upgradability

- the actual physical keyboard. If your goal is to be typing at a CLI prompt 80% of the time, and you want it handheld, that's a pretty sweet form factor, better than a tablet with a flipping keyboard.

Raspberry Pis are accessible, in your opinion? There's been a deep shortage for years.

I was able to get one last summer. Since then it has been even easier to get, albeit this was from Germany and it was not marked up in price but from a standard electronics retailer

Someone eventually manages to buy a RPi, but you were that lucky one as the RPis are still almost unobtanium almost everywhere. However, a RPi is just the most famous and heavily advertised one among the many Linux capable embedded boards out there, therefore pretty much every project using one can be converted to a similar board from another vendor.

Some alternatives:

https://mangopi.org/mcorer818

https://banana-pi.org/en/core-board-and-kit/

https://pine64.com/product-category/compute-and-ai-modules/

etc.

The Pi4 has been hard to come by, getting later. But 15 years ago there was _NOTHING_ like that. Now you can get Pis and all their clones, the odroid boards, bananas, all sorts

Ordering these things - or, at least, trying to - is beyond ridiculous now. I've been checking three or four sites every month or so for literal years now to no avail.

I've even tried shimmying in an order with an 'educational supplier' here in the UK who apparently have 'some', but only for schools, so they won't let me have one, despite how much material I buy from them each month for my non-educational purposes. I don't blame them, of course, but I do find it weird how there's apparently a heckload of demand and no uptick in production.

I'm curious which supplier that is?

Apologies for slow response.

Company is Kitronik. You need to have an existing educational account with them to order, i think.

Since PCB manufacturing has become super cheap, they could consider having a custom L- or U-shaped custom PCB around the battery box that serves no other purpose than routing signals and power to where they are needed and solder the different boards directly onto that. That would also allow to place supports for this large PCB where they are convenient instead of having to separately mount each of these small PCBs to the case.

With less cables in the way, they might get better airflow in the case as well.

Edit: I guesstimated the size and think that they could get five pieces of such a board for less than 20€, including delivery.

https://www.clockworkpi.com/uconsole

Those are cool.

It says the charging voltage is 8.2V, which is double the usual charging voltage for lithium cells.

Is it really safe to connect two 18650 cells in series like this?

I've read that you should manage each cell individually to reduce wear and/or risk of fire.

If the cells are matched it is less of a problem. But 2S battery management boards with additional features like low voltage cutoff and current limits are only a few euros in single quantities, so I see no reason not to use them.

Looks like he's using a BMS with a centre tap to balance the cells.

I looked up the module on AliExpress:

https://www.aliexpress.com/i/1005004814091297.html

Anybody care to have a look and comment?

I could only find a datasheet for the chip in Chinese, but from what I could get through google translate it looks like it can balance two or three cells in series.

I hope they used something other than PLA. I've made some really nice enclosures that warped over time from heat.

Have you thought of doing the same with an e-ink screen?

I would find it awesome to get work done while in nature.

It's just as awesome as you think. Sadly, those projects always end up getting cancelled.

What SoC is this using? For the life of me I can't find it in the article?

Edit: ok it's a CM4? on the back of the display...

For others searching for the secret sauce: BigTreeTech Pad 5

Reply on news.ycombinator.com