MRKane's final portable [KeyLime Pi-Rate Station]

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Final portable AKA "the only thread with Fluffy the WaxEye in it".

So I had no intention of entering in this years competition due to my situation in life, but after a discussion with a few friends they kind of twisted my arm with phrases like "You've got to do something" or "You can't just call it quits like that" or my personal favourite "It's only $2, do something basic with whatever you've got".

See I've been diagnosed with multiple sclerosis and am on some cutting edge treatment, but it's taken its pound of flesh meaning that it's time in my life to start shifting my hobbies and interests into something that I'll be able to keep doing. Hopefully it won't get too bad too quick but there's still a lot of other things I want to do so this hobby will have to get put into a box somewhere. Please don't cry me a pity party, I've got a terrible allergy to them.

So this year it'll be something within my ability, that's easy, and won't chew up my life. It's not going to be revolutionary, spectacular, or even a winning entry, it's just going to be..."something" :)

Remember these? I hated these. I mean what does this even do? How do you play this? Did anyone have a good time with these as a kid? All of mine had terrible valves that leaked water everywhere.
1622619238393.png

So I've got a Raspberry Pi 4, a little 5" screen with touch input, lots of bits and pieces, and the shoulders of giants to stand on with plenty of good documentation. I've mucked around with a bit of cardboard and wood and concluded that something similar in design to the DS (but without the hinge) would be the easiest thing to do. I'm fond of my DS that I got for $20 :)

PortableDesigning001.jpg

No promises here, I might push forward with it, it might get finished, or I might decide I'd rather have a holiday somewhere, but it's not going to be a contender for the competition anyway - and that's just life now ;)
 

fibbef

Wizardry V Completer
.
Joined
Nov 10, 2016
Messages
594
Likes
1,575
Portables
3
I'm really sorry to hear about your condition. But I'm glad to hear that you're choosing to quit this hobby on your own terms. I see what you wrote about pity parties, so I'll just leave it at that.

Let me know what kind of hobbies you take up next. I've got a tiny bit of experience in about a thousand different areas, so I'm always interested in other people's pasttimes.
 

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Thing is that's life, and it's something you die with, not from, so all things considered it's just something that re-frames priorities in life. Prior to this I'd happily survive on 4 hours sleep a night but now I need a full 8 hours so it's like I've been robbed of 4 hours of my day. This has really made me focus on priorities and start trimming out things so that I can achieve everything else. Family and community has won out here over hobbies that I've kind of mastered to a level where I don't feel I'm getting as much learning and growth from them as I was before.

As fate would have it I started building a CNC machine about two years ago, and finally brought it to completion recently so will be looking to start doing some stuff with that. It's also been a very long time since I did anything artistic, be it paint and canvas or with aforementioned CNC, and I've got this travel bug because there are so many places I've never been to yet, tracks I haven't walked, mountains I've not climbed, lakes I've not fished, and so many amazing dishes I've yet to eat!

I'm sure my hobbies will continue to grow and change, and I'll still be around, just I'll be spending less time over the soldering iron :)
 
Last edited:

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Standing on the shoulders of giants the PCB for the unit is going together very rapidly and easily, which makes me thing that I've done something wrong of course. I've turned to bits of left-over wood for the case prototyping too, just so I'm sure of the design :)

Sits in the hands well. Balance is great, functionality is still lacking.
Image from iOS (1).jpg

So I'm happy with the control layout, which as fate would have it is millimetre similar to the xBox 360 controller. I was kind of surprised by that, although feedback from people around me included all sorts of crazy designs where I could only say "no, that'd mean there's a gaping hole where that's supposed to be." and "So what do you think of the design little nephew? Yes, it can be a car. Yes, that's how portables drive along the carpet."

So I had a small nightmare with the redesign of the last portable and its shoulder buttons - things just didn't go great with balance and components under the hood getting in the way. This time I decided to try tact switches and 3D printed parts that flex to depress the tacts.

Thing is I'd be inclined to think that this'll only result in the plastic eventually fatiguing and breaking, so I'm 100% open to suggestions from anyone reading this about how best to do shoulder buttons. At the moment I'm thinking about simply sticking them through the back but might need to be 100% sure I'll not put batteries there for balance.

Image from iOS (3).jpg
 

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
The ongoing warpath...

So after playing with cardboard and mucking around with layouts I eventually settled upon something that felt good and fit the requirements I sought. Actually...I haven't mentioned what any of the features are...a working Pi Portable? Ok - lets go with that. I can add things on as I go.

My life revolves around MDF (glorified cardboard) when it comes to prototyping anything really - it's a little thin, but significantly more presentable than the box I tried to make that was about 30mm deep.
PrototypeLayout.jpg

So I've been bashing the PCB together with extreme prejudice, and standing on the shoulders of genius to cobble things together as easily as possible - a huge shout out to Dmcke5 here for helping me with the Arduino code and layout which will become the controller chip in the unit, and of course GMan for his assorted PMS systems. My aim here is to get everything made in the fab shop to make up for my declining ability to solder, and this is where I could do with a real help!

There are two chips that aren't available at JLCPCB which I'll have to hand-fumble if I cannot find an alternative:
-What could I use as an alternative to the PIC16F15324 chip which is used as part of the PMS? Actually I can get my grubbies on the straight 18 pin which should be easy soldering, so that problem might be solved if I can fit it on the board :)
-Is there a good alternative to the IP2721_MAX12 for negotiating 12v out of USB-TypeC?

At worst comes to worst I can hand solder these, but I'd prefer not to :)

So my cunning plan here is to include the boards I'll be using for the shoulder buttons, which will use pads from the PS2 (as I seem to have those lying around ATM: see ref to prior project), and on my sophisticated mockup the blue-tak buttons felt good mounted underneath. Highlighted are the two chips that I'll have to find a replacement for.

I opted to poke the thumbsticks through the board as JLC had bottom mount connectors that would appear to be suitable for the Switch thumbsticks, so I'm going to take any opportunity I can here.

Screen Shot 2021-06-17 at 9.07.43 PM.png


Alternatively I really like the Battle of Naboo N64 game and took a design style from that for the PCB...which nobody will see.

Screen Shot 2021-06-17 at 9.06.57 PM.png

As for the other part of my initial post...so I decided I should grab one of those water games and see what I could do with it.
It's plastic welded, and apparently too easy to "beat" plus there's one too many rings! It held Fluffys interest for about 30 seconds however.

FluffyWaterGame.jpg


FluffyWaterGame02.jpg


He probably thinks that they taste just like the SMD caps that he ate a few weeks ago. Apparently those were tasty.

I 3D printed the Lament Configuration, and it's become a new favourite around here, adding further proof that this tiny bird is actually only fulled by the desire to watch the world burn.

FluffyLamentConfiguration.jpg


Seriously though, Fluffy is a fantastic little pet and is much beloved :)

So that's about the state of things, I'll do my best to keep posting and pushing forward with the project, but as I said above I could really do with assistance here :)
 
Last edited:

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Well I've just send the PCBs off to JLC, but not without a mess of blues that I always seem to go through. Had I done this a week ago many of the components I wanted would have been in stock, and now I've got to purchase them elsewhere and the price is eye-watering.

Suffice to say, I've learnt that not caring if your board will work really helps to get it sent through, but cost wise this is it for me and if I can't make it work there won't be a second attempt.

I couldn't, however, send it away without adorning it with a name and a logo! So I asked around Discord, combined some favourites with my love of JET SET RADIO and here we have it :
PiRateStationLogo.png


Get down you funky things and understand the concept of love!
 

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Boards from JLC PCB showed up yesterday, and while I still cannot fault their design I've got to wonder what I was thinking when I made it as there are a couple of glaring oversights that I really should have been smarter about. Oh well, just going to have to make it work somehow.

I'm glad that the thumbsticks fit, and that everything lines up, of course the FFC connectors aren't mounted yet so I'm sure that won't go wrong...
BoardWithThumbsticks.jpg


Size wise it's really good to have had a mockup previously as looking at the components leaves me feeling that it's actually really large. I've also not trimmed the shoulder tabs off yet.
Image from iOS (5).jpg


Shoulder tabs trimmed, board sanded, tiny bird shooed from around work area. I'm already glad that I put 3 on as one of my cuts was a little lean. While the cut board will still work it's nice to have cat-swinging room.

The back is also a good detail of just what I couldn't get attached in shop, and that I'll have to do myself. I'm hoping to try and do as much of this as I can using solder paste and heat without knocking anything off. The headers to programme the PIF could have done with being "normal" sized, so there's another mistake, and the through-hole for the 5V line point looked plenty large when I dropped it in, but in revision it's a mite too small so I might try dropping the line onto the pad for the inductor.
Image from iOS (4).jpg


And yes, you bet your buns that's a volume dial and physically switching headphone jack, because: I have them in a bag so decided to make room, and couldn't really rely on things I'd never used before because if this doesn't work there won't be a "try 2" ;)

So at the moment I'm waiting for my laptop to come back from the shop - it's got a faulty battery and naturally the plans for the board are on it, plus yours-truly didn't think about notating the details for things like, polarity, chip orientation, etc. on the board itself as I struggled through getting this thing submitted amongst parts not being available and revising the design.
 

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Not sure how to put this: the PCB I had made has some serious underlying flaws and issues making it largely unusable for this particular task. While I could (and just might) end up going over all the components used to see what's wrong it's getting a bit beyond my ability to solder down tiny things. I spent a few hours last night mucking around with SMD resistors and think I managed to kill them...

Issues:
-Power system, zealously copied from GMans Wii PMS, doesn't work as expected. StonedEdge helped me figure out the issues with the BQ not talking to the PIC, but the TPS isn't converting the battery voltage to 5v
-ATMega32U4, zealously copied from the datasheets and using Dmcke5's controller code, talks to the programmer but won't be recognised as a controller.

I've spotted a few mistakes on the board that I didn't know about such as the difference between R22 and 22R resistor, or just grabbing the first resistor that looks correct instead of also checking the voltage, but beyond that it's well into the "I don't know" basket.

My knee-jerk reaction initially was to use a power bank power supply, and everything I've read about the Raspberry Pi 4 says that it's a hungry beast, but in my tests it doesn't appear to be so. Still, were I to tweak things I'd look at somehow increasing the amount of power I can provide to the unit as I'm an overkill kind of guy, and would like to have the shoulder buttons' pads on the main board somehow instead of making them floating. Almost warrants another revision...

So at this stage I've concluded that I really need some more eyes on the board and will look to upload the project onto GIT (or likewise) for people to give me a hand picking it over to see if I can somehow resurrect this thing, or even if I need to send through another order. I understand that this will invalidate me from the competition but from square one that wasn't the point of this project :)

Update:
After a night to sleep on it and cool the blood I've cleaned things up a bit, fixed the two resistors on the data lines for the ATMega32U4, and fixed the naming error that broke the BQ. I've not done anything beyond that, and flicked it up here:

I feel like this just became a community project of sorts, so have been playing with things taped to boxes to get a feel of the balance for the unit and have at least figured out where the batteries have to go to keep everything weighted well, and now I'm wondering about making a few more changes to the board to get ride of the "floating shoulder buttons" and just have pads on the back of the PCB. We'll see how I feel after a few stiff drinks tonight

Update 2:
Finally found the time to test the audio system, and while the Pi seems to recognise the attached PCM5102A DAC it's not piping out any sound, plus the boost components seem to make a lot of electrical noise which translates to the speakers. Bringing the grand total of "things that actually work" on this board to zero.

EDIT on Update 2: That's a lie. The screw holes are the correct size and the shoulder button pads worked just fine despite KiCad having a page of errors that seemed unfounded.

Update 3:
Isolated the PAM8304 amplifier to test putting some sound through that - there's a lot of electrical noise going through the system and I'm pretty sure that it's not just because of my Michel Mouse setup, so I'll need to find a good way to resolve that, and I've a suspicion that it could rhyme with "faraday bead". Open to suggestions here too. It's worth noting that given the music I usually listen to the electrical interference positively improved it in my opinion, but for the portable it's something I've got to fix.

Update 4:
The PWM system for the fan control works. Small mercy I suppose. I've gone about stripping down the project aggressively with the aim of completing something at least - y'all ready to get shunted back to 2013? That's where we're going!
.
 
Last edited:

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Hands up who wants an update! So after licking my wounds and totally not getting sidetracked making CNC carved models of Baby Groot I decided that I should stick to the spirit of the competition and actually make it look like I was trying to finish this competition as opposed to flicking it away with a "meh".

I'm open on suggestions on how to finish the carve, obviously it's out of a bit of 4x2 that I rescued from the neighbors skip.
GrootCarve001.jpg

This time I opted for a carrier board which I could attach known good working boards to. The likes of the PAM amps I have lying around, and the Arduino Micro I've been testing with, plus an over zealous stripping back of any functionality that I don't have to do and blind passing-off of function to the Pi itself. Audio will be provided simply by soldering wires to the 3.5mm jack for instance, and volume controlled through a shortcut command on the Pi...it's kind of an Apple feature.

By and large everything kind of the same, plus I've done my usual double-dipping for functionality and extras least they prove useful or necessary in some way. I've actually set this one up so that I can cannibalise components off of the failed boards to complete it and save a little money. I did a stupid amount of messing around with boxes, batteries, and insulation tape to figure out how I wanted the balance to be before employing the "well I've got these clips that I can bend into shape" solution and you can get the rest.

I'm also super-thankful for Dmcke5, RobertLong, StonedEdge, and TropicPug, for casting an eye over the design and picking up some mistakes I made - I'm really glad that they reduced the likelihood of me making another brick this time around!

The board was submitted for manufacturing about a week ago.

BoardScreenGrab.png

And finally I decided it was time to put all the tutorials I'd done on Fusion 360 to the test and see if I could actually model something in it. Which went about as well as you'd expect really.

No, not this angle...
CaseTop001.png



Could you believe this is the design I've got on paper?
CaseBottom001.png


So apart from being something that looks like the soviets won the console wars, it's actually a design that works around the core principle of having a case assembled from about 3 different parts so that it can have the Pi4 swapped out for any other SBC, hopefully with a change of no more than just a single printed part - or at least that's the idea. It's only had about an hour spent on it so far and most of that was figuring out depths for components. We'll see, and at worst I'll ship it all into blender and mangle it into shape.

So that's about it - no promises, but I'll do my best, just other priorities may intrude a little ;)
 

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
Right, project update time! There's been a lot going wrong in real life at the moment - NZ ended up under another lockdown and now we're running around trying to catch up with family that we've neglected for nearly a month; still the PCBs arrived yesterday and I managed to grab a day to work on the project! (what actually happened is that because of the MS I've cut my work hours down to a 9 day fortnight to help me keep on top of all the tasks around the house and get enough rest because if I think too hard I really do pay the price later so that totally won't happen after today...)

So this is it. I've run out of plastic. I do have more on order but when that arrives will be up to the post gods and in NZ that can be anything from this afternoon delivery to sometime next year.

CaseInHand.jpg


I had wanted to try to inject a little more style into this, with some cool design for the speaker ports but as my first complex foray into Fusion 360 I was running low on time and really needed to get a rattle on with things so simply dropped in a bar...which has really grown on me as a design aesthetic!

Naturally the inside is scarred to hell.

CaseInside.jpg


Removing the supports caused structural damage to the input sections as well as breaking off a few of the tabs intended for the screen. I can't help but feel that there's a lesson of sorts here pertaining to fragile design - but it's a prototype and that's what we learn from these things. I've also tapered the corners off more since this photograph as was unhappy with how sharp they were, and because the boards hadn't arrived at the time of printing (16 hours anyone?) had to manually cut holes for the power ports.

InsideCloseup.jpg


I'm 100% sure someone will look at this and say "you need to change X on your printer settings and it'll come out perfect!" which would frankly be awesome if you did. Don't hold back, please!

So with the case printing out I was able to get onto populating and proofing the board, which was actually really fun! It felt like I was putting together a kids kitset project with all the little header pins holding everything in place...and the frustrating as all hell 0.5mm pitch connectors for the joysticks, micro-USB and USB type-C connections...

So the first thing I did was mount the USB Type-C connection pins only, plug it in to check voltage, and then the tension in the cable flicked the board off my desk and tore the traces up. That's fine, it's why I ordered 10.

So on the top of the board I ran into a few issues right off the hammer. The tact switches which I had measured exactly, and required a 0.75mm hole, had a hole that I couldn't get a piece of 0.5mm wire through - it looks as though the plating thickened in a few places. This meant that there was opportunity for a great test involving a 70 year old drillpress and my smallest drillbit.

Drillpress.jpg


Which worked, don't get me wrong, but I felt that there was too much room for error on the board that would hopefully be proof, so I opted to painfully bend connections, and trim them, and solder them by hand as best I could. The tacts look like a cats breakfast but I guess they work. This is a photo with all major components mounted. Because I'm a contingency kinda guy I've got both a USB Type-C and USB mini on it, and thanks to one of the guys on discord mentioning the correct resistor values it appears to do a proper PD charge rate too.

Image from iOS (8).jpg


It's also worth noting that cannulating everything would have meant that I'd not have stubs sticking out into the areas where the control pads go, of course actually having the ATMega32U4 _on_ the board would solve all those problems too, but it didn't work on the last board and I'm through taking chances.

There's more going on with the board bottom that has changed my life after trying it out. As suggested by on of the guys on Discord (I forget who...feel free to comment so we all know!) I hand-cannulated the connections on the Arduino micro as I wanted to have some areas which didn't have a pin going through them so that I could get traces through - this turned out to be fantastic and in future if I do something like this again it'll be how I'm doing everything! Oh and you bet I like chunky headphone jacks (actually, I had it lying around so used it, and it's easy to solder)

Image from iOS (7).jpg


I'll note that I did end up thanking myself for putting socket mounts down for pretty much everything as I was able to test the charging by just wiring in a battery pack and didn't have to go as far as mounting the terminals.

Image from iOS (6).jpg

So with the majority of the board population out of the way, and a growing amount of neurological fatigue I decided that things were worthy of one last push to see if the control board would work as a controller board.

TLDR: It worked perfectly first time, and I've got DMCK5 to thank for that as I used his controller code for the project! Seriously I'm thrilled that I didn't have to have a month-long uphill battle to figure out just what had fallen over for no apparent reason and it's a really nice change!

Image from iOS (5).jpg

So that's about all of it for now, as I said I'll do my best to get the project complete, but no promises and I'm now at the stage where I just want this to be over and the bench cleared :)
 

MRKane

.
.
Joined
Nov 5, 2017
Messages
351
Likes
373
I reckon I owe you all an update as there's been a furious amount of development and problem solving on this end as I'm hell bent on getting this case perfect before I pay for someone else to do a good print of it, also given the timeline involved here I've really only got one shot at this, so no pressure...

So the big core thing I had to do was knock out the "back shell" of this handheld. The back shell had a couple of crucial elements, mounts, supports, and most of all also housed the Pi which is the powerplant here. The concept behind his design is if a Pi5 comes out with a different port arrangement I can simply tweak the ports and stands and get another one made to upgrade it.

It turned out to be a 38 hour print which meant that everything went through quite a thermal range during printing going from an apparent low of 2°C through to nearly 28°C as I've got a heat exchange system going from a conservatory I built through the house so there was no way to keep this stable. There were printing errors but f*ckit. First of all I had to peel the bloody thing, and you bet I screwed up there and I got lots of breaks when combined with said errors, but it served the necessary purpose which was to feel out my spacings.

002_CleanedShell.jpg


Removing the supports (and accidentally half of the case) took the better part of an evening and left lots of scrap. I don't let fluffy up while I do this as he's got a bad habit of eating things like this.

001_PiecesFromShell.jpg


Somehow all of the area with the PI contained in it managed to misprint. There's a delamination running through the whole thing (which you can also see on the right there) which time wise would coincide with the coldest part of the night I guess. Of course I printed it over two nights so it got hit with two "cold shock spots" overall. The skiff at the top was the result of me slipping with the scalpel because I was tired and already knew this case was junked so didn't care. The ports kinda fit.
003_MisprintsHappen.jpg


So the other ports seemed misaligned ever so slightly, and I discovered that this was due to two things: Me not adding a tolerance in Fusion 360 for the layer heights of PLA and also there being zits on the printed supports. You can see the second delamination here too. I hope that's enough ventilation for the cooling system, I'm not 100% sure it will be but honestly don't know how I'd check that...
004_MistakesHappen.jpg

So after concluding that I was 95% OK with everything I'd made I decided to do an intensive push to try and mop up that remaining 5% and go back to the front shell to really perfect everything.

What they don't tell you about neurological fatigue is that your concentration becomes a disposable resource, and you only start the day with a limited amount, which would be fine if I worked retail or something, but I'm a technical developer and I've just been told to learn C++ on top of my usual days work making all sorts of different software. Had to touch fish a bit to do things during lunch etc.

New print came out looking great, and is packed to the eyeballs with improvements over the last one.

Most obvious is the addition of the fuel gauge area which I've decided to just glue some black semi-transparent plastic behind and call it a day. In typical form I've grown rather fond of the "vertical slot" design style as it's just so jarringly different from the swish and stylish things that get commercially produced.
005NewTopPrints.jpg



Mounting screws now go all the way through the unit holding everything together, so I decided to see if clips could be employed here to aid in assembly and they work amazingly well, so much so that I turned around and really did put them everywhere I could.
008_ClipsAreYourFriends.jpg


The screen is held in place with clips which also double as a screw-mount to give me an extra point to hold the shell together around the screen. I'm one of those "how about more" kinda guys so it'll never go amiss.

010_ClipsDoubleAsMountingPoints.jpg

Ok, so successes aside, lets talk problems, and this is where I'd love some feedback from people!

So the thing that's been bugging me for the last two months has been to get the correct seating for the switch thumbsticks. SOMEONE NEEDS TO PUBLISH A DRAWING ON THIS BECAUSE THERE'S NONE ON THE INTERNET!!! _ahem_ ok. Right. I found some 3D scans that someone had done on thingiverse, and with a feeler gauge did my best to seat the thumbsticks perfectly. Naturally this threw out the design and the raised board for custom buttons means that the FFC cables will have to work for their dinner.

011_RaisingMainBoardMistake.jpg


With a bit of twisting and some tweezers I think I did manage to get them attached - it looks like it should just work, but I won't know for sure until I've fired everything up :)

012_SometimesTolerancesWorkInFavour.jpg


And now the power ports - without a 3D model of these from KiCad I can't perfectly place the ports so had to go off of guesswork. I'd probably be a clever monkey to try and get 3D models for the different ports so that I can position things exactly, but for the moment this is a point of contention.

007_PowerPortHoles.jpg

So the next drive is to get the case finalized within the next two days and get it sent to the print shop. At worst I'll make a video with this broken shell and call it a day because it's awfully close to being final, and even though it looks wacky as all hell it feels really nice to hold.

 
Top