Your PCB renders are so pretty, what application are you using?
1 Like
Its KiCAD, most of the previous renders have been standard, this last set though I used with Ray tracing turned on, sure comes up looking the goods.
1 Like
Wow it sure does! Where did you get those footprints? I’ve always have had a bit of fight with Kicad to get the right ones. Currently experimenting with LibrePCB which looks promising!
I like the fact that Kicad exports to FreeCAD, it’s becoming quite the opensource ecosystem!
I think that’s the main gripe with KiCAD is the libraries, lack of components and simple method of managing/updating symbols, footprints and 3d models.
As we progressed, any items that we couldn’t find and definitely need, I modelled myself following the dimensions from the datasheets and reference photos. We created our own shared library and keep updating our component list in there.
1 Like
Great effort! It’s a shame there’s no strong push to standardize this with Kicad, it’s probably not a very hot or exciting job when it comes to improving software.
A quick update, our pro-micro style test boards are almost done and we expect them back soon to test, once they are confirmed OK we will be shipping off the MMjoy board design for fabrication.(Yay!)
@Troll was kind enough to put me in touch with Mr. Mega_Mozg himself and he was gracious enough to allow us to pre-flash our boards with MMjoy2 prior to dispatch (once we are at that stage), which is a big plus for us & (I believe the novice) end users.
He has given us permission to place “Powered by MMJoy2” on the silk and we have also included the disclaimer “NOT FOR COMMERCIAL USE” to ensure that there is no confusion about these boards being specifically for personal use.
We believe that these simple aspects will make the boards that much more simplified for our target audience.
Now, it’s just a case of hurry up and wait… although there will still be some testing to perform as we have never run a card.with a maximised configuration, we want to make sure that all the connections will work in unison if the cards is populated to it’s maximum.
5 Likes
Awesome news! 
I’ll buy a couple!
1 Like
This sounds very interesting, I could probably use a couple too!
1 Like
A quick update:
things are grinding along in the background, albeit very slowly…
we have our test boards produced and they’ve now been shipped (lots of delays in China due to the obvious crisis), once they come back and we confirm that the production can handle it, we’ll ship off our MMjoy card design for fabrication (super exciting to condsider)
Might be another couple of weeks, but we’re getting there.
5 Likes
update: sample boards arrived today!
these are simple arduino-esque DIY boards (all 1206 SMD components & some pin headers) with a small daughter board that hosts 2 shift registers that can be left connected or seperated (via mouse bites) for mounting elsewhere or inside something smaller
we used this small project as a proof of capabilities of the manufacturer, to get a handle on the quality of the end product, a little bit of testing of these (at a glance everything looks spot on), a final triple check of the MMJoy boards then its a click & send for our initial test batch to be fabricated.
these will also provide a bit more practise of my SMD soldering; (both welding and hot air with paste)
pretty stoked they have turned out so nicely, gives a good vibe that the MMJoy boards are going to turn out great, the black PCBs are pretty slick, but I’m a bit nostalgic for the old school green.
6 Likes
a minor setback, as well as a live & learn scenario (& why we did non-critical boards for the test)
more of a learning curve with KiCAD than anything else & getting bitten by the app.
We encountered a little faux pas when we submitted this job & didnt realise it, even after we had received the boards. (still waiting for a few core components to arrive before we start assembling and testing)
It was only when reviewing the boards in the 3D viewer that one of my colleagues realised that there was an issue. Even looking at the real boards we didn’t notice it & in the PCB view KiCAD didn’t show the issue, but in the 3D viewer it was evident at a glance.
A limitation of KiCAD, possibly a newby mistake as well, but only 1 or the 3 boards on each panel is usable. Not a complete loss, it could have been worse, but still disappointing to have lost those boards to such a simple mistake.
Otherwise everything about the fabrication appears to be spot on.
1 Like
Oups!
I was afraid my PCBs would have some faults that I put there as well, but luckily they didn’t. Of course, they are a lot simpler than yours…
2 Likes
it was a simple copy/paste issue that bit us
screwed up the ground plane on the 2 additional boards on each panel
the original was fine, just not the additions
KiCAD obviously doesn’t support this (multiple boards - as it screws up all the net assignments)
Had we used a gerber file editor for panelisation - wouldn’t have been a problem
We are obviously well aware of this caveat now and will be including it in our QA checks prior to dispatch for fabrication
5 Likes
we’ve been waiting for some components on the slow boat from china
(specifically the ATMega32u4 MCUs)
these have only just arrived so that we can assemble the test boards,
we’re in lock down here so things have kind of come to a stand still.
once that has occurred, we will be submitting the MMjoy board for the initial batch.
2 Likes
A little update in the void…
we’ve been ordering some components we somehow missed
and also acquiring a few additional components to do some testing with.
with those additional delays we’ve taken a step back and looked at the design again and rethought about a few items.
here’s a snippet
please ignore the missing crystal, still waiting for the final component to produce the 3D object for this one
Firstly we’ve removed the reset switch, wasn’t really required, but you can still reset; we’ve labelled the 2 pins on the ICSP that can be used to reset (by shorting) which still allows for that function if required, we’re hoping you’ll never need it though.
You can also see we’ve stepped away from the micro USB and gone to a full sized USB B port.
the thought being you could mount the card inside a housing and have the port exposed and plug directly to it, or if you were against having anything physically interacting with the board, then the option is there to add a small USB B panel mount cable. like the one pictured below
(all of a $2 investment from Ali Express)
this will take any strain off the panel mounted USB B port and allow a sacrificial cable extension to suffer any potential (unintentional) physical abuse.
you might also notice the 4 sets of triple jumper headers, this is for the TLE 5010 / SPI Bus
depending on whether you want to use more of one than the other, initially we thought about 2 sets of 4, but then thought someone might have 5 TLE sensors which would render them being unable to use any SPI BUS devices
(also if anyone has any examples of devices they might want to use on SPI Bus please let us know so that we can look at the pinouts etc.)
so we broke the connectors down into pairs with a jumper for each pair.
Trying to provide a little bit of flexibility into the board whilst still retaining some simplicity into the config.
therefore each pair of connectors when jumpered is either configured for a TLE 5010 sensor, or for SPI Bus, (one or the other, not both) for each specific pair.
We apologise that its taking a while, but we will get there, & would rather get it right than have a dozen versions produced.
hope everyone is keeping safe.
cheers
2 Likes
Good decision!
One thing… In my previous TLE5010 projects I have encountered the issue with having to remove MISO line resistors on the sensor boards and put one on the main PCB board instead. Have you thought about this?
3 Likes
Hey @Troll we haven’t done enough with TLE5010 sensors yet to have encountered this issue
many thanks for the feedback
I’ve passed this along to the electronics guru component of the team.
We’re also thinking about a test unit to plug in each board so that it is maxed out with controllers/inputs
to do a complete QA check on each board once their built, for the elements which we cannot simply test via software.
if anyone else has any hidden gems they think might be worth considering,
please feel free to give us a hoy and we’ll look into ammending our design accordingly
thanks in advance.
cheers
I meant MISO, of course.
The thing is that using multiple TLE5010 sensors the resistance on the MISO line becomes too high, if each sensor has its own. The solution is to just remove them on the sensor boards and add one between MISO and VCC on the controller board.
2 Likes
Any update on these? They’ve pricked my interest as I’ve not had that fine soldering experience before, also I’ve not yet gotten an end design in mind and will likely change things around a few times before I am happy once I use a board (which could include adding more HATs than the initial design I’d draw up).
I’ve not worked before with micro controllers but how difficult would it be to flash the mmjoy away later if I want to write and load my own control set(I’ve got a programming background though higher level than driver / micro board controller programming).