Revise, revise, revise!

A project log for LayerOne 2017

Badge design for 2017 conference

charliexcharliex 01/23/2017 at 19:480 Comments

Since its so easy to rev boards nowadays I end up doing it more often than i'd like, but we made a nice black and gold board that everything works on :-

Haven't yet decided if we're sticking the module add on or bare LCD, these are very delicate displays and the flex will break even with light handling which is no good for a conference badge around peoples necks, having the module means it is easy to replace and someone else gets the headache of soldering it on, and it is cheaper than the parts , the sum really is greater than the whole.

I did revise this board immediately after it was submitted to the board house so that i could use the 415/405 variants of the STM32F4, something that happened in the past with the ten year anniversary layerOne proxmark board, darn VCAP1/2/3/4/5.

These LCDs seem like they are randomly thrown together, out of a batch of 20 i get 5 variants. Each with a slightly different controller so the host software has to recognise that and adapt them.

This is an 8 bit parallel version that is crazy cheap and with SDCARD however, i can't figure out a good way of driving it from DMA so i can't get the frame rate i need, the controller can support different modes but they're defined by the FPC connector itself so even though you can see the ID pins you have no hope of really reconfiguring it to a different setup, and even though the makers say they'll custom build ones for you, they never answer emails..

Finding the datasheets is generally straightforward, but unfortunately doesn't help.

This is a truly TFT8K24 revision.

The issue with 8 bit ones is that they have three other signals, chip select, data/command, read/write that also have to be set as the data is transferred, with the DMA its just CS, byte/byte/byte out as fast as you like, I'd be interested to hear if anyone has an idea (other than making a long buffer with all the patterns and mda that as 16 bit , since lots of RAM used there)

This is the bare LCD of the ILI9341 version, which only usually lasts a few bends before it'll break, so i have pads for the LCD and holes for the module board to save it. Note the part if pulled from an eagle lib has 4 missing pads for the XY touchpad..

Here is a version of the module with the GIANT mmc card and yet another variant this time with extra F_CS variant (it was also a DOA LCD) hence the poorly squiggled BAD on it.

Note this one supports two physically different LCD types.

I've also grabbed some of the popular 3.0" screens that are used on the raspberry PI, i found a deal on them cheap but not on 450 of them, it is always something. Those are also driven by SPI, more than likely 9341 variants. Some of them handily have the controller written on a sticker on the board, but it turns out not to be accurate so that has been a fun game.

Here is the battery all mounted up and tested. we've used this battery before , popular ipod type. Though we're using a different boost/charger setup this time. More on that later. This is more of a photodump log.

Fuzzy photo of a board being built

My workdesk tends to be a random explosion of things, i've watched all the various videos on doing horizontal or vertical layouts and all manner of 8 bullets, but who has the time or space... Top left scope has a CAN bus differential showing which the badge is decoding, i picked up a nice differential probe even though most scopes can do it easily, but i love single use tools.

That toorcon badge lower right is from years ago, why is it there? and an original sphero?

Also that was relatively tidy, look at it this morning! Some people are able to focus on one project at a time, I usually have a dozen or so going. I found a nicely made cheapish 30:1 worm gear drive on a 2 phase vexta motor so made a little turntable for it for some work related projects.

Now we just have to lock down the LCD and then I can focus on building lots of firmwares and other useful things I hope, its the opposite of most years where the software is ahead of the hardware, hopefully this time i won't be coding the night before L1.