This is quite exciting.
Proof of concept works up to 921600 baud over these many dodgy cables, hand soldered dead-bug prototyping on the 1206 caps and wire into wire into wire. I’ve finally started actual electronics work at actual work, and this “board” will be fabbed next week as a final product, first with a run of 3 reflow-skillet soldered prototypes, and then an initial run of 20 or so, and, depending how many they break, I’ll get more assembled.
For those that are wondering wtf it is, it’s a basic high-speed USB-RS232 converter, very very simple, but I haven’t been able to find one commercially that does 921600 baud, which I need.
Here is my 1st board reading for fabbing.
I’m getting quite nervous about making sure that all the footprints are correct, that the traces have correct thickness, that the impedance is enough (I have no clue about this), that the dimensions are right.
I just realised that I should put some sort of mounting holes on it! Could be important!
They are some Mill Max test pin headers, 2×4, with 1.27mm pitch, which is nice and tiny!
I’m glad I managed to get some samples, mainly because the minimum quantity I could find them in is 98, and at $11 each, that’s quite step, especially since I only really wanted 2 of them.
I’ll be using them as ICSP headers or maybe JTAG (JTAG fits on 8-pins, right??)
I’ve received more samples! This is fun!
- 2x 28-DIP ATmega328
- 4x 32-TQFP ATmega328
- 2x 100-TQFP ATmega2560
- 2x 44-TQFP ATmega644
- 2x 40-DIP ATmega644 (not pictured)
I’m kinda kicking myself for not having asked for some AT32UC3L’s too, since they are hard to come by in Australia and I don’t want to pay $40 shipping from the US. And, also, I’m going to be using them for my next project.
Good thing that I kept the 44pin->dip converter board samples that I received last year, since I can use them on the 644, though since I also have the 644 in a DIP package, it’s not that important.
I’ve been debating which LCD I should use in my project for a while now. One of the main contenders is the Nokia 5110, mainly because it’s everywhere, it’s cheap, and it’s easy to use! But the biggest problem I have with it is being able to find a reliable source of them, unmounted, and being able to mount them reliably, since they use that annoying little conductive rubber thing that once you take it off, it’s impossible to get back on correctly!
So in an attempt to find something better I turned to other sources, and found (or rather, rediscovered) New Haven Displays. And in particular, the NHD-C12832A in white. It’s transflective, small (41.4×24.3mm), runs off 3.3v, uses SPI (so, only 5 pins) and is relatively cheap at US$11. The biggest issue I had for my prototyping is that they wanted something like $80 for shipping! I ended up doing a group order through Mouser, and got them at about $20 each (with shipping), which is not too bad for this little sucker.
The code used to run it is based off LadyAda’s excellent ST7565 library (which is the actual chip running the thing). I had to modify a couple of things, specifically the hard-coded 8-line assumptions (the 128×32 only accommodates 4 lines – 4 pages), and a couple more little bugs.
I’ve been using the I2C RTC chip from STMicroelectronics, the M41T81, and I’ve managed to make a tiny little arduino library for it based on an existing library for the ubiquitous ds1307. I feel that it’s actually a better chip (on paper at least), it’s cheaper, does down to deciseconds and sets a bit when it has lost main power (so that you know when querying it).
I want to add very basic timezone support for it, since I’m dealing with UTC time internally for everything, and it’s nice to be able to read the time where you’re at. It’s going to be basic mainly because it will do dumb adds/subtracts, it won’t go into details of “is it the 28th of feb on a leap year?” which is enough for what I want.
Well… looks like this little ripper has been obsoleted!