I have been tinkering with the Proto1 board combined with an Arduino-DUE(clone) for a test-fixture redesign (it will replace the wonky WinXP-GUI computer - what a nightmare!). The Proto1 board and the DUE fit & work together very nicely.
My "WinXP" replacement project has some very modest needs: 5 RS232 ports, 1 LAN port, and a TFT touch-screen. So far - I've corralled all the necessary DUE libraries and somehow everything seems to be working well.
I should probably create another HAD project just for this test-fixture. Of interest to others might be a rudimentary button-GUI I'd written about 22yrs ago (ran under DOS & DJGPP!) - and was just able to port it to Arduino-land. It's running on the DUE now... and not too shabby! Response time is not super-snappy, but quite acceptable for this application.
I also tried the same GUI library on a Mega2560. Certainly it's slower, but tolerable (to me). The GUI could be used with the Mega2560, but my biggest concern the smallish RAM space. With the GUI, Ethernet, and softserial libraries loaded, I think there was about 5K RAM left. More than adequate for many applications - but not for this test fixture. If you don't need ethernet or softserial, using the Mega2560 might make perfect sense.
Back to the DUE:
I had a subtle but annoying problem with the DUE(clone) board I have... The analogRead() function just never gave proper results, even from the simplest example sketch.
After some probing & staring under a microscope, it seems that the resistor that connects 3.3V to the CPU's ADCREF pin wasn't soldered properly. One end was up in the air and not connected! This was very difficult to spot (and believe!). Here's a microscope pic:
The left end of that resistor is actually NOT connected to the pad. I think it would take some incredible soldering skill to do this on purpose! After a quick touch with the iron - the ADC inputs finally worked perfectly! The soldering quality on the rest of this particular DUE-clone is "dubious" at best. I guess you can't expect too much from clones.
I also tried the DUE's DAC0/DAC1 outputs. They work as expected. I will need some circuitry to convert these to "protected" 0-10V outputs, which is more or less the standard for industrial applications.
Built up the power supply section. Seems to work fine. Next... start on an Arduino-Mega2560 or DUE project.
Partly built up the first Proto1 board. I actually do need this stackup of "Mega2560 + Ethernet-V2 + TFT(SSD1963 chip) display" for a test-fixture project. The parts for the SMPS 24V->5/3.3V power supply have not arrived yet, so you can see that part of the board is still empty.
I ran in to some difficulties in using this board-stacking scheme - which I'm sure others have experienced.
Symptom: It's impossible, or darned near impossible, to actually plug anything into the "receptacle" side of these M/F stacking headers!
Analysis: CHEAP M/F STACKING HEADERS. While it's great to be able to buy a bushel of them for $5, when you look *really* close - often they aren't well made. I know my header repository has parts from at least 3 eBay suppliers. None of them were great quality. The main problem that I've found (with my stock of headers) is that the pin has been pushed into the housing a bit too far, or too hard. This results in the female "fork" side being pushed through the plastic housing a bit and thus are forced closed, or restricted from opening as they should when you try to plug in a pin. I was able to doctor my headers a bit to get past this problem by gently pulling the pins out of the housing by about ~0.5mm.
Solution: I would love to find a reliable source of quality M/F stacking headers, with 14-16mm tails. Any ideas? I can certainly have Samtec make them up, custom... $$$
Symptom: Difficult to get all the pins into all the headers, at the same time.
Read more »