DIYEFI.org Forum

Ok, briefs comments about the building here:

* The first thing I solder is the MCU and the crystal, they the hardest components. I reflow them in my oven (originally intended for food)
* Then, populate the PLL and crystal related passives, and the bdm header. 0805 are easy to solder by hand.

What I want, first of all, is to load the SM and verify the MCU is working. Once that miracle happens, I'll have the first freeems running in latin america, and the first firmware running on specific hardware :-)

http://img827.imageshack.us/img827/1087/1002720.jpg

http://img526.imageshack.us/img526/9559/1002736i.jpg

http://img132.imageshack.us/img132/8250/1002738kd.jpg

SM loaded!!
Now I have to load FreeEMS firmware, after buying a usb cable (I should have put a mini usb connector there). I never loaded freeems-vanilla, so its time to get Sean's loader, etc.

You can load freeems using the bdm too. but you'll not be able to see anything unless you connect an led to the pin that the xgate is twiddling.. I endors testing it through serial, though, as you need to verify that the chip is erase protected too.

Fred.

FreeEMS firmware loaded!

Next:
* populate some analog circuitry

I was going to say "next fix solder bridges before smoke comes out" but you've already done most of that.

Keep it up.

Fred.

In terms of supported software, I think that RPM is one of the first you might want to populate. Fred can you suggest what circuits might be best to populate to match the most mature parts of software?

He should populate inputs first and verify correct behaviour, one by one, if necessary or whatever, it's down to the board/schem design, not me. I'd be trying to verify numbers coming out of the software on the ADC channels vs expected numbers with various fixed value resistors. Datalog packets contain this. MTX can generate them, or he could quickly bring ECUManager up to standard and use that. EssEss's thing is lib like now, apparently. Unsure how to use it.

After inputs, RPM would be a good place to start, you'll want to check square waves of different voltages and finally the VR those things are designed for. Using Simple he can get a fuel pulse per input pulse and watch for excess outputs from noise, as noise is not handled. He'll want some debug code in place to align the input and output wave forms on a scope to verify the chip is handling it all correctly, or not.

Fred.

jharvey wrote:In terms of supported software, I think that RPM is one of the first you might want to populate. Fred can you suggest what circuits might be best to populate to match the most mature parts of software?

I might hold it until I get a better soldering iron. uMAX10 is a daunting package. If I install it, I'll let you know. Without doubts, its the most interesting input to have available for testing

Doing the uMAX10 first or near first might be a good idea. It's a bit hard to say. On one hand, more smt experience will probably produce a better result, however, doing that one last and breaking it after the rest of the board has many parts, can make for an expensive board to toss. I might encourage trying it earlier than later.

I soldered the uMAX10, and it wasn't so hard. However, Pin A6 isn't being pulled up correctly (0.04v) , so it can't leave the SM mode. I'm going to force the pin with an external pullup, altough the TA card doesn't need it...