Burnt out my MS3 daughterboard - Help diagnose

[gurov]

I seem to have burnt out my ms3 daughterboard.

It no longer initializes. I've ordered another ms3 board to replace this, but do not want to hook a new one back up to the wiring until i know at least a little bit more about what failed.

the last change i made was add control of the electronic thermostat to INJ H on the ms3x expansion, in BMW land this is a heater (about 12 ohm) that plugs into the thermostat to change its characteristics.

My wiring is done via the stock ecu's case and connector, i've cleared the path for this by taking control away from the stock ECU's FET array by desoldering those. There is no short to ground between INJ H and ground in this case with the stock ECU off. regardless of this, ms3x INJ drivers have short circuit protection.

I was changing the cranking RPM from 300 to 200 when the failure occured, but i suspect this was a coincidence.

At the time of the failure, the heater device mentioned above was unplugged. So INJ H was just hanging in the air, like it has been before.

The failure occured about 1 minute after starting the engine, and starting to idle it. I was trying to diagnose an issues where there's no fuel delivered right after starting to crank.

Things i have tried so far:

usb is detected (shows up on both windows and linux)
remove MS3 from car, onto the bench.
remove ms3x board
tried several computers for tunerstudio- windows/linux
tried flashing the firmware with/without bootloader jumper, both linux/windows flashers
tried another baseboard


What can i do to diagnose the CPU and ms3 components to attempt to figure out which component suffered a failure. Anybody I can send this board to for further investigation, so things like this could be prevented in the future ?

[billr]

What you described doing with the MS3X should not have taken out the MS3. What is the primary symptom now, no normal connection to TS? With it on the bench, only power and ground and com connected, same results? Have you removed the MS3 and checked the voltage at the 40-pin socket on the mainboard?

[gurov]

What you described doing with the MS3X should not have taken out the MS3. What is the primary symptom now, no normal connection to TS? With it on the bench, only power and ground and com connected, same results? Have you removed the MS3 and checked the voltage at the 40-pin socket on the mainboard?

i looked up the model numbers for the FETs that are on board, the have short circuit protection/overload protection etc. those should've been fine.

primary symptom is unable to communicate with the CPU and no start, no pump prime, no injector prime - dead.
can't flash it with/without bootloader.

there's 12v going to the voltage regulator on the baseboard, 5v coming out of it. another ms3 works and powers up on the same board and i can communicate with it

same no-go results with it on the bench, +12v, ground, USB comms, original ms3 = dead, another ms3 = alive.

I assume there's correct voltage on the 40 pin, cause another ms3 powers up and does the right thing.

It really shouldn't have taken MS3 out like this. I've done some really REALLY dumb s*** before debugging this thing on the bench, and it's always come back, but this time, on a car that ran a few minutes, it gave up the ghost ALL OF THE SUDDEN.

There was no typical burnt electronics smell either, I opened the box up quickly to smell it. Nothing looks burnt.

I've got a few MS2 boards floating around as well that aren't gonna be used, So i'll use those to test the baseboard utilities on the car, in case one of those components somehow got damaged. and my wiring is sending voltage down some line, But then again.

hmmm... i should also measure resistance between both db37's and the 40pin/every pin of the ms3 processor.

[billr]

Well, i agree with you about it probably not being the mainboard if a different MS3 works OK; trying one of those MS2s wouldn't hurt, though, for further confidence. Is this a fairly new MS3? If so, it also wouldn't hurt to contact your vendor about its seeming to fail. Have you been "bench testing" for just com with the MS3X removed? I assume you have done a careful visual check of the MS3 board using a magnifier and bright light. Note that there are intentionally several "solder bridges" between pins on the CPU chip.

[gurov]

Well, i agree with you about it probably not being the mainboard if a different MS3 works OK; trying one of those MS2s wouldn't hurt, though, for further confidence. Is this a fairly new MS3? If so, it also wouldn't hurt to contact your vendor about its seeming to fail. Have you been "bench testing" for just com with the MS3X removed? I assume you have done a careful visual check of the MS3 board using a magnifier and bright light. Note that there are intentionally several "solder bridges" between pins on the CPU chip.

this ms3 is not new at all, it was one of the first consumer ms3 daughterboards released. the board reads MS3 v0.1 (C) 2009. Have there been significant and potentially relevant changes since 2009 to the ms3 daughterboard ?

diyautotune sold it, but i doubt they have the electric engineers on staff who would be able to tell me more with little information, if they do, they're more than welcome to have this dead board to figure out why something like this can happen.

for comms to the cpu, i've tried to flash it with/without bootloader jumper on the bench - no go.

i took a glance at the board, and there wasn't anything visibly weird with it. I'll try the ms2 board, and see if that dies before plugging in another ms3.

[grom_e30]

possibly a dead processor?

if thats the case a few can replace it for you there the ones i know of.

James, jsmcortina in the uk
Jean Bélanger, racingmini_mtl in Canada
Peter Florance, usa.

[gurov]

possibly a dead processor?

if thats the case a few can replace it for you there the ones i know of.

James, jsmcortina in the uk
Jean Bélanger, racingmini_mtl in Canada
Peter Florance, usa.

well clearly it's a dead processor. but i want to know WHY it died, so when i got to replace it, i don't keep burning them out. sourcing another one is not the issue here, i already ordered another one.

i may have found a smoking gun, so to speak. i believe it may have something to do with my base board's fuel pump circuit.

i'm still trying to wrap my head around it, and doing more debugging.

[racingmini_mtl]

Sourcing a processor is not the issue. You need someone who can load the monitor on the new processor and this is only available from a few approved persons. Without the monitor, the processor is useless as you can't load and use the MS3 code.

Jean

[gurov]

Sourcing a processor is not the issue. You need someone who can load the monitor on the new processor and this is only available from a few approved persons. Without the monitor, the processor is useless as you can't load and use the MS3 code.

Jean

that's totally not where i'm going with this. why exactly would I go that way instead of ordering another ms3 board like i've already done ? i'm confused.

if the confusion is coming from me saying "processor" instead of "ms3 daughterboard" in my post above, i apologize for the confusion.

replacing just the processor is clearly way more of a pain in the ass than just ordering another ms3 board.

[gurov]

also, i think i found the issue that caused this. bear with me here.

the fuel pump driver on my baseboard was still the old transistor pair Q19 + Q2. R40 ( the 1ohm 1/2 watt resistor) looked discolored when i just desoldered it. like it had a bit of heat applied to it.

furthermore, when i went to test this baseboard with ms2 board, it seems i have burnt that guy's pin 33 out (it's now short circuited to ground). there should be around 1k resistance between that pin and ground, like there is on pin 34 ( fidle ), and confirmed by the other ms2 board i have that's not been in this baseboard.

i don't know why this would happen now, perhaps the components that were in the fuel pump drive finally reached their limit of abuse, and starter to short circuit things.

the spare ms3 board I have did not suffer this fate because I did not place a load on the fuel pump circuit to test it until i was sure (which i made sure via ms2 daughterboard sacrifice)

I've replaced the fuel pump drive stock setup with IRLZ44 mosfet, basically the same thing that the manual recommends for boost control.