AQ6 board seems fried - any suggestions?

[kirakurai]

Hi, my AQ6 board red power LED no longer comes on when connected to a battery. I'm not certain whether it is fried or not, and if it is, if it can be repaired. I would appreciate any suggestions with this.

I was trying to update the firmware via a suggested FTDI cable but was getting "Failed to get init ACK from device" on both win7 and win8. I tried a different FTDI cable with the same results and then I tried connected only GND, RX, TX pins, also with no success. The red power light came on properly every time as long as I was using a 12V 7.0AH battery. Next day, I tried using a different battery: 3S 3300mAh Lipo to continue debugging but the red light did not come on anymore. Even after switching to the old battery the red light remained off. I did not see any sparking or anything indicative of a short during any of these tests but it's possible a short may have occurred. One more thing I noticed, but thought it was by design, is that a power FTDI pin is shorted with the outer J3 connector, is this supposed to be the case? The soldered connections on FTDI and J3, J4 look clean.

I saw someone's post where they were able to repair a fried AQ board by replacing the RECOMs, is this worth trying? Is there some more tests I could do to narrow down my problem before replacing anything?

Thanks a lot,
Kira.

[Kisssys]

U4.JPG (100.21 KiB) Viewed 13984 times

The red power led represents the 3.3 volt line that powers the processors. You may have a bad recom but I would feel U4 and see if it is getting hot. It provides the 3.3 volts to the processor from pin 1 of U4. R13 feeds U4 with 5 volts. If you should have 5 volts on one side and only a little less on the other side. If you have a large drop then it usually is a shorted processor and you need to contact your vendor as to how to proceed. Having the volts volts of course confirms you have a working processor recom.

[kirakurai]

Steve, thank you very much for the diagram and your suggestions. The board's lights came back on today again (though I did not do anything to fix the board) but I measured all the pins to make sure they matched your diagram and the 1st pin on U4 is indeed 3.3V but the 4th pin measures 1.9V. Does this sound like a short?

One more thing that is strange is that if I do not put the J4 jumper on but leave the FTDI cable plugged in the board acts as though the jumper is on (red solid light on). If I unplug the FTDI cable, the lights blink again like normal. Should the FTDI cable alone be putting the board into bootloader mode?

Thanks,
Julia

[Max]

Hi Julia, do you have J3 closed AND power coming via the FTDI connector? It sounds like the board is getting power via the serial connection, which isn't good. If there's no jumper on J3, maybe a short? W/out a jumper, the upper/outer pin of J3 (square) should have continuity only to the power pin on the serial header. The other J3 pin connects to the onboard 5v circuit and should have continuity to, eg, the 5V solder pads on the PWM headers.

It should not be behaving like you describe simply by connecting the serial/FTDI. What serial adapter are you using?

-Max

[kirakurai]

Hi Max,

No, I don't close J3, only J4 (jumper is not shown in the pic) and I do have main power coming from 3S Lipo. The USB FTDI cable is the Flyduino's USB UART-CP2012 and it's supplying 3.3V. Here is a diagram of what I'm getting on the connectors. Steve is most likely correct in that I probably have a burned processor and may need to contact the seller directly to see if there are any repairs they may be able to do since the board is not completely dead. Is there anything else I could try?

AQ6 FTDI issue

BoardWiringFTDI-5.png (378.57 KiB) Viewed 14009 times

Thanks,
Julia.

[Kisssys]

Pin 1 is 3.3 volts to the processor. 1.9 volts on pin4 is about right. Pin 3 is the 3.3 volts out to the sensor group and you might check it.

Sounds like your voltages may be fine.

As Max mentioned its not good to supply 5 volts from the FTDI cable to the AQ, even worse would be to put a 3.3 volt cable on the AQ with the 5 volt ftdi jumper in place. Actually I don't think they put out 3v3 volts, just 3.3 volt logic level compatible.

[kirakurai]

Hi guys,
I spent another night debugging the board and double checking all the pins and connections I could think of. All of the voltages seem correct, even checked the RECOMs - fine as well. The board lights behave correctly now when switching modes with J4 jumper. Even tried checking soldering connections under the microscope - can't find any issues there either. However, I'm still getting:

stm32flash - http://stm32flash.googlecode.com/

Using Parser : Intel HEX
Using Parser : Intel HEX
Serial Config: 57600 8E1

Failed to get init ACK from device

This was tried with two different FTDI cables and on 2 different windows machines (Win8, Win7). I tried doing the firmware update on a MAC too but it would not show any msgs at all in the log window even though it was able to detect the board and connect to it when not in a bootloader mode. Also, connecting to the board using putty when not in the bootloader mode shows that some kind of data is being received by the comp which proves further that the board is capable of sending data over the link just fine.
I'm not sure what else to try short of re-soldering all the FTDI and jumper connections just to rule it out for certain , which is prolly what I'll be doing next... and I still do not know why the board was completely misbehaving when I first started this post :/ and now it seems to work short of " failed to get init ACK..."

Julia.

[chschmid]

Hi Julia

Did you set the serial speed to 115200?


Cheers
Christof

[Max]

Julia, are you sure nothing else has that COM port open while you're trying to flash? You have to manually disconnect from the AQ in QGC if you were already connected, or had tried to. QGC 1.2 checked this for you, but that's currently broken in 1.3 betas. And of course Putty, needs to be disconnected also. Sorry if this is obvious... it's just a common mistake so it needs to be asked.

You can also try using stm32flash.exe directly from the command line (you can find it in qgroundcontrol/aq/bin folder). Using that, you can connect to AQ board while it is in bootloader mode and get some basic info (IIRC, you just run "stm32flash COMx" where COMx is your port number, eg. COM4). Unfortunately the "Failed to get init ACK from device" error is a bit generic and gets thrown for various reasons, so it's hard to determine the exact problem simply based on that.

-Max

[kirakurai]

Christof, I've tried on every setting: 115200, 57600, and 38400 with each cable with no luck. Thank you.

Max, I make sure that the QGC is not connected to the board before trying to flash, I also restart it every time to be safe. I am using QGC 1.3 Beta 2 but it does not connect on its own and I do double check that it says not connected before trying to update, and of course I only fire up putty in isolated tests.
I've also tried to use it from command interface with the same error result.

I hope to get to re-soldering in a few days and will posts the results, though I have little hope it would fix it.
Thanks alot guys for your suggestions, I really appreciate it.
Julia.