Ultimate7e tutorial: Processor upgrade

Check all your solder joints with a magnifying glass. Not just for bridges but also for good solder on each pin. When using hot air and paste, I've had situations where a pin didn't get soldered properly.

To step through the address space, I believe it's in the "MCU Core" item under "Target". Can't quite remember and I'm not at home now to check.

Wasn't actually worried about the pins needed for programming cause we know that works. But beyond those and the USB, we would want to look at the LCD pins. Potentially also any others that the walkera bootloader might check for (if any like the power control pins). I'm not too familiar with the bootloader, but if everything is starting at 0x08000000 like PB said, then eliminating hardware issues will help.

Another check to do is the vreg. Make sure you are getting 3v3. The vreg and electrolytic cspacitors are pretty close to the processor. Could have taken some damage in the replacement. You won't be able to tell when using st-link if your programmer provides power during programming.

thanks for the pointers guys, I'll have a looke again this evening.

is there any chance I buggered something up when I progranmed the chip without boot being jumped? also how did that happen if you need to short boot to flash?

ok I double checked the soldering and there are no shorts that I can see, and the chip is definitely flashed properly, the 3.3v reg doesn't come on woth the power button, but when I jump the power pins it gives out a steady 3.3v. I tried pulling the battery and hooking 3.3v direct to the debug port gnd and + and that makes no difference, also there is good connectivity between the relevant pins on the mcu and the usb port.

when I go into the stm menu and choose core it opens a window with a load of little boxes with hex addresses, this is way beyond my current knowledge so i'm trying to read up on what it all means, any pointers what I should be looking for there?

this was the chips I purchased, are they the correct ones? >> m.ebay.co.uk/itm/2PCS-MCU-ARM-IC-LQFP-64...?txnId=1630237728017

so with all that it mind is there anything else to check? or do I give up and bin my ever faithfull 7e :/

I believe PB is right that you have a short on your boot0 pin to vcc. I tried shorting boot0 to vcc on my 7e and the power switch won't turn things on, but the power jumper does and it drops straight to stm's built in uart bootloader. When it is in this bootloader, you will not see anything light up on the lcd.

Use a multimeter to check the two boot0 jumper pins above the programming header. You'll probably find that it is shorted. Try tracing that cause it might not necessarily be a short at the processor pins. If your 7e was modded before, you could have wires connected that is causing the short. Take some clear close up pics and maybe we can see more.

You can also use a multimeter and probe the boot0 pin (pin 60) to the few adjacent pins next to it for shorts. Vss is only 3 pin away (pin63) and the short could be under the processor pins if you had too much solder paste.

You should be on the right track so don't give up!

nope no short across boot0 (now at least, i probably cleared it when troubleshooting earlier as i fluxed and wicked all 4 sides again etc before i posted here asking for help) and i'm about ready to give up on the poor thing

one question, when in the stlink software and you open target>mcu core, should you be able to run the code? also to get dfuse to see the tx is there any special method other than the usual deviation method?

time to talk to the other 2 guys as it's bad enough having to buy myself another 7e, i'm not having to be buying them one each too lol

As far as I understand - as soon as target is acquired by ST-Link software it stops running the code. To run the code while connected you need to connect from debugger, not from programmer (which ST-Link software is). No other method except usual Devo's method exist - to "see" the TX from DfuSe it should expose DFU device over the USB. For this bootloader should take control and for this it needs to see EXT button pressed.

How did you find that your boot0 is not pulled up to Vcc? By visual inspection or by using meter?

thanks for the info victor, as for boot0 i checked with a meter for continuity, both between the boot0 jumper points and also between vdd and the mcu pin itself. i'm going to wait until tomorrow for better light to get some good photos through the magnifying glass, but with what i've seen using my phone and blowing the pictures up all my solder points are good and there are no visible bridges. though as i had another go with the 1.2mm shoehorn tip drag soldering to be sure i had good connections i've now managed to pull a single trace out of position (pin 37 - PC6), thankfully there is still connectivity but it means there's no chance i can try and swap out the chip for another one now :/

re dfu mode yeah that's what i thought, i just wondered as it was essentially a bare chip if there was another method to get into dfu mode, but no matter what i try nothing is working.

Yea, when drag soldering you should not be too excited and press on the iron too much. It happened to me once. You can even upload the whole Deviation through ST-Link, but that's not the point - your boot loader does not work, so Deviation is not going to work either.

again many thanks for the info guys, looking through better light today it looks like boot0 may be bridged to pin61 under the mcu, though with my eyes I can't tell if it's a solder bridge or if it's just an existing trace that should be there, also pin 58 and 57 look to be bridged also under the mcu :/ any tips or removing bridges (if they are that is, going to check the photos of the bare board when I get back to the house) that are between trace pads behind the legs and under the mcu? as i don't want to be putting too much heat or pulling any more traces, though tbh I don't know what I'm worrird about as it's dead as it is lol

Desoldering wick can help. If there's not enough solder to wick anything, try adding more solder (yes, I know.. sounds wrong, but more wil also pull out what's already there).
Failing that, you can try placing a scalpel blade between the pins and heat them up.

There aren't any traces behind pins 60 and 61. Pins 57 and 58 do have traces, but should be covered with green solder mask so you shouldn't have solder behind either.

Here's the pic to see what's under the mcu:


Take a picture of the lifted pad you have, it may still be possible to remove the mcu, clean things up and start over. I've done it before, just need a bit of care, and it depends on the current damage. If not, what Arakon suggested is correct, just remember to add a good amount of flux and DONOT drag the wick across the pins, and DONOT pull it hard if it adheres to the pins.

ok I've just pulled the mcu and what I was seeing was the white mcu outline not bridges, I have another 3 mcu's here so I'm away to start fresh with a new one just in case I buggered the chip somehow. I've managed to reposition the pulled trace without snapping it so I'm lucky there, fingers crossed this works!

"she's dead jim"

replaced the mcu with another and this one won't even connect to the stm, the soldering is prefect, there is no bridges but when i plug into the stlink it unloads from windows lol. i've obviously killed something else on the board. i even put back on the original 7e mcu and no difference. and when pulling that back off i managed to pull off a trace properly though it looks like it doesn't connect anyway (pin14 PA0-WKUP)

so be warned people this isn't an easy swap, i class myself as a competent solderer, and i've had 2 extra modules, 2 x 3way switches and the range mod under my belt on this tx, and i have the correct equipment for the job but i've still managed to kill my poor little 7e! so be careful if you plan to try this