stm32f4discovery troubleshooting and repair

Hardware inside and outside of the ECU
Post Reply
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

stm32f4discovery troubleshooting and repair

Post by AndreyB »

http://rusefi.com/wiki/index.php?title=Hardware:stm32f4discovery_repair

I've just killed two more discovery boards within 5 minutes :( First one I should not have tried moving wires while everything was on, the 2nd one I have no idea why I got smoke - maybe it was already bad. This puts the grand total to 11.

Image

I really need some good news so let's see if we can fix at least some of them?

Out of 11, one says "bad A8" - I guess on this one just one timer capture is fried, at least it does turn on.
Four would not turn on at all - let's call this "Full darkness" scenario

Three of them LD2 is on but it's really weak - let's call this A drop of life

I am open to troubleshooting and repair ideas :)
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
Tambralinga
donator
donator
Posts: 174
Joined: Wed Jun 18, 2014 8:36 am
Location: Bangkok, Thailand

Re: stm32f4discovery troubleshooting and repair

Post by Tambralinga »

Check all diodes :D
Skype : shiva-p
+66 8 5357 5159
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

Tambralinga wrote:Check all diodes :D
I am electronically challenged, I need more specific steps :)
step 1) power discovery with 5 volts
step 2) ???
step 3) ???
step 4) ???
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

A lot of the dead bodies are dim or dark with JP1 and they light up as expected without JP1. What does that mean? Does that mean that these have issues with the main chip?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: stm32f4discovery troubleshooting and repair

Post by kb1gtt »

I see the regulator will provide about .6A before shutting down. I see D3 will drop 1V at .6A, which means it would dissipates .6W. I see the diode is only rated for .3W and that's when it has a good heat sink. In this design it does not have a good heat sink, as it has leads and is not connected to a GND plane or power plane. In this design based on gut feel, I expect it can only handle .2W before overheating failure.

Any how, the short version is that I expect a large number of failures to be caused by D3 failures, which are caused by an accidental short to GND. Some good news, is that D2 is the same diode, and is not used, so you have a spare. Once D2 is relocated to D3, most failures will be resolved.

I'm working on adding to that page. However it also requires me to study the circuit and see what I can identify as reasons for common failures, and what will happen when that failure happens.

In the mean time, you might want to pick up some of this smoke in a bottle, such that you can replace the smoke when you let it out :)

https://www.google.com/search?q=smoke+in+a+bottle&oq=smoke+in+a+bott&aqs=chrome.1.69i57j0l5.5137j0j7&sourceid=chrome&es_sm=122&ie=UTF-8#q=smoke+in+a+bottle&tbm=shop
Welcome to the friendlier side of internet crazy :)
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: stm32f4discovery troubleshooting and repair

Post by kb1gtt »

How many full dark do you have? Those can probably be fixed with the D3 thing.

I'm looking into the other weak points now. Perhaps will have some suggestions by end of tomorrow.
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

A lot of the dead bodies are dim or dark with JP1 and they light up as expected without JP1. What does that mean? Does that mean that these have issues with the main chip?

So far I've located one with failed D1, will try to play with that one. This one does not show voltage on top pad of D1 while all the others do show +5 there.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

I've fixed one of the them! The one where D1 was the issue. Took D2 from a donor board.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: stm32f4discovery troubleshooting and repair

Post by kb1gtt »

Hmmm, you don't happen to know anyone with a thermal camera by chance do you? I would wager a guess your dead players are largely caused by over voltage on the decoupling caps. The problem is that to find the shorted cap you need to remove them one at a time, then once you find the failure, you need to replace the ones you removed. This is yuck for labor. However if you had a thermal camera you could look for the hot one and remove / replace it. I guess you might be able to do the same by touching the board after it's been powered for a while.

Does the ST chip get hot? It's possible the ST chip is the failure, but it's most likely a cap. I wonder if there is some way to make your cell phone camera work as a low cost thermal camera. I know many cameras will make hot break rotors look like they are on fire. Perhaps a 0 ambient light and long exposure would work.

Image

Some low cost options noted here, but still kind of blah in the $ department.
http://www.instructables.com/id/How-to-Make-a-Cheap-Infrared-Camera-1/
http://hackaday.com/2013/03/06/a-real-thermal-imaging-camera-for-300/
Welcome to the friendlier side of internet crazy :)
andrew
Posts: 37
Joined: Sat Aug 02, 2014 1:34 pm
Location: Russia, Рязань

Re: stm32f4discovery troubleshooting and repair

Post by andrew »

Maybe connect high power(3v @ 1a for example) psu instead onboard power supply? Bad capacitors smoke or heat up.
User avatar
rus084
contributor
contributor
Posts: 678
Joined: Sun Dec 01, 2013 1:40 pm
Location: Russia , Stavropol

Re: stm32f4discovery troubleshooting and repair

Post by rus084 »

i think what wain chip on my discovery was destroed .
Because if i delete jumper of power main chip then debugger is work and 5>3.3 chip isn't hot .

May be try connect 3v and much ampers power for burn not working io port (fortunally only pin)?
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

Two of the boards top-right corner of the main chip gets warm/hot.

One of the boards I think R42 gets really hot. R42 is the one between LD6 and B2. It measures the same ~600 Ohm as on other boards, why does it get so hot?
Edit: actually R42 is the resistor for the blue LED. How could it be hot? With my fat fingers maybe it's something close to R42.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
rus084
contributor
contributor
Posts: 678
Joined: Sun Dec 01, 2013 1:40 pm
Location: Russia , Stavropol

Re: stm32f4discovery troubleshooting and repair

Post by rus084 »

600ohm = 600mv breakdowning on 1 diode . Because main chip is short circuit
andrew
Posts: 37
Joined: Sat Aug 02, 2014 1:34 pm
Location: Russia, Рязань

Re: stm32f4discovery troubleshooting and repair

Post by andrew »

Unsolder R42 and try again.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

Off-topic: just got three new boards and I've notices serial numbers (?) on the back

I've got
214190354
214190416
214190713

Jared, what is the lowest number on the dead boards I've send you? :)
The lowest I have is
2122xxxxx

Did they make 214190713-212122000=~2M boards? does this sound like a realistic number? They say there were about 3.5M Raspberry Pi sold.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: stm32f4discovery troubleshooting and repair

Post by kb1gtt »

board numbers are like this.

all say MB997 B-02

Here is the number that changes.
212310257
213180628
213130291
212431122
212310255
212430145
Welcome to the friendlier side of internet crazy :)
s_ali
Posts: 23
Joined: Tue Jul 08, 2014 3:57 pm

Re: stm32f4discovery troubleshooting and repair

Post by s_ali »

Hi mans !
lock at my fotoreport - manual :)
repear coments
repear coments
repear coments.jpg (197.72 KiB) Viewed 19660 times
==================

Step 1 Clear damged PCB
Step 1 Clear damged PCB
Step 1 Clear damged PCB
IMG_20150304_230239 - копия.jpg (89.38 KiB) Viewed 19660 times
==================

Step 2. To replace (or mix) old solder with alloy Rose
To replace (or mix) old solder with alloy Rose
To replace (or mix) old solder with alloy Rose
IMG_20150304_230919 - копия.jpg (120.5 KiB) Viewed 19660 times
==================

Step 3. Clear MCU pins
Clear MCU pins
Clear MCU pins
IMG_20150304_232316 - копия.jpg (79.12 KiB) Viewed 19660 times
==================

Step 4. MCU pins to correct (if it is necessary)
MCU pins to correct (if it is necessary)
MCU pins to correct (if it is necessary)
IMG_20150304_231850 - копия.jpg (70.05 KiB) Viewed 19660 times
==================

Step 5. Deleted oxydes from pins before solders
Deleted oxydes from pins before solders
Deleted oxydes from pins before solders
IMG_20150304_232051 - копия.jpg (104.43 KiB) Viewed 19660 times
==================

Step 6. Ethanol flux
Ethanol a flux
Ethanol a flux
IMG_20150304_232613 - копия.jpg (100.54 KiB) Viewed 19660 times
==================

Step 7. Air soldering :D
Air soldering :D
Air soldering :D
IMG_20150304_234726 - копия.jpg (83.32 KiB) Viewed 19660 times
==================

Step 8. To be happy :)
To be happy :)
To be happy :)
IMG_20150305_091913.jpg (114.41 KiB) Viewed 19660 times
=================
The end ;)
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: stm32f4discovery troubleshooting and repair

Post by kb1gtt »

Very nice, very nice. I have added a link to the wiki for this post. Wiki page found here http://rusefi.com/wiki/index.php?title=Hardware:stm32f4discovery_repair#Step_2_Check_the_target_ST_chip
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

I will continue from http://rusefi.com/forum/viewtopic.php?f=3&t=696&p=17471#p17469 (text slightly updated):
discovery which was inserted into Frankenso 0.4 inside a case under the hood in a warm garage has died (this one now has "1" drawn on it): a bunch of system LEDs blink randomly as if there is a power issue. Took another discovery which was in the garage as well and somehow 10 minutes later this one would not recognize ST-LINK as USB device while rusEfi seemed to be running fine (this one has "2" drawn on it).
Some new content.

Healthy board, diagnostics step 1:
Image

This one says "1" in black sharpie close to the ST logo. This one lights up four LEDs with mini USB and without JP1, same board with JP1 has zero LEDs emitting light. Next step?
Image

This one says "2" close to the ST logo. This one blinks LD1 off/red/off/red without JP1. With JP1, it looks like rusEfi firmware is running but this board is not recognized as ST LINK usb device. Next step?

Image





Note to myself: do not take pictures of black cables on black desk.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: stm32f4discovery troubleshooting and repair

Post by kb1gtt »

Would I be correct that you are using a laptop, the power cord probably has 2 wires with no ground? I suspect a GND loop issue.

I'm not quite sure what could be wrong with board 1, I would wager a guess it's USB power supply related. If powered via Frankenso and not USB, does it preform the same, AKA solid power supply vs USB power supply. I suspect the LD3985M33R is being overloaded and is short cycling. This can happen with a mild over current situation. If you have a hard dead short to GND, you have been blowing some combination D1 to D3. However if you have a slight over current, it allows the LD3985M33R to kick in the thermal limits, turns off, cools down turns on, over heats, etc which results in the blink power to the F4 and blinking 3V supply. I would bet that chip is getting hot when this is happening and if you gave it time, it would blink at a slower rate as it heats up the surrounding PCB. Don't touch the LD3985M33R you'll likely say ouch. What I don't know is what on the F4 is shorting to GND. Could be lots of things. Connecting to the board to the Frankenso power supply will likely burn what ever is causing a problem. With some luck it will be the audio codec chip or some other thing we don't care about. If it burns the F4, then it was broken any how. I might guess ESD for this, but could also be a GND loop issue with the USB cable.

Per board 2, this does seem odd. We know dead shorts to GND will blow D1 to D3 in various ways. In your case those all work, so the issue probably wasn't related to over current. Which leads me to think over voltage. The chances of a voltage surge coming from the Miata, past the F4 and breaking the STLink is not very likely, so I'm more inclined to think the surge came from the USB connector, or perhaps from a human ESD. I find it odd that LD1 is off while LD2 is bright. LD1 is a multi-color LED, connected to 3V , which we know is OK as it's powering LD2. The STLink can make that red, green, or 50/50 red green. However under no situations should that LD1 be off when you have 3V. So this LD1 must be broken, which I can only see happening via a voltage surge. The surge that took out the STLink probably is also what took out the LD1. I would most likely point a finger at the USB connection, as you may have GNDed your car via USB cable. I doubt a human ESD as I don't expect you could take out 2 chips in one event. You body just doesn't store enough energy.

If it's a USB GND issue, you might want to GND your shield before connecting the USB cable. This will allow the PC and vehicle to equalize before the 5V and data connections are made.
Welcome to the friendlier side of internet crazy :)
Rhinoman
contributor
contributor
Posts: 256
Joined: Thu Sep 24, 2015 2:14 pm
Location: Wiltshire, UK

Re: stm32f4discovery troubleshooting and repair

Post by Rhinoman »

I have seen a number of similar issues with grounds on the CANBus, laptop power supplies can cause big problems when they are plugged into the mains supply and running a ground wire between the laptop and the ECU is highly recommended, if you can then always run the laptop from its internal battery. One very well known vehicle manufacturer blew up a lot of ECUs by not using a ground wire of the CANBus of their diagnostic equipment.
The USB cable should ground the two devices through its screen, you could check that that is properly connected.
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

Both boards were damaged while running laptop from a battery, no wall power supply.

Pictures were taken while on a desk with a wall-powered desktop. What tests should I run to confirm which components are blown?
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
User avatar
kb1gtt
contributor
contributor
Posts: 3758
Joined: Tue Sep 10, 2013 1:42 am
Location: ME of USA

Re: stm32f4discovery troubleshooting and repair

Post by kb1gtt »

On board 1, for some reason I thought you said that the red green were blinking, sorry my bad. Those being on solid would appear to indicate the F4 is shorting a pin to GND which is holding the LED's on. I'm reasonably sure that F4 is toasted, probably it's ESD diodes. Can you connect your meter in series with the USB power and measure the current some how? Do you still have that hacked USB cable? You don't need it for USB coms just to get that 5V to power it. The current draw would help shed some light about what is happening. Perhaps you could jumper LDD with the meter instead of a jumper and see how much power the F4 is consuming. I seem to recall the 3v regulator could only produce around 100mA to 150mA. Also connecting board 1 to frankenso power supply would likely burn off what ever is broken, and make it really obvious what's wrong.

On board 2, can you measure voltages from each of the 3 pins of LD1 referenced to the GND pins in the corner of the PCB. That will help show what the STLink chip is doing. I'm pretty sure that chip is simply toast and probably. I'm not sure how esle to test the STLink chip. Board 2 can probably be programmed via different STLink and you can probably still use the F4 part of it.

I feel reasonably confident that each board has a broken large black chip. It's not like prior failure where you can swap D1 for D3 and revive it. I think the major chips are toast, and the board is probably not salvageable. What I wonder is why it happened.

I believe that with the Franekenso boards, Discovery failures have been relatively few. Is that correct? I wonder if there could be some issue with the Frank boards. I don't know of any, but perhaps there could be.
Welcome to the friendlier side of internet crazy :)
User avatar
AndreyB
Site Admin
Posts: 14292
Joined: Wed Aug 28, 2013 1:28 am
Location: Jersey City
Github Username: rusefillc
Slack: Andrey B

Re: stm32f4discovery troubleshooting and repair

Post by AndreyB »

kb1gtt wrote:I believe that with the Franekenso boards, Discovery failures have been relatively few. Is that correct? I wonder if there could be some issue with the Frank boards. I don't know of any, but perhaps there could be.
True, few discovery failures while on Frankenso board. I can recall the mystery of @s board killing discovery unites one by one.

With #1 board I am not 100% sure but it could be that the laptop was not even connected, just long USB cables from headlight area where the ECU was into inside the cabin, and cranking.
Very limited telepathic abilities - please post logs & tunes where appropriate - http://rusefi.com/s/questions

Always looking for C/C++/Java/PHP developers! Please help us see https://rusefi.com/s/howtocontribute
Post Reply