Bluepill faillure


Wim
 

Hi everybody,

I've been testing and running an Onstep conversion of an EQ5 mount for some weeks now.
Everything was working nicely for some time, I was able to shoot some images with decent results so was very happy with that.

However, a few days ago I couldn't start my Onstep anymore. I could see and measure the 5V output on the buck converter. The LED's on the power module and the RTC are on, but not on the STM32, wifi and USB module.

So I started measuring some voltages and it seems like the 5V was ok on all modules, but on the 3V3 I measured nothing.
I decided tot take the STM32 out and noticed something odd... The resistort R3 seems to have burnt up and is basically a shortcircuit now with the solder.

So, I wonder if anybody has a clue as to how this could have happened? I did change one thing to my setup shortly before this problem occured. I changed the gears to a 16/40 combination instead of 16/48. Offcourse I reflashed with these new values . I would imagine it might stress the motors and the drivers a bit more, but could this also be the cause of the STM32 faillure or do I have to look somewhere else?

Cheers,
Wim


Khalid Baheyeldin
 

The R3 looks to be modified to have a solder short instead of the resistor.
This is usually done if flashing fails due to an incorrect value resistor, as described in the troubleshooting section of the Blue Pill Wiki page.
So that may not be the issue.

Before you do anything, power up the controller again, and wait for 20 or 30 seconds, and see if anything is too hot to touch. If there are any, then disconnect power immediately.
If there is nothing that is too hot, then measure the voltages coming into the Blue Pill


Dave Schwartz
 

R3 is the resistor between the BOOT0 (center pin) and the actual BOOT0 input on the MCU. Some people have been saying that they need to short this resistor to get the BP to go into flash mode reliably. Is it possible that you did this and just forgot or, if you got it in a kit that it was done for you? Hard to imagine how that could melt the solder due to an overcurrent especially as there seems to be a lot more of it than there would have been initially due to the SMD placement.

If the WeMos power LED is not on, that's not a good sign because it is independently powered... its 5V comes from the common 5V trace that snakes around the PCB. The CP2102 power LED won't come on unless you plug in the USB cable... it is intentionally not connected to the PCB 5V power bus.

Was that the 3.3V measurement you made on the 3.3V pin of the Blue Pill and/or WeMos? Both of these are outputs from the module's own 5V to 3.3V regulator. If you have 5V on that input pin and not 3.3V on that output pin it would indicate the module regulator is blown. On the Blue Pill, the D1 red LED is powered from the 3.3V so if the regulator is gone you won't get the power LED coming on.

Nothing you can do with the motors or motor side of the driver (outside of a lightening strike I suppose) could affect the Blue Pill... the logic side and motor side of the drivers are pretty well isolated.

Speaking of lightening... its that time of year again (we've had a few in Ontario already). Lots of people on the LX200GPS group suffer failures every year due to nearby lightening strokes if cables are left connected. Could that be a possibility?

On 2021-04-29 3:14 p.m., Wim wrote:
Hi everybody,

I've been testing and running an Onstep conversion of an EQ5 mount for some weeks now.
Everything was working nicely for some time, I was able to shoot some images with decent results so was very happy with that.

However, a few days ago I couldn't start my Onstep anymore. I could see and measure the 5V output on the buck converter. The LED's on the power module and the RTC are on, but not on the STM32, wifi and USB module.

So I started measuring some voltages and it seems like the 5V was ok on all modules, but on the 3V3 I measured nothing.
I decided tot take the STM32 out and noticed something odd... The resistort R3 seems to have burnt up and is basically a shortcircuit now with the solder.

So, I wonder if anybody has a clue as to how this could have happened? I did change one thing to my setup shortly before this problem occured. I changed the gears to a 16/40 combination instead of 16/48. Offcourse I reflashed with these new values . I would imagine it might stress the motors and the drivers a bit more, but could this also be the cause of the STM32 faillure or do I have to look somewhere else?

Cheers,
Wim


George Cushing
 

Shouldn't be any power at the USB/RS232 adapter. I used this diagram to chase down a dead short in the RTC's female header. I've circled where the 5V is delivered to the STM32 and the Wifi module. To echo what Howard stated recently, 5V at the board level doesn't mean you have it a the component. Years of repairing mount electronic has taught me that 95% of the problems show up in the mechanical connections. They are a necessary evil, but keep them to minimum and on your boat solder everything..


Wim
 

Thanks for the hints guys.
Well I bought the kit from you Khalid, so if anybody know if it's removed on purpose then it's you problably.
I must say it does indeed look like the most logical explanation.

So right now I am definitely worried about the Wemos not coming on, because I did measure 5V on it's input pin. 
Since I'm not getting 3,3V on either the Wemos or the STM32 that doesn't looke great.

I do know the distinct smell of burnt electronics and I am pretty sure that nothing has been burnt up, nor has anything been too hot to my knowledge.

I will power up again tonight and check all modules for heating tonight. The 5V measurements I already did but I will write down what the 3V3 pins say. I believe it was basicaly 0 or a few mV...


Khalid Baheyeldin
 

On Fri, Apr 30, 2021 at 03:24 AM, Wim wrote:
Well I bought the kit from you Khalid, so if anybody know if it's removed on purpose then it's you problably.
I must say it does indeed look like the most logical explanation.
Then it must be me that did the solder bridge. I had one or two modules that didn't want to flash until I soldered them.
If I did that, then I tested the module to be sure it is 128K and working.

Let us know what you find out with power measurement.


Wim
 

I did some further tests , and it turns out that it is just the STM32 which has no 3,3V on the pin, it was about 18mV.
On the Wemos, I did measure 3,3V (and 5V offcourse). 
I'm going to try and see if I can flash it from Arduino to see if it's still ok. (I didn't have to flash it yet because you did it for me in the kit)

As for the STM32, I ordered 2 replacements (an heir and a spare ;) ). I guess it should arrive in about 2 weeks from China.
I can't really do anything else I think, until I get it.

One more thing I wanted to mention is that when I was measuring on the Wemos, I accidentally slipped from the GND pin to D4 next to it, and I saw the blue led blink when i did that... So, I'm hoping it means the Wemos is still alive at least.

Also , I pushed the reset button on the STM32, does that help or lock it up possibly?


Khalid Baheyeldin
 

On Sat, May 1, 2021 at 05:30 AM, Wim wrote:
I did some further tests , and it turns out that it is just the STM32 which has no 3,3V on the pin, it was about 18mV.
What about the LEDs on the STM32, and the 5V input pins (someone posted a diagram of how the wiring in a previous message).
In other words: does the STM32 get 5V input or not, and if so, does it light up?

I'm going to try and see if I can flash it from Arduino to see if it's still ok. (I didn't have to flash it yet because you did it for me in the kit)
Yes, please do that. That will tell us at least if the STM32 changes behaviour from Flash to Run position (power LED lights up, but PC13 does not when flashing, and both light up when in Run mode).

One more thing I wanted to mention is that when I was measuring on the Wemos, I accidentally slipped from the GND pin to D4 next to it, and I saw the blue led blink when i did that... So, I'm hoping it means the Wemos is still alive at least.
Perhaps that is true.
Also another mystery is why the WeMos does not light up and blink.
If the STM32 is bad, then the WeMos should just blink blue.
It means the input power to it is good.

It is not doing that, which says there is a problem somewhere.

Unless the WeMos is not flashed. I don't remember if I flashed it or not. Try flashing it and see if it comes up with no STM32 module in the PCB.

Also , I pushed the reset button on the STM32, does that help or lock it up possibly?
Unless you feel that it does not bounce back, I don't think it locked up.
If it does not bounce back, then it could be a problem.


Wim
 

Hi Khalid,

Just to make it clear, I have already used the controller for about a month with no problems.
The Wemos was indeed flashed by you , and it worked perfectly. I did some tracked photography, GOTO operations and guided photography with it already in the past month.

Now , I'm starting to fear for the Wemos too because it doesn't really do anything when I connect it to USB, although I can measure the 5V and 3,3V when it's plugged in, but no flashing LED.

Can you tell me where you buy them maybe ? I'm nog going to order a new one untill I have the replacement STM32, but I like to be prepared ;). I don't seem to find the exact same ones on Ali.

Now, the STM32, I measure 5V on the pin, but no LED's light up , not even the power led.
So that makes me fear the worst for that module, and I hope it will be fixed with a replacement.
Also , the reset button doesn't lock up or anything , so that's not going to be the problem either.

On the USB CP2102 I measure no voltages either, but I guess that's normal since they come from the STM32...


Wim
 

I do have another question about flashing the Wemos... 
Do I have to open Wifi.ino from the addons/Wifi directory? 

I read a bit about configuration settings, and I see some options which are not in the config.h if I do it like that , but they are in the config.h when I open Onstep.ino in Arduino...


Mike Ahner
 

On Sun, May 2, 2021 at 04:39 AM, Wim wrote:
On the USB CP2102 I measure no voltages either, but I guess that's normal since they come from the STM32...
The USB CP2102 is NOT powered from the STM32 or pcb, but only from the USB cable. This is by design to prevent backfeeding 5VDC at the same time as the main power is supply 5vdc.

One thing to do is to remove all the pluggable modules and check the 5vdc power from the voltage regulator along all the traces on the board. Make sure you have 5v at the socket pins for each module. Then power down and plug in one at a time. then power up and test voltage again. OnStep itself will not work without the Blue Pill and the RTC but you will more easily be able to tell if anything is causing a problem.

Also check your power supply to be sure it's not the problem, you can even use smaller batteries without the motors to test, especially C & D cell batteries. The motors need at least ~10 vdc to operate, but everything else is either 5v or 3.3v.

If the board was working for while, it could still be a bad solder connection. After pulling out all the pluggable modules, I would carefully verify the soldering on each module and the pcb to be certain. That's more common than you might think. I find it helpful to use a bright light and a magnifying lens of some sort.

-Mike


Khalid Baheyeldin
 

On Sun, May 2, 2021 at 05:39 AM, Wim wrote:
Just to make it clear, I have already used the controller for about a month with no problems.
The Wemos was indeed flashed by you , and it worked perfectly. I did some tracked photography, GOTO operations and guided photography with it already in the past month.
Good to know.

Now , I'm starting to fear for the Wemos too because it doesn't really do anything when I connect it to USB, although I can measure the 5V and 3,3V when it's plugged in, but no flashing LED.
The fact that nothing is lighting up (STM32, WeMos, ...) tells me that something bad has happened. Maybe a power short or surge that fried the components? I have never seen anything that will kill all modules on the board though.

The other possibility is that the PCB itself had a mishap. But you say that 5V output is good from the LM2596 module.

To me it is a puzzle.

This video shows what should light up when you power up.
Two LEDs on STM32, blinking LED on WeMos, then it goes steady.

https://www.youtube.com/watch?v=H34dbb7FO1E

Can you tell me where you buy them maybe ? I'm nog going to order a new one untill I have the replacement STM32, but I like to be prepared ;). I don't seem to find the exact same ones on Ali.
I get them from eBay. Nothing special about a specific model or anything. For example:

https://www.ebay.com/itm/142754701042


Khalid Baheyeldin
 

On Sun, May 2, 2021 at 05:49 AM, Wim wrote:
Do I have to open Wifi.ino from the addons/Wifi directory? 
Yes. You should not change any value in any file under addons.

I read a bit about configuration settings, and I see some options which are not in the config.h if I do it like that , but they are in the config.h when I open Onstep.ino in Arduino...
Use the Online Configuration Generator and it will generate a Config.h file for you.
You need the values from the spreadsheet.


John Petterson
 

On Sat, May 1, 2021 at 04:30 AM, Wim wrote:
I did some further tests , and it turns out that it is just the STM32 which has no 3,3V on the pin, it was about 18mV.
On the Wemos, I did measure 3,3V (and 5V offcourse). 
The 3.3V pin on both of these devices is an output voltage from an onboard regulator intended by the designers for powering devices attached to them.  So not seeing that on the STM32 may be an indication that the STM32 itself is failing.


Wim
 

I believe this is the case indeed John... Waiting for a replacement right now , and I hope the Wemos is still good...

@Mike : I doubt it's the powersupply , I've built a custom 18V batterypack , and I have a 19.2V laptop supply , both worked and both give the same results. (The +5V output is ok and can be measured on each +5V pin except the CP2102 for obvious reasons)


Wim
 

Ok , I did some other testing today...
I attached the Wemos to a micro USB powersupply , and it started blinking right away. Me happy.

Then I decided to take the STM32 module out , and plug the Wemos in and power it up. Again the Wemos started blinking (and kept doing so offcourse, because there's no STM32 running Onstep).



When I plugged the micro USB charger into the STM32 (which was removed from the PCB offcourse) , still nothing happened , not a single LED went on. 
So , I hope that it is the only part I need to replace, however I fear for the CP2102 also. 
When I plug the power supply on the CP2102 , also nothing happens. This would be bad because I fixed it with Epoxy on the PCB to be stronger.

I'm beginning to think I may have destroyed those modules when I tested a new USB mouse for my son... He bought one, and when he plugged it in, he got a mention that a USB device was drawing too much power form the USB port.
I also plugged it into my pc, and just maybe I did that while Onstep was connected too. Not sure if one faulty USB device could damage others on the same bus... But it may have been what happened a few weeks ago.

The micro USB cable I used for Onstep seems to be destroyed too, because when I connect it to a USB power supply, and the Wemos , nothing happens...

So the quest continues ...


Wim
 

Also , it would make sense that this is why the Wemos is still good, it had to have been plugged out if I was flashing the STM32...


Dave Schwartz
 

On 2021-05-05 5:18 a.m., Wim wrote:
Ok , I did some other testing today...
I attached the Wemos to a micro USB powersupply , and it started blinking right away. Me happy.
That is good... blinking in groups of flashes indicates that the WiFi software is running on it.

Then I decided to take the STM32 module out , and plug the Wemos in and power it up. Again the Wemos started blinking (and kept doing so offcourse, because there's no STM32 running Onstep).
Ditto.



When I plugged the micro USB charger into the STM32 (which was removed from the PCB offcourse) , still nothing happened , not a single LED went on.
So , I hope that it is the only part I need to replace, however I fear for the CP2102 also.
When I plug the power supply on the CP2102 , also nothing happens. This would be bad because I fixed it with Epoxy on the PCB to be stronger.
Why would you be using a charger?

The CP2102 is not powered from the PCB - only from the USB. Connect it to your PC with a known good cable (one which works for power and data with some other device) and you should see the LEDs come on for a while and Windows should recognize it and create a COM device you can see in the Device Manager. The LEDs go out after a while just because the Windows driver times out after a period of inactivity. If this doesn't happen then it is probably dead.

I'm beginning to think I may have destroyed those modules when I tested a new USB mouse for my son... He bought one, and when he plugged it in, he got a mention that a USB device was drawing too much power form the USB port.
That's unlikely. The CP2102 isolates everything from the PC except the data signals and there is no power or out-of-range voltages that could happen there (aside from lightening, or its poor cousin, static).
I also plugged it into my pc, and just maybe I did that while Onstep was connected too. Not sure if one faulty USB device could damage others on the same bus... But it may have been what happened a few weeks ago.

The micro USB cable I used for Onstep seems to be destroyed too, because when I connect it to a USB power supply, and the Wemos , nothing happens...
That's not good. Its hard to imagine how that would happen without some catastrophe which none of us can fathom from this distance.

So the quest continues ...


Wim
 

Hi Dave,

I used the charger because I didn't have another data micro USB cable until today. It was just a check to see what would happen if I put USB power on the Wemos.

So, I have found another micro USB datacable now, and I attached my PC to the CP2102... At first nothing seemed to happen, but then I saw the onboard LED light up extremely faint... This became brighter and brighter, and after about a minute, it was at normal brightness, but kept on burning, and no detection on the pc.
I guess this means its dead too. 

I ordered 2 new CP2102's yesterday, too bad I'll have to wait a bit longer now to test everything... I'm going to desolder the CP2102 today, and put the project in the fridge for a few weeks until the mods arrive.
I hope the bluepills will be ok. They seem to be difficult to come by I understand? However my order from about 10 days ago should be in local customs already...

Cheers,
Wim


Dave Schwartz
 

If that cable worked for a different device (was recognized and data transferred) then yes the CP2102 is dead too. This seems to have been a fairly catastrophic failure.

It has become a crap-shoot as to whether an STM32F103C8T6 module from eBay/AliExpress will be a 128KB or 64KB device. The C8T6 is supposed to be (by STM specification) a 64KB device but for the longest time they were all coming as 128KB, whether from STM or their licenced off-brand manufacturer CKS - I suppose it was cheaper to make 128KB devices and label them under both C8 and CB, which was the 128KB-spec'd device). They seem to have sorted this out and there are a lot more 64KB C8's in the mix now. Unfortunately, the source for the true CBT6 as well as the 256KB upgrade CCT6 (RobotDyn) is out of stock on both.

On 2021-05-06 4:17 a.m., Wim wrote:
Hi Dave,

I used the charger because I didn't have another data micro USB cable until today. It was just a check to see what would happen if I put USB power on the Wemos.

So, I have found another micro USB datacable now, and I attached my PC to the CP2102... At first nothing seemed to happen, but then I saw the onboard LED light up extremely faint... This became brighter and brighter, and after about a minute, it was at normal brightness, but kept on burning, and no detection on the pc.
I guess this means its dead too.

I ordered 2 new CP2102's yesterday, too bad I'll have to wait a bit longer now to test everything... I'm going to desolder the CP2102 today, and put the project in the fridge for a few weeks until the mods arrive.
I hope the bluepills will be ok. They seem to be difficult to come by I understand? However my order from about 10 days ago should be in local customs already...

Cheers,
Wim