Topics

STM32 upload firmware but not start


Alessandro Pensato
 

Hi all,
I am trying to build my setup with a STM32 pcb.
I tryed to upload the 3.16 and 4 firmware succefully, but after that the system does not start up.
On the same pcb I tryed the STM32 cpu uploading a blink example and it is OK
I tryed also the RTC module and it is OK
 
09/12/2020 11:44:03 
26.75C 
 
09/12/2020 11:44:13 
27.00C 
 
09/12/2020 11:44:23 
26.75C
 
I am trying to find infos to solve this issue, but i only found to check the STM32 and/or the RTC module

 


Khalid Baheyeldin
 

On the Wiki, there is a troubleshooting section in the STM32 Blue Pill page.

https://onstep.groups.io/g/main/wiki/6408

Have you read it and tried everything there that relates to your symptom?

Specifically, the I2C scanner and EEPROM wipe sketches?


davidgreenway@...
 

There are some, "a lot" 64k only STM32's Bluepill's in the market, there has been a long thread on the de-bug of this issue.


Alessandro Pensato
 

I2C Scanner
Scanning...
I2C device found at address 0x57  !
I2C device found at address 0x68  !
done

Blink OK

ll EEPROM
Init done
................................................................
EEPROM Wipe Complete


64k 
Checking across 40000 bytes of memory...
Success


128k
NO SERIAL OUTPUT


Could I have a fake 128k board?
 


Khalid Baheyeldin
 

On Sat, Sep 12, 2020 at 12:58 PM, Alessandro Pensato wrote:
64k 
Checking across 40000 bytes of memory...
Success

128k
NO SERIAL OUTPUT
This may mean that the STM32F103C8T6 chip on the Blue Pill has only 64K.
Or that the 2nd 64K are there but not tested.

Could I have a fake 128k board?
So far, there are no "fakes", as in imitations. The STM32F103 ARM Cortex-M3
is licensed to various companies, so they can be marked CKS, GD or STM.

The issue is that some chips has only 64K limit (which is the official limit for
the "eight" version of the STM32F103C8T6. Others have the additional flash
to make it a 128K. There is no way to tell beforehand.

There is a potential sure way to get a 128K board, although it is not tested.

The 3 button board here should work (Type C, 128K), with one or two mods.

https://www.aliexpress.com/item/32525208361.html

1. The 5V pin next to the USB connector has to be cut off before soldering the pin
headers.

2. After doing this, it needs to be tested and maybe a jumper made from the other
5V pin at the opposite end extended to the 5V next to the USB connector.

As I said, this was never tested, so I am not recommended this yet.

Another community memnber ordered the board to do such tests, but it takes 30 to
50 days for delivery though ...


Alessandro Pensato
 

I have two of these boards and none works.

If the project is based on this board, where we can find a board that is sure that work?

Il 12/09/2020 19:36, Khalid Baheyeldin ha scritto:
On Sat, Sep 12, 2020 at 12:58 PM, Alessandro Pensato wrote:
64k
Checking across 40000 bytes of memory...
Success
128k
NO SERIAL OUTPUT
This may mean that the STM32F103C8T6 chip on the Blue Pill has only 64K.
Or that the 2nd 64K are there but not tested.
Could I have a fake 128k board?
So far, there are no "fakes", as in imitations. The STM32F103 ARM Cortex-M3
is licensed to various companies, so they can be marked CKS, GD or STM.
The issue is that some chips has only 64K limit (which is the official limit for
the "eight" version of the STM32F103C*8*T6. Others have the additional flash
to make it a 128K. There is no way to tell beforehand.
There is a potential sure way to get a 128K board, although it is not tested.
The 3 button board here should work (Type C, 128K), with one or two mods.
https://www.aliexpress.com/item/32525208361.html
1. The 5V pin next to the USB connector has to be cut off before soldering the pin
headers.
2. After doing this, it needs to be tested and maybe a jumper made from the other
5V pin at the opposite end extended to the 5V next to the USB connector.
As I said, this was never tested, so I am not recommended this yet.
Another community memnber ordered the board to do such tests, but it takes 30 to
50 days for delivery though ...
--
Alessandro Pensato
mail: alepensato@...


Khalid Baheyeldin
 

On Sat, Sep 12, 2020 at 01:57 PM, Alessandro Pensato wrote:
I have two of these boards and none works.
They have CKS chips, which were verified to work by Dave Schwartz.

If the project is based on this board, where we can find a board that is sure that work?
It used to be that any Blue Pill ordered from China (via eBay or AliExpress) would work
since they have 128K, despite being advertised as 64K only.

In the past few months a few users reported that the boards are only 64K. We do not
know why. Maybe an old stock of chips that were tested only as 64K made its way to
the manufacturers, or something else.

At present, there is no way to tell whether a board is 128K or 64K ...

You can review this long discussion. No definitive conclusion.