wrong device marking STM32F100RCT6

[raducuradian]

Posted on March 23, 2014 at 07:57

I have chosen STM32F100RCT6 for a new project because of the large number of PWM channels it has. I was not able to activate the clock for timers 12-17, therefore I conduct some more experiments with this chip. Surprisingly, the MCU_DEVICE_ID located at 0xE0042000 shows that the chip is STM32F103, not corresponding to marking.

The devices I bought from Farnell. Does anyone encounters this kind of problem?Could it be a problem with a whole lot of ICs or only a few?I don't know how to go on with my project. Where should I order real STM32F100RCT6 chips. #wrong-device-marking

[Danish1]

Posted on March 23, 2014 at 11:50

I have had some items from Farnell which were not as ordered (sm resistors).

I assumed this happened because a customer returned them and somewhere along the line they ended up in the wrong bag. Farnell supplied correct replacements without question (not even asking for the wrong ones back). But if the part bear the wrong markings on them this is more serious. (Note: another manufacturer of arm microcontrollers once put an old part code into the ID of the first batches of their new part and issued an errata warning of their slip-up!) If Farnell have incorrectly marked parts, they should want to know about it. Your part should have batch / date codes on it. Ask Farnell to replace your part with one from a different batch.  - Danish

[raducuradian]

Posted on March 23, 2014 at 19:07

Thank you for your reply.

It's not an error of Farnell, it's a wrong marking from ST. The controller is not according to the marking.I struggled two days trying to activate the clocks for TIM12-TIM17 which exists on STM32F100RCT6.Then I found out by reading DEVICE_ID that my chip is not really STM32F100RCT6 is STM32F103. Of course TIM12-TIM17 are not existing at STM32F103.I have three more chips from farnell but I am nearly sure they have the same problem. Because it is unlikely this error to appear for one chip only.I need real STM32F100RCT6 and I don't know how to get them.

[Danish1]

Posted on March 23, 2014 at 19:49

Those ICs have a certain date / batch code on them.

It is unlikely that chips with a different date / batch code on them will have the same problem. Ask Farnell if they have any with a different date code.

[zzdz2]

Posted on March 24, 2014 at 09:32

I don't know how to go on with my project.

I'm just curious, why do you need that many PWM chans?

[Amel NASRI]

Posted on March 24, 2014 at 11:46 It seems that you are using STM32F100RCT6A which is based on STM32F103 die. STM32F100RCT6B contains the timers you need. -Mayla-

[Tesla DeLorean]

Posted on March 24, 2014 at 14:11

Is there a clear chart of parts vs peripherals, I find it nearly impossible to get clear answers from the Data Sheets and Web Pages.

I get vague descriptions of ''up to 16 timers'' or ''up to 512 KB flash'', and lower grade parts have some sub-set. What timers are present is more helpful than just the number, because the implementation is often sparse (holes in the count)

[raducuradian]

Posted on March 24, 2014 at 20:35

Thanks for your hint.

I didn't know about these significant differences between the STM32F100RCT6A and B. Could you please let me know where stands that the STM32F100A is based on STM32F103 die? Thanks

[raducuradian]

Posted on March 24, 2014 at 20:39

I need to control 24 high power LED drivers using PWM, each channel with independent duty cycle. 

[zzdz2]

Posted on March 25, 2014 at 15:45

I need to control 24 high power LED drivers using PWM, each channel with independent duty cycle. 

I suspected it can be something with LEDs, I wonder if it could be realized as LED matrix.It would allow to control 100+ LEDs with a single MCU.I suppose software sigma-delta modulation instead of hardware PWM would do.