STM32f030cct6 port pin failure

[karan]

Posted on April 18, 2017 at 12:47

we developed thermal printer solution using STM32F030CCT6 controller.The issue after 3 or 4 days of working on end user's site some of the pins failed (i mean showing zero resistance to ground).scenario is as below.

there are two cards one card is with stm32f030cc controller which handles some task and another card have printer

motor

driver ic(tb6562ang) which drive ltpd247ab printer mechanism.max switching freq of this signal is 300hz

there is 30 cm long jatag connector between this two card(for input to driving ic).there are 5 pins for motor driver ic in this 20 pin jtag connector.after few days of operation on of the pin from these 5 pin fail.I internally pull up this five pin and there is no external resistance connected on this pin.This is very argent as device is start to fail in numbers so any one please help me.......

#stm32-f0

[AvaTar]

Posted on April 18, 2017 at 13:31

This is very argent as device is start to fail in numbers so any one please help me.......

To be blunt, this is not the fault of anybody here on this forum. Any 'urgency' is solely on your side.

There is most probably no way around a hardware redesign, short of accepting those fails.

I strongly suggest a galvanic isolation of all motor driver hardware.

Or else, the usual hardware protection circuitry to protect silicon from the back-EMF of large coils.

In your place, I would consult an experienced hardware designer, or a hardware-design related forum. Most guys here, including me, are specialized on the software side.

[S.Ma]

Posted on April 18, 2017 at 14:00

The pins are probably stressed beyond the normal condition (electrical tables in the datasheet) with accelerated ageing.

Observe on oscilloscope the signals and dynamic current on the involved lines. HW debug with oscilloscope for voltage and current monitoring is required. Be careful to use the STM32 grounds as reference for the scope probes. Then, depending on what is observed, it will be possible weather SW can correct things...

[AvaTar]

Posted on April 18, 2017 at 14:30

The pins are probably stressed beyond the normal condition (electrical tables in the datasheet) with accelerated ageing.

Just 3 or 4 days are IMHO not sufficient to 'age a part to death' - except they exceed the spec. limits to a point just below breakdown.

Those pin fails are commonly related to back-EMF, especially with motor controlling circuitry.

But there are neither schematics nor layout and measurement data, so I don't lose myself in speculations.

[karan]

Posted on April 19, 2017 at 07:32

Thank you for your reply,

I am not trying to raise question to quality of controller.i am software developer but I have bit of knowledge about hardware(not pro) also and as i am involved in this project i just trying to resolve issue if i can. I know there should be free welling diode for back emf protection but as motor is driving through driver ic we not implement it.so there are probably two issue either some how back emf come to picture or 30cm long wire work as inductor.i try to monitor wave form with 5ms time window on cro but no spike is noticed(tested for two hour).I internally pull-up these 5 pins is it cause any problem???

[AvaTar]

Posted on April 19, 2017 at 09:13

A simple freewheeling diode might not suffice.

There are special suppressor diodes, that can absorb a specific amount of energy.

This protective elements need to be configured according to the (motor) hardware and the expected voltages/currents/durations/energies.

I internally pull-up these 5 pins is it cause any problem???

It should not, assuming the pull-up voltage is correct.

You might check that surges do not enter via the power supply.

Large currents (to/from coils) can create substantial voltage offsets if the ground plane is not very low-impedance.

Galvanic isolation tends to avoid those problems...