PWM using TPM low level voltage

[admin]

Hello there,

I'm using the TPM1CH3 of the S08EL32 to generate a PWM. The PWM is there the the low level votlage is 400mV and the high level voltage is 3V with 3.3V power supply whereas if I turn this pin into an I/O, then low level voltage is 40mV and high level voltage is close to 3.3V.

This makes me in trouble in order to drive a FET from the PWM output.

Is it a normal feature ?

Regards

Stephane

[Encoder1]

This is not normal, I never heard so.

In my experience with S08 family using PWM, 0 is ground and full scale is VDD with (very) few mV difference. I am using S08AC60 and S08H8: both uses S08TPMV3 version. S08EL32 uses S08TPM2 timer version but it shoudn't be any particular difference.

I am used to consider 0=Vss and full scale = Vdd since old MC68HC705P16 and successive MC68HC908 family.

Are you sure there is no static load or even a dynamic load from a low impedance RC filter?

Salvatore

[Stephman]

Well, I was indeed suspecting this is not normal. The TPM1CH3 output pin was connected directly to the gate of a FET. I've removed the FET, leaving thus the pin unconnected and I have the same behavior.

And I don't understand the difference when using the same pin as an I/O which has then proper voltages.

That's why I was thinking that the PWM module may not share the same output driver than the I/O

I do admit that I don"t know what to investigate further....

Stephane

[bigmac]

Hello Stephane,

My understanding is that the PWM output is derived from the TPM module, and the GPIO driver becomes disconnected when the TPM mode is enabled.  It is also possible that the maximum allowable TPM output current is different than the combined output current limit for all pins of 100mA

Perhaps you might check whether you get similar results with the other TPM channels and/or a replacement device.  It is possible that the particular device is damaged, and you will need to rule this out.

If you are attempting to directly drive a high power FET, keep in mind that the gate input capacitance can be quite large.  You need to include a series resistor to limit the output current to a safe value for the MCU.  If the presence of such a resistor results in slow rise time at the FET gate, when compared with the PWM period, this would indicate that an additional high current driver is necessary ot overcome the effects of the gate capacitance.  For 3.3 volt operation, I might suggest a current limiting resistor of say, 330 ohms.

Regards,

Mac

[Stephman]

Hi Mac,

I've finally found something wrong.

This is the first prototype. It has been soldered with a reflow oven but I had to redo some manual work and I've used a flux cleaner which apparently has put a small layer all over the PCB which is more or less conductive.

I've seen that when I've removed the MCU and seen voltages on tracks which were connected to nowhere !

I tried to clean the PCB with alchool and it has improved (reduced theses parasistic voltages) but stranges values are still present. The PWM is still 0.4 - 3V instead of 0-3.3V

As moreover this is a RF PCB (6 GHz PLL), there is no place for a such problem. I'm going to remanufacture a second prototype on monday (hopefully I have several PCB left...

I'll let you know

Regards

Stephane