HVP-KV46150M ADC converter not working

[jirictibor]

Hello,

I am having trouble with the ADC converter at HVP-KV46150M control board. I have two pieces of this board and problem is the same. It behaves that the EndOfScan interrupt occurs, but in result registers RSLTx are zeros. Converters must been configured properly, because when the exact same code is uploded to TWR-KV46F150M, the converter works normally.

There are the same microcontrollers on those boards HVP-KV46150M and TWR-KV46F150M, so could it seems to be some hardware issue? Could I check the controller's revision or production series somehow?

Has anyone already met with this kind of problem, or have any suggestions ?

If somebody want to try the code on own hw, it is attached in zip below.

Thank you for any advice.

Jiri

Original Attachment has been moved to: kv46_twr_adc-fmstr.zip

[jirictibor]

Hello people,

I've find solution for this issue. As I proposed, i was trying to run ADC at HVP-KV46 board. The problem was, that the ANALOG VDD is made on the motherboard (HVP-MC3PH). So the processor board only doesn't have analog Vdd. For the debuging purposes there is possible to interconnect two test points, then the analog and digital Vdd is connected and ADC works fine.

Sorry for the misdoubt.

Jiri

[jirictibor]

Hi xiangjun,

thank you for reply. I have already tried triggering by 100 milisecond PIT and result was the same. I think that FMSTR1_Poll(); function makes pretty enough long delay to completing conversion.

But as I said, exactly the same code was running with no problems at TWR-KV46F150M (exactly the same MCU, only another dev. platform.). So it could not be an software issue i think. This is why I have asked how to check the manufacturing series of the chip, because it seems to be an hardware issue.

Anyone has any experience with HVP-KV46150M control board please? Especially with ADC.

Thanks.

Jiri

[jirictibor]

Hello,

I have to add some important facts about this issue.

• End Of Scan interrupt is periodically occurring as expected, so I assume that ADC converter runs (SOMEHOW) but....
• Result register of every slot RSLTx is totally clear - 0x0000, so it leads to that conversion not be done properly (I assume, if there is some hardware issue on the board e.g. pin is not connected etc., RSLT register can't be totally clear, because of conversion noise or EMI etc.)

So where is the problem? Ones more, the same program runs perfectly on different development platform (TWR-KV46F150M) with the exactly same microcontroller. Can't see any errata or something about this issue.

Please help, we have bought HVP platform for 500€, but we can't use it.

Somebody from NXP has any suggestions? It must be already tested. ADC is the most significant periphery for motor control.

Thank you,

Jiri.

[xiangjun_rong]

Hi, Jiri,

I do not think the ADC module of KV46 has any issue, pls refer to the schematics of the HVP-KV46F150, and select an ADC analog channel, whose voltage is not GND, then you can develop code to test the analog voltage.

I tried to load/compile your code, but my PE version is not compatible with your code.

BR

Xiangjun Rong

[xiangjun_rong]

Hi, Jiri,

I think the code in int main(void) has issue:

  while(1)
  {
      FMSTR1_Poll();
      index++;
      CADC_DRV_SoftTriggerConv(adConv1_IDX,0);
  }

As you know that the DSC is too fast to trigger ADC if you set software triggering main loop.

I suggest you add the PIT beans, in the ISR of PIT, call the      CADC_DRV_SoftTriggerConv(adConv1_IDX,0);, while make sure that the PIT cycle time is greater than the total ADC conversion time of all ADC channels in one scan process.

Hope it can help you

BR

XiangJun Rong