TMS320F28379D: The threshold comparator output

Luke Chen

Hi champs,

According to ADC PPB Block Diagram in TRM, the ADCPPBxTRIPHI is connect to comparator's positive input and ADCPPBxTRIPLO is connect to another comparator's negative input.

When the value of ADCPPB1RESULT is between 0x0 and 0xFFFF, I conduct below experiments,

PPB1TRIPHI = 0xFFFF, PPB1TRIPLO = 0x0. Then both ADCEVTSTAT.PPB1TRIPHI and ADCEVTSTAT.PPB1TRIPLO bits are 0.
PPB1TRIPHI = 0x0, PPB1TRIPLO = 0x0FFFF. Then both ADCEVTSTAT.PPB1TRIPHI and ADCEVTSTAT.PPB1TRIPLO bits are 1.

Is the evaluation result reasonable? Looks like we need to swap the positive and negative inputs of the threshold comparators in the diagram?

Please advise your comments, thanks.

Regards,

Luke

over 3 years ago

0 Joseph Casuga over 3 years ago

TI__Mastermind 40140 points

Luke,

Yes, evaluation is reasonable and results are as expected. In the first case, no LOW trip is detected since TRIPLO is set to minimum (0) and ADC result is between 0 to 0xFFFF. Likewise, no HIGH trip is detected since TRIPHI is set to maximum (0xFFFF). On the second case, both LOW and HIGH trips were triggered since TRIPLO is set to maximum, meaning any ADC result value which is lower than 0xFFFF will trigger a LOW event. Opposite case will apply to TRIPHI which you set to 0. Hope this makes sense.

Regards,

Joseph

0 Luke Chen over 3 years ago in reply to Joseph Casuga

TI__Expert 5090 points

Joseph,

In the first case, the TRIPHI = 0xFFFF connecting to positive input of comparator, when ADCPPBxRESULT = 0x2000 connecting to negative input of comparator, the comparator should output 1, doesn't it?

Regards,

Luke

+1 Joseph Casuga over 3 years ago in reply to Luke Chen

TI__Mastermind 40140 points

Hi Luke,

I think you are referring to this diagram:

Ok, the diagram really does not represent the intended output of the comparator. We will fix the polarities. Thanks for the feedback.

Regards,

Joseph

0 Luke Chen over 3 years ago in reply to Joseph Casuga

TI__Expert 5090 points

Thanks Joseph, this diagram problem also happen on the TRM of F28002x, F28004x and F2838x.

Luke

0 Joseph Casuga over 3 years ago in reply to Luke Chen

TI__Mastermind 40140 points

Hi Luke,

The source diagrams for PPB is the same across these devices so they should be covered. Next TRM publication will have the correct diagrams.

Regards,

Joseph

0 Luke Chen over 3 years ago in reply to Joseph Casuga

TI__Expert 5090 points

Joseph,

One more thing needs your comment, if this is a comparator, the output should be a Hi/Low voltage level change, what's reason we mark the threshold comparator outputs 'Pulse'?

Regards,

Luke

0 Joseph Casuga over 3 years ago in reply to Luke Chen

TI__Mastermind 40140 points

Luke,

There is actually a version of PPB block diagram that did not make it to the F28379D TRM. We will get this updated to below diagram:

Evaluation of threshold limits is done per load of PPB result register and this makes the output as pulse. Also new diagram is does not put the sign on the inputs, which is just a symbol for the magnitude comparison (not an actual comparator with inverting and non-inverting inputs) where if PPBRESULT < TRIPHI, then output will be low pulse but if PPBRESULT > TRIPHI, then output will be high pulse. Hope the existing diagram as used in F28388D makes more sense.

Regards.

Joseph

0 Luke Chen over 3 years ago in reply to Joseph Casuga

TI__Expert 5090 points

Joseph,

What did you mean low pulse and high pulse? Is it really a pulse(LOW->HI->LOW or HI->LOW->HI) or just Hi / Low voltage level change please?

Regards,

Luke

0 Joseph Casuga over 3 years ago in reply to Luke Chen

TI__Mastermind 40140 points

Luke,

I think the output pulse for both magnitude compare is low->hi->low but checking with design team just to confirm. Will keep you posted as soon as i get some feedback.

Regards,

Joseph

0 Luke Chen over 3 years ago in reply to Joseph Casuga

TI__Expert 5090 points

Joseph,

Any feedback from design team please?

Regards,

Luke

+1 Joseph Casuga over 3 years ago in reply to Luke Chen

TI__Mastermind 40140 points

Luke,

Sorry for the late response. Confirmed that this is a low->high->low behavior from design team.

Regards,

Joseph

C2000™︎ microcontrollers

C2000 microcontrollers forum

TMS320F28379D: The threshold comparator output