F28069 ADC problem

Filipe Teixeira

Other Parts Discussed in Thread: CONTROLSUITE

Hello,

I am using the F28069 controlcard and the controlstick. Currently I am trying to sample the several channels with the F28069 ADC.

I execute the example Example_2806xAdcSoc supplied with ControlSuite.

I detected one error in the code, the access to the register AdcRegs.ADCCTL2.bit.CLKDIV2EN is EALLOW protected and it was not being access in the code as such, so the clock going to the ADC was the 80MHz.

A part from that, everything seems to be ok.

I call the device_cal function and the value of the registers ADCREFTRIM and ADCOFFTRIM are changed to the following:

ADCREFTRIM => 0x15AA

ADCOFFTRIM => 0x0000

This happens in the two F28069 that I have (the controlcard and controlstick). Is it normal for them to be equal on two different devices?

Then when I am converting, if I connect the analog input to 3v3 via a resistor I don't get 4095 on the conversion register. I get about 4010.

I have tried to set

AdcRegs.ADCCTL2.bit.CLKDIV2EN = 1;
AdcRegs.ADCCTL2.bit.CLKDIV4EN = 1;

and

AdcRegs.ADCSOC0CTL.bit.ACQPS = 63;
AdcRegs.ADCSOC1CTL.bit.ACQPS = 63;

and I still dont get 4095.

If I connect them both to GND, I get zero, which is ok.

If I connect one of them to gnd and another to 3v3, the one connected to ground gives about 56 and the one connected to VCC gives about 4019 with ACQPS of 8, and with ACQPS=63 it gives the values 0 and 4024.

Can someone tell me what is the maximum sampling rate we can have with the F28069 and if the calibration function of this device still has some errors? I would expect to be able to sample 5 channels in under 2microseconds...

Best regards

over 14 years ago

0 MatthewPate over 14 years ago

TI__Guru* 81440 points

Felipe,

These are TMX samples and as such are not bounded by a MIN/MAX(there is usually no MIN/MAX for TMX) error for most/all parameters in the datasheet. The gain and offset is one of these parameters, that can be out of tolerance to the datasheet TYP listed. What you have listed can be expected.

Once a MIN/MAX is published in the DS the part will be guarenteed to meet it.

Best,

Matthew

0 Mauro Fantina over 14 years ago in reply to MatthewPate

Genius 3800 points

Hi Felipe.

I have buy 28069 and I have the same problem.

I connect AN0 to 3.3V and I write the correct value to "AdcRegs.ADCREFTRIM.bit.BG_FINE_TRIM" till I read 4096 to the AdcResultReg.

so I write

EALLOW;
AdcRegs.ADCREFTRIM.bit.BG_FINE_TRIM = "correct value";
EDIS;

after calibration

Best Regards.

0 Filipe Teixeira over 14 years ago in reply to Mauro Fantina

Prodigy 235 points

Hello Mauro,

I have made more measurements and the errors seem to be in the ADCREFTRIM and ADCOFFTRIM.

Can you tell me what are the values stored in flash for the device you have? After calling the device_cal() function I get the values:

ADCREFTRIM => 0x15AA

ADCOFFTRIM => 0x0000

Is it the same on your device?

Best regards,

Filipe Teixeira

0 Mauro Fantina over 14 years ago in reply to Filipe Teixeira

Genius 3800 points

I verify.

The values are the same.

I left ADCOFFTRIM equals to 0.

ok.

0 Filipe Teixeira over 14 years ago in reply to Mauro Fantina

Prodigy 235 points

Hello Mauro,

I have done some tests and got the following values for the registers:

BG.Finetrim = 3;

OFFTRIM = 10;

And with these values I get a very good 0 to 3V3 range.

Best regards,

Filipe Teixeira

0 Mauro Fantina over 14 years ago in reply to Filipe Teixeira

Genius 3800 points

Hi Filipe:

you are right!

I have to calibrate even OFFSET.

I do some measurements and I find this values:

AdcRegs.ADCREFTRIM.bit.BG_FINE_TRIM = 1;

AdcRegs.ADCOFFTRIM.bit.OFFTRIM = 19;

see SPRUGE5A page 17

Best Regards.

C2000™︎ microcontrollers

C2000 microcontrollers forum

F28069 ADC problem