Hi,
I'm using a TMS320F28335 on an eZdsp evaluation board and experience some "trouble" with the ADC.
The device is configured as following:
ADCCLK= 25 MHz ; ACQPS = 1 ; CPS = 0 --> sampling time = 80ns.
Simultaneous sampling! Start-Stop-Mode.
The first conversion is started using SOCA (ePWM1) and following settings:
Conv0,-2,-4,-6 = 0, Conv1,-3,-5,-7 = 2 --> Sample ( ADCINa0 & ADCINb0 followed by ADCINa2 & ADCINb2 ) --> do 4 double-conversions (Maxconv = 7)
After this sampling is done, I save the values (about 45µs after SOCA in ePWM1_Int), change the Conv-registers to other channels and start a new conversion using SOCb (ePwm1) 125µs after sample No. 1. After the second sample I save the new values as well (in SEQ1_Int) and set the configuration back to it's state before the first sample (this includes resetting the sequencer) an start all over again. This seems to work fine so far.
So, now I have the following problem: the first two samples (in ADCRESULT0 and ADCRESULT1) don't fit !
Example:
4 Values(dec) ADCINa0 (regular battery): 2195, 2099,2099,2100 (Resultregister: 0,4,8,12)
4 Values(dec) ADCINa0 (regular battery): 2199, 2102,2103,2104 (Resultregister: 1,5,9,13)
4 Values(dec) ADCINa0 (tied to ground): 4, 4, 4, 5 (Resultregister: 2,6,10,14)
...
So it seems that the first two samples (2195 and 2199) of the whole sequence are containing errors and I wonder, where these errors come from. Can it be, that ePWM1_Int, which has the same "trigger" as SOCA (ePWM1.CMPA = 0), causes this or is it something else?
Thanks for any help!
Joe
