Quantcast
Channel: Raspberry Pi Forums
Viewing all articles
Browse latest Browse all 3892

General • Re: ADC creapage on the RP235X

$
0
0
I re-did the interval testing just to check for consistency, to make the data clearer with 12-bit readings, and alternated between reading ADC0 or ADC1 first, shown by '>'. I needed to use my RP2350B so cut that test short at three and a half hours so the growing creep seen isn't quite so large as with longer intervals but can still be observed.

I have taken the average as 24 as that's what I am mostly seeing here and in other testing. There is more noise being seen here which I guess is because the reads are not simultaneous. I expect the actual average is 24.5 which flips between the two, 24 and 25 -

Code:

                                ADC 0   ADC 1      DIVERGEAfter   0 mins, gap  0 mins     >  41      24      41  +17 __|After   1 mins, gap  1 mins        25   >  24      25  +1After   3 mins, gap  2 mins     >  24      24      24After   6 mins, gap  3 mins        25   >  24      25  +1After  10 mins, gap  4 mins     >  24      23      23  -1After  15 mins, gap  5 mins        25   >  25      25  +1After  21 mins, gap  6 mins     >  24      24      24After  28 mins, gap  7 mins        25   >  24      25  +1After  36 mins, gap  8 mins     >  25      24      25  +1After  45 mins, gap  9 mins        25   >  24      25  +1After  55 mins, gap 10 mins     >  24      25      25  +1After  66 mins, gap 11 mins        25   >  25      25  +1After  78 mins, gap 12 mins     >  24      24      24After  91 mins, gap 13 mins        25   >  25      25  +1  __After 105 mins, gap 14 mins     >  26      24      26  +2    |After 120 mins, gap 15 mins        25   >  26      26  +2    |After 136 mins, gap 16 mins     >  27      25      27  +3    |After 153 mins, gap 17 mins        24   >  28      28  +4    |After 171 mins, gap 18 mins     >  28      24      28  +4    |After 190 mins, gap 19 mins        25   >  29      29  +5    |After 210 mins, gap 20 mins     >  30      24      30  +6    |
As can be seen it's pretty much the same as originally tested and the alternating read shows the first clears the creep for the second read which immediately happens after the first.

For example, that first reading, where the RP2350B had been idling, not reading ADC, while I was doing something else -

Code:

After   0 mins, gap  0 mins     >  41      24
Both ADC0 and ADC1 had likely crept up to "41" and the read of ADC0 shows that, but reading also clears the creep so a read of ADC1 is then back to the creep-free "24" average it usually is, what would have been expected.

Statistics: Posted by hippy — Fri Nov 22, 2024 1:04 pm



Viewing all articles
Browse latest Browse all 3892

Trending Articles