SOLVED: How can I reduce ADC jitter

[Demon]

I would try the following:

1/. Use the Fixed Voltage Reference (FVR) instead of Vdd as your ADC reference voltage.
2/. Add some capacitance - say 100nF - between the pot wiper and ground.

Cheers
Barry
VK2XBP

1. Yeah, FVR is on my TO-DO list (Richard mentioned it up there).

2. Does it matter if the cap is at the switch or up against PIC (like decoupler caps).

3. Temp is also on my TO-DO list.

[richard]

also try using the fastest conversion clock that does not go Outside the recommended TAD time.
you are using the slowest possible clock
if you use the fvr don't let the wiper volts exceed the fvr output volts

[Aussie Barry]

if you use the fvr don't let the wiper volts exceed the fvr output volts

What happens if the wiper voltage exceeds the FVR voltage?

[Aussie Barry]

It shouldn't really matter if the cap is at the PIC end or the Pot end, as long as the distance between both is not too great.
Note the silicon errata document (Item 1.3) regarding the use of FVR as ADC reference voltage:
https://ww1.microchip.com/downloads/...-80000702J.pdf

Cheers
Barry

[Demon]

It shouldn't really matter if the cap is at the PIC end or the Pot end, as long as the distance between both is not too great.
Note the silicon errata document (Item 1.3) regarding the use of FVR as ADC reference voltage:
https://ww1.microchip.com/downloads/...-80000702J.pdf

Cheers
Barry

Right now I'm testing on 16F18855, but I'll face the same problem on final version with 16F18877.

I'm still trying to figure out how to see the Revision using MeLabs USB programmer. This is the most I can see.

[Ioannis]

I do not think you can see the Revision of your chip with Melabs USB programer. Only with Microchips IDE and a Pickit or ICD programers.

1. Can you test with increased conversion time as the Errata says?

2. Can you test with Vdd reference instead of FVR to be sure of the cause of this jitter? In this case use a linear 5V regulator to have the cleanest Vdd power supply.

Ioannis

[Demon]

I do not think you can see the Revision of your chip with Melabs USB programer. Only with Microchips IDE and a Pickit or ICD programers.

That's what I'm afraid of. I never got my PK4 to work right.

1. Can you test with increased conversion time as the Errata says?

I could try that.

2. Can you test with Vdd reference instead of FVR to be sure of the cause of this jitter? In this case use a linear 5V regulator to have the cleanest Vdd power supply.

Ioannis

I was using VDD before, jittery. I'm using FVR x 4 right now, still jittery.

And 7805 can't handle 3A anyways. That's why I switched to TPS56637 design.

[Ioannis]

Is it possible to just feed a clean 5V to this chip, just to test the ADC?

Ioannis

[Demon]

Have to go to restaurant for wife's birthday. Gonna try 2 bench power supplies later.

[Demon]

Can't edit post 23:

I took OUT the manual averaging routine; it doesn't go to the extreme ends (rarely get completely to 0 or 1024 because of jittter).

I'd like to try the ADC² average a try. Hardware should be faster than anything I can write in PBP.