LC filters on Vcc and ADC input?

Go To Last Post
6 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've read the docs on several of the Atmel chips. The ones with an AVCC pin show a 10uH coil and a 100nF cap. I understand this is to clean up the analog input for use with the ADC. However, I'm using an Tiny84 with no AVCC pin. Should I still use an LC filter on the VCC pin?
And now the ADC sampling circuit. I'm using a 470k + 10k voltage divider to sample my battery voltage. The ADC will use the 1.1V internal reference. Should I use a series inductor between the divider and ADC pin with a 100nF cap from the ADC pin to ground? Would this actually help clean up the ADC signal to any appreciable level? The battery voltage only changes a little under load. The load is switched at 244Hz.

Jim M., Rank amateur AVR guy.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

LC reduces ripple on AVCC. If you dont use this as a ref, I dont see how it helps. Batteries dont have ripple, so I cant see how it helps in that type of circuit either. The 1.1v ref is +-5%, so you need to calibrate it against a known V.

Imagecraft compiler user

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Should I still use an LC filter on the VCC pin?

On Attiny84 you don´t have AVcc, it's all together. So you should use it only if you have ripple on your power supply or you have some load's switch, you need to see your power supply.
Quote:
The ADC will use the 1.1V internal reference. Should I use a series inductor between the divider and ADC pin with a 100nF cap from the ADC pin to ground? Would this actually help clean up the ADC signal to any appreciable level? The battery voltage only changes a little under load. The load is switched at 244Hz.

A capacitor after your voltage divider may help, but is good to "see" your Vcc when you switch your load, maybe it produce some noise on it.

Regards,

Bruno Muswieck

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

bobgardner wrote:
LC reduces ripple on AVCC. If you dont use this as a ref, I dont see how it helps. Batteries dont have ripple, so I cant see how it helps in that type of circuit either. The 1.1v ref is +-5%, so you need to calibrate it against a known V.

But if the batteries have a load being switched 244 times a second, there can be noise in the battery supply due to the hard switching of the PWM. But there are 2 regulators and plenty of filtering between the 2 regulators and at Vcc and I am also using a 100nF cap at VCC any way. So scratch the LC filter at VCC.

However, I was thinking that the LC filter would help filter out high frequency noise in the ADC sample created by ringing at the turn-on / turn-off events.

Yup. I know about the tolerance of the 1.1V Vref and have provisions for calibrating. But good note.

Thanks.

Jim M., Rank amateur AVR guy.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

brunomusw wrote:

A capacitor after your voltage divider may help, but is good to "see" your Vcc when you switch your load, maybe it produce some noise on it.

I've put a scope on it while on breadboard and there is definitely noise. However, in a breadboard things are always different with respect to noise.
I may just leave the pads/tracks in place to use an LC filter, then test after a prototype is assembled. I can always just use a 0 ohm chip in place of the inductor if it doesn't help. Right?

Jim M., Rank amateur AVR guy.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
I can always just use a 0 ohm chip in place of the inductor if it doesn't help. Right?

You can always, but may be no effect..
Try to see what noises you have, you said noise, but just noise is vague... Even in breadboard maybe some frequencies are from some source of swicth..
But if your concern is only about read the battery voltages, may be you're too concern about it...

Regards,

Bruno Muswieck