Comparator Input Impedance

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

Hello all,

This must have come up before, but darned if I can find a definitive answer. What is the input impedance of the comparator? I'm using an ATmega162.

I have looked at the data sheet, not there, hard as that is to believe! Thanks.

Greg

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

Since I'm not an EE, that never bothered me. "So high that it doesn't make a difference" is good enough for me. ;) Curious: why is it important?

We can see that ADC is 100Mohm typical. What is the input impedance of a hi-Z digital input? We ain't got no number for that either.

But we do have a number for 1uA input leakage current max at 5V. Can we then imply that the input impedance is 5Mohm or greater?

Now, we also have some leakage numbers for the analog comparator. How do we read that, though, with the spec at 5V/2, and +/- limits of 50nA? Does that imply an input impedance 20x higher than a digital input, 1uA/50nA? Or is it 10x higher, 'cause we are applying 5V/2?

In any case, it would be "high", right? Some dozens of Mohms maybe. Once you get that high it would have to be a pretty wimpy signal for it to matter.

Pop an email to avr@atmel.com and post back the response.

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

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

The comparator inputs are CMOS gates. Thus, the "impedance" is really high (almost infinite). The dominant effect on external circuits is leakage current. This is really hard to spec because it is so small and it can flow either direction, depending on whether internal devices connected to ground on that pin have more leakage than the internal devices on that pin connected to Vcc.

So, go with the leakage specs. And, remember that leakage does not define impedance.

Jim

 

Until Black Lives Matter, we do not have "All Lives Matter"!

 

 

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

The resistance part will be something like 1e12 Ohms. The leakage current is a separate effect. There is also a capacitive component (probably a couple of pf) and a series inductance (most likely tens of nH), depending upon the package.

But as pointed out already, its easy to build circuits that make the input impedance unimportant.

--
"Why am I so soft in the middle when the rest of my life is so hard?"
-Paul Simon

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

Quote:
But as pointed out already, its easy to build circuits that make the input impedance unimportant.

Thanks to all who responded. I will be using it to look at the output of a magnetic pickup. This is part of the ignition system on a 1982 vintage motorcycle. The pickup connects to an electronic ignition module of indeterminate specification. I don't want to probe the module for fear of killing it and I don't want to have to track down another one. That being the case I just want to be careful not to be injecting any problematic current into the thing. A high impedance comparator makes that much easier. I figured it was pretty high, but didn't want to design for 1M ohm only to find it was actually 50K Ohms.

Greg

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

Quote:

The resistance part will be something like 1e12 Ohms.

It isn't clear from the block diagrams, but I think you have the digital input gate "load" as well, in parallel.

Now, Atmel has thought of this, and on the late-model Mega88P there is a DIDR1 to disable that "load".

But I don't think you have that on the now venerable Mega162.

Lee

You can put lipstick on a pig, but it is still a pig.

I've never met a pig I didn't like, as long as you have some salt and pepper.

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

The leakage specs are usually for the hole temperature range. As with typichal diode leakage at moderate temperature leakage will be orders of magnitude lower.

The disabeling of the digital inputs is not to remove the little bit of loading from the inputs, but to cut down current consumption if the input voltage is just in the transition region of the digital signal. The switching of the digital input could also have some minimal feedback to the input. So most likely they don't seperate the input from the digital input stage, but just remove power from the input stage.