Same wire going to adc and comparator kills interrupts

I have the output of an op amp go directly to the BBB adc and the values are read correctly.
But, the same wire is going to a comparator to generate interrupts.
It looks like the comparator is not firing all the way to ground because of the adc drag.
I thought of placing a 1K resistor from the adc input to ground but that does nothing.
Anything higher is going to bother the adc. I’m stumped.

I have the output of an op amp go directly to the BBB adc and the values are read correctly.
But, the same wire is going to a comparator to generate interrupts.
It looks like the comparator is not firing all the way to ground because of the adc drag.
I thought of placing a 1K resistor from the adc input to ground but that does nothing.
Anything higher is going to bother the adc. I’m stumped.

Well, it depends on the opamp. Some opamps are not able to drive all the way to 0V, but as long as it goes below the logic low for your I/O input, it shouldn’t matter. If you really want an opamp that goes to zero, choose one that is capable to operating rail to rail. Also, the opamp feedback circuit is important because you are going to want some hysteresis or your GPIO input might see some noise close to the trigger point.

Regards,
John

The lmv762 comparatorworks 100% if I sever its input from the adc - it's only when the adc is tied to it that it does't go to ground at the output.

From: maxmike <maxmi...@gmail.com>
Reply-To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Date: Monday, January 19, 2015 at 5:10 PM
To: "beagl...@googlegroups.com" <beagl...@googlegroups.com>
Subject: [beagleboard] Same wire going to adc and comparator kills
interrupts

I have the output of an op amp go directly to the BBB adc and the
values are read correctly.
But, the same wire is going to a comparator to generate interrupts.
It looks like the comparator is not firing all the way to ground
because of the adc drag.
I thought of placing a 1K resistor from the adc input to ground but
that does nothing.
Anything higher is going to bother the adc. I'm stumped.
Well, it depends on the opamp. Some opamps are not able to drive all
the way to 0V, but as long as it goes below the logic low for your I/O
input, it shouldn¹t matter. If you really want an opamp that goes to
zero, choose one that is capable to operating rail to rail. Also, the
opamp feedback circuit is important because you are going to want some
hysteresis or your GPIO input might see some noise close to the trigger
point.

Regards,
John

--

For more options, visit http://beagleboard.org/discuss

---

You received this message because you are subscribed to the Google
Groups "BeagleBoard" group.

To unsubscribe from this group and stop receiving emails from it, send
an email to beagleboard...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

The lmv762 comparatorworks 100% if I sever its input from the adc - it's
only when the adc is tied to it that it does't go to ground at the output.

Well, the datasheet shows VO has a worse case voltage range from 250mv to
VCC-0.35v. Is this consistent with your measurements?

Regards,
John

Being a software guy I tend to be quick to use a sledgehammer, but if push came to shove would a buffer before the comparator screw things up?

Hi software guy,

why don’t you drop the hardware and use the sampled ADC value to trigger the interrupt?

The interrupt has to occur waaaaaaay before the adc has a chance to measure the value that caused it.
I.e. it has to be very fast.

O

Well, the datasheet shows VO has a worse case voltage range from 250mv to
VCC-0.35v. Is this consistent with your measurements?

Yes - actually I’m seeing 3.3V to 0.

Yeah, it’s that or parallel the entire path from signal source. I have to say, these TI adc’s seem to have too low an impedance.