Interfacing GPIO with circuits at different ground potential

LeafLabs Garden » Support » Maple Hardware Support

Interfacing GPIO with circuits at different ground potential

(14 posts) (4 voices)

Started 5 years ago by structuresound
Latest reply from poslathian

structuresound
Member

I suppose this is a more general question about microcontrollers, but I thought I'd see if anyone here has ideas about it.

I'm looking at ways of scaling the voltage coming out of a pin to interface with analog circuits running at higher voltage, in this case 9v supplies running at virtual GND at 4.5v in some cases I need the PWM to swing from 4.5 to 9v.

one solution I've come up with is to run a voltage divider being 9v -> 100k -> output -> 100k -> PIN. So when the pin is low, the output is 4.5 and goes up from there. Because of potential current draws I'm connecting the output to the non-inverting input of an opamp. I think this should make the current a reasonable amount for the GPIO to handle, at least in output mode (maybe there's a problem when set to input or at boot) and it seems to be working. My question is, is there a better way to do this (interface a microcontroller pin with a circuit at a different ground potential)

one of the advantages of doing it my way as I see is that if I make the voltage divider this way with V+, then I don't have to AC couple the circuit and thus get very stable DC voltages when I run the PWM through a lo-pass etc. Maybe I should just buffer the PWM first, however if I do that then I need to AC couple it so that the LOW state doesn't hit the low rail of the opamp (causing the output to swing up to V+).

Ideas? Possibly a good start to a general discussion about simple ways of dealing with GPIO as control source for low current (i.e. not motor control) situations.

Posted 5 years ago #
gbulmer
Moderator

Can you explain a bit more about what your trying to achieve?

I'd recommend getting hold of a copy of the "Art of Electronics" by Horowitz and Hill
(http://www.amazon.com/Art-Electronics-Paul-Horowitz/dp/0521370957)
Maybe from a local library. They cover interfacing in depth.

It sounds like you might get the effect you want by switching the pin between OUTPUT (Push-Pull) HIGH, and OPEN_DRAIN, HIGH, and use suitable pull-up resistor+maybe a zener diode. How much current is being sourced (or sunk) by the Maples pin?

Posted 5 years ago #
okie
Administrator

structuresound, is this an audio application?

You could use a rail-to-rail opamp to avoid having to AC couple the PWM. I've been using the TLV274 for audio stuff using the Maple. I think it won't misbehave too much with rail-to-rail input and output.

Posted 5 years ago #
gbulmer
Moderator

Okie - If it's only level translating PWM, a comparator would do, wouldn't it?

Posted 5 years ago #
poslathian
LeafLabs

I am not sure of the pwm performance over open drain (although it does work). I imagine with a big enough resistor and pulling up to a high voltage you would start to get a distorted square (depending on the speed, obviously). But this might be the easiest way to get started.

An opamp circuit sounds like the way to go here, as okie mentioned, if you are going to run the pwm fairly fast (10's of KHz) youll need a rail to rail opamp otherwise you wont get a nice square output.

[ EDIT - NOT this
http://imgur.com/zBi11.jpg

Posted 5 years ago #
okie
Administrator

gbulmer, for sure--a comparator would work. I am also doing filtering, and I used one of the opamps in the quad-opamp chip for buffering PWM.

Posted 5 years ago #
okie
Administrator

poslathian, that circuit's (http://imgur.com/zBi11.jpg) output will never change. It's a comparator that compares Maple's output to 9V/2, and Maple's output is always at or below 3.3V. Make the resistor divider give 3.3V/2 and it should work nicely.

Posted 5 years ago #
structuresound
Member

@Okie, yes it's an audio application. Eventually controlling SSM2164, works really well I think. I've thought about looking into those "rail to rail" versions, I'll also order a TLV274.

@ poslathian, I'm running it as fast as possible (or at least without a scalar on my clock). Open drain would be when the pin swings low, correct? Although the output seems perfectly square on my scope, I also don't care that much because I want to end up with DC anyway.

@gbulmer I'll look at how much current is going into out of the pin, I think it's <1mA, at least HOPE so!

Posted 5 years ago #
structuresound
Member

Schematic, works well on breadboard so far.

http://www.structuresound.com/eric/pwmLPF.png

Posted 5 years ago #
poslathian
LeafLabs

@okie - oops, I left out the feedback trace from output to negative input. I was actually going for a unity gain amp re-centered at 4.5V, which I still think I got wrong. I need to just stop it...

Posted 5 years ago #
structuresound
Member

@ poslathian is that possible without AC coupling or a clamp circuit? I'd be curious if one can do that with a single opamp stage!

Posted 5 years ago #
poslathian
LeafLabs

I certainly thought so, I think you can do this:
http://imgur.com/fE2hz.jpg

which also works for non-inverting if you want to go this route. I use this circuit when I want to rebias -1.5V,1.5V audio signals to 0-3.3V ADC signals centered around 1.15V

Posted 5 years ago #
structuresound
Member

@post

looks like what you've drawn basic inverting summing amplifier configuration, I never thought to use that to add dc bias since in normal audio applications you usually want to AVOID that. Good idea, I'll try it out.

Posted 5 years ago #
poslathian
LeafLabs

yea, make sure to subtract off the offset in digital to get everything 0 centered again! (or just adjust all your DSP to account for it....which you probably dont want to do)

Posted 5 years ago #

RSS feed for this topic

Reply

You must log in to post.