Testing Salvaged Motors

[darren adcock]

OK after more tinkering and reading (definitely don't understand it all) but managed to get rid off all high frequencies being audible through my sound system (whether this would be different with a louder sound system i'm unsure), only thing that is left is a low chopping type sound. Here is the schematic of what I have done, maybe I've made this issue worse, maybe I've added too many components or some are redundant, but tried to follow decoupling advice in to air on the side of caution, the decoupling on the op amp made a really sudden difference. Plus i can now have the ground part of the patch cable attached to the circuit without the screeching sounds, so i'm getting somewhere. View attachment 42063

 

[hevans1944]

It sounds like (literally) that you have discovered the wonderful, mysterious, tear-your-hair-out, world of ground loops. Entire volumes have been written and published on how ground loops are formed and methods to defeat them. As you have discovered, sometimes removing a "ground" connection actually helps. Sometimes it doesn't.

The problem is this: the little "ground" symbol is supposed to represent an equipotential surface. There should be NO potential difference between any two points labeled as "ground," but in the real world there always is. It may be anything from barely measurable picovolts to several hundred volts, depending on the circumstances. For audio ground loops, the voltage levels are generally in the microvolt to millivolt range.

There are many types and many causes of ground loops, but the most often encountered are "common mode" ground loops. A pretty good article that discusses ground loops, and how to prevent them, can be read here.

Often the problem arises when heavy-current equipment, such as a motor, shares conductors, especially "ground return" conductors with sensitive electronics. Separating the two "grounding" paths will help to eliminate this common mode noise source. A so-called "star grounding" system is also mostly effective when properly implemented. Do some reading on this for more ideas, but trial-and-error is also an effective troubleshooting technique when pursued logically and systematically. Bypassing the power supply leads at, or as near as possible to, your electronic components is always a good idea.

And, yes, moving the circuitry from a solderless breadboard to a strip-board or even a full-fledged printed circuit board can often do wonders to reduce noise. At the very least it will look tidy.

 

[darren adcock]

It sounds like (literally) that you have discovered the wonderful, mysterious, tear-your-hair-out, world of ground loops. Entire volumes have been written and published on how ground loops are formed and methods to defeat them. As you have discovered, sometimes removing a "ground" connection actually helps. Sometimes it doesn't.

The problem is this: the little "ground" symbol is supposed to represent an equipotential surface. There should be NO potential difference between any two points labeled as "ground," but in the real world there always is. It may be anything from barely measurable picovolts to several hundred volts, depending on the circumstances. For audio ground loops, the voltage levels are generally in the microvolt to millivolt range.

There are many types and many causes of ground loops, but the most often encountered are "common mode" ground loops. A pretty good article that discusses ground loops, and how to prevent them, can be read here.

Often the problem arises when heavy-current equipment, such as a motor, shares conductors, especially "ground return" conductors with sensitive electronics. Separating the two "grounding" paths will help to eliminate this common mode noise source. A so-called "star grounding" system is also mostly effective when properly implemented. Do some reading on this for more ideas, but trial-and-error is also an effective troubleshooting technique when pursued logically and systematically. Bypassing the power supply leads at, or as near as possible to, your electronic components is always a good idea.

And, yes, moving the circuitry from a solderless breadboard to a strip-board or even a full-fledged printed circuit board can often do wonders to reduce noise. At the very least it will look tidy.

Cheers Hop, Identifying the problem is a huge help. Yeah, yesterday was a long day of which mostly felt like I was chasing my own tail, certainly frustrating but good to learn on the fly.

First thing I'll do today is make a block diagram of how the system is connected now and how I intend on it being connected eventually.

Then re-read the article and identify where problems may arrise

I enjoyed the article you have shared and will delve into it further when at my workshop, I especially like it's candid mythbusting. I've done a lot of sound engineering for gigs over the years and couldn't begin to say how many times I've had a guitarist tell me all kinds of woo woo about guitar cables and demand i give them all kinds of cable types, ground lift thingamajigs....

cheers again

Darren

 

[darren adcock]

Hi, so completely baffled by grounding, or maybe I've just read too much and am a bit overwhelmed by information. Anyhow, I endeavored to draw a block diagram of how I have it connected now and how Intend on having connected eventually when built. Hope the diagrams make sense, MS paint an all

Main things i'm confused by;
So i read that I should keep signal ground and power ground seperate, however I read also that these eventually have to connect. So my first though was to find a way of isolating them, but then how do they eventually connect? My second thought was to try and make sure they have as short a path to ground as possible, but this seemed counter intuitive to my first thought (these thoughts are based on what I've read today)

In the second include block diagram I can't see the difference from how I have it set up now apart from the ground would have a shorter return path.

Even though I may have eliminated the problem by removing signal ground to the VCPWM motor driver has the problem actually gone away or will it come back to haunt me, I fully expect these grounding issues can become nightmarish! unless dealt with properly t begin with

anyway, gonna have a break then gonna continue reading for an hour or so, but here's the block diagrams so people can see what I've done and am trying to do, maybe the grounding problem is glaringly obvious

 

[darren adcock]

Or the more i think about it, the cv signals don't need the ground to go with them? I remember looking into if i could have built my modular system with Banana plugs and sockets and it was doable, older modular's had banana plugs and sockets. So maybe these signal ground paths are unnecessary?

 

[darren adcock]

As promised here's a small video of the prototype module and synced motor, just tied a zip tie to the shaft for demonstration here. When I have the full system and built mechanical sound sculptures made I'll share that. But anyhow it works, cheers for helping, I'm not foreseeing many problems, but i'm sure they'll appear as I expand on this.

 

[hevans1944]

Congratulations, Darren, on a job well done!

So maybe these signal ground paths are unnecessary?

No, signals are always represented by the presence of a voltage (or a current) between two points in space. Often one of these two points is called "ground" or "signal common" or "analog common" or some such name and is shared with multiple signal sources via a common, low impedance, connection such as a "chassis ground" or even "Earth ground". The key concept here is a common reference point for the signal or signals, however that is accomplished.

The alternative concept is a "floating" differential signal source, which has a lot to recommend it in terms of the ability to suppress "common mode" noise while allowing the differential signal to be processed. IIRC, differential mode signals were quite common for microphone inputs, leading to the development of the ubiquitous "XLR" 3-terminal quick-disconnect connector:

 

[darren adcock]

Thanks Hop