High amperage 12v power supply but Ammeter problem

[dclevy1]

So I need help because I'm stumped I had a box where I would deliver 12 volts to a subwoofer amp for a project and the box was great but I took it apart and made a new form factor for easy installation in a media Shelf . Upon finishing the box and getting everything wired and testing it the ammeter is now acting up which was working perfectly fine before but now the amperage jumps around and is negative and I have no idea why it's doing this. There's absolutely nothing conneced to it.


Sent from my iPhone using Tapatalk

[Kalvin]

Swap the sense polarity?

[dclevy1]

Swap the sense polarity?

You're talking about the thin wires that hook up to the shunt


Sent from my iPhone using Tapatalk

[Twoflower]

Did you really fixed the 12V high current wires with a zip-tie to the shunt? That's a very bad idea. Especially in a wooden box. Use the right material or you might start a fire! And the soldering in the first picture won't do either!

Edit:
The longer I look to the pictures the worse it gets.
1. Are the PSUs rated to run in parallel with out a resistor to prevent unwanted currents because the PSU modules are not 100% identical?
2. The color-coating on the shunt might changed the value of the shunt.
3. The uninsulated crimps on the high voltage side don't look very trustworthy
4. I'm not sure if it is a good idea to use uninsulated high voltage parts to a silver coated surface (hint these colours might contain metal flakes)? And yes I saw the useless plastic washer (the screw renders it useless) and the distance of the contacts to the surface is to small.

[dclevy1]

Did you really fixed the 12V high current wires with a zip-tie to the shunt? That's a very bad idea. Especially in a wooden box. Use the right material or you might start a fire! And the soldering in the first picture won't do either!

Edit:
The longer I look to the pictures the worse it gets.
1. Are the PSUs rated to run in parallel with out a resistor to prevent unwanted currents because the PSU modules are not 100% identical?
2. The color-coating on the shunt might changed the value of the shunt.
3. The uninsulated crimps on the high voltage side don't look very trustworthy
4. I'm not sure if it is a good idea to use uninsulated high voltage parts to a silver coated surface (hint these colours might contain metal flakes)? And yes I saw the useless plastic washer (the screw renders it useless) and the distance of the contacts to the surface is to small.

1. I don't need criticism on the project. I need help with the issue

2. The PSUs are 100% identical

3.Just because you do like the look of something doesn't mean it's not functional.

4. The only thing that's acting as a plastic washer is for the output that's on the outside of the box.

5. Again I need help with the issue not a negative  OCD comment




Sent from my iPhone using Tapatalk

[jaycee]

Silver paint often has metal particles in it. You have most likely ruined that shunt resistor by spraying it, not to mention you will have created leakage currents all over the place.

As an aside, no PSU is 100% identical on its output. Unless the datasheet for the PSU specifically says you can parallel them without load balancing resistors, you can't.

[Someone]

5. Again I need help with the issue not a negative  OCD comment

You need help with lots of things, including the common mode range of the ampere meter. The problem is clear to those of us who know what we are doing but when this is obviously going to be accessible to non technical people and operated for long periods with little supervision you need to address the significant safety issues before proceeding.

[Someone]

For posterity when the OP takes the images down, this is the connection method to the current shunt. Cant be sure what the attachment method is and if thats solder or something else like conductive epoxy.

[jaycee]

For posterity when the OP takes the images down, this is the connection method to the current shunt. Cant be sure what the attachment method is and if thats solder or something else like conductive epoxy.

It looks like solder to me, but obviously the soldering iron was unable to put enough heat into the terminal of the shunt to solder correctly. The correct way to make the connection wouldve been to use ring/eyelet terminals, crimped and soldered to the wire, and then bolted to the shunt. What is pictured is insufficient, will have considerable contact resistance at high current, will get hot, and will probably start a fire.

[dclevy1]

Silver paint often has metal particles in it. You have most likely ruined that shunt resistor by spraying it, not to mention you will have created leakage currents all over the place.

As an aside, no PSU is 100% identical on its output. Unless the datasheet for the PSU specifically says you can parallel them without load balancing resistors, you can't.

So you think the shunt is ruined by the paint. Load balancing resistors can you tell me more? I am a novice obviously with electrical wiring and stuff but I do know something. And can understand very quickly


Sent from my iPhone using Tapatalk

[dclevy1]

5. Again I need help with the issue not a negative  OCD comment

You need help with lots of things, including the common mode range of the ampere meter. The problem is clear to those of us who know what we are doing but when this is obviously going to be accessible to non technical people and operated for long periods with little supervision you need to address the significant safety issues before proceeding.

Oh yeah I totally understand but the comment was not helping at all that he posted. Sounded like a personal preference comment


Sent from my iPhone using Tapatalk

[dclevy1]

So so far when I picked up is I need to change the connection method to the shunt. As far as the main negative wire and actually mount the wires properly to it


Sent from my iPhone using Tapatalk

[CJay]

Did you really fixed the 12V high current wires with a zip-tie to the shunt? That's a very bad idea. Especially in a wooden box. Use the right material or you might start a fire! And the soldering in the first picture won't do either!

Edit:
The longer I look to the pictures the worse it gets.
1. Are the PSUs rated to run in parallel with out a resistor to prevent unwanted currents because the PSU modules are not 100% identical?
2. The color-coating on the shunt might changed the value of the shunt.
3. The uninsulated crimps on the high voltage side don't look very trustworthy
4. I'm not sure if it is a good idea to use uninsulated high voltage parts to a silver coated surface (hint these colours might contain metal flakes)? And yes I saw the useless plastic washer (the screw renders it useless) and the distance of the contacts to the surface is to small.

1. I don't need criticism on the project. I need help with the issue

2. The PSUs are 100% identical

3.Just because you do like the look of something doesn't mean it's not functional.

4. The only thing that's acting as a plastic washer is for the output that's on the outside of the box.

5. Again I need help with the issue not a negative  OCD comment




Sent from my iPhone using Tapatalk

Regardless of if you like what twoflower said or not, they're all valid points.

That wiring and box is a mess.

Your power supplies look like they're (HP?) server PSUs that have been stripped out of their cases, they may be OK to parallel, difficult to know unless we know where they're from though.

If they are server power supplies, why did you strip them out of their cases? Those cases are ideal for EMC shielding (which is one possible cause of your problems) and prevent 'random' stuff falling into the PSUs which makes them a lot safer, plus they contain the shrapnel when they go bang (which they can do, sometimes quite spectacularly)

To parallel power supplies they have to have been designed to be paralleled so unless that's explicitly specified in the data for those supplies then you have got potential problems (pun intended), it might even be the cause of your ammeter jumping around as they both try to set their output voltage to what they think is correct, try disconnecting one and see how the ammeter behaves then.

Was your ammeter specified to work with that shunt? How are you powering the ammeter?

Can you try moving the shunt sense wires (the thin ones) away from other load carrying wires, including mains wires?
You're probably dealing with millivolt signal levels and if they're close to wires carrying current then it's possible that's causing problems with 'pickup'

Load balancing resistors are wasteful but useful, you add a low value resistor in series with the output of each PSU, the resistor 'soaks up' and wastes the voltage difference between the outputs as heat but that poses another problem, you have to get rid of the heat somehow.

Spraying everything silver is a bad idea, makes it easy to make a mistake when you're fault finding and at best lead yourself up the wrong path, at worst you end up with a pile of scrap components, as others have pointed out, metallic paints often have metallic flakes in them, bit of a giveaway in the name really.

Uninsulated mains connections are a *really* bad idea, it's a moment's work to short them out with a loose screw, screwdriver, small rodent, your finger or worse, someone else's fingers, insulate them properly and not with the black snot you've used for the 12V outputs.

The current shunt *really* needs to have crimped and bolted connections, again others have said that but again, millivolt signals and you've got *really* poor connections on a high current component, at best it won't work properly, at worst it'll arc and burn up the wiring.

[Kalvin]

Swap the sense polarity?

You're talking about the thin wires that hook up to the shunt

Initially yes, but it looks like you may have some other problems with the construction as others have pointed out.

[dclevy1]

Did you really fixed the 12V high current wires with a zip-tie to the shunt? That's a very bad idea. Especially in a wooden box. Use the right material or you might start a fire! And the soldering in the first picture won't do either!

Edit:
The longer I look to the pictures the worse it gets.
1. Are the PSUs rated to run in parallel with out a resistor to prevent unwanted currents because the PSU modules are not 100% identical?
2. The color-coating on the shunt might changed the value of the shunt.
3. The uninsulated crimps on the high voltage side don't look very trustworthy
4. I'm not sure if it is a good idea to use uninsulated high voltage parts to a silver coated surface (hint these colours might contain metal flakes)? And yes I saw the useless plastic washer (the screw renders it useless) and the distance of the contacts to the surface is to small.

1. I don't need criticism on the project. I need help with the issue

2. The PSUs are 100% identical

3.Just because you do like the look of something doesn't mean it's not functional.

4. The only thing that's acting as a plastic washer is for the output that's on the outside of the box.

5. Again I need help with the issue not a negative  OCD comment




Sent from my iPhone using Tapatalk

Regardless of if you like what twoflower said or not, they're all valid points.

That wiring and box is a mess.

Your power supplies look like they're (HP?) server PSUs that have been stripped out of their cases, they may be OK to parallel, difficult to know unless we know where they're from though.

If they are server power supplies, why did you strip them out of their cases? Those cases are ideal for EMC shielding (which is one possible cause of your problems) and prevent 'random' stuff falling into the PSUs which makes them a lot safer, plus they contain the shrapnel when they go bang (which they can do, sometimes quite spectacularly)

To parallel power supplies they have to have been designed to be paralleled so unless that's explicitly specified in the data for those supplies then you have got potential problems (pun intended), it might even be the cause of your ammeter jumping around as they both try to set their output voltage to what they think is correct, try disconnecting one and see how the ammeter behaves then.

Was your ammeter specified to work with that shunt? How are you powering the ammeter?

Can you try moving the shunt sense wires (the thin ones) away from other load carrying wires, including mains wires?
You're probably dealing with millivolt signal levels and if they're close to wires carrying current then it's possible that's causing problems with 'pickup'

Load balancing resistors are wasteful but useful, you add a low value resistor in series with the output of each PSU, the resistor 'soaks up' and wastes the voltage difference between the outputs as heat but that poses another problem, you have to get rid of the heat somehow.

Spraying everything silver is a bad idea, makes it easy to make a mistake when you're fault finding and at best lead yourself up the wrong path, at worst you end up with a pile of scrap components, as others have pointed out, metallic paints often have metallic flakes in them, bit of a giveaway in the name really.

Uninsulated mains connections are a *really* bad idea, it's a moment's work to short them out with a loose screw, screwdriver, small rodent, your finger or worse, someone else's fingers, insulate them properly and not with the black snot you've used for the 12V outputs.

The current shunt *really* needs to have crimped and bolted connections, again others have said that but again, millivolt signals and you've got *really* poor connections on a high current component, at best it won't work properly, at worst it'll arc and burn up the wiring.

Switching sense wire didn't work

The paint is non conductive

Removing one or the other 12v sources didn't work.

And I'll work on getting the shunt properly terminated

The shunt is rated to work with the ammeter. I have it powered through a buck converter changing 12v to 5v to the meter


Sent from my iPhone using Tapatalk

[Cyberdragon]

Is there any adjustment trimmer on the meter? If not then it's either getting noise in the 5v supply or the meter just plain broke for some reason. If you have a sensetive DMM you could check if there actually is any leakage across the shunt.

EDIT: Wait a minute, is that a 4 wire volt meter? It's probably leaking current back into itself, thus causing reverse biasing of the shunt. Check the voltage across the shunt and across the leads of the volmeter that go across the shunt (disconnect them first), it should be 0. If it's not, the voltmeter might need recalibrating or replacing.

[dclevy1]

Did you really fixed the 12V high current wires with a zip-tie to the shunt? That's a very bad idea. Especially in a wooden box. Use the right material or you might start a fire! And the soldering in the first picture won't do either!

Edit:
The longer I look to the pictures the worse it gets.
1. Are the PSUs rated to run in parallel with out a resistor to prevent unwanted currents because the PSU modules are not 100% identical?
2. The color-coating on the shunt might changed the value of the shunt.
3. The uninsulated crimps on the high voltage side don't look very trustworthy
4. I'm not sure if it is a good idea to use uninsulated high voltage parts to a silver coated surface (hint these colours might contain metal flakes)? And yes I saw the useless plastic washer (the screw renders it useless) and the distance of the contacts to the surface is to small.

1. I don't need criticism on the project. I need help with the issue

2. The PSUs are 100% identical

3.Just because you do like the look of something doesn't mean it's not functional.

4. The only thing that's acting as a plastic washer is for the output that's on the outside of the box.

5. Again I need help with the issue not a negative  OCD comment




Sent from my iPhone using Tapatalk

Regardless of if you like what twoflower said or not, they're all valid points.

That wiring and box is a mess.

Your power supplies look like they're (HP?) server PSUs that have been stripped out of their cases, they may be OK to parallel, difficult to know unless we know where they're from though.

If they are server power supplies, why did you strip them out of their cases? Those cases are ideal for EMC shielding (which is one possible cause of your problems) and prevent 'random' stuff falling into the PSUs which makes them a lot safer, plus they contain the shrapnel when they go bang (which they can do, sometimes quite spectacularly)

To parallel power supplies they have to have been designed to be paralleled so unless that's explicitly specified in the data for those supplies then you have got potential problems (pun intended), it might even be the cause of your ammeter jumping around as they both try to set their output voltage to what they think is correct, try disconnecting one and see how the ammeter behaves then.

Was your ammeter specified to work with that shunt? How are you powering the ammeter?

Can you try moving the shunt sense wires (the thin ones) away from other load carrying wires, including mains wires?
You're probably dealing with millivolt signal levels and if they're close to wires carrying current then it's possible that's causing problems with 'pickup'

Load balancing resistors are wasteful but useful, you add a low value resistor in series with the output of each PSU, the resistor 'soaks up' and wastes the voltage difference between the outputs as heat but that poses another problem, you have to get rid of the heat somehow.

Spraying everything silver is a bad idea, makes it easy to make a mistake when you're fault finding and at best lead yourself up the wrong path, at worst you end up with a pile of scrap components, as others have pointed out, metallic paints often have metallic flakes in them, bit of a giveaway in the name really.

Uninsulated mains connections are a *really* bad idea, it's a moment's work to short them out with a loose screw, screwdriver, small rodent, your finger or worse, someone else's fingers, insulate them properly and not with the black snot you've used for the 12V outputs.

The current shunt *really* needs to have crimped and bolted connections, again others have said that but again, millivolt signals and you've got *really* poor connections on a high current component, at best it won't work properly, at worst it'll arc and burn up the wiring.

So I got the shunt properly terminated with large ring terminals as well as I stripped all the paint from the shunt. I also move some wires around from touching especially the sense wires for the ammeter.Even though this did not fix the issue I did actually find the cause of the issue it was a fact that I was not running the ammeter on a separate power supply I was running it off of the server power supply voltage. This alone was making the a meter jump around and read amperage from whatever device. If y'all could explain the reason why that would be awesome.


Sent from my iPhone using Tapatalk

[dclevy1]

Is there any adjustment trimmer on the meter? If not then it's either getting noise in the 5v supply or the meter just plain broke for some reason. If you have a sensetive DMM you could check if there actually is any leakage across the shunt.

EDIT: Wait a minute, is that a 4 wire volt meter? It's probably leaking current back into itself, thus causing reverse biasing of the shunt. Check the voltage across the shunt and across the leads of the volmeter that go across the shunt (disconnect them first), it should be 0. If it's not, the voltmeter might need recalibrating or replacing.

I actually found the issue read my previous comment


Sent from my iPhone using Tapatalk

[floobydust]

I think the problem is too much common-mode EMI between the ammeter power supply and the main 12V SMPS.
The shunt floats compared to the ammeter's power. You could prove this by powering the ammeter from a battery instead.
I can't see if the ammeter power and shunt need to be referenced together.

You could tie one leg of the shunt (-) (psu end, not load end) to the ammeter power DC(-), or try add 22nF capacitor across there.

[dclevy1]

I think the problem is too much common-mode EMI between the ammeter power supply and the main 12V SMPS.
The shunt floats compared to the ammeter's power. You could prove this by powering the ammeter from a battery instead.
I can't see if the ammeter power and shunt need to be referenced together.

You could tie one leg of the shunt (-) (psu end, not load end) to the ammeter power DC(-), or try add 22nF capacitor across there.

I see I will try this as well and I don't understand a lot of lingo and abbreviations you guys use. I am a beginner with this stuff as you can see. As well as I wanted to know what can I use to make sure the a meter is showing closest to accurate readings because I think I messed it up when I was screwing around with the potentiometer calibration thing in the back of the unit.


Sent from my iPhone using Tapatalk