High Pitch Whining on Daisy Chain Powering

Started by Baran Ismen, December 13, 2023, 01:45:46 AM

Previous topic - Next topic

Baran Ismen

Hi everyone.

Here's the new deal. I've given a try on some of the pedals I've bought (finally, been gathering for some time) and connected 4 of them and powered them with a single 4-to-1 power cable distributor with a 9.5V 4.5A DC adapter (a solid one from Verifone).

And I've seen that the Mooer X2 should be isolated, otherwise, when connected with other pedals it produces a feint but present high pitch noise. I've put a ferrite core around the cable, no luck. It precisely sounds like this

I've read somewhere that it's kind of common for DSP-powered pedals, and the solution is either using an iso brick, or a separate adapter for those that produce noise or simply to break the ground loop.

Question time;

1- What's this problem called, what's the root cause and how should I google it in detail? Ground loop? High-pitch whining guitar pedal? Girlfriends cry after refusing her dinner-out offer?

2- Is it possible to DIY an isolated power brick?

3- What makes an isolated power brick "isolated"? Separate transformers for each line, or separate filtering caps, or what really? Assuming that the grounds are merged somehow and somewhere, what causes this ground looping in practice?

4- I see some power bricks on Aliex which are as small as a hand palm, how is it possible? They're fed with 5V USB-C type cable even! A common transformer is used and distributed to each channel or they are fake af?

5- Would it be possible to make a DC distribution circuit with different channels, caps, and such (say 10) and use this Verifone adapter to power it (actually distribute its current over a circuit, not via daisy cables)? Connections from each pedal will be individual to this unit anyway.

6- I've seen that VoodooLabs made a 1-2 cable with ground lift. Even though I couldn't understand the logic behind it, it seems logical. I wonder whether it's possible to do the same with more than 2 cables.


Rob Strand

#1
Most generic switch mode power supplies do like to operate under light loads.   It affects the way the switch works and you can get low/audio frequency ripple coming through.

One work around is to add a dummy load.

9V @ 4.5A is way over rated for a group of pedals.  You might need a load of 0.5A before the ripple gets back to normal, 0.1A if you are lucky.   Another issue with excessively rated power supplies is by their very nature of design they tend to produce more noise, in all their evil forms.

To make matters worse.  If the PSU does goes into a low frequency ripple mode.  The chances of it affecting a multiple pedal connection skyrockets since the low frequency ripple ends of pulsing ripple currents through all the grounds.

If you have a well behaved PSU then it's highly likely you can connect multiple pedals without using power supply isolation and without problems.   What you can't do with non-isolated supplies is connect up negative ground and positive ground pedals off the same supply.    One out here is to wire the positive ground pedal with a flipped ground so it becomes negative ground but you can still get issues if the pedal PSU bypassing and filtering is bad.

The isolated pedal are very simple.   Typically something like:

One switch-mode PSU with multiple outputs (not unlike what you see on on PC power supplies).  Each output might produce 12V DC to 15V DC.  Then that DC output is passed through a linear regulator.   LM317 and smaller SMD packages being very common.

If you don't take all precautions you can still get some problems with noise because the isolated PSUs have some coupling through the common switch-mode.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Baran Ismen

Quote from: Rob Strand on December 13, 2023, 03:32:37 AMMost generic switch mode power supplies do like to operate under light loads.   It affects the way the switch works and you can get low/audio frequency ripple coming through.

One work around is to add a dummy load.

9V @ 4.5A is way over rated for a group of pedals.  You might need a load of 0.5A before the ripple gets back to normal, 0.1A if you are lucky.   Another issue with excessively rated power supplies is by their very nature of design they tend to produce more noise, in all their evil forms.

To make matters worse.  If the PSU does goes into a low frequency ripple mode.  The chances of it affecting a multiple pedal connection skyrockets since the low frequency ripple ends of pulsing ripple currents through all the grounds.

If you have a well behaved PSU then it's highly likely you can connect multiple pedals without using power supply isolation and without problems.   What you can't do with non-isolated supplies is connect up negative ground and positive ground pedals off the same supply.    One out here is to wire the positive ground pedal with a flipped ground so it becomes negative ground but you can still get issues if the pedal PSU bypassing and filtering is bad.

The isolated pedal are very simple.   Typically something like:

One switch-mode PSU with multiple outputs (not unlike what you see on on PC power supplies).  Each output might produce 12V DC to 15V DC.  Then that DC output is passed through a linear regulator.   LM317 and smaller SMD packages being very common.

If you don't take all precautions you can still get some problems with noise because the isolated PSUs have some coupling through the common switch-mode.


You mean 4.5A is way too much and unnecessary? I'll have 10 pedals I guess, maybe half of them are digital or dsp based, and they need some extra mA unlike a TS or muff.

FiveseveN

Quote from: Baran Ismen on December 13, 2023, 01:45:46 AMto break the ground loop
It's not a ground loop. Not every source of noise is a ground loop and unless you're Sunn O))) and use 20 amps at once I can almost guarantee that any ground loops you create by daisy chaining won't be a problem. They have to be long and carry high currents to become an issue (e.g. between multiple amps).

Quote from: Baran Ismen on December 13, 2023, 03:42:31 AMthey need some extra mA
Hundreds, not thousands. Take the guesswork out by noting each pedal's current draw from the manual or the enclosure, then add them up. What's the total?

First of all, you should not expect any random PSU to have the adequate level of regulation and filtering for audio applications.
But if it's just the one Mooer pedal that's causing noise it's probably intermodulating (heterodyning) its internal buck converter frequency with the PSU's switching frequency, resulting in a product in the audio band. Powering it separately, ideally with a linear PSU could be the simplest solution. You might also get away with more involved filtering (not just a ferrite bead).
Quote from: R.G. on July 31, 2018, 10:34:30 PMDoes the circuit sound better when oriented to magnetic north under a pyramid?

antonis

Quote from: Baran Ismen on December 13, 2023, 01:45:46 AMIs it possible to DIY an isolated power brick?

Yes.. :icon_wink:

Quote from: Baran Ismen on December 13, 2023, 01:45:46 AMWhat makes an isolated power brick "isolated"?

Seperate transformer secondaries..
(see Link above..)
"I'm getting older while being taught all the time" Solon the Athenian..
"I don't mind  being taught all the time but I do mind a lot getting old" Antonis the Thessalonian..

Rob Strand

Quote from: Baran Ismen on December 13, 2023, 03:42:31 AMYou mean 4.5A is way too much and unnecessary? I'll have 10 pedals I guess, maybe half of them are digital or dsp based, and they need some extra mA unlike a TS or muff.
That could change what I've said.

Some digital pedal themselves put noise back on the supply and that could cause issues for the pedals which are primarily analog (including pedals BBD).

Sometimes is pays to work out what's going on like pull one "angry" pedal at a time to see which pedals are causing trouble.  If you pull a sensitive pedal it can also stop the noise but that one is more a victim than an agitator.  If you have angry pedal then that's a good reason to run it from another power supply.  You can also separate sensitive pedals.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Baran Ismen

#6
Quote from: Rob Strand on December 13, 2023, 04:54:04 AM
Quote from: Baran Ismen on December 13, 2023, 03:42:31 AMYou mean 4.5A is way too much and unnecessary? I'll have 10 pedals I guess, maybe half of them are digital or dsp based, and they need some extra mA unlike a TS or muff.
That could change what I've said.

Some digital pedal themselves put noise back on the supply and that could cause issues for the pedals which are primarily analog (including pedals BBD).

Sometimes is pays to work out what's going on like pull one "angry" pedal at a time to see which pedals are causing trouble.  If you pull a sensitive pedal it can also stop the noise but that one is more a victim than an agitator.  If you have angry pedal then that's a good reason to run it from another power supply.  You can also separate sensitive pedals.


I've got still 2 pedals on the road, I'll calculate once they arrive and let you know then but I don't think it'd exceed 1.5A's in total.

Just took a quick look that the most consuming is the mooer x2, which requires 300mA, there are 3 other small ones that requires around 150mA, rest is analog, like ts9, small stone, tuner etc.

Rob Strand

Quote from: Baran Ismen on December 13, 2023, 05:57:50 AMI've got still 2 pedals on the road, I'll calculate once they arrive and let you know then but I don't think it'd exceed 1.5A's in total.

Just took a quick look that the most consuming is the mooer x2, which requires 300mA, there are 3 other small ones that requires around 150mA, rest is analog, like ts9, small stone, tuner etc.

So you could have enough load to meet the minimum required for a 4.5A supply but if you remove one the high current pedals from the rig then it could cause a whine problem to start-up as the load is no longer enough.    This is the problem with switch modes.   I'm a big fan of using smaller rating supplies and leaving a dummy load on there.

That being said.   Evil noisy pedals need to have their own supply.  They are just going to cause problems.

I think for rigs without digital pedals you don't really need an isolated supply.  However, things have changed, there are plenty of digital pedals out there.  You can't guarantee they won't pollute the power.  If you just want a plug-in and work solution with doing much thinking you can see the appeal with those isolated supplies.  But if you want to do a little work you might get away with two cheap generic supplies.

Switchmode wall-warts running at light loads is always going to be an issue with low drain rigs.  You pretty much need a dummy load to make sure they are always working well.   Easy to do, just one of the multiple outputs into a box with a resistor in it.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Baran Ismen

Quote from: Rob Strand on December 13, 2023, 02:04:21 PM
Quote from: Baran Ismen on December 13, 2023, 05:57:50 AMI've got still 2 pedals on the road, I'll calculate once they arrive and let you know then but I don't think it'd exceed 1.5A's in total.

Just took a quick look that the most consuming is the mooer x2, which requires 300mA, there are 3 other small ones that requires around 150mA, rest is analog, like ts9, small stone, tuner etc.

So you could have enough load to meet the minimum required for a 4.5A supply but if you remove one the high current pedals from the rig then it could cause a whine problem to start-up as the load is no longer enough.    This is the problem with switch modes.   I'm a big fan of using smaller rating supplies and leaving a dummy load on there.

That being said.   Evil noisy pedals need to have their own supply.  They are just going to cause problems.

I think for rigs without digital pedals you don't really need an isolated supply.  However, things have changed, there are plenty of digital pedals out there.  You can't guarantee they won't pollute the power.  If you just want a plug-in and work solution with doing much thinking you can see the appeal with those isolated supplies.  But if you want to do a little work you might get away with two cheap generic supplies.

Switchmode wall-warts running at light loads is always going to be an issue with low drain rigs.  You pretty much need a dummy load to make sure they are always working well.   Easy to do, just one of the multiple outputs into a box with a resistor in it.


Wait, you're saying that I should have a bare minimum load for an adapter otherwise it will squeal like a pig? Now that's interesting! I've always thought keeping the current limit well above of the requirement would be harmless, but never thought there also should be a minimum load.

I've got another 9V adapter which is rated for 2.5A I suppose, I'll try with that as well.

Rob Strand

Quote from: Baran Ismen on December 14, 2023, 12:03:20 AMWait, you're saying that I should have a bare minimum load for an adapter otherwise it will squeal like a pig? Now that's interesting! I've always thought keeping the current limit well above of the requirement would be harmless, but never thought there also should be a minimum load.

Yes, switch-modes can do weird stuff with light loads.    They work with light loads but they are not always quiet - the ripple can drop into the audio band.   Plenty of examples on this forum and many issues solved with dummy load - 220 ohm resistor for 1A to 2A 9V supplies. 

The old transformer wall-warts never did such a thing - whole different way of working.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Baran Ismen

Quote from: Rob Strand on December 14, 2023, 12:07:48 AM
Quote from: Baran Ismen on December 14, 2023, 12:03:20 AMWait, you're saying that I should have a bare minimum load for an adapter otherwise it will squeal like a pig? Now that's interesting! I've always thought keeping the current limit well above of the requirement would be harmless, but never thought there also should be a minimum load.

Yes, switch-modes can do weird stuff with light loads.    They work with light loads but they are not always quiet - the ripple can drop into the audio band.   Plenty of examples on this forum and many issues solved with dummy load - 220 ohm resistor for 1A to 2A 9V supplies. 

The old transformer wall-warts never did such a thing - whole different way of working.


So I should short one of the empty power leads with a 220ohm resistor between + and - tips, then?

Is there any estimation or calculation for finding a good spot for a power supply mA rating by the way? Say, my consumption is 1.2A and my adapter should be minimum X and maximum Y rated..

Rob Strand

Quote from: Baran Ismen on December 14, 2023, 12:32:59 AMSo I should short one of the empty power leads with a 220ohm resistor between + and - tips, then?
Yes that type of thing.   

QuoteIs there any estimation or calculation for finding a good spot for a power supply mA rating by the way?

Say, my consumption is 1.2A and my adapter should be minimum X and maximum Y rated..
Not really.   A crude rule of thumb is to have a load at least 1/10th the rated current.  That ends up with quite a lot of power dissipated.   Many power supplies are quiet with 1/50 to 1/20 the rated load.   There *are* power supplies which are quiet with no load.   Sometimes that's because the power supply has a dummy load built in, recent rules put an end to this solution.  Other times the switch-mode is a type which can run down to low current without a problem.

Switchmodes are complicated, and not all the same, so very difficult to make predictions and generalize.   If you know the possible problems you can try some stuff to work around those issues.
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Baran Ismen

Quote from: Rob Strand on December 14, 2023, 12:41:45 AM
Quote from: Baran Ismen on December 14, 2023, 12:32:59 AMSo I should short one of the empty power leads with a 220ohm resistor between + and - tips, then?
Yes that type of thing.   

QuoteIs there any estimation or calculation for finding a good spot for a power supply mA rating by the way?

Say, my consumption is 1.2A and my adapter should be minimum X and maximum Y rated..
Not really.   A crude rule of thumb is to have a load at least 1/10th the rated current.  That ends up with quite a lot of power dissipated.   Many power supplies are quiet with 1/50 to 1/20 the rated load.   There *are* power supplies which are quiet with no load.   Sometimes that's because the power supply has a dummy load built in, recent rules put an end to this solution.  Other times the switch-mode is a type which can run down to low current without a problem.

Switchmodes are complicated, and not all the same, so very difficult to make predictions and generalize.   If you know the possible problems you can try some stuff to work around those issues.


So this clearly explains the reason why I'm having a squeal when I power just a single pedal and when added more, the noise is decreased.

Thank you for detailed information, Rob, helped greatly! I think I'll give a regular type of adapter.

I think by wallwart, or linear, you mean this type,

What I have now is this I assume this one is SMPS, right?

Rob Strand

#13
Quote from: Baran Ismen on December 14, 2023, 12:56:27 AMI think by wallwart, or linear, you mean this type,

What I have now is this I assume this one is SMPS, right?

Yes, that it.    The old "linears" have conventional iron-core transformers inside.  They are a lot heavier for the rating.   The ratings of linear types tend to be low for the size/weight compared to switchers.  Common ratings 200mA, 300mA, 500mA, 1A.

The flat one you have is a switcher.

You can also get wall-wart looking switchmodes.   Basically the same as phone chargers.  They weigh nothing and have ratings 1A or 2A, maybe upto 3.5A or 4A for larger ones.

https://www.auselectronicsdirect.com.au/9v-dc-1a-power-adapter-with-reversible-2.1-dc-plug?gclid=EAIaIQobChMIqIDehKOOgwMVqKRmAh2M6Q7QEAQYAyABEgLTkPD_BwE
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.

Baran Ismen

Quote from: Rob Strand on December 14, 2023, 01:16:31 AM
Quote from: Baran Ismen on December 14, 2023, 12:56:27 AMI think by wallwart, or linear, you mean this type,

What I have now is this I assume this one is SMPS, right?

Yes, that it.    The old "linears" have conventional iron-core transformers inside.  They are a lot heavier for the rating.   The ratings of linear types tend to be low for the size/weight compared to switchers.  Common ratings 200mA, 300mA, 500mA, 1A.

The flat one you have is a switcher.

You can also get wall-wart looking switchmodes.   Basically the same as phone chargers.  They weigh nothing and have ratings 1A or 2A, maybe upto 3.5A or 4A for larger ones.

https://www.auselectronicsdirect.com.au/9v-dc-1a-power-adapter-with-reversible-2.1-dc-plug?gclid=EAIaIQobChMIqIDehKOOgwMVqKRmAh2M6Q7QEAQYAyABEgLTkPD_BwE


Being a wall-wart makes a SMPS more reliable, then?

Rob Strand

Quote from: Baran Ismen on December 14, 2023, 02:57:33 AMBeing a wall-wart makes a SMPS more reliable, then?
The leaded ones like you have tend to be more reliable than the wall-wart style.    However they do tend to have higher current ratings.   

Reliability is strongly dependent on the brand and running them below maximum current.    The brand of electrolytic capacitors and temperature have a big impact on reliability.   You can get 1 year life to 10 years life!
Send:     . .- .-. - .... / - --- / --. --- .-. -
According to the water analogy of electricity, transistor leakage is caused by holes.