SMPS considerations for guitar pedal effects

[Potato Cakes]

Hello, everyone,

Some time ago I shared my struggles with using switching power supplies with some pro audio circuits I was building and how in the end it was a capacitor multiplier circuit that was the answer to removing the high levels of noise and oscillation from the audio path. So now I set out to build a couple of specific 9V and 12V power supplies to use with some guitar effects pedals that I'm using as studio effects since they sound really good and I can control them either via a USB editor or MIDI. Using the same filtering, I get a constant tone with most of them. I did a comparison with an old Danelectro branded wall wart and that seems to take care of the noise, but that is only for one pedal and that particular power supply does not have enough draw for some of the pedals being used. My question is what new thing am I now missing when filtering and SMPS to use with unbalanced audio gear like effects pedals?

Thanks!

Paul

 

[Audio1Man]

Hi Paul

Many of the SMPS need a minimum load or they PULSE OFF & ON. Most SMPS are never clean enough for AUDIO USE. Make sure to add a fixed load and filtering for your needs.

Duke

 

[Potato Cakes]

I think you missed the part regarding the use of a capacitance multiplier in the first post. The SMPS is being properly being loaded. There is something else in addition to the filtering that I'm missing. The filtering being used has proved very successful with balanced pro audio gear using various SMPS units. I just don't know what I'm missing when trying to use this power scheme with guitar effect pedals.

Thanks!

Paul

 

[john12ax7]

Can you post a schematic for your capacitance multiplier?

One thing to keep in mind is often guitar pedals will have worse PSRR than pro audio gear,, so you might need things to be cleaner coming in. But I would think if you start with a decent SMPS (they can vary a lot) and use a min load and RC or LC filter it should work fine.

 

[Potato Cakes]

Attached is the schematic. This was made with a lot of help from @thor.zmt. I had some very small boards of this made to use in future projects. This combined with a SMPS that is not clean enough for audio use yielded an exceptionally low noise floor in a full 1073 type project. Without this board the noise and oscillations are ridiculously high and completely unusable in pro audio circuits. So I used this same circuit with a small 9V SMPS but I got varying levels of oscillation/hum depending on the type of pedal. So now I'm back to scratching my head.

Thanks!

Paul

 

[ccaudle]

used this same circuit with a small 9V SMPS

At some point you will have to get an oscilloscope so you can tell what is happening.
The small 9V SMPS would likely have a different behavior than whatever SMPS you used for the bigger projects.
If you post a link to the SMPS datasheet and where/how you added the load resistors we can take a look and see if anything jumps out you may have missed.

 

[Potato Cakes]

This is the power supply used:

https://www.mouser.com/ProductDetail/418-CFM12S090-T

 

[Tubetec]

In the modern age many pedals require vastly more current than the typical 9v at a few 10s of mA like the old days .
Multiple Smps sharing a funky loopy ground circuit tend to cause horrendous noise ,
it can range from just about usable in a live situation to total and utter garbage .
Adding a new pedal into the mix can change the situation entirely , and even with true bypass of the electronics the garbage still remains there .
Guitarists can be quite hard headed about their sound , they generally wont have much of an understanding or even care about the noise , when you as the sound engineer flag up an issue you will frequently be met with a F**k you very much , and keep your hands off my set up . Of course you can try and mitigate with noise gates and filters or some kind of post process denoise algo but thats attacking problem arseways and not addressing the root cause . Its no reflection on their musicianship ,but they can take it very personally when all your trying to do is get them the best sound you can . Frequently the most tallented can be the hardest to deal with ,as an engineer or producer its very easy to end up in conflict and the show starts to fall appart .
The technicallities are the easy part , juggling personalities and fragile eggshell egos is by far the more challenging part of the job .

 

[Potato Cakes]

Fortunately these power supplies are for me. I only care if I get them to be quiet.

 

[Tubetec]

My experience tells me multiple SMPS's in guitar pedal setups are especially problematic .
If a pedal isnt being used on a particular take , it has no reason to be in the circuit , that will certainly change the sound the guitarist expects to hear , that can easily have them throwing hissy fits .

 

[Tubetec]

When Kiefy Richards tried to tell Chuck Berry his sound wasnt hitting the mic right , kaboom the session collapsed cause Chuck was being a hard hat about it .

 

[Potato Cakes]

I may try these next:

https://www.mouser.com/ProductDetail/418-CFM36SB120-T
One of things that was pointed out during my searches was that the use of a Class II transformer was very beneficial in this application. The Danelectro wall wart that had no noise used a Class II and the one other wall wart that created noise did not.

Thanks!

Paul

 

[thor.zmt]

So I used this same circuit with a small 9V SMPS but I got varying levels of oscillation/hum depending on the type of pedal. So now I'm back to scratching my head.

Missing earth.

First, fit a differential mode filter using "Y" capacitors (2,200pF or larger "Y" rated, film preferred) between L/N & PEN (Mains Earth). Only use formally "Y" rated capacitors with UL please.

Second, fit a 100 Ohm resistor in parallel with 100nF and a anti-parallel pair of 1N4001 or similar diodes between +9V (ground) and PEN (Mains Earth).

Note, this earth connection is NOT a safety earth and should be substituted a mains earth connection in the system.

Thor

 

[ccaudle]

Missing earth.

The data sheet indicates it is rated Class II, i.e. no protective earth connection needed (at least for safety reasons).

First, fit a differential mode filter using "Y" capacitors

What about using a pre-built module? For DIY seems a lot easier. You spend about US$5 but saves the hassle.
Something like this for differential filtering:
Schurter input power filter module

There are a few things missing from the original post that might help diagnose what is happening.

get a constant tone

Frequency content of the tone would be useful (including spectrum). That is something you can get from most DAW software, you don't have to have an oscilloscope or dedicated spectrum analyzer. Fundamental frequency, waveshape and/or harmonic content can give some ideas about where the noise is originating.

One of things that was pointed out during my searches was that the use of a Class II transformer was very beneficial in this application.

Primarily for reducing 60Hz noise, but you have not indicated whether your noise problem is 60Hz or not.
The power supply you indicated you are using is rated Class II and only has line and neutral input connections, so I am not sure why you pointed that out in the context of using a different power supply. You also indicated 9V in the original post, but that alternate supply does not have a 9V model available.

The SMPS is being properly being loaded.

It would be easier to help troubleshoot with actual component values and a schematic.
How did you determine what "properly loaded" is for this particular supply? The datasheet is pretty vague. The "typical" switching frequency is listed as 65kHz, but only at full power (i.e. 36W). No indication what happens to the switching frequency at different loads.
Have you tried changing the static load to see if the interference frequency changes?

 

[thor.zmt]

The data sheet indicates it is rated Class II, i.e. no protective earth connection needed (at least for safety reasons).

That is strictly for electrical safety. Not for audio quality and freedom from hum.

I suggest reviewing this:

iFi-audio - TechNote "Why does my system hum?"

And this:

FAQ "Groundhog"

I hope this is clear enough, it is intentionally not very technical.

* for disclosure, until 2019 BC (Before Covid) I was head of R&D at ifi, almost all of their products are my designs and most technical notes were written by me and then dumbed down by S&M (Sales & Marketing, not the S&M you are thinking). All their "Accessory/Fixer products" work and are based on "sound" electronic design (pun intended), however they are severely overpriced, based on what amounts to stolen (from me) Intellectual Property and I do not recommend purchasing them. They can be in most cases easily be replaced with DIY versions.

What about using a pre-built module? For DIY seems a lot easier. You spend about US$5 but saves the hassle.
Something like this for differential filtering

Sure, to me I'd just pull some capacitors from my part shelf. Using a canned module is a great solution, especially if you mount everything together in a nice case.

There are a few things missing from the original post that might help diagnose what is happening.

From my experience, no. And trust me, I have a lot of experience, as my old company sold low noise SMPS including aimed at pedals.

These not only have -120dBV noise on the DC output, but also a special winding scheme for the switching transformer that allows for a very low value "Y" capacitor compared to "cooking SPMS, after we encountered similar issues as the OP in the field with a first version using a more "bog standard" design (like the OP is using).

Thor

 

[Potato Cakes]

Missing earth.

First, fit a differential mode filter using "Y" capacitors (2,200pF or larger "Y" rated, film preferred) between L/N & PEN (Mains Earth). Only use formally "Y" rated capacitors with UL please.

Second, fit a 100 Ohm resistor in parallel with 100nF and a anti-parallel pair of 1N4001 or similar diodes between +9V (ground) and PEN (Mains Earth).

Note, this earth connection is NOT a safety earth and should be substituted a mains earth connection in the system.

Thor

I'll give this a whack later this evening.

And thank you very much helping me sort out my SMPS woes. Power supplies are my worst nightmare. Audio nightmare that is.

Thanks!

Paul

 

[Potato Cakes]

Missing earth.

First, fit a differential mode filter using "Y" capacitors (2,200pF or larger "Y" rated, film preferred) between L/N & PEN (Mains Earth). Only use formally "Y" rated capacitors with UL please.

Second, fit a 100 Ohm resistor in parallel with 100nF and a anti-parallel pair of 1N4001 or similar diodes between +9V (ground) and PEN (Mains Earth).

Note, this earth connection is NOT a safety earth and should be substituted a mains earth connection in the system.

Thor

Just to be sure I'm reading this correctly, is the attached photo what you are meaning in the second step?

Thanks!

Paul

 

[eopett]

Hi Paul. SMPSs are great and also can be a big head ache at the same time. Your definitely going to want to consider using some inductors and low pass filtering especially. And obviously using some capacitors (fairly small value ones) to help with noise. It’s not that difficult but of course having an oscilloscope will help tell you about what frequencies your seeing as the noise. A minimum load is always something to consider. Also, determining how much of a load and choosing the right power supply will be better and save you money. If your running a big supply at 1% load it will be a little unstable generally. If you only need 50 mA get something for 100-200mA. But at the same time when it comes to these ac-dc switchers do not cheap out. You get what you pay for.

Check out schematics for mixing consoles and audio gear you might be able to get some filtering ideas just to see some approaches.

-Eric

 

[thor.zmt]

Just to be sure I'm reading this correctly, is the attached photo what you are meaning in the second step?

Correct.

The 100R/100n allow a earth reference to be established and common mode noise to drain to earth.

The diodes are not strictly neccessary, they prevent the resistor from going up in smoke, if for some reason larger fault currents present themselves.

Thor

 

[Potato Cakes]

Gotcha. I got it wired and tested for voltage but won't get to connect to any pedals for further testing till tomorrow. I will report back.

Thanks!

Paul