DC-DC Converter Pedal PSU whine

[mmlee]

Hi,
I'm working on a tiny PSU to fit into a pedal board that I wanted to power with my USB battery pack (phone charger thing) and also for using all those 5v USB plug in phone chargers around.

I got a couple Murata MER1S0509SC 5v to 9v DC-DC converters (as that's what I could find locally) and I've tried it out on the breadboard using my phone charger.  The charger puts out a steady 5v's under load, the DC converter puts out around 12v with no load and around 9v under load depending on the amount of pedals, seems to drop .1v the more pedals that are daisy chained.

The problem is I'm getting a high pitched whine through the amp.  The switching frequency of the DC-DC converter is 62kHz, have I made a mistake here?  I thought that'd be fine as it's above 20kHz.  Also, the max ripple & noise is stated at 20mVp-p, have I seriously underestimated this? Would putting it in it's enclosure help? as said it's on breadboard with lots of long crocodile clips attached to jumpers and stuff (probs could have clipped straight onto the pins now I think about it).  Am I just going to have to accept I need to use some extra filtering? didn't really want to use any components other than the converters.

I've powered a few different pedals with it thinking it could be possible heterodyning with modulating effects but it's happening with simple IC & tranny based drives and fuzzes too.

Any help is appreciated.

Cheers,
Marcus

[drolo]

I know this is not really going to help much since you already bought your converters but I had very good luck with these:

http://www.bgmicro.com/pm6078-004.aspx

I made a PSU with 16 of them and could not notice any noise or whine with any pedals, digital or analog. Just hooked up the converters in parallel to the power supply and the outputs straight to the DC jacks, no filtering. They are quite cheap (except for the shipping maybe depending on where you are located ...)

62kHz should normally be ok as it's above the audio as you mentioned but the noise you are getting might also be a product of the 5v power supply's switching frequency plus the converter's (not sure it's true though)
Have you tried with a different power supply or perhaps with your 5V battery pack?

Other than that you might need to try filtering the output but you might lose some voltage doing that.

[mmlee]

Thanks for that, only $.30! Very cheap, I am in the UK though and like you said, I've already bought the converters.

I've only tried the one psu, I've got a couple other different ones so I'll give that a go tonight, along with the battery pack.

If that doesn't work I guess I'll try a 100ohm in series and then 100uF (or more) to ground.

Is there any way to check the output noise from the 5v psu?

I can definitely rule out the lack of shielding then?

Marcus

[drolo]

Not sure about the shielding, except if you are running the power wires close audio wires.
Not sure how much 62kHz can radiate. It's usually worse the higher you go. Try moving things around to see if it makes any difference.

For filtering try 22 Ohm and 470uF instead. Same cutoff frequency but smaller voltage drop :-)

[mmlee]

Just to keep info together, just looked up typical switching frequencies of these 5v chargers and a couple articles state they are typically 50kHz or more, 33kHz in some older designs.

Marcus

[samhay]

>the DC converter puts out around 12v with no load and around 9v under load...

This is quite common. If you connect an LED (and associated resistor) to the output, you will have both a handy indicator that it is working, and likely enough load to keep the output close to 9V.

>For filtering try 22 Ohm and 470uF instead.

I would not expect a 1/4W resistor to last very long here with the current draw of a few pedals + some inefficiency in the DC-DC converter.

>typical switching frequencies of these 5v chargers ... 50kHz or more, 33kHz

Heterodyning between the charger and DC-DC converter might be the culprit. If your charger is whining at 50 kHz, this will beat agaist the 62 kHz DC-DC converter to give you noise at 12 kHz and 112 kHz (which will only bother the neighbourhood bats).

If you can get a low dropout 5(ish)V regulator between the charger and DC-DC converter, you may be able to filter out the 50 kHz and thus prevent the heterodyning. Alternatively, try another changer. At the price we (don't) pay for them, I would expect the quality to vary from crap to marginally ok. Also, a 33 kHz switching charger will have advantages in this application...

[mmlee]

Thanks, Sam. I guess I'll find out tonight when I try out the battery pack.  Rather than a 5v regulator, would a filter between the charger and converter not work? (Providing the voltage drop doesn't cause it to go below 4.5v, the converters minimum voltage input).

Also, how did you work out the frequencies caused by the heterodyning?

Marcus

[samhay]

No problem.

As a first approximation, the heterodynes (new frequencies) are simply the sum and difference of the 2 frequencies that are mixed.
If you have 62 and 50 kHz, this gives 62 + 50 = 112 kHz and 62 - 50 = 12 kHz. I'm glossing over a few things and there may/probably will also be multiples of these frequencies - think harmonics in e.g. a distortion device. The wikipedia page is a good first point of reference: https://en.wikipedia.org/wiki/Heterodyne

The problem with a RC filter is that the voltage drop is sensitive to load (remember V = IR), so it isn't a great solution for a power supply that may have to deal with quite a range of loads (number and types of pedals). If you're thinking that you will just use a smaller resistor to minimise voltage drop, you may quickly discover that the size of both the required reistor (maybe multiple watt) and cap (1000's u) befomes problematic.
If you built this into a pedal and you knew the expected current draw (which probably won't change much), then you can probably make a RC filter work.

[mmlee]

Update:

Just tried some different chargers.

Samsung eta0u70xbe - whining
Samsung adaptive fast charger - whining
Apple model a1401 - no whining
EasyAcc pb10000cf  no whining

Success, so I thought, just don't use the Samsung chargers.  Turns out quite a bit of noise is also being introduced, like a white noise "shhhhhhhhhh" kinda thing.  Also, a lot more stomp switch clicking/popping, from 3pdt true bypass, buffered and optical bypasses.

Marcus