Suggestions for noisy benchtop power supply

[wayfaerer]

Hello, I have a new benchtop power supply and have been using it to test an overdrive pedal I'm making. Unfortunately the pedal has that high pitched whine I sometimes get with pedals when I use a cruddy power supply. The whine goes away when I switch to battery power or a decent wall wart.

The power supply is from one of those no name Chinese brands. You know the kind... the volt and amp ratings are a little too high for the size and weight of the unit, the price is a little too low, etc.

The pedal, which uses all four stages of a TL074, already has a 100uF capacitor across the DC input as well as two 100uF and two 100n caps on the virtual ground rail (made with 3.3k resistors from the + and – rails). I guess I'm just wondering what I should do now... or rather what would you do (returning not an option)?

• Try to fix the problem by updating the pedal... I've tried adding an extra 220uF across the DC in but it didn't do a thing.
• Mod the power supply by adding a big ass cap across the output, or perhaps something else?
• Accept that benchtop power supplies (or at least mine) are noisy and don't worry about whine when working on pedals.

[antonis]

3..!!!

[ElectricDruid]

Big caps across the power supply don't actually qualify as an RC filter on their own, because there's no "R"! Ok, the wire always has *some* resistance, but we'll assume that your wire isn't corroded and your solder joints are good, in which case - it's negligible.

You'll get better results from the big caps by putting a series resistor from the power supply input to the top of the cap. Unfortunately that produces a voltage drop across the resistor, so small values are best to minimise this effect - 47R to 100R are pretty common. You can work out what the cutoff is with the usual 1 / (2 Pi R C) equation. Or stick the numbers into an online calculator, like this:

https://electricdruid.net/rc-filter-calc/?f=&r=47R&r_series=3&r_errors=1&c=220u&c_series=1&c_error=10

This shows that 47R/220u gives a cutoff at 15Hz, which should provide a heavy dose of attentuation for any high-pitched whine. It'd even knock mains hum a bit!

HTH

[iainpunk]

if you have some experience, id make a secondary voltage regulator box, that takes in a higher and noisy input voltage and regulates that down to a noise-free correct output voltage. there are numerous options for this, like the LM7809 voltage regulator.

cheers

[wayfaerer]

You'll get better results from the big caps by putting a series resistor from the power supply input to the top of the cap

Thanks... problem solved with a 47 ohm resistor! No more high pitched whine. I'll incorporate this into future designs as well. As for the power supply itself... I'll leave it be since in some ways it's more useful as a worst case scenario than a best case one.

Thanks again!

[anotherjim]

That noisy supply is actually a friend. It gives you a typical worst-case condition your project may have to deal with.

[Rob Strand]

I guess the question is did the bench supply whine since new?  (more than likely  )

A lot of SMPSs start to whine when the output caps dry out.   
On PC power supplies that can happen within 2 years from being new!.

Here's a pic of some caps in the process of bulging and other's have already popped,



Good idea to replace the caps before something pops as the PSU doesn't always survive when
the caps totally fail.

[anotherjim]

^ To be fair, it probably wasn't a good idea to mount the caps so close to something that can get hot.

[Rob Strand]

^ To be fair, it probably wasn't a good idea to mount the caps so close to something that can get hot.

True, but it's cool to see the attrition.   It's actually not too smart to block the airflow around the heatsink like that either

Eventually you get the same if with caps out in the open, especially if they are cheap caps carrying a high ripple current.

I was really impressed that the electrolytics caps on the controller board from our ceramic stove-top were in good condition after 15 years.  ESR tested good as well.   Not a nice environment camped next to hot plates all your life.    I think they were Samwha and Elna long life parts.   The board wasn't exactly high powered so the heat came from outside and not inside.

[Andrekp]

I'm only in the experimenting stage so far, but if you have a battery powered tool (like a drill with a big 18v battery on it), you can buy adapters from the store named after the River that clip onto it just like the tool would and have two wires coming out for the two poles of power.  From there, it's a simple thing to rig up a LM317, or whatever, buck board to create power for pedals or testing pedals.

I've gone so far as to do the breadboarding of the buck board circuit, and I works great, but I have not tested it with a pedal yet.  It should just work like a giant battery though.  Then you just charge it up every so often with the tool charger.

[crane]

Just my 2 cents - If whining is what I think it is (high frequency, not 50/100/200  or 60/120/240) Hz - most likely a "BIG ASS" electrolytic cap won't help. You might have better luck experimenting with smaller (1uf and less) film/ceramic caps.

[cab42]

if you have some experience, id make a secondary voltage regulator box, that takes in a higher and noisy input voltage and regulates that down to a noise-free correct output voltage. there are numerous options for this, like the LM7809 voltage regulator.

cheers

I did that to the noisy and unregulated 12 power supply for my desktop test amp, and also added another regulator to get 9v for my test rig

[ElectricDruid]

Just my 2 cents - If whining is what I think it is (high frequency, not 50/100/200  or 60/120/240) Hz - most likely a "BIG ASS" electrolytic cap won't help. You might have better luck experimenting with smaller (1uf and less) film/ceramic caps.

+1 agree that some big electrolytics don't perform that well at high frequencies, which is why you often (ok, "sometimes") see a 100nF ceramic in parallel with them. You'd think that adding an extra 100nF to a 200uF cap wouldn't change anything, and in terms of cutoff frequency it doesn't, but the properties of the two caps are complementary, so one filters noise that the other misses. For "best practice", yes, a 100nF ceramic in parallel is a good idea.