Is the output supposed to pulsate?
If not, then keep increasing the resistor value until it stops pulsating.

Edit: Wait a minute. Can you please post your circuit diagram?

 

Sorry. I need to see the entire circuit diagram of the power supply.
I need to see where the wiper sits in the control loop. You want to be stabilizing the feedback signal.

 

Okay, I've got a 10k resistor where the 1k was. Smooth up and down, full range, no spikes, either lightly loaded, or with several amps of load. When I roll thru quickly, it may go high for a few tens of milivolts, but settles back within a second or so.

 

Okay, I've got a 10k resistor where the 1k was. Smooth up and down, full range, no spikes, either lightly loaded, or with several amps of load. When I roll thru quickly, it may go high for a few tens of milivolts, but settles back within a second or so.

Sounds good.

 

Sorry. I need to see the entire circuit diagram of the power supply.
I need to see where the wiper sits in the control loop. You want to be stabilizing the feedback signal.

Sorry, I've searched for it, no luck. It's a "black-box" smps from china, so very little data. I obtained a few for experiments, and to build a quick-n-dirty (well, not too dirty) 10-90v bench supply, as it also features current limiting. It looks like this one:

https://www.amazon.com/Yeeco-Adjust...qid=1615062293&sourceid=Mozilla-search&sr=8-6

 

My updated schema, just for completeness:

 

Just a note.
I dont like the idea of just scrapping the part and getting a new one entirely, although that may be part of the solution. That's because i nave encountered this problem several times in the past and i had found that even if you use a high quality pot there is always the chance that it will become defective and then your nice regulator will put out 20 volts when you have it set to just 5 volts. You dont know it until something blows out.
I was doing a power supply where i wanted to get really good regulation and prevent it from going over voltage for any reason that might come up, so i went with the dual regulation scheme where we have two different regulating circuits the second one being set just slightly higher than the first.
It takes a little more effort to adjust the two pots but it gives some piece of mind knowing that if one circuit gives out the second will kick in, and you will know it because next time you go to adjust it one of the pots will be unresponsive or cause jitter on the output when you turn it. You could also add some LEDs to indicate a failure which would be nice i think.
Yeah, a little more circuitry, but if it is a bench power supply it is worth it i think because you are going to use it to test some possibly much more expensive equipment.
.

 

Is there any point to a 10-turn wirewound pot? Isn't the resolution of a carbon or metal-film potentiometer infinite, but that of a wirewound is limited to a discrete number of turns of the wire around the former?

 

if you use a high quality pot there is always the chance that it will become defective and then your nice regulator will put out 20 volts when you have it set to just 5 volts.

Try two pots in parallel worked by the same shaft. i.e. a stereo pot. If one wiper disconnects, the odds are the other one won't. Add a cap as well if you want to. If the impedance of the following circuit is high (i.e. the non-inverting input of an op-amp) then the voltage will not change if only one wiper connects to the track instead of two.
ALso, a stereo pot is much more stable on a pcb than a mono pot unless the mono pot has a bracket.

 

I do believe 10 turns refers to shaft rotations not the number of windings.

It does, but there is a finite number of turns of wire, and that means only a finite number of wiper positions. It's like a multi-position switch (but with an awful lot of positions). The wiper of a potentiometer with a carbon track can take ANY position on the track.

 

Yea, I guess I deleted a little too late...sorry

After thinking about it, I figured you already knew that.

 

Yea, I guess I deleted a little too late...sorry

After thinking about it, I figured you already knew that.

That's ok - perhaps I didn't make it clear enough which "turns" I was referring to.

 

Try two pots in parallel worked by the same shaft. i.e. a stereo pot. If one wiper disconnects, the odds are the other one won't. Add a cap as well if you want to. If the impedance of the following circuit is high (i.e. the non-inverting input of an op-amp) then the voltage will not change if only one wiper connects to the track instead of two.
ALso, a stereo pot is much more stable on a pcb than a mono pot unless the mono pot has a bracket.

That would work in designs that only depend on a voltage signal adjustment. For those that depend on a resistance adjustment the loss of one pot would cause the resistance to go up by 2 times which would cause the voltage to either go way up or way down.
Still, for the voltage adjust mechanism i think it would work.

 

That would work in designs that only depend on a voltage signal adjustment. For those that depend on a resistance adjustment the loss of one pot would cause the resistance to go up by 2 times which would cause the voltage to either go way up or way down.
Still, for the voltage adjust mechanism i think it would work.

True. The potentiometer has to be acting as a potentiometer not a rheostat.
If the power supply is designed with a reference voltage to ground, the pot across the reference voltage, with the wiper going to the non-inverting input of an op-amp, with the op-amp driving the power stage and feedback to the inverting input, then it's a valid system, and probably quite robust.
If you can accept a control that is ever-so-slightly square-law, then a resistor between wiper(s) and ground will send the output to zero if both wipers lose contact with their tracks - like when you accidentally drag the power supply off the bench by its cable and it lands on the pot.

 

...
If you can accept a control that is ever-so-slightly square-law, then a resistor between wiper(s) and ground will send the output to zero if both wipers lose contact with their tracks - like when you accidentally drag the power supply off the bench by its cable and it lands on the pot.

Every-time I pick up the iron, I'm entering a door to a puzzle/maze that has multiple outcomes... How fun!
Thank you folks for answering my questions like philosophers. It's the only approach that counts.