Multi turn Pot failures

[xavier60]

I'm having trouble with the cheap 3590S 10K Pots from ebay sellers. After some use, the resistance element goes open circuit.
This can be bad when the Pots are used in a bench power supply as the voltage set point will jump from 0V to full reference voltage(5V) as the wiper passes over the break. Three have failed during the last 6 months.
Has anyone had the same type of failure?
I'm now using "BOCHEN" branded WXD3-12 10K Pots, hoping that they are better.

[Yansi]

I have never seen one of these fail.

Also note, that if the PSU goes berserk when the pot is open circuit, then I'd say the designer of that circuit shall think twice if his design is correct. Pots failing open is I'd say quite common type of failure mode for them.  Even a "scratchy" and worn potentiometer may cause spikes in outptut voltage, that may damage the connected load.

Always think when designing with any pots or trimpots about making the design fail safe when the trim/pot goes open circuit.

[xavier60]

I have never seen one of these fail.

Also note, that if the PSU goes berserk when the pot is open circuit, then I'd say the designer of that circuit shall think twice if his design is correct. Pots failing open is I'd say quite common type of failure mode for them.  Even a "scratchy" and worn potentiometer may cause spikes in outptut voltage, that may damage the connected load.

Always think when designing with any pots or trimpots about making the design fail safe when the trim/pot goes open circuit.

Yes, after the first Pot failed, I add protection to the micro-controller board, https://www.eevblog.com/forum/projects/linear-lab-power-supply/msg2373084/#msg2373084
The CC setting Pot failed last. It happened over night while the PSU was on and sitting idle. The break occurred at the high end of the range which I rarely use.
 

[Yansi]

To the control board I have added a dual comparator which monitors for sudden increase of wiper voltage for the CV and CC Pots via RC differentiators.

That unfortunately is not the correct fix to the problem; what would be a correct step is to change the pot role in the circuit in such way, that the open circuit wiper terminal automatically puts the output voltage to zero (or any controlled variable to zero, otherwise safe state). For example to wire the potentiometer in a different place of the feedback loop that he impact of it going open circuit gets the opposite effect.

Note that in some circuits this may not be possible, which indicates such topology is not fit for the purpose and different one shall be chosen.

And btw, cool PSU build, I like when people have the fine skill to work with some nice metal, instead of just the dumb 3Dprinter everything.

[xavier60]

I'm using them as potentiometers, dividing the 5V reference rather than as VRs. I already had a bleed resistor from the CV Pot's wiper to ground in case the wiper lost contact with the resistive element.
But this precaution doesn't help when the resistive element goes open. When I dismantle a failed Pot, I find a break in the resistance wire that makes up the compound helix resistive element.
The failed Pot acts like a switch. When the wiper is below the break, 0V is selected. When the wiper is moved above the break, the full 5V reference is selected.
The hair thin resistance wire is surprisingly strong.
I'm not going to abandon the design, if not for the failing Pots, it works perfectly.

[Conrad Hoffman]

You're buying junk electronic components on eBay. Use Bournes (or a couple other quality vendors) from an authorized distributor and you won't have that problem. If you have space, you can do a resistor network that you adjust by clipping out resistors. Cheap, highly reliable and won't change with vibration.

[Le_Bassiste]

You're buying junk electronic components on eBay. Use Bournes (or a couple other quality vendors) from an authorized distributor and you won't have that problem.

  +1

[Leo Bodnar]

There is a good reason military loves and uses them over conductive plastic or ceramic.
Leo

[bd139]

They're damn expensive though. I have about 6 here used for various RF tuning things and they were £15 a pop from RS

Another option is use two cheaper single turn pots (Alpha etc) for separate coarse and fine controls. This is actually more convenient from experience for voltage or current setting. Getting from 0 to Vmax on a 10 turn is laborious when you have to do it several times a day

[jbb]

Coarse and fine pots work OK, but they can be very annoying when you want to slowly sweep the voltage.  I often start at one voltage, twist the fine pot all the way over, don't reach the target and have to (very carefully!) adjust the coarse knob.

[bd139]

Use both hands

[jmelson]

I'm having trouble with the cheap 3590S 10K Pots from ebay sellers. After some use, the resistance element goes open circuit.
This can be bad when the Pots are used in a bench power supply as the voltage set point will jump from 0V to full reference voltage(5V) as the wiper passes over the break. Three have failed during the last 6 months.
Has anyone had the same type of failure?
I'm now using "BOCHEN" branded WXD3-12 10K Pots, hoping that they are better.

We have gear with lots of 10 turn trimpots and some panel pots that have been used (and abused) for DECADES!  Generally, any failed pots show signs of mechanical abuse from misaligned screwdrivers, etc.  These are all US-made name-brand units.

Jon

[BravoV]

We have gear with lots of 10 turn trimpots and some panel pots that have been used (and abused) for DECADES!  Generally, any failed pots show signs of mechanical abuse from misaligned screwdrivers, etc.  These are all US-made name-brand units.

Yep, example, a 10K 15w 15 turns and 0.02% linearity. 

[John B]

The genuine 3590S aren't that expensive, especially when purchased from somewhere like LCSC. However you're limited to only a few values, like 5k or 10k. I guess it's about supply/demand - the 3590 are the most common ones, some of the special size/shape/values can get crazy expensive.

The ebay ones are fake/counterfeit. The turning mechanism has a totally different feel ie harder to turn, more binding.

[xavier60]

The genuine 3590S aren't that expensive, especially when purchased from somewhere like LCSC. However you're limited to only a few values, like 5k or 10k. I guess it's about supply/demand - the 3590 are the most common ones, some of the special size/shape/values can get crazy expensive.

The ebay ones are fake/counterfeit. The turning mechanism has a totally different feel ie harder to turn, more binding.

LCSC's photo's show signs of the Pots being genuine. Amongst other observations, I think I can see the "C" stamped on the 3rd pin which none of my ebay Pots have.
I have also noticed tight spots. Maybe this leads to the broken resistance wire.
The  "BOCHEN" branded WXD3 series POTs are also very cheap so are they genuine or a counterfeit of something else? They do feel very nice.
And this. Bochen also make 3590S series Pots,
 https://www.alibaba.com/product-detail/RoHS-BOCHEN-3590S-2-103L-10K_60238488487.html

[Vovk_Z]

I had problems with some of 1k..10k 3590S 10-turn potentiometers too. Have bought em at about $2.5-3.0 at local shops. It's like some virus eat them.

[xavier60]

I had problems with some of 1k..10k 3590S 10-turn potentiometers too. Have bought em at about $2.5-3.0 at local shops. It's like some virus eat them.

How did they fail?

[Vovk_Z]

They start to "jump" in resistance (for 10-20%) at some point.

[H713]

I've had good luck with used 10 turn pots. I'm not brave enough to buy the knock-offs. The genuine helipots are really nice. I will say, I think I've only replaced one 10 turn pot.

My main bench supply (made with modified Power-One Linears) uses a pair of concentric wirewound Clarostat pots for fine and course adjustment. They're the ducks guts but not the easiest to find.

If you have resistive tracks going open (you mentioned having a carbon track go open?!) take a look at how much current you're putting through the pot. They aren't rated for all that much and I don't recommend pushing them.

[maxwell3e10]

Even Bourns can have bad batches, I suppose. Just replaced a bunch of 3590S on several power supplies that were about 10 years old and hardly used. The wiper was loosing contact almost everywhere on the track. The power supplies were from a reputable US company, so I assume they had genuine parts.

[sleemanj]

I use the wxd3 ones in a few things of my own for years, and sell them, I have not had a problem with them other than some scratchyness on the oldest one, and none of my customers have ever reported an issue either, I expect I might have sold a hundred give or take over the years.

[xavier60]

I use the wxd3 ones in a few things of my own for years, and sell them, I have not had a problem with them other than some scratchyness on the oldest one, and none of my customers have ever reported an issue either, I expect I might have sold a hundred give or take over the years.

That's good to hear. The WXD3 Pots in my bench supply have so far survived a year of heavy use.
A few drops of  PAO oil permanently fixes scratchy Pots. But nothing will fix the broken resistance wire problem that I was having with the knock off "Bourns" Pots.

[jonpaul]

Hello: Used 10T posts from Beckman and Bourns since 1970s,

The wire slide elements get oxidized, they can be opened and lubed with contact cleaner, DeOxit D-100, etc.

The dials fail mechanically and can be repaired.

The Chinese clones are junk, as always poorly made to appear the same but every material and component is the lowest quality.

Suggest genuine used Bournes or Beckmann, I find them on ebay and at Ham Radio fleas.

"You get what you pay for" "Buyer Beware"

DONT BUY CHINESE JUNK AND EXPECT PROFESSIONAL QUALITY

Kind Regards,

Jon

[wizard69]

This may sound hilarious but it always pays to read the manufacture technical docs.    For example some trimmers have surprisingly low usage figures.   I'm talking 5 or maybe 20 rotations.

Back in the day we use to send a bunch of A&B servo drives back for repair.  It was common for the drivers to come back with all new trimmers.   Even more conventional potentiometers have life spans (the number of turns) that vary dramatically.

By the way, buy name brand if you expect high performance.   Your still need to read the manual though.

[David Hess]

I'm having trouble with the cheap 3590S 10K Pots from ebay sellers. After some use, the resistance element goes open circuit.

...

Has anyone had the same type of failure?

That was a common problem even 40 years ago with "off brand" multi-turn wirewound potentiometers.  Either design the circuit to be tolerant of that failure or buy better parts like from Bourns.