Not only brush wear but worse, localized heating because of higher resistance. And so it can lead to insulation burning at higher loads. So there must be adequate overload protection included. So while AC voltage regulation can work very well, it does have limitations that lead to it not being commonly used.

Thanks for those comments. I'll have to rethink this thing. Originally it was more a passing fancy but after talking it out I see it may not work as well as originally thought. That's unless maybe I can limit the rotation to once every minute or so, as that would still be able to follow most of the line fluctuations we have here. It would only be used for 10 minutes, then off for four hours, then back on for 10 minutes, then off for four hours, then on for 10 minutes, then off for maybe 12 hours or more. At one movement per minute, that would mean 30 movements for one day, and maybe I could go every 2 minutes to reduce that too.

I'd look into something like a variable drive (for AC) but I am afraid that would regulate voltage not power. I really need to regulate power as I am doing some very important experiments. I wonder if a VFD could be modified without too much trouble to regulate power.

 

Utilities routinely use tap changers to hold the voltage in a window as the power demands shift. I watched that with a voltmeter at a rural campground one HOT afternoon on a very slow day. the voltage would be stepped up in what seemed to be 2 volt steps, until the system power limit was reached and exceeded a bit,then it would drop down about 8 volts for a while, then the cycle would repeat.

 

Utilities routinely use tap changers to hold the voltage in a window as the power demands shift. I watched that with a voltmeter at a rural campground one HOT afternoon on a very slow day. the voltage would be stepped up in what seemed to be 2 volt steps, until the system power limit was reached and exceeded a bit,then it would drop down about 8 volts for a while, then the cycle would repeat.

Yeah, that might work. I'd have to get a special transformer, unless I could install several permanent taps on a variac. Ever try that?

 

Yeah, that might work. I'd have to get a special transformer, unless I could install several permanent taps on a variac. Ever try that?

I have thought about it. I have a large (8 inch diameter) 220 volt variac with a burned section. I could solder to the last good turn and then a few other points and have a tapped transformer. It would be a challenge to solder to the top of one turn only (#14 wire) without creating a single shorted turn.

 

I have thought about it. I have a large (8 inch diameter) 220 volt variac with a burned section. I could solder to the last good turn and then a few other points and have a tapped transformer. It would be a challenge to solder to the top of one turn only (#14 wire) without creating a single shorted turn.

Oh yes I could imagine. What I was thinking was, create an insulating ring the same diameter as the core out of maybe a fiberglass board, then using brass screws to go through the board and have the tips contact any single winding, then the upper part of the screw can be the terminal for wiring, or have the PC board etched with solderable pads that each screw goes through along with an extra nut. They may have to be small diameter, M3 maybe, but that would be ok for my purposes. The screw tips would be ground or filed to a little blunt cone shape with the blunt end maybe a tiny bit curved. The ring would be mounted in some way so it stays stationary with respect to the core.
I probably won't get around to doing this but maybe in the future sometime.

 

Oh yes I could imagine. What I was thinking was, create an insulating ring the same diameter as the core out of maybe a fiberglass board, then using brass screws to go through the board and have the tips contact any single winding, then the upper part of the screw can be the terminal for wiring, or have the PC board etched with solderable pads that each screw goes through along with an extra nut. They may have to be small diameter, M3 maybe, but that would be ok for my purposes. The screw tips would be ground or filed to a little blunt cone shape with the blunt end maybe a tiny bit curved. The ring would be mounted in some way so it stays stationary with respect to the core.
I probably won't get around to doing this but maybe in the future sometime.

That is an interesting concept indeed. I suggest attention to the contact resistance when evaluating it. Point contacts are not usually suited for carrying much current.

 

That is an interesting concept indeed. I suggest attention to the contact resistance when evaluating it. Point contacts are not usually suited for carrying much current.

Yes, I was thinking blunt tip. If I went to a knife type contact it would be harder to construct I think. The variac is 20 amps but I only need around 4 to 5 amps max. The actual variation would be about 250 watts to 500 watts, with 120vac nominal input, output maybe 100vac max.

 

My concept had been to mask the adjacent tops that are flattened for brush contact, and then to tin just the one turnfor most of it's flat bare copper length. Then the attachment lead which would be about #18 stranded, and pre-tinned could be held in place and soldered to that tinned segment of the single turn. Probably the masking material would need to be teflon, although actual glass would be much stiffer. I think that a section of glass could be ground to a wedge shape to seat between turns and mask rather well.

 

There are many technologies that have been developed for "high-cycle" rotary controls, depending not just on cost but on linearity requirements. A highly linear rotary control which is just about standard for the thrust lever on jet engine controls for an aircraft is known as an RVDT (rotary variable differential transformer), it comes with the need for a little circuitry to effectively "demodulate" and compare the outputs. Other less linear analog techniques include moving a vane through a slot optocoupler, or adjusting the proximity of a magnet to a Hall effect sensor. There are also digital approaches, but if you're replacing an existing control you're probably constrained to using something like decoding an absolute Gray code parallel optical encoder, this is perfectly linear but the cost depends on how many channels of resolution you are going for, and you might consider passing the digital decoded output into a D/A converter to give you an actual voltage. Any of these approaches assume that you at least have the option to accept a voltage (or ratio of voltages) instead of a variable resistance. Good luck!

 

There are many technologies that have been developed for "high-cycle" rotary controls, depending not just on cost but on linearity requirements. A highly linear rotary control which is just about standard for the thrust lever on jet engine controls for an aircraft is known as an RVDT (rotary variable differential transformer), it comes with the need for a little circuitry to effectively "demodulate" and compare the outputs. Other less linear analog techniques include moving a vane through a slot optocoupler, or adjusting the proximity of a magnet to a Hall effect sensor. There are also digital approaches, but if you're replacing an existing control you're probably constrained to using something like decoding an absolute Gray code parallel optical encoder, this is perfectly linear but the cost depends on how many channels of resolution you are going for, and you might consider passing the digital decoded output into a D/A converter to give you an actual voltage. Any of these approaches assume that you at least have the option to accept a voltage (or ratio of voltages) instead of a variable resistance. Good luck!

We are talking about high power transformers. 500 watts minimum.

 

One thing several have missed is that the application is in an existing video game console. So it needs to replace a fairly small and cheap variable pot.

???

 

One actual written source I found regarding potentiometer specifications is from Digikey. https://www.digikey.com/en/articles/the-complete-guide-to-potentiometers. While not a formal standard, it supports MY interpretation, below.

The basic mechanical life of a potentiometer is defined as the number of rotations (full cycles covering 90% of the resistive track and back) where the potentiometer maintains its specifications. The full cycle from 0 to 100% is chosen because it puts the maximum amount of wear on the mechanical wiper (this is a mechanical life test). Presumably, with your average pot, the wiper wear will be the most likely failure mechanism, thus a primary specification. To put this in perspective, a typical 1" diameter panel pot probably has about 1.5" of resistance track, so one cycle is 3 linear inches of wear. Times 15,000 cycles says the wiper will wear out in 45,000 inches of travel over the resistive element. At this time, some aspect of the total specification will be exceeded, probably contact resistance or rotational noise. Of course there will be some wear of the resistive element, but not enough to cause it to go out of tolerance (which is measured end to end). The life test will also be conducted on a statistically significant number of samples and the results based on this statistical sample, not just a single unit taken at random.

In your application of gaming, the rotational travel may be limited to, say, a single 1/2" portion of the resistive track. Under these conditions you could expect the wiper to mechanically wear out in 270,000 cycles. However, the resistive element is unlikely to withstand that same number of wear cycles concentrated in only 1/6th of the total track. So, depending on exactly how the potentiometer is manufactured and the exact conditions it is used in, you could expect more than 15,000 cycles but less than 270,000 cycles of operation in gaming use becore the pot needs replacement.

I'm sure the manufacture, during the life testing, measured the amount of wear of the resistive track, but since the test is defined as "meeting specification", the details of failure mode are not revealed on a spec sheet. As others have said, if you want to know the expected life of just the resistive element or the failure mode of the life test, you will have to contact the manufacture.

Hope that helps.

 

???

Hi,

What exactly are you questioning here? He @MisterBill2 just said the potentiometer has to be small enough to fit into the original package. We were talking about things that are too large for that so it was a reminder.

 

AN interesting possibility would be an optical arrangement in which a mechanical shutter varied the light on a photo-resistor. Making that small enough for a game control would be a mechanical challenge, but it could remove the electrical noise from mechanical movement It might be simple with some types of control sticks, though.

 

AN interesting possibility would be an optical arrangement in which a mechanical shutter varied the light on a photo-resistor. Making that small enough for a game control would be a mechanical challenge, but it could remove the electrical noise from mechanical movement It might be simple with some types of control sticks, though.

A roller on a cermet resistive element in a clean and sealed enclosure is also very quiet and very long lasting - and expensive. And they already exist. The same roller technology has been used with linear autotransformers for decades. You guessed it - expensive and large. An optical vane moving through an optical channel is stable, noise-free and highly linear. They have been used with precise pressure sensors for decades. Again, not cheap.

All very interesting, but does any of this have anything to do with the OP question?

Does anyone have any experience on actual correlation between cycle life testing in the factory and actual life of the potentiometer.

 

AN interesting possibility would be an optical arrangement in which a mechanical shutter varied the light on a photo-resistor. Making that small enough for a game control would be a mechanical challenge, but it could remove the electrical noise from mechanical movement It might be simple with some types of control sticks, though.

Yeah that would be cool, although there would be some issues trying to get the resistance value right. For sure, SMD parts would make this smaller. One of those light sensitive resistors might work, the kind used in audio work. There's a slight delay in the change of resistance with light, but not long enough to matter for a potentiometer.

 

REturning to the original questions in post #1, the only life tests I have read about for variable resistors cover the full rotation.
My suggestions about alternative schemes were not to suggest ALTERNATIVE CREATIONS, such as some engineers have produced. A "one-off" device that would not be practical as a standard product would be a possibility.