If you have a traditional regulated power supply that you want to make adjustable, you’ll have somewhere in the circuit a feedback line driven by a potential divider across the output. That divider will probably incorporate a variable resistor, which you’ll adjust to select your desired voltage.
The problem with using a standard pot to adjust something like a power supply is that a large voltage range is spread across a relatively small angle. The tiniest movement of the shaft results in too large a voltage change for real fine-tuning, so clearly a better means of adjustment is called for. And in many cases that need is satisfied with a ten-turn potentiometer, simply a pot with a 10 to 1 reduction drive built-in.
[Dardo] had just this problem, and since 10-turn pots are expensive to buy and expensive to ship to his part of the world he built his own instead of buying one.
His 10-turn pot is a fantastic piece of workbench improvisation to deliver the goods from what is at hand. A capstan and shaft from an old radio tuner drives a small rubber belt and pulley from a CD-ROM drive. This in turn has a gear which drives a much larger gear coupled to the shaft of a standard potentiometer. All the drive components are mounted on a frame made from an old aluminium heatsink.
To be strictly accurate it’s a 13-turn pot rather than a 10-turn one, but the principle is the same. And it’s an extremely effective component, especially for the price. We like it!
If you are inspired to put a 10-turn pot in your supply then this project provides a blueprint. It’s worth saying though you could take a look at an off-the-shelf component instead.
This one goes to 11.
It’s hard to get a ten-turn pot in the Argentine?
Probably. I heard from a guy from Argentina that there was a time (not too many years ago) when was nearly impossible to find 1k resistors!
Shipping times to Argentina are between 1-4 months. Even for small parcels. If you use a carrier you can only use it 5 times a year. Also most things you get in stores are 2-20x pricier than in the rest of the world. A cell phone you can grab for $400 on ebay, on a retail store in argentina you would pay $1200 or through some shady reseller for around $1000. Taxes are also high.
I ship to Argentina and Brasil a lot. They are the worst to get in to, even with FedEx expedited handling. Customs forms and invoice copies have to be perfect and include data we don’t need for anywhere else. Long delays and duties to be paid by the recipient.
I have heard that the cost of the Customs system comes from the fees, so there is incentive to catch every cent. Delivery time is not important.
Well, you can still buy them ready-made: http://www.vishay.com/search?query=turn+dial&searchChoice=part
and they’re not even that expensive.
But the problem here is, that standard potentiometers are simply not precise enough to be used with such a gearbox knob.
You’ll have to buy a precision one, that’s likely as hard to get as the knob.
this
I was wondering this too; good ole’ snap, krackle, and sometimes even pop when heard on audio once it has been in service for a year or two.
Actually I think that offloading user hamfistedness to another shaft, probably improves longevity.
AB made 10 turn pots with the standard composite resistive layer similar to the Type J. They used a ball planetary system to do the reduction while keeping the size of the pot about the same size as a standard panel mount pot.
Yup, I’ve seen those in a Heathkit VTVM. They’re not much good as far as I’m concerned. They slip and don’t provide a good contact across their range. At least that’s been my experience.
I have found them in tektronix stuff so they cant be too bad. I think they are mostly used as a vernier since they can slip once their extents are reached.
Exactly this… the whiplash and overshoot in a traditional pot will be a limiting (but certainly not prohibitive) factor in such an arrangement. It is a clever hack though, and does provide the user with needed functionality that exceeded what was previously available.
Why not use two pots in series in a 1:10 ratio?
The smaller one is just for fine-tuning the total resistance of the series, so the bigger one does 0-90% over the turn and the smaller one does +0-10% to that.
You could even get clever and drill a hole through the shaft of the bigger potentiometer, and attach the fine tuning knob in the middle.
Its’s going to depend on how the pots are wired, if they just use it as a variable resistance with two wires you can do that. If you have it wired up as a voltage divider it will be more complicated. There are single pots that do this. They have two pots on a single shaft internally, one coarse and one for fine. You feel the knob get stiff when you make coarse adjustments and then go the other way and it is lighter for fine. Its hard to explain.
“If you have it wired up as a voltage divider it will be more complicated.”
Not much. One pot is wired to itself and the other as the divider. If the fine tuning pot is R1 and coarse adjustment pot is R2, R1 is wired to itself and R2 gives out the divided voltage, then the output will be N*R2/(n*R1+R2) where N, n, are the wiper positions from 0…100%
It’s not strictly linear for the fine adjustment, but as long as R1 is small compared to R2, like 10:1 then it won’t make much of a difference.
You beat me to it. Similar to old style tuning, just connect a smaller variable cap in parallel with a main one.
Yah, that’s sort of what I was thinking but instead of the big pot I was picturing a many-throw switch and fixed resistors. Either way I guess…
What I like about that, and I’m guessing you were thinking the same is that you don’t give yourself a cramped wrist when you twist it from one end of the range to the other too many times in a short period.
Another way to achieve this might be a gear setup very similar to what he did but extend the pot’s shaft all the way though and put a second knob directly on it.
Do you mean a Fine and Coarse adjustment? Lots of cheap power supplies use those. I hate them.
Best post title of 2017 so far! Keep it up.
:)
They used to make a knob with a built in reduction mechanism. I dont remember who made it though.
I got a lot of 10, 10k, 10 turn pots from ebay that were very nice for under $15 shipped a year back for a project
They used a shaft that moved a wiper along a coil of resistance wire, they were very sensitive and did not drift noticeably over time
China man …
Problem was not the source but the destination. Shipping to Argentina is limited and slow. Shipping to USA and having a friend carry it and drive from USA to Argentina would have been faster and cheaper.
Does that mean there is a great market opportunity buying stuff from China in the US and taking it to Argentina to sell?
Or are there some big tariffs or something that make this impractical?
30-40 years ago thousands of *people* disappeared in Argentina. I’m sure it will be nothing for them to disappear your shipment as well … That is the problem. Unlimited and unchecked government power, bribery and stealing.
Maybe I should’ve submitted my post here to Hackaday:
http://miscdotgeek.com/bitx40-vfo-alternatives-and-more/
I builit a similar drive, although I must admit mine is not nearly as pretty. It takes a 3/4 turn pot and gives it a 20:1 ratio, and makes a very nice difference. The real trouble with using gears is backlash. It can be done better with pulleys, or even better, with a 10 turn pot :)
Backlash and that rubber band, yuck. LM317’s only take a two term hookup. That’s easy to do with two pots.
I have some ten turn pots only a centimeter in diameter and a 15 turn pot that is grapefruit sized. The weird one is a dual 75 turn pot. Hamfests! Old science gear. The turn counting dials are a giveaway.
Why not using a tiny microcontroller + 12bit DAC Output (eg. MC4802) for output and a rotary encoder as input?
Or just a standard off the shelf digital resistor divider. Or just a 10 turn pot. My bet is if he couldn’t source a 10 turn pot in the first place, sourcing these alternatives won’t be too much easier. Sometimes you work with what you have.
Make absolutely certain that the micro which you choose runs Windows10.
I use PIC16F628 for this, Software is written in JALV2in addition to the rotary encoder I can set the Voltage via serial,
My HaD 1K challenge did something like that except I use a sigma-delta DAC built out of an integrator with a $0.20 opamp and a PID feedback loop. The DAC is around 13.5 bits and easily go higher resolution, but limited by the 10-bit AVR ADC. The adjustment is in 10mV per step and 20 steps per turn. Too lazy to try to add velocity sensing as I was down to 30 bytes left allowed by the contest.
https://hackaday.io/project/19078-adjustable-linear-bench-supply-in-1k
just use a huge in diameter knob over it
That idea made me think.
Take a plain one-turn pot with a knob.
Take an old CD. Drill finger-width holes at regular intervals along it’s radius.
Actually, I probably wouldn’t drill them all across a single radius line. That would make it unbalanced. The point is that they would be at some even intervals distance from the center. I’d try to spread the holes at evenly spaced degrees so that it remains balanced.
Attach the CD to the knob.
For the finest tuning turn the CD by it’s edge. For coarser tuning stick your finger in one of the holes and turn it that way. For the coarsest tuning just ignore the CD and turn the knob directly.
For bonus points use one of those clear spacers that come in big CDR packs or peel the foil off of a CD to make it clear. Paint or etch a radius line on it. That’s your dial marker. Now you can write voltages or whatever else the knob controls on the surface behind the CD.
I suppose you could just cut out a clear acrylic circle with a laser cutter but you lose all your creative re-use points when you do that.
Another alternative is to glue knobs to the CD rather than drill finger holes in it.
If you do place markings underneath it with a line you could also paint the back leaving out a ‘pie piece’ around the line so that you only see the markings that are near where you are currently tuned. Functionally it’s no better but it might look nicer.
Some older test equipment and radios used a large disk with a small knob that just grips the edge. Turn the large disk for fast tuning, turn the small knob for fine tuning.
Go old school. Stretch a piece of nichrome on a meter stick and slide an alligator clip…..
I build a dial cord mechanism to make a 3/4 turn pot into a multiturn pot. Like an old radio tuner. I even put a pointer on it and a scale.
A good 10 turn pot has a lot of linear area on the slider. I suspect that while a home brew 10 turn pot would be acceptable someplace where you want to zero in on a value, such as setting the output voltage of a power supply, but not be so good at repeatability. That is cranking it all the way one way than say 5 turns the other and getting the exact same value every time. I would just assume have a multi position switch with 3.3, 6, 6, 7.5, 9, and 12 and perhaps a position with a pot, but 99% of the time what I want would be one of the switch positions. Easy to set the output without screwing around with putting a DVM on it. There are also chips that are digitally programmable pots. I am not sure one of them and an analog regulator would be better than a unipolar D/A and a power opamp.