Go Down

Topic: 40 v digital potentiometer (Read 3429 times) previous topic - next topic

cpp333

What I am trying to do - I have a dc-dc converter with potentiometer attached to it. I found a couple of digital potentiometers and played around it and go it working. But the issue is, the digital pot, is upto 5v, and my dc-dc converter goes from 12-50V.

so my question is, is there any other way to control the dc-dc converter, by removing the pot present on it and adding a digital control using micro controller?

I was considering adding a motor and 3d print an attachment so fit the pot, and control it. But that might not be accurate.
I have searched and found the max digipot is 36V. Does anyone know any digipot for 40V or 50V?


Any help will be appreciated.

Thank you

MarkT

No way to tell, you haven't told us which converter.

Usually a digital pot is not the answer, unless you are doing an audio volume control.  Usually
such a pot is simply outputing a control voltage, a DAC is the way to generate an analog
control voltage, not a digipot.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]

cpp333

What I am trying to do - I have a dc-dc converter with potentiometer attached to it. I found a couple of digital potentiometers and played around it and go it working. But the issue is, the digital pot, is upto 5v, and my dc-dc converter goes from 12-50V.

so my question is, is there any other way to control the dc-dc converter, by removing the pot present on it and adding a digital control using micro controller?

I was considering adding a motor and 3d print an attachment so fit the pot, and control it. But that might not be accurate.
I have searched and found the max digipot is 36V. Does anyone know any digipot for 40V or 50V?


Any help will be appreciated.

Thank you

Wawa

#3
Oct 07, 2016, 01:45 am Last Edit: Oct 07, 2016, 01:46 am by Wawa
I think most people have failed using a digital pot for a DC/DC converter.
Search for it in the box on top of this page.

I suppose you have a voltage feedback on your DC/DC converter.
You could try replacing (or adding to) the "feedback resistor to ground" with a digital pot.
The feedback pin of a DC/DC converter shouldn't have more than a few volts on it.
Leo..

TomGeorge

Hi, 
Have you actually measured the voltage on the pot, not the output of the DC-DC.
Check the voltages from gnd to each of the three terminals on the pot.

Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

Coding Badly


@cpp333, please do not cross-post.  Threads merged.


DrAzzy

Best I know is the 36v one, I was looking at much the same thing last year, for the same exact purpose.

The 36v ones worked okay, though the wiper resistance is high enough to be annoying.
ATTinyCore for x4/x5/x61/x7/x8/x41/1634/828/x313 megaTinyCore for the megaavr ATtinies - Board Manager:
http://drazzy.com/package_drazzy.com_index.json
ATtiny breakouts, mosfets, awesome prototyping board in my store http://tindie.com/stores/DrAzzy

Southpark

#7
Oct 07, 2016, 05:44 am Last Edit: Oct 07, 2016, 08:40 am by Southpark
What I am trying to do - I have a dc-dc converter with potentiometer attached to it.
Maybe need to indicate.... attached in what way? For example, I could attach a screw-driver to dc-dc converter.

If your digital potentiometer has unexpected constraints, such as the end-terminals must be within 0V and 5V, then it might be necessary to do some voltage reduction and/or voltage level shifting in order for the potentiometer to work within its own required operating limits.

There are digital potentiometers that don't have this particular 0 to 5V limitation, but don't know if there are any potentiometers that allow up to +40 Volt on any particular end terminal.

MrAl

#8
Oct 07, 2016, 09:38 am Last Edit: Oct 07, 2016, 09:43 am by MrAl
Hello there,

I found this to be a problem also when i was looking into using a digital pot because the pots on power supplies wear out with constant use and have to be replaced, and who wants to do that all the time.  Also, it would be nice to be able to adjust the supply with a digital control.

The problems started to mount fast.  first, the voltages are limited as you found out already, that was a big factor.  Second, the resolution of most of the pots i found in the time allotted was not very good either: 256 steps in most cases.  That's not very good for a digital control.  My power supplies have to go 0v to at least 30v, but say they only had to go from 0 to 25.6v, that would mean that using a pot with 256 steps would only at best do a resolution of 0.1 volts, which doesnt cut it for me as i want minimum of 0.010 volts (10mv) but ideally 1mv resolution (0.001v).  Why bother if i cant get what i need.  So other ways have to be found.  If you dont mind the not so good resolution though you might try replacing the lower resistor in the feedback divider as someone already mentioned although you have to be careful there too.

All controllers have a feedback input to sense the output voltage, but some also have a reference input.  If you can change the reference, you can change the output, it's as simple as that...well sort of.  The reference can be anything from about 1.2v to 5.0v with the typical being 1.24v and 2.5v.  Thus, if you use a DAC to control the reference you might have something, because you can get high resolution DAC's like 12 bit that would work nicely.
Unfortunately that's not the end of the story either though, because you cant decrease the reference to 0.010 volts for example and expect decent operation.  So that has limitations too.

Another idea is to roll your own digital pot using analog switches.  You can switch in and out several resistor values and when arranged in a binary fashion, the binary code will determine the feedback resistance.  There may still be a problem with finding analog switches that can go over a certain voltage however, and the resolution will again be limited unless you dont mind using a lot of resistors and analog switches.

Another idea yet is to use a small signal transistor in the feedback path.  This would be controlled by a DAC and amplifier.  The drawback here is you would be forced to measure the output and make adjustments to the input to the transistor.  That means you'd have a secondary feedback system, and unfortunately that would mean a slower system because it could only respond as fast as the digital system could make measurements and make corrections to the transistor drive.  Lucky though it may not have to be that fast as the original system will still be fast, and perhaps an FET would be a better choice than a bipolar.

Possibly another idea would be to use a multiplying DAC, but of course the max voltage would probably be limited again there too.

Reed relays?  Maybe, if arranged in a digital fashion with appropriate binary weighted value resistors like the analog switch idea.  8 relays would only give us a resolution of 256 steps however.  The reliability might come into question too here depending on the quality of the reed switches used in the relays.

Another idea might be to add a small bias to the feedback signal, which would mean it would adjust the output voltage.  This might work and would only require small voltages less than 5v, but would have to be tested carefully for stability.

One last idea i can think of offhand is to use an audio level control analog opto coupler.   That is an opto isolator that is made for audio compressors and expanders and the like.  The internal LED is driven with a current signal, and the output is a resistance...an actual Ohmic resistance, that varies with the level of the LED current.  That would vary the feedback resistance although most likely not in a linear fashion and probably not too temperature stable, so again a secondary feedback system would have to come into play.  Also, we'd have to check the max voltage for these devices to see if they can work up to the required range of 30v or higher.

So there are many alternatives, but none of them seem very good really as they bring in complexity and uncertainty in an application that is usually well understood by everyone.  Perhaps it just requires more thought and some clever invention.


Southpark

Yep! At the moment, it looks like digital potentiometers (or most of them) have certain drawbacks - depending on what people want to do with them, for now. Hopefully the GEN 3 or GEN 4 potentiometers are high fidelity, and can be used in almost the same way as classical potentiometers - eg... none of the crazy voltage limitations on the end terminals.

MrAl

Hi,

I found a couple with 1024 steps, but the package type is TSSOP which is too small.  Also, if i go through that much trouble i think i want at least 4096 steps or even more.

MarkT

And I'm going to repeat myself:
Quote
No way to tell, you haven't told us which converter.
There's nothing really to say without that information.
[ I will NOT respond to personal messages, I WILL delete them, use the forum please ]

Go Up