Turn a log pot into a (kind of) linear pot

[osterchrisi]

Hey there!

I made a terrible mistake in a circuit that I have to fix now: I ordered 250 wrong potentiometers (logarithmic scale) but they should be linear. Ordering new ones is not an option as this would take 8 weeks to wait for but we start production next week.

Is there ANY way of hacking a logarithmic pot into a linear-like one? The resistance value is 100K but I also wouldn't mind loosing part of this range, to like 50K or whatever.

Any clue anyone?


//edit:
Mhm, thought I attach some more info: The log pot is actually responsible for setting the speed of a 555 timer (you know, between Trigger and Discharge). So maybe there are more possibilities of changing the logarithmic behaviour of the setup?

 

[shrtrnd]

You'll only compound the problem by using the wrong pot.
Can't you swing a deal with the supplier to swap the audio pots you got for linear ones?
You may not think you've got an option, but if you don't fuss up to the issue, you may
lose your job before the replacement pots get there. Cause it's going to cost a lot more
to recall and retrofit the product, than getting the correct pot.
SOMEBODY out there has 250 linear pots on the shelf right now.
Get busy finding them, and rushing them to your business.
This seems bad right now, it'll get worse if you don't take steps to correct the mistake.

 

[jackorocko]

This certainly isn't a solution, But if you mechanically attach one to another that are in the exact opposite arrangement, you should have a somewhat linear pot. But that creates more problems then it solves considering you now have a shortage of pot's.

You should listen to shrtrnd and suck it up. What country are you in? Maybe someone here can help find or already knows of a supplier that might be able to help you. Only reason I can see for you not wanting to get linear pots is that you can't return the 250 you got.

 

[osterchrisi]

No no, the problem here is another one: This is a pot with a switch, the PCBs are designes especialle to fit this pot into there. So as I said, ordering similar ones would take 8 weeks.

Thanks for the advice though

 

[OLIVE2222]

Here under a article they can maybe help

http://www.elby-designs.com/documents/tailoringpotentionometers.pdf

Olivier

 

[osterchrisi]

Wow thanks! That really looks promising!
Gives me more guidance in what I tried to experimentally test today - I will try more things tomorrow, thanks a lot again!

 

[jackorocko]

Here under a article they can maybe help

http://www.elby-designs.com/documents/tailoringpotentionometers.pdf

Olivier

Guess that is pretty cool, now the question is where does the OP need his linear range Figure 6 shows that a 10% fixed resistor will give you a close approximation to a linear pot in middle. I have definitely bookmarked that page.

it is a shame we can't make this a sticky since it is a third party site. Such valuable information.

 

[(*steve*)]

And it is done (kinda) here.

 

[shrtrnd]

I hope some of you guys aren't designing my next consumer product buy.

 

[jackorocko]

I hope some of you guys aren't designing my next consumer product buy.

"If you don't have the time to do it right the first try, when will you find the time?"

 

[osterchrisi]

I tried around a bit now and have really strange results:

When I connect the log pot left leg to Discharge and wiper to Threshold/Trigger I get a really nice linear scale - but 'reverse action', i.e. fully CCW is highest speed and fully CW is lowest speed.
If I then replace the left leg of the pot (Discharge) with the right one, I get a log scale but with 'right direction', i.e. fully CCW lowest speed and fully CW highest speed.

Anyone understand that?

 

[shrtrnd]

Yeah, it's called a logarithmic pot in a linear pot application.
You got a lot of good advice for trying to make your circuit work here. We think in
terms of adapting, modifying, and just plain MAKING things work. It's probably why
you came here asking for advice.
Unfortunately, you are designing/building for a new product. And because you made
a mistake, you're willing to risk your company rep and the ire of potential customers
by jury-rigging a fix.
I completely understand the advice you got here, and under normal situations it would
seem to me to be pretty ingenious advice. But in your situation, new product and all,
I'd advise you to bite the bullet, take some heat for a mistake, and get it right.

 

[osterchrisi]

Yeah yeah yeah, I totally got that. I myself come from the DIY-scene and learned electronics myself, now I started a company a few months ago. I much appreciate the advice given here but 'tailoring' the pot with resistors didn't really work out for me unfortunately. Although the Elby Designs paper is possibly one of the greatest concerning that topic.

And I totally also get your advice about re-ordering, I would possibly advice the same thing to anyone else but here it's really not an option because if we don't start selling/shipping next week we're basically bankrupt, I can't wait another 8 weeks for a lin pot...

I'm now actually thinking about simply using the log pot in a linear application and have the behaviour 'in the wrong direction'. Is this 'using a log pot in a linear application' a known thing I've never heard of?

 

[OLIVE2222]

is the design working well with a linear pot?

 

[osterchrisi]

Well that's the strange thing I don't understand: I breadboarded it now again and with the log pot in 'reverse lin' setup it works perfectly fine (apart from the wrong direction).

With a lin pot I get a log response.

But: On a board I soldered earlier with the same pot I'm speaking about (the one with a switch I use) but then the linear version (I only got that one from the distributor to check my footprints) it also works fine...

Big confusion... I'm talking about standard basic 555 square wave oscillator here.

 

[daddles]

Well, I certainly understand the urgency and the danger of going bankrupt -- I certainly hope it doesn't happen. However, you've been a bit remiss in asking for help here. You haven't shown a schematic of what you're dealing with nor pictures of the parts (pot, PCB, etc.) involved, along with a clear description of exactly what the problem is and the needed design fix. Note I'm not chiding you, but you've lost some precious time. A well-stated problem with good documentation often leads to a dead-on fix when a knowledgeable person sees the problem (of course, there's no guarantee a fix exists, but to paraphrase an old programming proverb, "all bugs become shallow with lots of eyes"). A good problem description (say, in a PDF) should take less than an hour to produce and might generate a work-around for you where mere words haven't sufficed. And, at the very least, you could also post the PDF on e.g. Electro-Tech and AAC and get hundreds of good eyes looking at the problem for you. That's due diligence in my book.