From: Spehro Pefhany
Subject: Re: Don't do this! Just don't!
References: <firstname.lastname@example.org> <email@example.com>
X-Newsreader: Forte Agent 1.93/32.576 English (American)
Date: Sun, 19 Jan 2003 23:13:20 GMT
NNTP-Posting-Date: Sun, 19 Jan 2003 18:13:20 EST
On Sun, 19 Jan 2003 23:23:28 +0100, the renowned "Frank Bemelman"
>"John Larkin" schreef in
>> OK, every once in a while we need to have a trimpot crank in a little
>> bit of DC offset into an opamp or something, and there are those nice
>> +-15 or whatever power rails handy. But I see so many presumably
>> intelligent engineers do something like this:
>> 100r O <-------- small offset
>> This is a *very*bad*idea*. I recent sold a bunch of
>> amplifier/discriminator boxes because somebody else did this, among
>> other dumb things.
>> So quit doing this. Just quit it.
>Ha! I never use pots ;) Well almost. Sometimes for gain.
>But if I had to make an offset circuit, my first prototype
>would use the circuit above. And if the prototype didn't
>reveal any problems, it would end up in the final thing.
>But I would choose a smaller ratio between resistor and pot,
>to make sure the pot has a large enough range. 10K and 1K pot
>I'm looking forward to hear why it is bad practice ;)
With the values shown, it would (in Spice simulation anyway) have
+/-75mV on the element ends. Say one 10K resistor was low by 1% and
the other high by 1%, you'd get about 600mV different at the
mid-point.. 4 times the range of adjustment. The thing is, that
typically two resistors from a batch are pretty closely
matched, and the +/-15V might be close enough matched that it works
well enough to get shipped. The worst kind of engineering mistake.
Note also that the 100R might have a tolerance of 20% or 30%, in
comparison to even cheap 1% resistors, so you've got to make the
adjustment range proportionally greater.
This is much better and costs only pennies more:
1K0 O <---- small offset (+/-68mV)
Now a 1% mismatch in the 10K resistors will only cause about
a 750uV change in the voltage at the center point of the pot
Also, if the pot element is 20% low, it will only reduce the
range of adjustment by about 2%.
The only downside is the increase in "CRV" due to the
higher pot element. If the adjustment affects the gain,
this might not be acceptable and you'd have to use some
other method, buffer it, or just drop everything above
by 10:1 in resistance and live with the higher power
"it's the network..." "The Journey is the reward"
firstname.lastname@example.org Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com