From: Spehro Pefhany
Subject: Re: Don't do this! Just don't!
X-Newsreader: Forte Agent 1.93/32.576 English (American)
Date: Sun, 19 Jan 2003 22:15:22 GMT
NNTP-Posting-Date: Sun, 19 Jan 2003 17:15:22 EST
On Sun, 19 Jan 2003 21:12:19 +0000, the renowned
C.D.H.Williams@exeter.ac.uk (Charles DH Williams) wrote:
>In article , Jim Thompson
>> John, Why don't you elaborate for the students lurking here, and
>> explain what the "gotcha" is?
>He means that the small (few millivolt) offset this arrangement is
>supposed to create is only as stable as the 15V rails.
Also the tempco of the element unnecessarily comes into play
relative to the resistor tempcos. And the element tolerance, which
is often huge, affects the adjustment range, meaning they have to
make it even worse than it would otherwise be.
>My pet hate is this one:
> +--------- voltage reference
> 1K O <----+
> T |
> | |
> 0V ----------+
>It works just long enough to pass QC and and after a few months the
>wiper-track contact starts fluctuating. There was a digital panel meter
>sold in the eighties by RS Components (not Radioshack, the UK firm) that
>used this. You could change the reading by several percent by shouting at
It should not be that bad, but using it and expecting much better
than 1-2% isn't a good idea, even with a good pot designed so it's in
at, say, 90% +/-5% of rotation. Did they use a carbon pot too?
The very top and bottom of the ranges are always troublesome. I also
spec a procedure to seat the wiper in before calibration to make it
stay put a little better.
I hate dorks that don't bother designing the circuit so that it has
the minimum adjustment range that always works, and compensate for it
by putting a 22 turn pot in there (with the same one-turn element that
a 10 cent cermet trimpot has, just with a bit better wiper that's
geared down). Hilarity ensues (not).
"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