[sdiy] really basic question .. tolerance

[John L Marshall]

I agree totally. Probably an RMS improvement. But, what is this where the
tolerance adds? That implies to me that two 1% will accumulate to 2% error.
I don't empirically see how that can possibly happen.


Take care,
John
www.sound-photo.com
---------------------------------------------------------------------------

----- Original Message -----
From: "harrybissell" <harrybissell at prodigy.net>
To: "John L Marshall" <john.l.marshall at gte.net>
Cc: "Tim Ressel" <madhun2001 at yahoo.com>; "diysynth"
<diy_list at analog-synth.de>; <synth-diy at dropmix.xs4all.nl>
Sent: Saturday, January 18, 2003 2:27 PM
Subject: Re: [sdiy] really basic question .. tolerance


> Tell ya what John...
>
> I'll buy your arguement that it can't be worse than the original
tolerance...
>
> if you buy mine that its likely to be BETTER than the original
tolerance...
>
> H^) harry
>
> John L Marshall wrote:
>
> > Tim or Harry or Peter,
> >
> > Please show me an example of two 1% resistors in series or parallel
where
> > the tolerance exceeds 1%. I heard this rule many times but I can't see
it.
> >
> > Complex circuits are another story. I can see how errors accumulate
through
> > gain stages.
> >
> > I need empirical evidence to be convinced.
> >
> > Take care,
> > John
> > www.sound-photo.com
>
> --------------------------------------------------------------------------
-
> >
> > ----- Original Message -----
> > From: "Tim Ressel" <madhun2001 at yahoo.com>
> > To: "diysynth" <diy_list at analog-synth.de>; <synth-diy at dropmix.xs4all.nl>
> > Sent: Saturday, January 18, 2003 1:32 PM
> > Subject: Re: [sdiy] really basic question .. tolerance
> >
> > > Joerg,
> > >
> > > That is an optimistic way of looking at it. However
> > > the worse-case is still the sum. If I'm worrying about
> > > tolerances, I always look at worse-case.
> > >
> > > --tim
> > >
> > >
> > >
> > > --- diysynth <diy_list at analog-synth.de> wrote:
> > > > Hi all
> > > >
> > > > > tolerance are connected in series that the result
> > > > will be out of
> > > > tolerance?
> > > > >
> > > > > Add two 100k resistors 101k + 101k = 202k still 1%
> > > >
> > > > Sure ? I allways thought, the resulting tolerance
> > > > gets better with the
> > > > factor t_res=t_single/sqrt(n), where n is the number
> > > > of used resistors, and
> > > > t_single is the tolerance of the single resistor ?
> > > >
> > > > Joerg
> > > >
> > > >
> > >
> > >
> > > __________________________________________________
> > > Do you Yahoo!?
> > > Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
> > > http://mailplus.yahoo.com
>