[sdiy] Speaking of OTAs/Matched Pairs????

[Rob]

Hehe. Thats considering that they think its easier to test for matching
rather than just adjust doping for a tighter spec, use deviations to
determine the number of units that *might* fail,  and *hope* they all pass..
;)

Its 2001, not 1971.  ;)) It doesn't pay to test everything when the doping
is able to be closely monitored.

I would say just buy a handful of 3086s, test a few, and if they all line
up, you could prlly figure that they will all be good as long as they are
from the same lot/nearly the same lot.

>From talking to my sources at RCA (who invented the thing ;))  they say that
typically these days they will be really close to being matched as is since
they come off the same line and the manufacturing is much more accurate than
it was when they were introduced.

Heck, I hear ppl at work talking about seperate transistors matching out of
the box these days, so, its only stands to follow that units in the same
package would be really close.

Rob

----- Original Message -----
From: "Colin Fraser" <colinf at ntlworld.com>
To: "'Jim Patchell'" <patchell at silcom.com>
Cc: <synth-diy at node12b53.a2000.nl>
Sent: Monday, March 12, 2001 6:55 PM
Subject: RE: [sdiy] Speaking of OTAs/Matched Pairs????


> > -----Original Message-----
> > From: owner-synth-diy at node12b53.a2000.nl
> > [mailto:owner-synth-diy at node12b53.a2000.nl]On Behalf Of Jim Patchell
> > Sent: 12 March 2001 21:08
> > To: Kevin Peek
> > Cc: synth-diy at node12b53.a2000.nl
> > Subject: Re: [sdiy] Speaking of OTAs/Matched Pairs????
> >
> >     Arrays like the CA3046 are actaully pretty well matched.  Even on
> > the speck sheet, typical matching is 500uV (worst case is 5mV).  The
> > CA3086 does not garentee any matching (it is the same part, just not
> > tested).
>
> If the CA3086 is just a 3046 without matching guaranteed, isn't that
likely
> to mean that matching is guaranteed to be outside the spec for the '46 ?
>
> i.e. 3086 = 3046 that failed the matching test
>
> Colin f
>