Path: ...!eternal-september.org!feeder2.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Cursitor Doom <cd999666@notformail.com>
Newsgroups: sci.electronics.design
Subject: Re: Curve Tracers
Date: Wed, 20 Nov 2024 18:00:34 -0000 (UTC)
Organization: A noiseless patient Spider
Lines: 57
Message-ID: <vhl842$70jp$1@dont-email.me>
References: <vhkhcg$2ip8$2@dont-email.me>
	<o50sjjpmjmrdbvs9po8inhcdn961a0923g@4ax.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Injection-Date: Wed, 20 Nov 2024 19:00:36 +0100 (CET)
Injection-Info: dont-email.me; posting-host="05ffc764ebf22792ae96d10cef977184";
	logging-data="230009"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX18yGD3uG8gQmcUbAf8YYNJq+o+leCFTkp8="
User-Agent: Pan/0.149 (Bellevue; 4c157ba)
Cancel-Lock: sha1:Z/1pb+Wvkt91ehd1og7HqG11BYY=
Bytes: 3502

On Wed, 20 Nov 2024 07:40:08 -0800, john larkin wrote:

> On Wed, 20 Nov 2024 11:32:32 -0000 (UTC), Cursitor Doom
> <cd999666@notformail.com> wrote:
> 
>>Gentlemen,
>>
>>Curve tracers reveal useful info about the dynamic characteristics of
>>semiconductors and make designing for same much more predictable and
>>dependable than relying on spice models and simulation alone. But
>>they're typically rare beasts and expensive to come by and boat anchor
>>varieties are seriously heavy and bulky.
>> I think therefore that a curve tracer would make an excellent project,
>>using the X&Y inputs of a scope as the display. Has anyone here
>>attempted this? I'd be interested to know what the main challenges are
>>likely to be.
>>
>>-CD
> 
> I've considered it. It would be much more valuable if it traced
> capacitances too.

A VNA is best for that. I've been using my big HP one this afternoon to 
label some random bunches of caps and inductors whose values I couldn't 
read. Trouble is, there doesn't seem to be much agreement on what 
frequency to test these devices at. I've heard 100khz, 1Mhz, 10Mhz and 
100Mhz mentioned from different sources. I'm pretty sure the 100Mhz 
testing is for RF specified devices only, though. From what I've seen, 
there's no reliable alternative to testing each batch for oneself, because 
(certainly with ancient NOS leaded (as in non-SMD) stock) you have no idea 
what frequency the factory tested them at and different manufacturers in 
different countries at different times used different methods!

> The display should be on a computer and the data archived.

For people such as yourself, certainly. For me as a hobbyist, a CRT is 
fine.

> I often just set up a breadboard and test parts. Last week I blew out
> some power mosfet gates. Some have internal zeners; this one doesn't.

I can kill a MOSFET at 50' just by looking at it (certainly in winter). In 
fact I shock myself from all the static I build up and it ain't funny. 
Having dry skin is great for when I touch HV by accident (happens quite a 
lot) but the flip side is I accumulate and hold static charge like no one 
else I know.

> Most mosfets drain avalanche, but the voltage is never specified, in
> fact deliberately hidden. I have to measure that.  A good curve tracer
> should measure that too.
> 
> Currents should go down to picoamps.
> 
> The fact that there aren't many curve tracers for sale suggests a small
> market.

But a market, though small perhaps, there must be.