Tek 2465A diff amp nonlinearity

[Walter Harley]

When I feed a 1V sine wave into both ch1 and ch2 of my Tek 2465A, and set it
to display the difference (invert ch2 and add the channels), I do not see
the expected flat line. Rather, there are some bumps.

So, I presume this means that one or both input amps are not quite linear.

My questions are:

1. How much nonlinearity should I expect, in a working and well-tuned 2465A?

2. Anybody got any pointers to where in the service manual I should look,
for diagnostics and troubleshooting of this? (The service manual I have is
a not-so-good photocopy so it's very hard to search through without at least
a head start, and I haven't spent much time with it.)

3. If I wanted to buy a used differential instrumentation amplifier, ideally
with bandwidth >= 100MHz, ground-referred common mode range >= +/-15V,
input-referred wideband noise <= 250uV and differential gain >= 20, is there
a "classic" model that I should search for? (Sort of like the 2465 'scopes
are a "classic" - something that would be easy to find used, easy to find a
manual for, and that would be pretty reliable.)

Thanks!

-walter

 

[Jorgen Lund-Nielsen]

Walter Harley wrote:

When I feed a 1V sine wave into both ch1 and ch2 of my Tek 2465A, and set it
to display the difference (invert ch2 and add the channels), I do not see
the expected flat line. Rather, there are some bumps.

So, I presume this means that one or both input amps are not quite linear.

My questions are:

1. How much nonlinearity should I expect, in a working and well-tuned 2465A?

Have you tried to play with the var.gain at both channels to match
the gain of them exactly?
At what frquency did you do this? (Phase match maybe...)
And at what V/Div of the Channels did you do this measurement?
Did you raise the input sensivity to see that behaviour and maybe
overdriven the input amplifiers? (The "cancelation" of the signals by
"inverted adding" is done after a few amplifier stages which can of
course saturate than and can do this in different matter (overdriven
Amplifiers are not specified exept the recovery time)

Just a few ideas.

Jorgen
dj0ud

 

[Walter Harley]

"Jorgen Lund-Nielsen" <jorgen.lund-nielsen@xxxxxxxxxxdesy.de> wrote in
message news:dcskj8$u2q1$1@claire.desy.de...

Have you tried to play with the var.gain at both channels to match
the gain of them exactly?

Yes - sorry, maybe I should have been more detailed. The residual is not a
sine wave, so it is not just a phase or amplitude discrepancy. It looks as
if the two channels track well over most of the input range but have
differences at several small, distinct ranges. So the residual is a flat
line with a couple of triangular spikes at irregular locations

At what frquency did you do this? (Phase match maybe...)

Something low, like 1kHz.

And at what V/Div of the Channels did you do this measurement?
Did you raise the input sensivity to see that behaviour and maybe
overdriven the input amplifiers? (The "cancelation" of the signals by
"inverted adding" is done after a few amplifier stages which can of course
saturate than and can do this in different matter (overdriven
Amplifiers are not specified exept the recovery time)

I think I did that initially and then realized it and backed off and still
saw the problem. But let me re-check and make sure that wasn't what was
happening.


Thanks,
-walter

 

[Jim Yanik]

"Walter Harley" <walterh@cafewalterNOSPAM.com> wrote in
news:eICdnfvW2PckCWzfRVn-tQ@speakeasy.net:

When I feed a 1V sine wave into both ch1 and ch2 of my Tek 2465A, and
set it to display the difference (invert ch2 and add the channels), I
do not see the expected flat line. Rather, there are some bumps.

So, I presume this means that one or both input amps are not quite
linear.

My questions are:

1. How much nonlinearity should I expect, in a working and well-tuned
2465A?

2. Anybody got any pointers to where in the service manual I should
look, for diagnostics and troubleshooting of this? (The service
manual I have is a not-so-good photocopy so it's very hard to search
through without at least a head start, and I haven't spent much time
with it.)

3. If I wanted to buy a used differential instrumentation amplifier,
ideally with bandwidth >= 100MHz, ground-referred common mode range >=
+/-15V, input-referred wideband noise <= 250uV and differential gain
= 20, is there a "classic" model that I should search for? (Sort of
like the 2465 'scopes are a "classic" - something that would be easy
to find used, easy to find a manual for, and that would be pretty
reliable.)

Thanks!

-walter

The manual should have the common mode rejection ratio specs.
The input of the 2445/65 series is not intended to be equal to a true
differential amp.
TEK still makes differential amps for their scopes,probably listed under
accessories.

--
Jim Yanik
jyanik
at
kua.net

 

[Walter Harley]

"Walter Harley" <walterh@cafewalterNOSPAM.com> wrote in message
news:B5ydnTkjisbzsW_fRVn-rg@speakeasy.net...

I think I did that initially and then realized it and backed off and still
saw the problem. But let me re-check and make sure that wasn't what was
happening.

Well, I must have been doing something funky, because now it's not doing it
at all - it's behaving just fine. The only distortion is from clipping, and
even that only when I crank the gain way up.

Thanks anyway...

 

[Walter Harley]

"Jim Yanik" <jyanik@abuse.gov.> wrote in message
news:Xns96A87A45E7B95jyanikkuanet@129.250.170.83...

The input of the 2445/65 series is not intended to be equal to a true
differential amp.
TEK still makes differential amps for their scopes,probably listed under
accessories.

I see a lot of diffamp plugins, but the only standalone diff amp I see
popping up on eBay is the P6046 differential probe.

Looking at the pictures, I don't quite understand how that probe is supposed
to work - it looks sort of like the spacing between the two input pins is
fixed?? I was envisioning something that would have inputs for two separate
probes, rather than something that had a single probe with two pins. I want
to be able to do things like comparing signals at various points on the left
and right channel of a stereo amp.

Thanks,
-walter

 

[Jim Yanik]

"Walter Harley" <walterh@cafewalterNOSPAM.com> wrote in
news:9L2dnUK97q3LD27fRVn-sg@speakeasy.net:

"Jim Yanik" <jyanik@abuse.gov.> wrote in message
news:Xns96A87A45E7B95jyanikkuanet@129.250.170.83...
[...]
The input of the 2445/65 series is not intended to be equal to a true
differential amp.
TEK still makes differential amps for their scopes,probably listed
under accessories.

I see a lot of diffamp plugins, but the only standalone diff amp I see
popping up on eBay is the P6046 differential probe.

look at TEK's website;www.tek.com.

Looking at the pictures, I don't quite understand how that probe is
supposed to work - it looks sort of like the spacing between the two
input pins is fixed??

Yes.

I was envisioning something that would have
inputs for two separate probes, rather than something that had a
single probe with two pins. I want to be able to do things like
comparing signals at various points on the left and right channel of a
stereo amp.

I do not see what use a differential probe would be for that use? What
useful info do you expect to get from it? No stereo amp will be matched any
better than the ordinary non-differential scope inputs inverted and added.
And you would have to feed the -same- signal to both channels.

Thanks,
-walter

BTW,some other company(not TEK) used to make a power supply/"mainframe" to
allow one to use a 7A22 or 7A13 differential plug-in as a standalone
instrument,with a single-ended 50ohm output BNC to feed a regular scope
input.I can't recall the specific company,though.

--
Jim Yanik
jyanik
at
kua.net

 

[tekamn]

Walter,

as Jim mentioned: The P6046 probe alone won'T help much. This probe is
designedto work with a diff amp unit like the 7A13 or the 7A22 or
similar. The 2465A scope definetely has no such input. So if you really
need to do differential masurements: A 7A13 with a 7313/7603/7613
mainframe will get you more out for the solution.With the siganl
through output of the 7xxx-scope you can still connect to your 2465A -
but I guess that's only necessary if you wantthe cursor meas to do ;-)

hth,
Andreas

 

[Jim Yanik]

"tekamn" <and7@bigfoot.com> wrote in
news:1123486184.688254.324550@g47g2000cwa.googlegroups.com:

Walter,

as Jim mentioned: The P6046 probe alone won'T help much.

This probe is
designedto work with a diff amp unit like the 7A13 or the 7A22 or
similar.

No,it's not.
It can work with any single-ended scope input.

The 2465A scope definetely has no such input. So if you really
need to do differential masurements: A 7A13 with a 7313/7603/7613
mainframe will get you more out for the solution.With the siganl
through output of the 7xxx-scope you can still connect to your 2465A -
but I guess that's only necessary if you wantthe cursor meas to do ;-)

hth,
Andreas

--
Jim Yanik
jyanik
at
kua.net