Maker Pro
Maker Pro

Problem with Tektronix 475 "ADD" function

I own a Tektronix 475 that works well, but requires some realignment.
So far, I've been able to recalibrate the channel preamps and adjust
the input attenuators. Individually, each channel now produces an
accurate trace; however, I have a troublesome problem that I can't
quite figure out.

If I feed the same signal to both channels, invert Channel 2 (press
"INVERT") and then add it to Channel 1 (press "ADD"), I get a
flat trace that's offset by nearly -6 volts. The flat trace is good,
but no matter what I try, I can't get rid of this offset.

I've traced through the block diagrams and schematics and at this
point have pretty much ruled out any problem with the preamp board. I
fear the problem is down stream, probably in the signal combination
circuitry prior to the final vertical amplifier. Is integrated circuit
U370 the source of the problem? Is there an adjustment I've missed?

My eyes have grown weary tracing schematics, so I could really use some
sage wisdom here to save my eyesight.

-Dave Drumheller
 
This has got to be a DC offset in the inverter stage itself, that is
even though it night not be a seperate stage, whatever stage does it.

This offset is being compensated for by the front panel position
control, so after to get that straightened out you'll have to reset
it's range of course.

If the problem is indeed in a vertical amp IC, you probably can't get
it, but you may be able to swap it with the channel that doesn't
invert. You would actually be adjusting out a problem, but if it works
it works.

There is a guy in this NG who can probably give you loads of more
specific info named Jim Yanik. Watch the thread, if you get a reply
from him it will probably be quite enlightening. You might also search
the group for his name and find something useful.

Another piece of information that would be helpful is whether the 6
volt offset is at a certain scale, and reads 6 volts at different gain
settings, or is it simply at a certain position on the screen, i.e., if
you change the range does indicated offset or the trace position remain
constant ? This will indicate whether it's before or after the
attenuator. Also does the trace move when you use the variable control,
that is bringing it out of the calibrated stop ?

If you can't solve this problem without parts and can't live with it
there are surplus places that might have one with serious screen burn
or a bad CRT or something. One such place is Electronic Surplus in
Cleveland Ohio. If they were open 24/7 guys like us could spend days
there looking at all their cool stuff. If you talk to the right person
there they will probably know if they have what you need. They also
used to have quite a selection of Tek plugins, but IIRC the 475 doesn't
use them. If it did it would simply be a matter of sliding in another
vertical amp. I don't have their number handy and at the moment I dunno
if they got a website. You may be able to get some NLA parts via mail
from them but it's nowhere near as much fun as being there. If you're
near Ohio it might be worth a visit if you have the time. If near
Florida you might want to check out a place called Skycraft in the
Orlando area. I would definitely call first, but if they have what you
need a visit would be interesting if it is feasible for you. In either
place I suggest you take your gold card, you will find some goodies
believe me.

JURB
 
J

Jim Yanik

Jan 1, 1970
0
[email protected] wrote in
This has got to be a DC offset in the inverter stage itself, that is
even though it night not be a seperate stage, whatever stage does it.

That's my opinion.It could be a invert balance adjustment,or a BNC that is
generating a DC offset(I've seen a few of these).
I doubt it's in the channel switch itself,that's just a fancy diode
gate,IIRC.
This offset is being compensated for by the front panel position
control, so after to get that straightened out you'll have to reset
it's range of course.

If the problem is indeed in a vertical amp IC, you probably can't get
it, but you may be able to swap it with the channel that doesn't
invert. You would actually be adjusting out a problem, but if it works
it works.

IIRC,the 475 vert preamp uses those TEK-made spider ICs,155-0078-xx,and
their gain and current supply resistors do go out of spec,they can cause
gain,offset,or compression problems.I'm working strictly from
memory,though,I have no access to schematics anymore.
There is a guy in this NG who can probably give you loads of more
specific info named Jim Yanik. Watch the thread, if you get a reply
from him it will probably be quite enlightening. You might also search
the group for his name and find something useful.

Thanks for the recognition! (He's emailed me already.)
Another piece of information that would be helpful is whether the 6
volt offset is at a certain scale, and reads 6 volts at different gain
settings, or is it simply at a certain position on the screen, i.e., if
you change the range does indicated offset or the trace position remain
constant ? This will indicate whether it's before or after the
attenuator. Also does the trace move when you use the variable control,
that is bringing it out of the calibrated stop ?

When working on DC offsets,I would not use any input signals.
If you can't solve this problem without parts and can't live with it
there are surplus places that might have one with serious screen burn
or a bad CRT or something. One such place is Electronic Surplus in
Cleveland Ohio. If they were open 24/7 guys like us could spend days
there looking at all their cool stuff. If you talk to the right person
there they will probably know if they have what you need. They also
used to have quite a selection of Tek plugins, but IIRC the 475 doesn't
use them. If it did it would simply be a matter of sliding in another
vertical amp. I don't have their number handy and at the moment I dunno
if they got a website. You may be able to get some NLA parts via mail
from them but it's nowhere near as much fun as being there. If you're
near Ohio it might be worth a visit if you have the time. If near
Florida you might want to check out a place called Skycraft in the
Orlando area.

You might run into me there! I love that place.
Their stock of TEK instruments has declined drastically lately,though.
I wonder if the local test instrument refurb shop bought them all up?
 
Jim and JURB,

Ok, I've checked the value of the offset that appears when using the
"ADD" function and it is a constant one division (5 subdivisions)
regardless of the vertical scale deflection setting. (Of course, both
channels are always set to the same vertical deflection). So does this
suggest that the problem is in the channel preamps or a following
stage?

Also, I poked around the vertical switching circuit on the input side
to U370 (monolithic switch) and found all DC voltages to be consistent
with the schematic, thus suggesting that there's nothing wrong with it.

Suggestions?

Thanks. This is very helpful.

Best,

-Dave
 
J

Jim Yanik

Jan 1, 1970
0
[email protected] wrote in
Jim and JURB,

Ok, I've checked the value of the offset that appears when using the
"ADD" function and it is a constant one division (5 subdivisions)
regardless of the vertical scale deflection setting. (Of course, both
channels are always set to the same vertical deflection). So does this
suggest that the problem is in the channel preamps or a following
stage?

I'd say that the problem would be in the preamps.
I can't recall if there's a spec on offset for the "add" function.
The "invert" function for Ch2 is done by one of those spider IC's I
mentioned,and IIRC,there's a DC balance adj.for the inverted Ch2.(I could
be wrong;no schematics for reference)

It might be that the Ch2.invert-add function is not spec'd for DC offset.
You would have to center the trace with no signal input,then make your
measurements.
 
D; regarding your statement:
Also, I poked around the vertical switching circuit on the input side
to U370 (monolithic switch) and found all DC voltages to be consistent
with the schematic, thus suggesting that there's nothing wrong with it.

If current setting or limiting resistors go out of spec, rather then
open, you may well see no change in voltage readings. Even if they go
completely open you still might not see any significant change. What's
more any change you see might well be less than the device to device
tolerance. A voltmeter is not the solution to everything. Even it's
drop from a power supply isn't an accurte indicator whether the porper
current is flowing or not.

There is a definitive way to check this without ridiculous test
equipment, but it involves math and alot of thinking. Thus I'll go into
it another time. It's late and I gotta work.

JURB
 
J

Jim Yanik

Jan 1, 1970
0
[email protected] wrote in
D; regarding your statement:


If current setting or limiting resistors go out of spec, rather then
open, you may well see no change in voltage readings. Even if they go
completely open you still might not see any significant change. What's
more any change you see might well be less than the device to device
tolerance. A voltmeter is not the solution to everything. Even it's
drop from a power supply isn't an accurte indicator whether the porper
current is flowing or not.

There is a definitive way to check this without ridiculous test
equipment, but it involves math and alot of thinking. Thus I'll go into
it another time. It's late and I gotta work.

JURB

I don't trust those voltages TEK supplied on their schematics,anyways.
I often wrote down my own readings on my personal set of manuals,along with
part numbers of some components.
I often wish I had had storage space for them when I left TEK,but alas...
 
Top