The AM Forum

Greetings, All!

I have a 1970's vintage Dumont 190 o'scope (Navy AN/USM-281E) that I've been
using for several years; this week it has developed a bizarre fault that perhaps someone
in the AM world will recognize.   When using channel 1 (it's a two-channel, dual-trace scope),
the trace has a slight curvature near the LEFT-HAND side of the screen.  The degree of curvature
changes as the vertical input (range) switch is changed, but not uniformly (i.e., the curvature is
larger for some input ranges but does not seem to get larger as the vertical sensitivity is increased).
At higher sweep speeds (e.g., 1 u-sec/div), the curvature becomes a glitch at start of the trace that
expands into a decaying oscillation as the sweep time becomes smaller.  This looks just like
something is ringing somewhere in the circuitry.  Channel 2 also has a slight curvature to the trace,
and the ringing is there at high sweep rates, but is hardly noticeable compared to ch. 1.  I am not an
expert on o'scopes by any means, so this one really has me stumped.  Any ideas will be greatly
appreciated!

NOTE TO ALL:  I erroneously stated that the curvature was at the right-hand side of the screen, but
I meant LEFT!!  (See Above).  Guess I never did really learn which was which!  SRI.

73,
Ken  W5KFS
ex-WB6VHE

first I would check the resistor values in the vertical deflection amplifier, and check for unwanted signals on any bypass capacitors. Could a blanking signal be getting into the DC circuit via a leaky cap?  The ringing is a tough call.

Is this only with signal or without? Can you set it to dual channel, superimpose the traces, set to alternate, and use a second scope to compare the waveforms from both channels?

Ringing usually will occur with magnetic deflection used in TV's.   The resistors and caps that were affixed on the TV's yoke coil prevented ringing.   A scope uses electrostatic deflection.  The sweep waveform should be a sawtooth.  If the sawtooth rise slope was not a straight line the sweep across the screen would not be linear.  Your problem of a curvature in the trace on the right side of the screen is a (I'm thinking) hard one to figure.

Rereading ur description of the problem I would begin in the sweep circuit.  Look for the usual type problems, leaky caps, bad filter caps. 

Inject a sinewave into the vertical input.  See if the waveform is linear across the screen.

Being 4:30AM, my brain is beginning to short circuit and this is the best I can think of right now.

I'm sure more help will be posted later today.

Fred, KA2DZT

Ken,

If it were the USM 281C I could offer a lot more help since I am much more familiar with Tektronix analog scopes.

With scopes, the first thing you want to do is make sure ALL of the power supply voltages are correct and clean.  With that out of the way I would look at the channel switching circuitry since the problem is with both vertical channels but it differs somewhat between them.  With a timebase/sweep issue I would expect the problem to be the same for either channel.  Also try some different triggering modes and settings to make sure it isn't a problem with the triggering circuit/auto sweep generation.

 

Since the vertical and horizontal drive circuits in most scopes are so seperate from each other I would suspect it's something common to them both since the effect is obviously sync'ed between them. That pretty much nails it down to the power supply as the most common element.

At the beginning and ends of the trace one or the other horizontal plates is at it's extreme potential possibly loading a regulator a bit more than at mid screen. That's where I would start.

The next possibility would be a leaky blanking circuit. I've seen failures of blanking circuits where the unblanked retrace deviates significantly in the vertical. 

Just thinking out loud. Of course, if you had a service manual it's easier to check for all the proper waveforms than seek out what might be wrong, but that doesn't sound like an option for this.  I just picked up a Navy scope made by Symetrics. No docs available of course. It has a trigger problem that appears to be just a poor design. I've already put more hours into it than it's worth.

Good luck with Dumont.

Thanks to all for the suggestions!  They are all much appreciated.
I will dig into it and post any observations for your perusal.  Please see
the change in my original post....I still can't tell left from right after all
these years....or maybe I am getting senile and used to be able to tell
the diff, but not any longer!!

73!

If you shop here, you'll always know your left from your right:
http://www.thelefthand.com/   :)