Need help with HP 1725A resurrection

[RacerX]

Hi folks:

After a long absence from electronics work I am easing my way back in and am trying to get my old HP 1725A scope going again.

I have had this one for a long time and even though it is pretty old now it is a great scope and 275Mhz is nothing to sneeze at.

The scope powers up and the controls seem to work in the expected manner.    I am able to get both the A and B traces on the screen and move them around with the position controls as expected.    Without question the controls were a bit dirty, but seemed to clean up pretty quickly and seem stable now.   I do have plans to spend some time with contact cleaner and contact grease as soon as I can manage it.

However, I am having some problems that seem to be beyond the normal dirty switch contacts and such.

If I use the scope's calibration signal (3Vpp @1khz) I get this on the screen at .2ms horizontal:



The trace is folded over such that the rising portions of the trace are BELOW the falling portions.    Hardly useful!

If I change to .1 ms horizontal we get this:



What a mess that is...     

Has anyone seen anything like this before?     It does the exact same thing on both the A and B channels.

An interesting data point is changing the sweep to 1 second or slower results in the two split parts of the trace sweeping across the screen and then sweeping back in the other direction at the same intensity.   This behavior is not visible at sweeps faster than 1 second.   Obviously the 1khz signal has no hope of displaying properly at these sweep rates, but I don't think it is supposed to be sweeping backwards!

I have already popped the hood and checked all the voltages and they are within spec according to the service manual.   Before I get too deep I wanted to post here in hopes that someone has seen behavior like this and can give a nudge in the right direction...   Feels like it has to be something fairly simple...   I hope!


Thanks!
Jim

[g0rsq]

first trace looks like a "Compensation" error.

Where is the 0V level displayed on the scope in relation to the trace pictured?

What probes are you using to connect to the calibration output, and what settings are on the scope?
You should be using high impedence probes (1Mohm or greater) and DC 1Mohm input on scope. Not 50ohm input.

You should then be able to adjust the probe compensation to get a good square wave without the overshoot.
 
The second image looks like two traces displayed at once. This is most likely a triggering issue, and you should be able to eliminate the second trace by careful use of the triggering settings. Use "normal" not auto trigger, and adjust the level untill you see a stabilised trace.

The HP 1725 is a nice scope, not as good as the TEK's of this era, but certainly very capable.
However it is also very complicated to set up for a beginner, as it has many odd functions such as dual timebase etc.

The image you describe sounds like a "flyback" problem, and I can notice it on my 1725 in slow scan speeds, but not as obviouse as you describe.

Mine needs a good service, especialy in the HT circuits as the intensity is very poor.

Of course all the issues you describe could be faults, but without knowing how you are making the measurements, and the scope setings it is impossible to be certain.

Peter

[RacerX]

Hi Peter:

The probe is a Tek P6104, 10Mohm

I am using the DC input, timebase at .2ms, normal triggering, sweep mode MAIN.   No other switches pressed on the unit.  Adjusting the compensation on the probe has little/no effect.

In the pictures I attached I had not put the trace on 0V before taking the picture.  But if you align the trace on 0V before probing the calibration post then 0V appears to be in the proper location, which is the horizontal portion of the upper one of the split traces.   The lower one is way out of whack.

With the probe not on the calibration post, there is one bright dot at the far left of the screen.   With it on the post, but not sweeping (i.e., adjust the trigger level so there is no sweep) there are TWO bright dots at the left of the screen.    When I increase the trigger level to cause a sweep, it looks like the pictures in the initial post.

These are the things leading me to believe there is a fault in the unit...

[KJDS]

There'll be a little screwhead visible on the probe near where it connects to the scope. The screw adjusts the compensation.

Twiddle it and you'll lose the peaks shown in the first trace.

[RacerX]

There'll be a little screwhead visible on the probe near where it connects to the scope. The screw adjusts the compensation.

Twiddle it and you'll lose the peaks shown in the first trace.

Already tried that...  it had little to no effect

[dfmischler]

Just take a piece of wire small enough to fit in the center of the BNC connector, stick it in there and touch the other end to the calibration output.  This will allow you to decide if the weirdness is in the scope or the probe.  At 1KHz it will look fine on the scope if the scope is OK.  Of course, if you still see the same thing this won't tell you if the problem is in the calibrator output or the scope input.

[luke.dft]

Also try out an external calibration signal source.

[g0rsq]

The description of the dots when the trace is not triggered is not normal.

This is caused by the gate control circuit not blanking the beam if it hasnt been triggered.

You also described seeing the trace on the return sweep, which should also be blanked, again suggesting problems with the gate control circuit.

I noticed the same behaviour on my 1725a, but was able to adjust it out. If I turn the beam intensity up I can still see these effects so I think there is still a problem that I must look at again.

To check the input channels for compensation faults you will need to connect the inputs up to a function generator (or pulse generator) with a 50 ohm output. Use 50 ohm coax cables and connect to the inputs, which are turned to DC 50 ohm input. This should eliminate probe compensation issues,

If everything is correct, then you should see no overshoot in the waveform. If you still see the distorted waveform, then something else is amiss in the Y-axis circuits.

I dont quite follow your description of the waveform, relative to the zero volt line.
If I recall correctly the calibrator output should be giving a -ve 3V pulse output, so should not be showing any trace above the zero volt line.

Peter

[RacerX]

The description of the dots when the trace is not triggered is not normal.

This is caused by the gate control circuit not blanking the beam if it hasnt been triggered.

You also described seeing the trace on the return sweep, which should also be blanked, again suggesting problems with the gate control circuit.

I noticed the same behaviour on my 1725a, but was able to adjust it out. If I turn the beam intensity up I can still see these effects so I think there is still a problem that I must look at again.

Will take a look at the gate circuit.

What did you do to adjust it out on your scope?   

I dont quite follow your description of the waveform, relative to the zero volt line.
If I recall correctly the calibrator output should be giving a -ve 3V pulse output, so should not be showing any trace above the zero volt line.

You are correct, it is -3v so if the base line is centered on 0v it should trace below the line, not above.    And, well you see in the pictures what happens on this scope....

[RacerX]

OK I am feeling a bit silly now...   seems the probe is bad, and in fact, has failed completely at this point!    I get nothing on the scope with the probe connected so it now resides in the trash bin.

The image below was made using a lead from a multimeter instead of the probe:



Looks pretty good, with it set to 1 volt per division the deflection is 3 volts under the 0v line as expected and I am seeing the expected number of pulses on the screen at what looks like in the ballpark of the proper width.

EXCEPT!   we now have NO vertical traces.   ugh.

Good news is the double tracing is gone, as is the double dots on the left side.   So I guess that mess was a symptom of the failing probe but seems we may still have a fault in the scope with the missing vertical traces.

I will have new probes in my hands soon.

[ahnuts72]

They aren't there because the beam is moving to fast in a vertical
direction to leave a trail in the phosphorous.
My 475a looks pretty much the same on a really fast rising signal.

[dfmischler]

EXCEPT!   we now have NO vertical traces.   ugh.

It looks fine to me.  Just crank up the sweep speed and you will be able to see the signal rise (or fall) at some point.

[RacerX]

It looks fine to me.  Just crank up the sweep speed and you will be able to see the signal rise (or fall) at some point.

I would agree except increasing the sweep speed has no effect.   Up until the point where a single pulse completely fills the screen there is still no vertical trace.    Increasing the volts per division (less vertical distance for the beam to move) does not help either.

This makes sense if the signal rise and fall is really that fast...   

I will assume that the basic fundamentals of the scope are working at this point and focus on cleaning the contacts on the controls since some of them still seem a little noisy.     Will re-visit the troubleshooting if issues remain after that.

Thanks everyone for the help so far!

[dfmischler]

Mine's Tek, not HP, but the rise time on my old scope's calibrator signal is on the order of 1 microsecond.  You might need to turn up the intensity quite a bit, too.