Tek 2440 repair help needed

[cncjerry]

I picked up the subject scope for a great deal and it seems to be almost working.  It halts asking for self cal and external cal. I can exit out of the diagnostics and have had success getting traces.  It seems to have lost its calibration and as others have done, I figured I would replace the Dallas NVRams and battery.  I order the NVRams and sockets and wanted to see what battery it has and found it doesn't have one.  Where the battery should be there is a factory shorting wire.  So I guess this is a newer version of the scope and I can't find any newer versions of the schematics.  Any ideas on this one?  I've searched for other service manuals without luck.

Thanks

Jerry

[David Hess]

If your 2440 has the Dallas NVSRAMs, then the battery circuits were not needed and left out but the same printed circuit board was used.

The service manual from Artek Manuals shows the later Dallas NVSRAM version but really there is not much to know.  Just change the Dallas NVSRAMs and then do the external calibration procedure which is not difficult.  Or do not do the external calibration procedure and live with reduced accuracy.

[cncjerry]

David, thanks.  This all makes sense.

Are the scope ref1, ref2 traces, etc, stored in the NVrams?  They seem to work.  The only thing that doesn't are the cal settings.

I have new NVRams comings

By the way, I don't know how I ended up with two threads on the same topic.

Jerry

[David Hess]

Are the scope ref1, ref2 traces, etc, stored in the NVrams?  They seem to work.  The only thing that doesn't are the cal settings.

Yes, all of the stored data and calibration data is contained in the NVSRAMs and they comprise all of the RAM for the processor.

[cncjerry]

David, does your 2440, I assume you have one, maintain a square wave as you dial lower in time base?  Mine switches to something like a bad ramp at 200ns, I think it was.  All higher/slower time base settings are ok.

[tautech]

By the way, I don't know how I ended up with two threads on the same topic.

Report yourself to the moderators and ask them to merge the threads.
Provide a link to the other one too. 

When done I'll delete this post.

[David Hess]

David, does your 2440, I assume you have one, maintain a square wave as you dial lower in time base?  Mine switches to something like a bad ramp at 200ns, I think it was.  All higher/slower time base settings are ok.

Two things:

The calibrator output tracks the sweep speed so at faster sweep speeds, the calibrator produces a higher frequency square wave which is not very square at the highest frequencies.

The jump at 200ns is caused by changing acquisition modes to support equivalent time sampling.  This could mean:

1. The default calibration needs to be reloaded and the automatic standard internal calibration done.  There is a procedure in the service manual for this.

2. The CCD calibration needs to be done.  This should not be necessary.

[cncjerry]

David, thanks, I understand the functioning of the calibrator circuitry.  Today I took the cover off and probed with another 1G+ scope on the calibrator and I see the same thing on that scope.  I see a decent square wave up until I hit 200ns and then instead of even a semblance of a square wave (bandwidth limited) I see more like a series of "U"s like UUUUUUU.  Here's a picture from another site that has the same problem.

Just wondering if this is normal or not for this scope?  The square wave looks ok until the junction of pins 8 and 6 on U831C.  Then I get almost the exact same display (on a separate scope) as below.  Am I fixing a design flaw?  A wentzel squaring circuit would have corrected this.

Thanks again.  I Have the new NVRams coming and I plan to replace the one that holds the calibration as the reference memories seem ok right now.

Jerry


P.s. I took this image from another site.  I see the same thing.

[David Hess]

For the calibrator output that looks normal to me.  I am not sure why Tektronix allowed it to operate at such a high frequency but it does serve as a frequency reference.

[cncjerry]

I guess all the 2440s are that way?

I was thinking of modifying the circuit, which can be done pretty simply as a little add on to the side of the scope where that board is located.  I could just copy the input from the TAPR TADD-2 which uses  a Wenzel(sic) sine to square wave followed buy a driver.  This would make a near perfect calibration signal at 50 ohms up to and way past their 5Mhz upper end.

One thing I noticed though was that the signal starts to jump once the scope is warm.  It is only the calibrator, not the timebase clock which drives to calibrator, that is jumping, almost like the scope is losing the trigger.  I was looking at it with another scope.

I'm trying to figure out so how you would normally catch that jump. I guess I would have to record a fairly long stretch and try to find what is dropping-out or look for a runt or something.

Thanks for the input.

Jerry

[cncjerry]

I looked at the incoming signal again, TP100 on the attached.  When the time base is switched low enough to set the calibration to 5Mhz, the signal at TP100 is pretty much a triangle.  On the incoming side of R885 (to the left of TP100), it is still a square wave.  The time constant on the filter circuit consisting of R885 (100ohms) and C883(100pf) doesn't seem like it should cause all that.  So I wonder if U831D has a problem?  Given that other scopes exhibit the same symptoms, it would be a pain to get U831 out, and who really gives a rats ass about a calibrator when I have more accurate calibration signals from Rb to Cs, I think I'll just leave it alone.

But then again, I have a hell of a time leaving anything alone that can be better!

Thoughts?

Jerry

[David Hess]

I guess all the 2440s are that way?

All of the 24xx DSOs are that way and the 24xx analog oscilloscopes are similar if not identical.

One thing I noticed though was that the signal starts to jump once the scope is warm.  It is only the calibrator, not the timebase clock which drives to calibrator, that is jumping, almost like the scope is losing the trigger.  I was looking at it with another scope.

I do not remember if the 24xx DSOs do it but the 24xx analog oscilloscopes tie the calibrator output to the sweep in such a way that jitter in the calibration signal is displayed if measured with another instrument.

I looked at the incoming signal again, TP100 on the attached.  When the time base is switched low enough to set the calibration to 5Mhz, the signal at TP100 is pretty much a triangle.  On the incoming side of R885 (to the left of TP100), it is still a square wave.  The time constant on the filter circuit consisting of R885 (100ohms) and C883(100pf) doesn't seem like it should cause all that.  So I wonder if U831D has a problem?  Given that other scopes exhibit the same symptoms, it would be a pain to get U831 out, and who really gives a rats ass about a calibrator when I have more accurate calibration signals from Rb to Cs, I think I'll just leave it alone.

But then again, I have a hell of a time leaving anything alone that can be better!

Thoughts?

The circuit was just never intended for that kind of high frequency fidelity.  The output edges are at the mercy of the edge fidelity from the clock source and the series feedback which controls the level is slow compared to the edge speed.  This is not the type of circuit to use for fast edges.

Some of the earlier Tektronix calibrators are designed in a completely different way and produce clean fast edges.

[cncjerry]

David, thanks for the clarification.

Not to bore you with this, but I took the time and loaded that circuit into LTSpice. I did the best I could with the models and it looks like, according to LTspice at least, that this circuit should have significantly higher bandwidth than what we are seeing. I see the correct output into 50 homes of .2 V and .4V into 1Meg ohm and I see some rise time and fall time increase at 5Mhz (in my model) but nothing like the distortion we see in the live circuit. 

When I probe the live CA3046 amp at for instance pin 10 and 7,  I start to see high frequency drop off adding rise time of about 40ns to the 10ns in the square wave from the time base.  I then see even more rise time (and fall) added as it goes through the next transistor so that once it hits pins 8 and 6, it is pretty much a 5Mhz triangle at that point. 

I thought I had some CA3046 chips around here and if I find one, I might breadboard the circuit to see what maybe a new version of that chip might produce.

Again, thanks for the pointers and clarification.

Jerry