11301 - basic analog scope troubleshooting help needed

[cncjerry]

Interesting, TerraHertz, it might take me a day or two to set that up as my other analog scopes are 3 floors away, but that is an cool test.

In the pictures above, I had two pretty nicely locked displays that I couldn't repeat without a lot of work. Actually gave up earlier today. So I think it has a nasty triggering problem.  It shouldn't be amplitude as the vertical amps have been tested in my DSA602.  The backplane though could be screwing with the amplitude but it looks OK on the scope.

Thanks for the input.

[cncjerry]

OK, so how do you figure this one out?  Setting aside some of the autoset calibration issues, I can now reliably trigger a signal and get rid of that jitter.   The main trigger light is on when it is jittery and also on when I tweak it very carefully so I can't figure it out.  The scope has at least 3 levels of trigger sensitivity plus AC, DC, HF reject, LF reject, etc. etc.  It is pretty complete.

The pictures don't do it justice with the blooming,  especially on the 7603 equivalent picture, and I agree it still has a problem but it could be all calibration.   Of all the pictures, the best on-screen display is with the 7603.  It is pretty amazing how good it looks as I used a blue translucent custom fit filter since it came with an RF shield I had to pitch. 

I'm not suggesting it doesn't have a problem, but for what I might use it for in the future twiddling with the trigger is tolerable.

I know I'm not the first person that wanted the schematics for the 11000 series.

btw: This is a calibration signal at 1Mhz listed as 'fast' on the 11301 calibration menu.  The 7603 and 3054 were on a longer cable than the 11301.

[TerraHertz]

By the way, why are those test 'square' waves so terrible? Are you running a source cal output via coax directly to the scope channel input BNC, without a terminating 50ohm pad? Thus getting lots of reflections in the coax. All that ringing can cause triggering instability if you aren't careful with trigger mode and level.

This makes me wonder if maybe some of your test setups before were using cal signal paths with poor shielding/grounding/termination, and the ringing plus some external noise source (RF in the lab?) adding HF noise, caused triggering jitter.

If you are used to using low BW scopes (100MHz) and now trying a 500MHz one, part of the problem may be that you aren't used to the extra weirdness that happens at higher frequencies.

With edges that have a lot of ringing, you should be using HF DC trig mode, ensure the trigger edge is on screen, and trigger at a point on the edge that is reasonably free of timing ambiguity. A nice clean rising/falling part of the edge.

[StillTrying]

I don't think there's enough evidence to say it's a triggering problem, and we won't know until you get some very clean test signals.

I'm sure the strange dots on the falling edge in #17 are due to the strange ringing shown in #26.

[cncjerry]

Most of my equipment goes to 3Ghz+ with my analyzers up to 20Ghz and I've worked with RF for almost 50yrs.  These signals are what I have handy as I moved the scope to a place where I can work on it but I moved most of my test equipment out of my shop.  So I am using the square waves from the calibrator on the 11301 which I think is 450ohms.  The cable going to the 7603 and the 3054 is 75ohms and pretty long, maybe 4 meters, as the 7603 is far away.

This signal is called a fast square wave on the 11301 menu opposed to the other signal on the prior thread.  I have a feeling though no proof that the other signals are DAC generated consequently you can see the steps.  I am pretty sure it is the same on m DSA602 where you can see the steps as well.  Some of my Tek scopes can't drive a 50ohm load from the calibrator so I've been keeping the scopes in 1Meg or higher impedance, just using cables.    But I've tried looking at signals with proper probes and connectors with the same results.  I have an issue with strong FM signals that always show up around 100.3Mhz but the jitter on the 11301 cant be measured with the 11301 and I don't see it (the jitter) on any of my other scopes though I can see the 100.3Mhz on the 3054 if I zoom way in.

The only way to get a clean trace out of the 11301 is to set the triggering right on the hairy upper or lower edge which isn't always correct trigger setting and the Autoset sets it higher or lower than zero when in AC and that produces a jittery trace.

As sson as I get some coffee in me I will run a sign wave into it from one of my generators to see if I can get a sine wave to lock.  the other isse is that it isn't completing the Enhanced Accuracy calibration and throwing an error "Main amplifier failed".  This could be related to the overall brightness of the signal as it uses photo diodes to align signals to the graticule lines in EA mode.

The EA problem could be overall calibration as well.  This showed up after a lot of messing around with the CRT alignment which is still inconsistent.  When the scope powers on, I have to reset the astig, brightness and focus for the character displays which should be stored in EEPROM which I think is rotten as it isnt holding.  The battery looks and tests fresh.  I've looked for a typical EEPROM to replace without luck.  According to the manual, a lot of the scope trace signals are set by a DAC on the scope logic board.  So the issue could be no calibration data on a scope that is more complex than the typical analog scope. 

I'll get better signals up a little later.

Thanks for sticking with me.

Jerry

[Johnny10]

On the DSA 602 I believe they self-calibrate isn't that the purpose of enhanced accuracy?
Not sure what happens with the 11301.

I have seen choppy captured signals on the 7000 series from the digitizer which are removed with the averaging function.

[StillTrying]

As a random thought, are we sure it's not something to do with the 11301's Y delay line.
Does CH2 show the same ringing on fast edges.

[cncjerry]

I drug my HP 8648C signal generator (not my best, but somewhat portable) down to the workshop and the results are below.  I am starting to wonder, and think that I am pretty sure, that the basic scope doesn't have a problem other than the calibration data.  If you look at the first trace it has the typical wide jitter (as I've been calling it) and the second is nice and crisp depending on how the trigger is set.  I took the same signal over to my 7603 and I can get an equivalent wide trace on it depending on trigger setting so maybe since I haven't been working with a analog scope in a while, this is normal?  Can someone with an analog scope see if you can get a trigger with a wide but locked trace?

It has other problems as the calibration is way off.  I might be able to fix that by just hand editing the calibration data text file as it is pretty self explanatory, instead of going through the PC based software routines.  As far as the Enhanced Accuracy (EA) this could be just an issue with a photodiode failing.  I read where it has 4 photo diodes (at least) in the corners and I think it uses them to time the signal against the graticule.  It could also be doing it with the touch screen detectors, no idea really, but during the process it puts a trace on the screen that moves back and forth over the graticule line.  I have a feeling that the focus is off for the test lines as they are really fat and that is why the scope isnt finding the dip when it passes over the graticule?  all speculation at this point.

Autoset brightness is off as though it completes successfully, there is no display until I crank up the brightness (which is a digital encoder).  I remember seeing a calibration routine for it.  But the Autoset displays the jittery, wide trace, not the sharper trace as when I tweak the calibration.

So now I'm not sure.  Most people would be happy with the lower trace.   I went down the trace brightness path as depending on the brightness setting you can get a wide trace that looks exactly like the jitter trace and since the scope had all kinds of focus and brightness issues when I got it, I thought it was all related to brightness.

So now I don't know what is going on.  Does it have a problem or is it all operator?  (I'm fine with the latter accusation).

If you play with the trigger on your analog scope, can you get a wide trace that is locked?  Is this typical?  Another thought, is since I am setting the trigger pretty precisely, is it possible it is on one side or the other of the jitter trace so it is only picking up some of the signal traces and not all of them?  Think of the setting only picking up the peak or trough of a wave where the overall envelope would be finer?  The trigger when in fine mode (little button by the knob) has about 3 - 4 settings where the trace is locked in a thinner display.  The display goes from real dim (locked and thin) to bright and thin to bright and wide depending on how the trigger is set. 

Thanks,

Jerry

[cncjerry]

All the channels have the same issue.  By the way, I was wondering what that loop of cable was inside the frame.  It looks like 100ft of shielded cable in a nice loop.  I figured it was a delay line of some type.  I was going to scavenge that as well as all the connectors if I decided to junk the scope.

[StillTrying]

On your last 2 photos the triggering levels seem to be +/-2.7 Divs, right on very the peaks of the sine wave, I'd expect an analogue or digital display to be jittery with the trigger there, because a very small amount of noise on the signal = a large difference in time.
I always try to put the trigger level on the fastest moving part of a waveform, = center for a sine to get the least jitter.

[WastelandTek]

All the channels have the same issue.  By the way, I was wondering what that loop of cable was inside the frame.  It looks like 100ft of shielded cable in a nice loop.  I figured it was a delay line of some type.  I was going to scavenge that as well as all the connectors if I decided to junk the scope.

delay line

[cncjerry]

So I think the autoset is setting the trigger closer to the peak vs where I set it is more in the signal so maybe it is ok.

I went back and tried to look at a slower signal and it is pretty tough to get it to trigger on a slow sine wave.  I think I can get it to trigger on a slow square wave because they have high freq. components.   On a slow sine wave it just jitters all over the place.

There's a 11301 for sale on ebay for a reasonable price and I am thinking of purchasing it.  I would really like to have a 400Mhz analog scope, been looking for one for a while.

Anyway, I wish they had schematics on this beast.  I just don't have enough doc to figure this out.

[cncjerry]

Well, I am officially giving up until by chance I either find another one or schematics.  I found that yes, I can get a decent display but when you lock it there, the counter reads garbage.  Hitting autoset, though there is no display, the counter reads the correct frequency.  I know there could be separate problems, could all be just calibration, but I think by playing with it more I am just optimizing around the base problem.

As far as calibration, you need very specialized signal normalizers; time markers and levelers; and a host of other equipment that I doubt even Tektronix has at this late date.  I was able to locate the calibration software and believe it or not, the proper rs232 cable.  I'm able to run tests, load and save the calibration, etc.  I tried loading the default factory calibration data but other than changing some measurements, it had no impact on the base triggering (or whatever) problem.

I'm going to keep an eye out for another one and use the amps in my DSA602 as planned.  It's too bad as I think these were pretty capable lab-grade bench scopes in their prime.

Thanks for all the pointers along the way.  By the way, I was thinking about posting some pictures from my DSA602 if anyone is interested.  If you don't know, that machine takes the same vertical amps.  The 602 is an interesting scope.  It seems to have pretty high precision capabilities for its age, lots of match functions including convolution, FFT functions with the inverse, filtering, etc.

Jerry

[StillTrying]

"Well, I am officially giving up until by chance I either find another one or schematics."

I wouldn't blame you, I hate it without schematics even on something much simpler than this.

According to the wiki the 11302 is nearly identical, no schematics in the manual, but some board diagrams.
http://w140.com/tekwiki/wiki/11302

[cncjerry]

Yes, I have the 302 manual and used it prior to purchasing ($12) the 11301 service manual.  No real help.  I might go back and try that XY idea someone had on page 1 just for the heck of it.  I might also be able to sell the software in this one as it is the latest on 8 or so ROM chips.

Setting aside the calibration (which is tough to do) you would think I should be able to trace the signal into the functional blocks.  I guess I can pull the cable coming out of the delay line to see if it is stable.  That would be pretty easy.  This would tell me if the backplane (where the vertical amps are connected) is OK.  I've probed around on the Timebase board and at one point I thought I saw an issue with the 1T DOCXO as when I unplugged it, the horizontal ramp (but not the trace) became a lot more jittery.  Then I started thinking that if that was the case, then maybe the 1T feature is bad and I can just feed it a stable 10Mhz signal easy enough to see if that cleared up the issue.  But then I found that playing with the triggering gave me a decent trace.  But now I am pretty sure that the jitter reduced trace is related to tracing only a smaller number of repetitive waveforms resulting in a more stable main trace.  This would also reduce the counter value, which I am seeing, so that's when I decided to give up.

It's not like I need the beast but it is hard to put down!

[StillTrying]

But now I am pretty sure that the jitter reduced trace is related to tracing only a smaller number of repetitive waveforms resulting in a more stable main trace.  This would also reduce the counter value, which I am seeing, so that's when I decided to give up.

That sounds very much like the trigger level is still on the very edge/peak of the waveform. Can you get the trigger DC coupled and near the center on the waveform. I can reproduce a fat 100kHz sine by triggering at its peaks.

It's not like I need the beast but it is hard to put down!

It would be nice to get working, don't buy another one!

[TerraHertz]

Could the trigger circuit have lost its DC path, leaving only some HF AC path? So the actual trigger comparator sees nothing but some spikes derived from fast edges in the signal?

That would result in symptoms like being able to (sort of) trigger solidly on square waves, but poorly on sine waves.

Not having the schematic is a real bitch.

[cncjerry]

I'm trying to think on what side of the delay line is the trigger taken, according to the wiki, I think it is before the delay line:

The purpose of delay lines in oscilloscopes is to allow observation of the leading edge of the trigger event. In the vertical signal path, before the delay line, there is typically a trigger pick-off which supplies an undelayed copy of the vertical signal to the the trigger and sweep circuitry. Trigger and sweep circuitry need about 60ns to react when presented with the trigger event. Without a delay line, the trigger event would already have come and gone before the scope can trigger and sweep. By sending the input signal through a delay line, the scope will have triggered and begun sweeping by the time the trigger event emerges from the delay line. Thus, the trigger event is drawn on the screen where the operator can see it, photograph it, or record it by other means.

So if I look at the signal at the input to the coil of wire and maybe some of the amplifiers around there, maybe I can find something that looks funky?  The delay line is easy to find; the problem though is it is buried in the frame and I can't see the ends of the cable to get the color code.

[StillTrying]

Yes, the trigger is before the Y delay lines, it gives the X timebase a few ns to start up and a few ns of trace before the actual triggered edge is drawn on screen.

I'm going by what's on the screen in the last 2 photos, the trigger says, "Main Trig= -2.74 acdiv",
"Main Trig= 0.1 div dc" would seem better for that displayed sine!

I suppose if an earth connection had come off one end of the delay lines it might cause that fast edge ringing on both channels, but I don't know much else about them.