Tektronix 2465B oscilloscope teardown

[tonedeak99]

Thank you AMR for the info. Ill see if I can lube it next time I take it apart.

Mark your suggestion was genius. My DMM is only accurate at 60hz so I set the generator to that frequency and immediately noticed that with the amplitude maxed out it was over 7.07v. So I lowered it to give me that value and the waveform was spot on 20v on the scope!  Im thrilled that my scope is well calibrated but Im mindblown that my generator and other scope were both uncalibrated in voltage by the exact amount.

[AMR Labs]

Mark your suggestion was genius. My DMM is only accurate at 60hz so I set the generator to that frequency and immediately noticed that with the amplitude maxed out it was over 7.07v. So I lowered it to give me that value and the waveform was spot on 20v on the scope!  Im thrilled that my scope is well calibrated but Im mindblown that my generator and other scope were both uncalibrated in voltage by the exact amount.

As a future reference, and as a quick test, I would think that you should be able to use DC voltage all the same to check the accuracy of the vertical inputs. Just switch input mode to GND, center the trace on the middle (or bottom) baseline, select the appropriate V/div range that will give you best resolution, then switch input mode to DC and apply the known accurate DC voltage measured with your DMM, and count the number of divisions the trace goes up or down the screen from the baseline depending on the applied polarity. If for example you chose 5V/div (with a x1 probe) then inputting 20V DC should raise the trace by exactly 4 divisions. Just make sure that you really switched to DC (1Mohm input), and NOT the 50-Ohm input option (also DC coupled) as the input current could blow or overheat the internal load resistor. 20V/50R = 0.4A ^2 x 50R = 8W = PUFF!!

[MarkL]

Ok, I have some ideas for you to check for the dimming.  It could be the intensity pot input which goes to the CPU, something in the CRT circuitry, or something in between which includes the DAC and display sequencer.

I would suggest looking at two different points to figure out which section to concentrate on.

1) The first is to verify the INTENSITY input control  This is fairly easy and you can do it from the front panel.  Get into Exerciser 01 and turn the INTENSITY knob so that the INTENSITY knob status line is displayed (should start with "17").  See table 6-9 for the meaning of these fields, but what you're looking for is any variation in the displayed numbers as the dimming is occurring.  If there is, the CPU thinks it's reading changes coming from the INTENSITY pot and the problem is on that side.  Hopefully the problem with dimming will not get in the way of seeing the status display.

I'd suggest reading the section on DIAGNOSTIC ROUTINES starting on 6-10 through (at least) Exerciser 01 to learn how to operate the diagnostic menus.


2) The second place to check is after the CPU, DAC, and display sequencer to see if the CRT is being instructed to dim.

Set up the scope for the test waveforms for the A1 board, which immediately precedes the A1<6> schematics in the manual.  Make sure the readout is completely off.  There is a very small range at 12:00 that this occurs.  If it's not off you'll see lots of junk on the test waveforms as the beam turns on and off when the CRT makes dots on the screen to form the characters.

Check signal VZ on the anode of CR966 (diode further from back edge), schematic <6> loc 8L, waveform [65].  The upper level of VZ is the brightness.  Watch this to see if it correlates to the dimming.

Check signal VQ on the anode CR972 (diode closer to back edge), schematic <6> loc 8L, waveform [66].  This is related to the brightness, but is probably not the cause.  Check for correlation here.


3) (Ok, so I added a third test..)  If you find that VZ or VQ correlates to the dimming, you can take a look at the output of the DAC on the cathode of CR354.  It has a range of about 0 to 2.5V which you should be able to verify as you turn the INTENSITY knob.  The DAC and associated sample and hold circuitry are a known source of problems.  You could even do this test before #1 or #2 since it's easy and targets a known problem area.

[tonedeak99]

Thanks AMR for the tip in Vdc amplitude testing. I think that’s how I checked my other scope when I got it.

Mark sorry for the delay in testing but for some strange reason the scope wasn’t dimming. Today it finally started acting up again and I tried your first test. It passed. The EXER 01 values for pot 17 never changed. Unfortunately I couldnt tell if the trace was dimming since the trace disappears once you enter diagnostics. But I tried to go into your test once I noticed the trace acting up and dimming or disappearing.

Sorry if this question is a dumb one but for test 2 do I use a second scope or the same scope to see the waveforms?

[David Hess]

The z-axis hybrids which control intensity are known to fail, but it could be a different problem so exclude everything else.

[MarkL]

Thanks AMR for the tip in Vdc amplitude testing. I think that’s how I checked my other scope when I got it.

Mark sorry for the delay in testing but for some strange reason the scope wasn’t dimming. Today it finally started acting up again and I tried your first test. It passed. The EXER 01 values for pot 17 never changed. Unfortunately I couldnt tell if the trace was dimming since the trace disappears once you enter diagnostics. But I tried to go into your test once I noticed the trace acting up and dimming or disappearing.

Sorry if this question is a dumb one but for test 2 do I use a second scope or the same scope to see the waveforms?

Sometimes it's ok to use the same scope to measure itself, but given the issues with dimming which may blank the display just when you need it, I would use a separate scope to observe #2.  #3 is the DAC brightness output and would normally be a static DC voltage when the INTENSITY knob is not being moved, so you could either watch it on a DMM or on a second channel on the other scope.

Because the symptom is now occurring more infrequently, you should monitor both points (VZ and DAC output) at the same time to increase your chances of seeing it.

Test #2, monitoring VZ and VQ, is directly after the z-axis hybrid (U950) that David mentions.  If it's bad you should catch it misbehaving, but the display sequencer (U550) precedes it and could also be involved, so you'd have to narrow it down a little further before reaching any conclusions.

I should also mention that you should be very careful probing around the hybrids.  Some of the pins have higher voltages than just plain TTL, and they are very unforgiving if you accidentally short the wrong two adjacent pins with a probe.  If possible, it's safer to probe a thru-hole component on the same node, or if your probe comes with those little protector caps that only allow you to contact one pin at a time, I recommend you use it.

[tonedeak99]

Ive been ready to do test 2 and 3 but the trace has been working fine. I think it’s related to the ambient temperature. Lately it’s been warmer outside and therefore inside too.

[tonedeak99]

Test #3 passed with flying colors.

test #2 created the 2 expected waveforms but the trace never dimmed so test was inconclusive.  I adjusted the intensity and noticed the amplitude changing accordingly.  Even opened the window right next to scope to let in some cooler air.   Not sure what could be the cause yet but will try again tomorrow.

[MarkL]

Test #3 passed with flying colors.

test #2 created the 2 expected waveforms but the trace never dimmed so test was inconclusive.  I adjusted the intensity and noticed the amplitude changing accordingly.  Even opened the window right next to scope to let in some cooler air.   Not sure what could be the cause yet but will try again tomorrow.

On test #3, meaning the dimming issue occurred but no changes in the intensity DAC output?  I guess you didn't get a chance to probe VZ at the same time.

If the DAC output remained fixed during the dimming issue, we're left with: display sequencer (U650) --> z-axis hybrid (U950) --> CRT driver board.  I'm not saying those chips are necessarily to blame; just that those are good places to divide the problem.

I prefer to catch a problem red-handed, but if you want to jump ahead a little and it appears to be temperature related, one thing to try is to reseat the z-axis hybrid.  Remove it, clean the contacts gently with isopropyl, and put it back.  Make sure it's lined up exactly before tightening the nuts; you don't want to crack the ceramic.

[tonedeak99]

Hey Mark.  Actually it hasn't dimmed again. Sorry if I wasn't clear but it didn't dim during my Test #3 either.  I have had all 3 tests going on simultaneously.  CH1 on the VZ, CH2 on the VQ and DMM probes on test #3.  I called #3 a pass because it was a range b/w 0 and 2.5 and changed smoothly and linearly as I turned the intensity knob.  Sorry wasn't aware that for test #3 I also needed to check it during dimming.  Will repeat it if this dimming ever comes back.  I'm dealing with the anti-murphy law.

[Giovanni62]

Could you post the schematic of your dummy load?

Alright, so I changed all the 50 ohm resistors out with 100 ohm ones on my dummy load. Hooked it up and the supply was 100% in regulation. Put it back in the scope and it started right up. All rails are in spec and there's no visible ripple on +5VD (at least at 50mv/div).

This thing is gaslighting me!

Nearest I can figure is maybe one of cables that goes from A1 to the supply wasn't secure, though I swore I checked those.

I'll check deeper tonight, measure the ripple on all the rails. But yeah, there you go. Weird.

Edit: Oh, here's the load I made. Pretty simple and in the end worked great!






Sent from my Tablet

[AMR Labs]

The schematic of the PS dummy load is available in the service manual of the scope. Not sure exactly at which page/section (don't have a manual in front of me), but its easy to find.

[factory]

There is no schematic for the regulator dummy load (I've looked in the 2465 & 2465B manual PDFs), but there is one for the "primary test load" for the inverter board.



And it mentions a 2 ohm 25W resistor, for the test load of the +5VD supply.



The five rails between +15V and -15V on the regulator board, have a 100 ohm load for each rail.



David

[MarkL]

I'm not sure this is the best place to bury something like this, but this thread does seem to have the right readership...

Attached is a utility that will allow people to restore their calibration data to NVRAM on 24xx A/B version scopes using EXER 02 data.  In short, you will need an EPROM burner and a 27512 EPROM or equivalent.  There are detailed instructions in the README.

For years it's been recommended that people take a video of EXER 02 data to backup their calibration data, but without a GPIB interface it's largely been left unsaid how to get it back in, particularly on models that have a separate battery and SRAM.

It will not work on 2445/2465 plain models, as those use the GI ER1400 EAROM.  Fortunately, the ER1400 is much less prone to failure.  However, if someone is having an issue with a ER1400 scope, please post the problem.  A similar utility could be developed for the ER1400, but right now I'm not seeing much need.

Please consider this utility a last resort when there's nothing else to lose.  It's job is to irrevocably write over your calibration data.

[AMR Labs]

This utility will also work for SN50K+ 2465B scopes? (with surface mount A5 board and NVRAM). I was loosely following a thread on Tekscopes2 which I think is about the same utility, and was under the impression that only B version scopes with SN under 10K (thru-hole A5 board) would be the only ones that can be recovered with this utility. Can you please clarify?

[MarkL]

This utility will also work for SN50K+ 2465B scopes? (with surface mount A5 board and NVRAM). I was loosely following a thread on Tekscopes2 which I think is about the same utility, and was under the impression that only B version scopes with SN under 10K (thru-hole A5 board) would be the only ones that can be recovered with this utility. Can you please clarify?

Thank you for pointing that out.  I can only say it works with thru-hole A5 boards that use a 27512 EPROM in U2160.  I did not take into account the different memory model used in the SN50K+ B scopes (SMD A5).  The D27011 paged EPROM (U2360) operates quite differently, and the memory control PLD U2250 is also different.

Although a 27512 would plug into the U2360 socket, I doubt the utility would work as written, and may even conflict on the bus with the SRAM chip.  I would need to find an SMD A5 board to create a version to run on those boards.  With any luck it could be made to work on both SMD and thru-hole so there wouldn't be two versions floating around.

From the schematics I've looked at so far, the SMD boards always use the self-contained Dallas module.  If anyone knows differently, please say.  So, it should always be possible to write the EXER 02 data directly to the module before installing it in the scope.  The module will be out anyway, assuming it is a replacement operation because the battery died.

I will look to buy an A5 SMD board when I can find one for cheap, or if anyone has a working spare to loan me in the US for a few weeks, I promise to return it unscathed.  Please PM me.  I also have an extra thru-hole A5, and can trade if you are interested.

Another path... I have some ideas on what should work, if anyone wants to work with me and try some images.  You need to have an alternate way to restore your cal data, such as a GPIB port.  I can pretty much guarantee your cal data will get trashed until we get to the right solution.  PM me if interested.

Also, it is the same utility posted on TekScopes and TekScopes2.  I was following up with the posts from the OP on those forums.  Maybe I should have kept it in one place.

[MarkL]

I said...

...
I will look to buy an A5 SMD board when I can find one for cheap, or if anyone has a working spare to loan me in the US for a few weeks, I promise to return it unscathed.

Actually, a broken one is ok too, as long as it boots.  I know these SMD boards can have analog and other issues from leaky capacitors, but it doesn't matter to get the NVRAM utility written.

[AMR Labs]

Thank you for pointing that out.  I can only say it works with thru-hole A5 boards that use a 27512 EPROM in U2160.  I did not take into account the different memory model used in the SN50K+ B scopes (SMD A5).  The D27011 paged EPROM (U2360) operates quite differently, and the memory control PLD U2250 is also different.

Yes that was what I understood from the exchange at Tekscopes2.

My 2465B 50K+ has been operating on a FRAM for several years now, so I kind of solved the NVRAM rat race issue. Plus I still have the original NVRAM chip, plus the fresh one I used at first to copy everything into, which both are now safely stored away.