Tektronix 7623A Oscilloscope - Repair Log

[sean0118]

Hi All,

I picked up this Tek 7623A a little while ago, it came with both 7B50 horizontal and 7A18N vertical modules.

It was completely dead (no power LED etc), seems to be a common issue with the 7623A so I thought it might be worth making a repair log (I might also need some help from the Tek experts). There's not many photos available online so I've taken lots:

Photo album: https://imgur.com/a/ySsIr3Z





Board locations:


A5 Vertical Amp board:


A6 Horizontal Amp Board:


A8 Z-Axis Board (CRT intensity):


A14 Storage Board:


A12 Low Voltage Regulator Board:


A11 Low Voltage Rectifier Board:







The first issue I found was one of the micro switches used in the mains power switch was broken, the switch was always open and the little plunger had lost it's springiness). It should have been an easy fix, except one of the holes they used for the mounting rivet is also the rivet which hold the switch together. All the replacement micro switches available already have a rivet or plastic clip in that location. The black switch in the photos below is the replacement one, in the end I just used one mounting screw with a star lock washer, I'm not sure if it will stop it rotating long term, but seems to work fine so far.










It still didn't power up after replacing the micro switch. It took me a little while to find the best place to measure the power rails, it turns out it's actually on one of the Z-Axis board headers. There was also an unpopulated fuse footprint for F813 on the rectifier board, I thought maybe it was removed in a previous repair attempt, but seems it's not actually connected to anything, I think it might just be for holding spare fuses.

After measuring the voltage rails on the Z-Axis board I found the +15V and +5V were missing. In hindsight I should have measured this un-powered using the resistance to the chassis GND, because it turns out the +15V was shorted to GND, lucky it's current limited. After isolating different boards I found there was a shorted tantalum capacitor on the Vertical Amp board. The board is interesting construction as the output transistors and some other components are soldered using point to point construction. The output transistors need to be de-soldered before the board can be removed. Once the board was off I replaced both tantalums with some 15uF electrolytic capacitors. This fixed both the +15V and +5V rails, all other rails measured within spec as well.

Header on Z-Axis board to measure power rails on:






Shorted tantalum shown with red arrow below:




Replacement caps on Vertical Amp board:







So at the moment it powers up, the CRT is either completely green or blank depending on the storage settings, there's no trace. I initially suspected the vertical or horizontal amp had a fault, but testing the waveforms going into the CRT deflection plates roughly match those shown in the service manual (with the 4V 1khz cal feed into the scopes input). I'm currently looking at the Z-Axis board which controls the CRT intensity, the military service manual mentions a test location which is meant to vary in voltage when the front panel intensity knob is rotated, however mine is consistently at 12V, the metal transistor package on the left also heats up a lot.

Maybe even more worrying is someone has cut a trace going to the header for the front panel intensity knob. There's a mod wire and resistors on the back of the board connecting it instead. I'm not sure if this is a factory mod or a repair attempt? I also can't find a full schematic for this board in the service manuals, only a partial one. So I'm going to spend a bit of time trying to reverse engineer it today and test the components as I'm pretty sure there is a fault somewhere on the board.



Intensity test location on Z-Axis board:


Cut trace and other mods on Z-Axis board:

[Zenwizard]

Looks like a fun project. I will help out where I can I have done a 7623A but I have been in a number of its cousin the 7603. One thing to note it does not look like there will be an issue in the 7623A cause the black connector behind the voltage test point is not the same size however in the 7603 If the beige and black connectors are switched it is total death to the scope. You send 130 Vdc in to most of the TTL and custom tek IC's. It is a bad day in the lab

Zen

[sean0118]

One thing to note it does not look like there will be an issue in the 7623A cause the black connector behind the voltage test point is not the same size however in the 7603 If the beige and black connectors are switched it is total death to the scope. You send 130 Vdc in to most of the TTL and custom tek IC's. It is a bad day in the lab

That would be a bad day in the lab. While the cable plug has one extra contact the header on the PCB is the same size, so they could be swapped over very easily.

The two pin beige connector next to the blue one doesn't have pin one marked on the PCB. You just reminded me to check the manual and it's mounted backwards! I had been putting it back how it was when I got it, but apparently that was the wrong way. It only has +5V present, but might explain why one of the transistors was getting to hot to touch. The scope was broken when I got it, some of the screw heads are stripped out, so I suspect someone has attempted to repair it before. I'll go through and check all the other headers now.

I tested all the transistors on the Z-axis board, just measuring the internal diode drops with a dmm they seemed okay. I haven't looked at the HV board yet, I was going to look at that next, but I'll investigate that backwards header first.


These are the only other teardown photos of a 7623A, but it does help confirm the 2 pin beige header was backwards:
http://www.johnlooyestijn.nl/electronics/slides/slidesTekScopes.htm

[med6753]

If you decide to re-cap the rectifier assembly adapters are available from Ebay to allow easy mounting of the new capacitors. But getting the old capacitors unmounted takes some brute force methods. This is a 7603N but your 7623 should be similar.

[TerraHertz]

med6753 that's neat. I'll bear those cap adapters in mind next time I'm in that situation.

Another thing to consider - if you're going to that much trouble to restore some old gear, why not also substitute an IEC socket/filter for the annoying hard-mounted mains cord. If there's room on the backpanel and space inside. Looks like there is on that unit.

Makes such an improvement.

[vk6zgo]

med6753 that's neat. I'll bear those cap adapters in mind next time I'm in that situation.

Another thing to consider - if you're going to that much trouble to restore some old gear, why not also substitute an IEC socket/filter for the annoying hard-mounted mains cord. If there's room on the backpanel and space inside. Looks like there is on that unit.

Makes such an improvement.

My 7613 has an IEC socket as standard, but the Gorillas have been at it, with the lead to the fuse in the Neutral line, instead of the Active.

When the 'scope died, it also lost the +15v & +5v supplies, but I read somewhere that Q931 tended to fail.
Sure enough, it had, so I replaced it with the nearest spec transistor I could find.

Lazily, instead of tracing each connector & cable individually, I relied upon the pictures I had previously taken, & plugged them back in the "logical"  places.

On turn on, I now only had -50v!!!

OK, time to stop being lazy, so I unplugged all the connections between the LV rectifier board & the regulator board.
Turned on again, --------zilch, not even the fan!

Started chasing through the LV rectifier PCB, found nothing on the fuse on the board--- what the heck?
Realised that was just an unconnected spare fuse, so rethought things, removed the screw in fuse---O/C.(it looked very old, so I think it was just a physical failure)
Replaced it with the spare, now all the rectifier outputs are OK.
At this point, I also realised the "gorrillaisation"which had been performed on the IEC .

Now it looks like, before I connect up the LV regulator, I will have to look at Q931 again, & also check those caps on the vert amp that sean0118 found on his 7623.

I don't think (or, I hope), I have connected the -130v in the wrong way, or it would probably have killed the -50v as well.
The latter appeared in the right place, which is why I am optimistic.

[med6753]

med6753 that's neat. I'll bear those cap adapters in mind next time I'm in that situation.

Another thing to consider - if you're going to that much trouble to restore some old gear, why not also substitute an IEC socket/filter for the annoying hard-mounted mains cord. If there's room on the backpanel and space inside. Looks like there is on that unit.

Makes such an improvement.

I would if I had a suitable punch. A drill and hack saw is a no go. 

[Zenwizard]

Yea P1171 and the connector behind it was what I was mentioning extreme care is needed with those 2 that they are in the correct spot and aligned the correct way. I made that mistake ONCE. Having -50 is good if I remember correctly that is the reference supply for the scope. I had a 7603 where the voltages especially the 130 Vdc were all over the place. and were reading high. 130 volt was 149. It ended up being a burned out driver transistor for the pass element. The pass element was not even turning on. So no regulation. There could defiantly be an issue in the HV supply. The tube lighting up and just showing green could be the flood guns starting. If that is true then the screen is WAY to bright and the flood guns need to be adjusted down. Ran in to this on a 577. But if it is just flood guns then there is no beam. Which could be sweep or HV. Also the scope will not sweep with out a B time base in it. So with no plugins a dark looks like dead unit is actually normal and expected behavior.

Zen

[sean0118]

If you decide to re-cap the rectifier assembly adapters are available from Ebay to allow easy mounting of the new capacitors. But getting the old capacitors unmounted takes some brute force methods. This is a 7603N but your 7623 should be similar.

Thanks, that might be a good upgrade for long term reliability.

Realised that was just an unconnected spare fuse, so rethought things, removed the screw in fuse---O/C.(it looked very old, so I think it was just a physical failure)
Replaced it with the spare, now all the rectifier outputs are OK.
At this point, I also realised the "gorrillaisation"which had been performed on the IEC .

Now it looks like, before I connect up the LV regulator, I will have to look at Q931 again, & also check those caps on the vert amp that sean0118 found on his 7623.

The fan is on the primary side of the transformer, so the broken screw in fuse would explain that. I forgot to mention, the 130V fuse on my regulator board fell apart when I removed it too, the epoxy had dried out so much the ends weren't held to the glass at all.

Having -50 is good if I remember correctly that is the reference supply for the scope.

That probably explains why only the -50V is adjustable then, I was wondering about that.

There could defiantly be an issue in the HV supply. The tube lighting up and just showing green could be the flood guns starting. If that is true then the screen is WAY to bright and the flood guns need to be adjusted down. Ran in to this on a 577. But if it is just flood guns then there is no beam. Which could be sweep or HV. Also the scope will not sweep with out a B time base in it. So with no plugins a dark looks like dead unit is actually normal and expected behavior.

I did open the HV supply, but only tested the transistors, looked for shorts etc. I don't have anything that can measure HV. That is interesting regarding the flood guns, where are they adjusted from?

You also answered another question I had, what happens when no plugins are inserted.

Also, I had another look at the manuals and found the schematic for the Z-axis board. The dc values seems a bit off, but they change with the intensity knob anyway. But more importantly, I'm not getting any waveforms on the input to the z-axis board or the A or B outputs to the HV supply. Are these generated in the time base module? I might be able to *borrow* a second one to test.

Also, there's a few coax connections missing from the z-axis board. Most of them seem to go to the Readout Board which mine doesn't have installed (I think it's an optional extra?). But J1135 is also missing, it's labeled "X/Y Inhibit" going to J503 on the horizontal amplifier. Is that coax optional?

[med6753]

Also, I had another look at the manuals and found the schematic for the Z-axis board. The dc values seems a bit off, but they change with the intensity knob anyway. But more importantly, I'm not getting any waveforms on the input to the z-axis board or the A or B outputs to the HV supply. Are these generated in the time base module? I might be able to *borrow* a second one to test.

Yes, comes from the time base plug-in. Without that signal you'll have no trace. Check and make sure the plug-in is properly seated.

[Zenwizard]

Also it goes with out saying make sure the time base is in AUTO mode not NORM and that it is not in X-Y some of the B time bases have X-Y. you can also put a 7000 series scope in X-Y mode with two A plugins one in the vert slot and one in the HORZ slot. This will at least drive the tube directly to see if there is any signs of life. But auto mode with any coupling should at least give a sweep regardless of trigger.

[sean0118]

Also it goes with out saying make sure the time base is in AUTO mode not NORM and that it is not in X-Y some of the B time bases have X-Y. you can also put a 7000 series scope in X-Y mode with two A plugins one in the vert slot and one in the HORZ slot. This will at least drive the tube directly to see if there is any signs of life. But auto mode with any coupling should at least give a sweep regardless of trigger.

I didn't realise the 7B50 horizontal module has X-Y, but turns out it does! It's button to enable it is labelled 'Amplifier' while the one for regular operation is 'Time Base'. It's good to know but didn't fix the issue.

I just aquired a Tek 7603 as well, which has a 7D01 logic amplifier in it. Keep an eye out for a repair thread on this soon, because it also seem to be broken. If I place in the vert and horizontal modules from the 7623A it sort of gets a trace, but it's at about 45 degrees with no waveform on it. But I need to investigate that one further.


Back onto the 7623A I measured both Sweep and Sweep Gate on plug P53 on the Main Interface Board. I'm not really sure if they are correct but they are shown below. Sweep Gate is used to generate 'Z Axis Out' on the A2 Logic Board. Probing this very carefully and with the rectifier module pulled out to provide space, I found Sweep Gate came onto the board and was present on the collector of U99B, and also the emitter of U99C, but disappeared all together from it's collector!

So I pulled out U99  (which is a CA3046 NPN Transistor Array). I found I can measure the base emitter and base collector diode drops on some of the internal transistors but not others, including U99C. I'm fairly confident this is the reason for the loss of trace, but I don't have a replacement CA3046 to test yet.

The funny thing is, I removed the Main Interface and Logic Boards earlier and tested all of the transistors. I re-seated the ICs and applied DeOxit, but I didn't think to test them. 


Main Interface P53 Pin 4 Sweep Gate:


Main Interface P53 Pin 2 Sweep:


A2 Logic Board:

[sean0118]

I replaced the CA3046 with an LM3046, and I now have a trace! 

I think the vertical gain is a bit unstable, might just be a calibration issue. Feeding in the 4V 1kHz ref wanders around a bit, will be on 4V then a few minutes later expands to 5V. I guess it could be the ref itself, needs more investigation.







A2 Logic Board:


A1 Main Interface, A3 Trigger Selector and A4 Vertical Interface:


Higher res photos: https://imgur.com/a/Hopvil1

[med6753]

Good deal! 

[sean0118]

Thanks, it seems to be working pretty well now.

I think the vertical amp issue was caused by the rotary switch on the vertical module, the trace seemed  to jump around a bit when putting pressure on it. I suspect the contacts inside are dirty from not being used.  It seems to have gone away now that it has been used a bit. I tested the ref and it seems spot on.

The storage function seems to work, but only intermittently. I was measuring the voltages on the storage board and was only getting 12.95V (way to low) on the FAST test point, when I pressed the fast button the voltage went to the expected 125V, when I released it the test point stayed at 125V.

There's also voltage test points labeled FGA and FGK, I assume these are for the flood gun anode and cathode voltages. I checked these with the scope in a few different modes and they matched the service manual values.

I suspect the storage issues might also be caused by dirty contacts in the front panel buttons. Some DeOxit might help, but I think I will just call this one a success and move on for the moment, I have about 20 other repair projects to move on to.

But I hope this thread helps others repair their Tek 7000s, I've seen a lot of threads with unresolved power on issues which might be similar. 

[mansaxel]

But I hope this thread helps others repair their Tek 7000s, I've seen a lot of threads with unresolved power on issues which might be similar. 

Yes, it's a common deal with Tek scope supplies that there's one voltage that's king. On my 5440 it is -30V and it's dead as a dodo. Issue's been traced (via Kelvin-style 4-wire resistance mesurements) to a dead short Tantal cap (what else?) that's going to be exchanged come Monday, when the Mouser shipment arrives. 

9. Capacitors are Murphy's foot-soldiers.