Tektronix 7603/7613 Repairs

[dave_k]

I recently acquired 4 Tek 7000 series scopes (with plugins), a pair of 7603s and a pair of 7613s. The idea is to repair/restore these in my spare time, so updates might be slow.

Starting with a 7603 and the first unit I check is 100% dead. Others who have repaired these have replaced the main filter caps in these scopes. I figured that is a good place to start before checking the regulator board.

Rather than replace the original Mallory electros with NOS ones I elected to use modern equivalents. Most (if not all) modern electros are snap-in PCB mounting using 10mm spaced leads. The original caps had 4 leads around the perimeter with a 5th leg in the middle. Over lunch one day I came up with a neat answer to this square peg/round hole dilemma.


The original cap is on the right, replacement with adapter on the left. PCB was made by iTead (US$1 per board) and the mounting pins are from Harwin (p/n H2101-01).


Completed replacement of all 6 caps. I am still waiting on the delivery of a smaller adapter board for the 1 electro with a different footprint.


The PCB pins were selected to space the adapter board away from the main PCB as seen here.

Finally, the original caps were used to link PCB tracks together (to allow other tracks to pass underneath the cap). I have seen other repairs where wire links had to be installed along with bodging in the replacement cap. This adapter board neatly solves that problem too. 

To be continued...

[Rupunzell]

Very nice. This would be a good product offering.

Bernice

[PaulAm]

I use a similar adapter when repairing 465s.  On those scopes if you use very long leads instead of pins, you can thread the new cap into position without having to disassemble the scope.

Always use 105C caps from a reputable manufacturer if available.  And then check them before installation anyway.

[dave_k]

Very nice. This would be a good product offering.

Bernice

I am happy to make the gerber files available to anyone who wants them...

[dave_k]

UPDATE: I have received the smaller adapter PCB from iTead.



Now the rectifier assembly is complete.



Next step is to begin looking at the regulator board. More to come....

[dave_k]

It's been a while but I have had a little bit of time to work more on the 7603.

The regulator board checked out OK, except for one bad transistor (Q931) which was stopping the +15v rail from working. The original part is a 2N5859, best replacement I could find was a 2N5320.

So with the rectifier and regulator boards checking out it was time to try a first power up. Still nothing, no power light because something is still loading down the +5v rail. After some more probing it turns out the culprits were on the main interface PCB.



2 out of the 3 larger electros measured dead short (never seen an electro do that before!) It's not the easiest PCB to remove so it too was totally re-capped. I also learned the Main Interface PCB is a 5-layer board - high tech for a early 1970's design.



Finally the unit powers up and produces something on the screen.



Next step is to remove all the front panel controls for switch cleaning, etc. before trying to solve the issues with the fuzzy and wonky display (possibly a problem with the horizontal amp).

[dom0]

Nice work.

The blurred display could be caused by a number of things. I'd first check supply ripple - wideband, not with a multimeter. Trace the signal from readout board (easier to access than the main board) to the deflection plates to see where the noise is added ; might be an oscillating amplifier stage. Somewhere I read about a troubleshooting technique for all the Tek scopes with differential signal paths: short the diff pair at every stage from the horiz/vert amp going backwards and see where the problem starts showing up. Never used it myself, though.

[dave_k]

Well, the cause of the display problem was an anti-climax! Only one of the 2 horizontal plates was connected.



So now I have another problem to solve. From this you can see the trace coming from the plug-in is displayed nicely, but the readout generator is kinda 'ghosting'.



I hope the problem is a out-of-spec passive component and not U510 (Tek 155-0022-00 channel switching IC).

[AF6LJ]

Very nice. This would be a good product offering.

Bernice

I agree, I was thinking about having to make some of those for another resteration project I plan on getting on next year. Those are nicely made.

[AF6LJ]

Very nice. This would be a good product offering.

Bernice

I am happy to make the gerber files available to anyone who wants them...

That would be great.

[AF6LJ]

Very nice work, you are making good progress.

[c4757p]

I have a scope from the same family that might need the PSU caps replaced too - I'd also like the Gerbers

[dave_k]

Okay .. I'll post them here

[c4757p]

Thanks dude!

[AF6LJ]

Thank You Sir 

[dave_k]

Time for a brief update regarding the fuzzy readout; the characters appear fuzzy only in the vertical direction, as if there is some sort of oscillation occurring on the vertical drive from the readout board.

I did some more high-level troubleshooting and swapped out the horizontal amp PCB from the other 7603. The result was everything now sharp and in focus. So now to solve the question, how does a kaputt horizontal board cause a vertical issue with the readout?

[AF6LJ]

I am just wondering out loud.....
Does the fuzziness of those readout characters change with sweep settings.

[DimitriP]

Bad ram  in the character generator circuit will cause fuzzy readout ...if that lives on the Horizontal board you might have a wiiner.

[dave_k]

I am just wondering out loud.....
Does the fuzziness of those readout characters change with sweep settings.

No .. the sweep rate of the readout is generated independently.

[dave_k]

Bad ram  in the character generator circuit will cause fuzzy readout ...if that lives on the Horizontal board you might have a wiiner.

There is no RAM .. it's a 100% analogue design.

[DimitriP]

Bad ram  in the character generator circuit will cause fuzzy readout ...if that lives on the Horizontal board you might have a wiiner.

There is no RAM .. it's a 100% analogue design.

Not quite 100%
From the readout board, the it connects to horizontal board using J508, that would be a good neighborhood to take a look at.
 
As for the the  readout board, there is an adjustment for a
row match R2183,
Column match R2214
and vertical separation R2291
that might be worth fiddling with

[AF6LJ]

I am just wondering out loud.....
Does the fuzziness of those readout characters change with sweep settings.

No .. the sweep rate of the readout is generated independently.

Good to know, I might have to download the manual for this scope, these things are really interesting to me..

[miguelvp]

I am just wondering out loud.....
Does the fuzziness of those readout characters change with sweep settings.

No .. the sweep rate of the readout is generated independently.

Good to know, I might have to download the manual for this scope, these things are really interesting to me..

I have one of each.

The 7613 I got with a 5110 for $50 for both in really bad shape:

But they did clean up pretty good:


The 7603 came from a friend and needed some repairs:
https://www.eevblog.com/forum/repair/repairing-tektronix-7603-oscilloscope/


So far I have 12 7000 series plug-ins and 3 for the 5000 one.

Btw, analogs rule, free running I got  609,190 waveforms per second on those 40+ year old scopes, more than the pathetic 5,000 wfps with memory depth set to the minimum 14K points on my Rigol DS2000 (sure it can go up to 50,000 but only at 700 points memory depth), on normal operation with any decent memory depth well, it tanks bad, but it's really useful since it's in memory so not really a complaint just to point it out.

https://www.eevblog.com/forum/testgear/first-personal-'scope-purchase/msg681691/#msg681691

[timb]

I recently acquired 4 Tek 7000 series scopes (with plugins), a pair of 7603s and a pair of 7613s. The idea is to repair/restore these in my spare time, so updates might be slow.

Starting with a 7603 and the first unit I check is 100% dead. Others who have repaired these have replaced the main filter caps in these scopes. I figured that is a good place to start before checking the regulator board.

Rather than replace the original Mallory electros with NOS ones I elected to use modern equivalents. Most (if not all) modern electros are snap-in PCB mounting using 10mm spaced leads. The original caps had 4 leads around the perimeter with a 5th leg in the middle. Over lunch one day I came up with a neat answer to this square peg/round hole dilemma.


The original cap is on the right, replacement with adapter on the left. PCB was made by iTead (US$1 per board) and the mounting pins are from Harwin (p/n H2101-01).

Very clever solution, but one question: Do all four of the ground pins actually need to be connected from the cap to the board? The only reason I can see for that is if Tek used the caps to hook separate ground runs together.

I've replaced a lot of these caps on TM 500 power modules. On these four pin caps, the three inner pins have 10mm lead spacing, so I simply replaced them with new snap in style caps with that spacing, only using two of the holes on the board.

On the TM 500, the front side of the board connects all four ground pins together with a solid plane, while they do route off independently on the back side.







Bonus Picture: 30 Years, Twice the Capacitance, Half the Size!




Sent from my Tablet

[dave_k]

Very clever solution, but one question: Do all four of the ground pins actually need to be connected from the cap to the board? The only reason I can see for that is if Tek used the caps to hook separate ground runs together.

.. and that is exactly what is happening on the rectifier board in the 76x3 scope. The cap is used to bridge 2 parts of a ground plane together so that a track may run underneath. I have seen other repairs where wire links were installed to make the bridge. I think the PCB solution looks neater