Tektronix 2232 scope.... in need of a doctor..

[David Hess]

Well that's quite interesting ! So the sizing of filter caps, as I learned it at school, is mostly relevant when used behind the main bridge rectifier in a linear power supply.. but for "local" decoupling on individual boards, ESR is the most important factor. I thought that's why there was this common practice of associating a big electrolytic cap with a small 100nF "plastic" cap, for better high-frequency/transient response ?

These are very different capacitor applications.  Input capacitors are selected for energy storage to produce enough holdup time between power line cycles.  Decoupling capacitors are selected for impedance which primarily depends on ESR.  Output capacitors need low (and often controlled) ESR but may need to be selected for ripple current in high current applications.

I guess using a tantalum kinda make for an "all in one" compromise : ESR low enough to be able to get rid of the extra 100nF cap, along with a capacitance high enough to be able replace the old electrolytics. OK, so basically in modern designs they use tantalum to kill two birds with one stone, as well as increasing long term reliability I guess.

Solid tantalum capacitors have better high frequency characteristics than aluminum electrolytic capacitors however they are still usually used in combination with ceramic or film decoupling capacitors and in practical designs, the distance between the decoupling capacitor and circuit is very important anyway.  Notice on the storage board that only one bulk decoupling capacitor is used per supply voltage while lots of small ceramic decoupling capacitors are distributed around the board.

Tektronix used aluminum electrolytic capacitors here because they were cheaper and sufficient.

Well, come to think of it. Since you said the ESR decreases as the capacitance increases, does that also hold true for tantalum ? In this case wouldn't it be worth using 22uF or 33uF caps instead of 10uF ? I mean, the price difference is not that great, it's not a factor... so no need to be cheap here. The 35V ones are expensive, yes, but the 16V ones, that we are going to use here, are much cheaper:

There is no reason you cannot do that except for the increased cost.  The larger capacitance part will also have a higher ripple and surge current rating leading to higher reliability in some specific cases but probably not in this case where the voltage derating is enough.  There might be some performance benefit since this board is a mixed signal design but I doubt it.  In my experience, the 2232 achieves unusually low noise in digital storage mode anyway; it is not unusual to see +/-0.5 LSB or better performance.

[Vince]

Notice  on the storage board that only one bulk decoupling capacitor is used per supply voltage while lots of small ceramic decoupling capacitors are distributed around the board.

I was goingto say they were miniature tantalum, being yellow and in a dipped package, but of course it would not make sense. BOM is there to remind me...
So, these tiny cute yellow dipped are ceramic not tantalum. I always thought Tantalum had the monopoly for yellow dipped packages, somehow ! LOL [hiding in the corner]

Tektronix used aluminum electrolytic capacitors here because they were cheaper and sufficient.

I guess even expensive gear need to be pragmatic and count beans if the company making them is to make a profit...

There is no reason you cannot do that except for the increased cost.

I just received the news caps in the mail. That was quick, happy with their service... 
I actually ordered both 10uF and 33uF, not knowing what your reply might be...
In the end I ended up using the 10uF ones, simply because they had the correct 5.08mm pitch, and had a lower profile which made them unlikely to ever kiss the case while opening the scope, which looked like what caused the death of that electrolytic. The 33uF ones had their legs spread too far apart, and bent outward at sharp angles. Bending them back to a narrower pitch proved to be ugly and stressed the leads. So, 10uF it would be.  Replaced all 6 of them of course.

Then powered up the the scope, gave it a quick sanity check in both analog and storage mode... had somehow to again tweak a couple trimmers there, nothing that worried me that much (vertical a tad, and horizontal offset a fair more bit, not sure why). All was fine. Then screwed the back plate to the underside of the digital board, and secured the digital board back down to the chassis, quick check again, still fine, then put the scope cover back on, been a long time ! Getting there... I thought. what a fool. Then checked the scope again, for the 3rd time in 5 minutes... still fine in analog mode... then switched to store mode, uh what's going on ?! No !  Moving the trace up and down produces a funny result !  It does move no worries but there is some "ghosting" for want of a better term. It produces a vertical "comb" pattern, like a paint running on a wall or something ! Weird.

Made a clip of it to illustrate :

https://youtu.be/YKeMm_kueGk

Suddenly it reminded me seeing a similar behavior a year ago, when I acquired that scope and played with the store mode to get a feel for it.

Then a couple minutes later, the "comb" was gone and the trace was fine again, phew... then 2 more minutes later.... back to troubles : the trace this time had completely vanished/disappeared from teh screen ! Shows up perfectly fine in analog mode, no worries, but as soon as I switch to storage mode, the readout show up just fine no problem there, but the signal trace is nowhere to be found !  As if it's brightness control were turned all the way down, and/or the scope was not triggering properly... except I don't think it's any of that since the "trigger'd" LED on the front panel is appropriately lit (and it just uses the same settings as the analog mode anyway, which does work/display the trace just fine).

So... I think... this scope.... despite all the time and effort I have put in it in the last 6 weeks... does NOT like me ?!  Can a scope ever show GRATITUDE to its loving and caring owner ?  .... grrr.... or maybe I am misinterpreting his wil... maybe he does love me, so much, soooo much that he is afraid that I might stop looking at him the moment he is healthy again ?! I didn't think psychology had a place in electronics...

Anyway, I am puzzled ! Will I ever see the end of this repair ? 

I am trying to see the bright side of things  (scope SMPS fixed, "working" again)... but I am finding it hard presently...


Sooooo.... back to school again. What could be wrong ?

Logic has it that the problem is related to the last thing I touched... the caps ? Hmmm... don't think so, because of everything you said, and also of course because as I said I tested storage mode right after replacing the caps, and all was fine.

So my second best hypothesis would relate to that X offset trimmer. I had to adjust it, quite a bit (was a full division off) but I didn't get WHY it was that I had to touch it... why would it have moved during my repair ? I didn't touch it or even think about him, and I don't quite see how I could have moved it by accident !
I also found it strange/worrying that I had to rotate that trimmer all the way to the left in order to adjust the readout&trace, and even then it was still a tad not quite there yet, so had to finish it off by slightly adjusting the analog control counter part on the front panel, which I didn't judge quite normal...

So me thinks, maybe the trace disappearing might be the X offset circuitry going bad. Maybe that trimmer is defective, or has bad solder joints, goes open circuit and that makes the trace fly off the screen all the way to the left or right.

Will have to open the scope again and check the trimmer up close. Will also check the schematic to see how the X offset circuitry is put together, see if my theory would be possible at all.

While I am doing that, if anyone has any ideas about this strange "comb"/ghost pattern when I move the trace up and down ? Maybe a know problem, or a common issue in digital scopes at large, of that era, all brands included ?

So again.... NEVER a dull moment with that scope  !!!   

[Vince]

Below an extract of the circuitry around the X_OFFSET trimmer (sheet #16). Indeed looks like an open circuit/nad connection there, would send it off scale, as it's connected simply across a 5V reference.  Simple enough to trouble shoot then... I should be able to get a voltage ranging from 0 to 5V as I move the wiper.

Then that voltage goes to sheet #19 (attached below), to the "X VECTOR" input circuitry. Basically it just goes to an op-amp input so I guess the troubleshooting ends there !   

Might check a bit further down that vector chain though, while I am at it, as later in that pipeline there is yet another trimmer : "X VECTOR".

I I have not fixed it by that point, then I could unfold the story a bit further : after the XOFFSET, after the X VECTOR, there is the X amplifier section, with a differential pair which I guess produce the actual signals driving the CRT deflection plates, or quite close to that anyway (going back to the analog/main board at least, which we know work fine). So I guess that's my whole "X"/horizontal chain right there... just need to follow it until I find where my trace is hiding !   

[Vince]

OK I did it again.... 5AM ! This scope just won't let me go to bed, what a selfish piece of .... gear.

Made some progress I think.   X offset trimmer seems fine in fact, red herring I think. Now I think the problem is more to do with the vertical side of things :

I have got the trace BACK !  I got it by turning the ANALOG vertical position control almost all the way !  The Vertical offset trimmer on the digital board won't do anything, probably because the trace is so far away !  Way out of adjustment range.  So, I fed the scope with a sine wave (5Vp-p, 1KHz). Then probed the input signal on the digital board, with my old Hameg scope (which I isolated of course), right at the ribbon cable connector/input of the front-end. I put the alligator clip of the ground lead, near the connector, on the chassis. Then probed one of the signals in the differential pair... any of them should show my sine wave anyhow. Not too clued about a potential DC offset the differential lines might sit on, with respect to ground, but I figured it did not matter : I should still see a bloody sine wave in there. I set the Hameg to AC coupling to make my life easier, no sine wave to be seen. So switched back to DC and set vertical gain so I could see the entire/absolute maximum voltage range the differential line might present, ie +/- 8,6V, since that's what the analog rails are on the scope.

So, result ? Still no sine wave. Instead, a DC signal sitting at +4V, with random square pulses (impossible to trigger/stabilize the trace) going about 1 or 2V (can't remember) negative relative to that 4V base level. So, no sine wave, but DC with some (presumably, for now) digital crap finding its way on top of it !
So if the front end of the digital board sees basically 4V DC, I guess that's way out of range and why the trace was way of the screen, and why the offset trimmer on that digital board, couldn't get the trace back in sight.   So it's starting to make some sense now, what do you think ?!....

So I am going to bed now, but tomorrow I guess the next course of action will be to figure out if the problem comes from the digital board/front end, or if it's the main analog board sending crap though that ribbon cable !  Some service manual reading is on the cards I think...

This scope is showing it's age I think, it's like an elderly.... no matter how easy he takes it, it always hurts one way or another. A joint aching, legs suddenly not responding without warning  and falling in the stairs...  it's just a random succession of misery no matter how much you baby it, no matter how careful you are with it...

Still, I don't give up, I don't give up.... not sure now if I will ever manage to get it to work 100% ever again, getting depressed a bit, but I will keep trying, if just because it's a good for developing  trouble-shooting / repair skills which will be useful to fix other gear I plan to buy later on... and because I can't afford to buy another 2232... which most probably would need fixing as well anyway, so might as well fix mine !

Oh boy.....

[Vince]

7 AM, still up and thinking of that scope... this thing will kill me no doubt.

Just had a quick look at the schematic for the analog front end of the digital board, attached below.

I see that there is a 4.3V Zener diode at the input, reversed biased between ground and the 8,6V rail. So by the looks of it, presumably, it's there to provide a DC bias for the differential signals to ride on.  So, that would explain the 4V DC I am seeing on my Hameg when I looked at the differential lines at the connector. It's normal then. So the crap/negative random pulses that ride on top of that 4V DC level... is supposed to be my sinewave ? Gee... takes smore than imagination to believe these square looking pulses could be my sien wave. OK, I guess if the sine got amplified and/or clipped, somehow, it could make the sine look squarish. But at least it would remain a periodic signal and I should able to trigger on it.. but that's not the case, these pulses look random and jump all over the place.

[tautech]

Get some sleep man or you'll be no good to anyone. 

[David Hess]

Then checked the scope again, for the 3rd time in 5 minutes... still fine in analog mode... then switched to store mode, uh what's going on ?! No !  Moving the trace up and down produces a funny result !  It does move no worries but there is some "ghosting" for want of a better term. It produces a vertical "comb" pattern, like a paint running on a wall or something ! Weird.

The "comb" is completely normal when the oscilloscope is using equivalent time sampling at faster sweep rates like 0.1us/div.  Instead of waiting for a full display record, the oscilloscope updates the display on every acquisition.

Then a couple minutes later, the "comb" was gone and the trace was fine again, phew... then 2 more minutes later.... back to troubles : the trace this time had completely vanished/disappeared from teh screen ! Shows up perfectly fine in analog mode, no worries, but as soon as I switch to storage mode, the readout show up just fine no problem there, but the signal trace is nowhere to be found !  As if it's brightness control were turned all the way down, and/or the scope was not triggering properly... except I don't think it's any of that since the "trigger'd" LED on the front panel is appropriately lit (and it just uses the same settings as the analog mode anyway, which does work/display the trace just fine).

So me thinks, maybe the trace disappearing might be the X offset circuitry going bad. Maybe that trimmer is defective, or has bad solder joints, goes open circuit and that makes the trace fly off the screen all the way to the left or right.

It has to be something which affect the storage display and not the readout.  See if the reference display works; you will have to save something to it though if the reference memories are empty.

Under the Advanced Functions menu, there is a diagnostic menu for testing various functions.

[Vince]

SUCCESS !!!! 

After some sleep... things became clearer, I guess...

The more I fiddled with the scope, trying to make sense of its behavior, the more crazy it acted !  It got to the point where the readout itself, not just the trace, would get shifted upward by several divisions, and at some point the readout would form a CIRCLE and dance all over the screen, psychedelic !  Has to be seen to be believed ! LOL  The time base/sweeping sometimes also decided to go haywire.

Anyway, at this point it was acting sooooo weird, that I decided to stop wasting my time chasing red herrings all over the place... I thought it must be something stupid simple, a bad connection somewhere.  I eventually found the faulty bugger : J9011 !

It's the big connector near the hinges in the middle, with nothing on it but 8 black beefy wires.  One side of the top part of it, was cracked and so was not making a firm/good contact anymore. Still, on that type of connector it should not matter this much, since the top part is mostly useful during the assembly of the connector, as it pushes the wires into the terminal so they can cut through the wire insulation and get to the copper. But once this is done, contact is made and the top part could as well be discarded ! Well, exaggerating a bit of course, but that's the idea. So I still wasn't sure that this connector was the problem, especially since all these thick black wires sure had to be ground, so why would it make the scope go crazy if a wire or two, out of 8, were not making perfect contact ? Hmmm.   So, removed the top part of the connector altogether. Then realized that the wires were NOT at all firmly in place ! Even though they were sitting inside the terminals, the insulation of the wire had become quite hard/stiff, due to the heat generated by the board I guess. So although the wires were in place... it didn't take much from my part for them to lift out of the connector !!! 

STILL ! If these were all ground, it should still be making enough of a contact for the board to work, I thought ! So... could it be that, maybe.... just maybe.... these black wires were NOT all grounds, could there be actual SIGNALS in there as well ?!  Highly unlikely, but at this point it was my best lead so I followed it... looked up the schematic, it's connector J9011.  Still no signal there, but MORE than ground though ! this connector carries all 3 the power rails ! Why would Tek make them ALL black ?! Just so people would make mistakes like I did ? Was color THAT expensive an option ?! 
Anyway, now knowing that all the supplies were in there, of course all was starting to make sense at long last !

I could not reassemble the connector of course, and I didn't want to anyway, as I had seen that the stiff wires were the real root cause of the problem, more than the connector itself.  So I didn't want to be dicking around, I needed to make a reliable connection to all these wires if I wanted to make further trouble-shooting any meaningful.  So, I didn't cut corners : I pulled all the wires from the connector, and soldered them straight to the connector terminals, on the underside of the board. 

That fixed the problem, scope works like a charm  !  And with such a good connection, I now have the peace of mind that I am no longer at the mercy of some random glitch in the future ! At least not due to this crappy connector...

HAPPY I AM, fixed a last !!!!!!   

I would of course like to replace that connector if it's a standard item one can buy off the shelf, but I can't find J9011 in the parts list!  Not in the A10 section any how.  Maybe in some other place but god knows where....

But for now, it works just fine. The original connector was NOT removable any how (the removable part is on the main/analog board, not on the digital board), so having the wires soldered to the board actually doesn't change a thing to the original setup really, from a practical point of view....

So if you know/recognize this type of connector and know where to get one, I would replace it, but for now I will call it done and put the cover back on ! Crossing fingers that this time, THIS TIME, putting the cover back on will not give rise to yet another weird problem !   

[WastelandTek]

will you promise to get some sleep now?   

[Vince]

Yes I will ! 

This 2232 put me under a lot of pressure because it s my main scope, it was vital for me to fix it really as fast as possible.

But now that this is done, what a relief !!! 

I will just need to buy 4 little plastic knobs for the front panel, to replace the ones that disintegrated when I tried to pull them off, and I will be done I think 

I can now move on to the next repair : moving up league in complexity : the TDS 544A ! This is my second scope, the big boy, so I definitely need to fix it too, but the pressure is not as great as the 2232. I will start working on it in the coming days (another thread soon to come then  ), but this time I will not loose sleep of over it. I will take it easy. Will work actively on it because I am in vacation so have lots of time, but I won't let it pressure me. If I see that it's starting to drive me nuts I am starting to loose sanity... I will just take a little break and switch to some other gear repair or project, have a few on-going things, always something to do in the lab 

[Vince]

Well I am stupid... I just said the problem was with the wires, so replacing the connector won't get me very far. I would need to replace all the wires so that the new connector would have a good "bite" on some fresh/softer wire insulation. But replacing the wires would mean replacing the connector at the other end of the cable, on the main board... this one I don't know how I could replace it, tooling wise...
So I would have to re-use this connector, by cutting all the wires and soldering new/fresh wires onto it... might look ugly ! Or maybe not, especially now I have bought plenty of heat shrink  tubes. Color as well ! hmm.... replacing the wire would also allow me to use color wires to show clearly where the grounds are, and differentiate the various rails. That would be a nice upgrade for future servicing... me thinks ?!

So well yes, might do that at some point... but not now. Finding a new connector might proved challenging I don't know. I don't want to spend my holly days on this ! So I will put the cover back on for now, move to the next repair/project, and when/if I can find a replacement connector, I will re-open the scope to do the upgrade. Will be quick any way, once the cover is off, there isn't really much that needs to be disassembled to perform this little job, if anything. Just removed a few screws to be able to get the underside side of the digital board to do the soldering, and that's it.

[tautech]

Be sure to take measurements before you put the cover on.
Pin pitch, height, OD, ID etc.
And pics.....if your memory is anything like mine. 

[Vince]

Yes good idea !

I was relying on the fact that I would eventually find the technical details of this connector somewhere in the parts list, but if that fails then yes I guess that if I take all necessary measurements maybe I could find a replacement if I spend enough hours reading dozens of connector datasheets...

Either that or... I could just remove the connector from that board altogether, so I have it handy ! After all it's not even being used anymore now ! Just sitting there on the board doing nothing...

[Vince]

OK found it ! The block diagram and schematics all talk about 'P' or 'J' 9010.... but on the actual connector itself is printed "W" 9011... "W" that's how they refer to "W"ires ; they don't supply the connector itself, but only the cable sub-assembly, made in-house. This one I could at last find int he parts list ! :      It's an 'IDC' type connector, made by 3M, and wire gauge is 18. Pitch is 0.156 .... I guess unit is imperial/inch, meaning it would be 3,96mm which appears indeed to be a standard pitch, and consistent with what I measured with the calipers.

Problem ? Spent some time on Farnel and Mouser but I can not find anything that looks close to that ! Looks like these types of connectors are put in the "Wire to Board" category. In this category there really isn't that many 8 way 3,96mm pitch connectors... 

Surely it must be somewhere, I just can"t find it that's all... might post on the forum to get help identifying/finding this little bugger. I can't possibly be defeated by a bloody connector eh ?! How sad ! 

So looks like I have no choice but indeed leave these wires soldered directly to the board and put the cover back on to the scope. At least I tried...

[Vince]

Oh no !

NOOOOOOOOOOOOO !!!   

Yes, you guessed it !  Put the cover back on, and the problem in store mode (only) are back again !  Weird behavior involving vertical offset on way or another, and signal trace not showing up, just a flat line.

At that point I was close to either committing suicide.. or murder, couldn't make my mind up so I decided instead to get back to work...

Again I voted for a bad contact somewhere. At least I knew the power supply wires were fine since I soldered them straight to the board... but what about the connector on the other end of the cable, on the bottom/analog board side of things ? Looked in perfect condition, but wiggled the wires anyway, no joy.
Unplugged/plugged all the ribbon cables from the hinge side of the board, no joy.

So I thought hell, some manage to fix expensive gear that display weird faults, just by re-seating some chip... so I thought why not, worth a try at this point, I am so upset with this thing that I am willing to try anything. So did that. Started with the 2 EPROM chips for the CPU... just for heck of it, because obviously the CPU side of things was working just fine anyway. Next on the list, and more likely to be involved in this analog related series of events/problems : the highly sensitive stuff... the 2 ADC hybrids, and the all mighty acquisition chip where all the black magic happens. All 3 in fragile/brittle ceramic packages, all with with lots of pins that hold them firmly in place... in short, a recipe for disaster !  But I had to try something ! So as you can imagine  I used extreme precaution, went very progressively and slowly, and lost a couple pounds worth of sweat in the process ! But... all went well, miracle, what a relief !

And it was worth it because the problem is fixed (again), back to life ! 

... until I put the cover back on I guess ....

But before I do that, there is still a little problem to fix, something new. Not too worried about it since the scope works fine basically, but still I would like to try and fix it if possible.

Problem : at power up, the scope always displays an error message:

- - - - - - - - - - - - - - - - - -
POWER UP FAILURE
PRESS SETUP BUTTONS TO CONTINUE

CMOS: recovered
- - - - - - - - - - - - - - - - - -

Now I am not sure about the CMOS thing... will have to check the manual.
The backup battery is still good, 3 Votls no problem, so the SRAM should be alright even when the scope is powered off.

I get that message each and every time I power up the scope, even though I have now fixed all the previous problems.
No fault code shows up on the 7 segments display on the digital board.

So I ran the various self tests. See below. It reports an error about the trigger system, but maybe it's just because you need to apply a signal to the inputs of the scope so it has actually something to trigger from ! LOL So I will check the manual to verify this and see what signal I am supposed to feed exactly, so that the trigger test can be performed.

That being said... when I selected the A/D tests, it DOES tell you that it requires an "appropriate" (their wording not mine...) ramp signal... so if they take the time to ask for it in the A/D test, why wouldn't they also ask for it in the Trigger test, hmmm....

Anyway, I am off to read the service manual to check for that.

Oh, and the memory tests are funny : there are 3 tests : ROM0 and ROM1, both passed (I does not much more than scanning them quickly to verify the checksum) and an NV RAM test, which report : "sorry, NV RAM not implemented" .  You bet !  No NVRAM in my scope, just a stand alone battery !  Does that mean that some revisions of the 2232 did come with a fancy NVRAM ? I guess so...

I notice there  sme jumpers on the CPU side of things :

- WDT : "Watch Dog Timer " I assume.
- RESET : I suspect I might have to play with this one to clear that bloody CMOS error... will have to read the details in the manual
- NORM : Normal running mode ?
- DIAG : DIAGNOSTICS mode ?

The latter does not have a jumper on its pins, so I assume it works in conjunction with the "NORM" jumper. Jumper must be set to either setting.
I wonder what magical things I may have access to to if I power up the scope with the jumper moved to the DIAG position, mmmm..... 


Any way, all thoughts are welcome, as usual ! 

[David Hess]

Oh, and the memory tests are funny : there are 3 tests : ROM0 and ROM1, both passed (I does not much more than scanning them quickly to verify the checksum) and an NV RAM test, which report : "sorry, NV RAM not implemented" .  You bet !  No NVRAM in my scope, just a stand alone battery !  Does that mean that some revisions of the 2232 did come with a fancy NVRAM ? I guess so...

Page 6-17 of the service manual says that the "NV Ram" test was not implemented.  On the 2232, the NVRAM is implemented on the storage board and all 2232s include it.  On the 2230, the NVRAM was included as part of the RS-232 and GPIB option boards.

The same area of the manual discusses the power up tests and what display they produced.  I think yours means that an error was found in the NVRAM but recovered.  You could try erasing the contents of the NVRAM at the Reference - NVMem menu.  Also go to Advanced Functions - Save Setup menu to configure the startup as default or power down.

Or just disconnect the battery for a bit so that the oscilloscope powers up and reformats the memory.

[Vince]

Hi Dave,

Scope is fixed, errors cleared. As expected it was minor.

Unplugged the battery as you suggested, worked for the CMOS error.

For the Trigger test, it was just me not understanding the instructions given on the CRT : set trigger MODE to "BOTH".  There is no such thing as a "MODE" control in the Trigger section of the front panel, and if there was, then a "BOTH" option would have me puzzled, how could you possibly trigger from both channels at the same time... I guess with some imagination, we could trigger on some logical combination of the signals: "Ch1 OR Ch2", but that's far stretched and best left to digital power houses like the TDS 500/700 series... they have this kind of stuff.
So, had to be either "BOTH" Input channels activated, or "BOTH" time bases. Tried both options (pun intended), turned out I satisfied it by enabling both time bases. No trigger error anymore.  Read the manual section 6 as you did, and it's not clearer there either.  Also what's not clear is the A/D test. CRT instructions ask for an "appropriate" ramp signal, without giving actual amplitude and offset nor frequency.... and the service manual doesn't give any more detail either. So... I would love to feed you with whatever signal it is that you require for this A/D test, my dear little scope.... IF ONLY you would be so kind as to tell me WHAT IT IS exactly that you want !  LOL

Anyway, I put the cover back on.... and it STILL works !!! 
...though it did scare me a little (he likes to scare me, you must have figured this by now...), because before putting the cover back on, I checked/adjusted one last time the Y_OFFSET trimmer of the analog front-end, as a better safe than sorry measure, so that the "ground" levels were matching perfectly between analog and storage mode. But once the cover was back on and I gave the scope and quick test, horror.. the trace in store mode was a couple small divisions below zero ! NOOO !!!  But... after a couple minutes warming up (?), it progressively settled back to where it should be ! 

So that's one fixed scope, I say ! 

About the damaged connector, I posted on the forum about it, and no luck ! So indeed looks like it's hard to find...  I took close up pictures, and took all necessary measurements... in case I come across something suitable in the future.

And still 4 little knobs to replace on the front panel and that will be it.

On to the next project... and a BIG thank you to all who helped me through this 7 weeks, 6 pages long repair !   

Learned a lot in the process, and even got to build a cute little dim bulb tester ! 

[tautech]

And still 4 little knobs to replace on the front panel and that will be it.

Do a Google search for the Tek part # as there's a few sources for the common Tek knobs.

[Vince]

Parts list says 366-2049-01 .

Checked the Sphere and Qservice. The latter has it listed but none in stock, as for the former, that have a different part number for what looks like the same knob, but at 15 USD EACH? and I need 4 of them, plus expensive shipping to France + duty taxes..... that's gonna be for X-mas....

So I will get some knobs, now I know the can be found.... but not right now..... not at this price   

[tautech]

These ?


http://www.repair--parts.com/Industrial-Metal-Machines-Parts-/CNC-Equipment-Parts-/CNC-Wood-Router-Parts-/6-pcs-tektronix-tek-knob-cover-366204103-366-2041-03.php5
Email them.

[Vince]

Yes that looks like it... will e-mail.  Weird site though, looks like a '90s website lost in cyber space... the guy at the other end is probably dead by now.... or am I just being a bad mouth... will see ! 

[Vince]

E-mail sent... felt like sending a bottle to the sea !!! ....

[Vince]

Just read the ad more closely... it mentions "QService" !!!  So, this mailing-list ad must have been how tehy started teir business 20 years ago.... but now they have there own website... can't believe ther server hosting this mailiing-list still has not shut it down, must have been inactive for many many years now, probably !

Still, was worth a try....

So I know now that I can search for 3 different part numbers which all correspond to that same knob :

366-2049-01 or -03 or  366-2041-03

On Qserice I see there is also the this knob :

http://www.qservice.tv/vpasp/shopexd.asp?id=1902

Looks like the same one I need, but without the "bar". I guess I could use that as a fallback if need be. I mean I don't care about the bar, I have never looked at it in practice ! I couldn't care less. When I grab these knobs I hardly look for the bar.. I just get hold of the knob and turn it in whatever direction, and whatever amount, is required to get the trace where I want it to be ! I don't care about the bar...

Better no bar than no knob at all, IMHO !

[David Hess]

Also what's not clear is the A/D test. CRT instructions ask for an "appropriate" ramp signal, without giving actual amplitude and offset nor frequency.... and the service manual doesn't give any more detail either. So... I would love to feed you with whatever signal it is that you require for this A/D test, my dear little scope.... IF ONLY you would be so kind as to tell me WHAT IT IS exactly that you want !  LOL

Page 6-20 describes the test signal.  25 kHz 6 divisions and then double the amplitude to overscan the display.  Set the timebase to 10us/div.

[Vince]

Oh no, typical ! That was on the last page ! The only one I did not read ! All teh others tests were described one by one, foloowing the order in the Diag menu, so when I saw that the A/D one was skipped, I thought well too bad, and that's it, I did not read the last page thinking maybe it may talk about it  !    Bad boy Vince, bad boy....

I just printed the procedure and rushed to the bench to try it out, of course !

Procedure says you need a x10 attenuator which I don't have, a 50 ohm terminator which I don't have either, and a Function Generator with minimum 99% linearity, which I probably don't have since my Generator is an old 1970's basic analog Philips unit ! Never had a very stable amplitude nor Frequency. As for linearity well....

Still, it was either that or nothing. So I fed its 50ohm output straight to the scope, no attenuator, no 50ohm load. Adjusted the frequency as close I could to 25kHz, using the 2232 cursor measurements because I don't have a counter either ! LOL  Yes, still lots of gear to get for my lab...
At least I could adjust the amplitude just right.
Well I guess the attenuator they ask for might be because the  FuncGen tehy ask to use, a Tek FG503, might not have a built-in attenuator to output the required amplitude. But my Philips unit has both 20 and 40 dB attenuators available, so no worries, I switched the x10/20dB one, that will do.
Can't do much about the 50ohms terminator though... I just don't have one.

Then I ran the A/D test and hoped for the best ! Worked just fine !  On both channels, I get either 0 missing code, or one single missing code, about 50/50% distribution. I put that on the horrible output of my generator. Even by eye, the waveform was jumping all over the place (frequency not stable) and signal was very noisy (due to the required very low amplitude I assume).   I turned on the averaging which cured both problems, giving a stable and clean signal.. but I guess the A/D test processes the "raw" input data, not the averaged signal... So in these conditions I find it quite excellent to only have one missing code, and only part of the time ! 

So looks like this scope is working like new ! 

I am so glad to have it back, and feeling better than ever ! 

BTW Dave, I gather from your previous posts here, that you have several 2XXX scopes ? My 2232 is my only one of this era so I can't compare, but I am wondering about the big time base knob : I find it extremely "stiff"/hard to turn.. especially compared to that of my old Hameg scope which is absolutely feather light in comparison, like your typical detented rotary encoder, say, to give an idea.  So the Tek being so stiff, makes the knob feel a bit "spongy", every time I turn it, I am scared that it will (again) slide on its shaft and the dial loose track of things again ! Admittedly that knob has not one but TWO screws, 90° apart, to secure it to the shaft... but still, even though I tighten them quite firmly, it still doesn't feel "right".  Is this stiffness normal on these scopes ? Or is it just the switch getting stiffer due to age, and I could make it a bit smoother with some lubricant/contact cleaner spread liberally in it ? Or is it just the knob that could need to be replaced altogether ?... now that I know of Qservice and Sphere, I guess if need be I could order a "new" one.
I just wish Tek had machined a key on the shaft, or splines, or something.. ANY thing.... but not relied on a smooth/slick shaft 

5 AM again ! LOL   but this time it was not because I HAD to, it's because I CHOSE too, quite a difference ! 

Going to bed anyway...