Tektronix 2215 Scope repair

[Vince]

Hello everyone.... oops, I did it again.... I bought yet another old CRT scope ! I just can't help, I kinda have a fetish for these things 

It's a 2215 , will grow my little collection of Tek scopes. The problem is that as time passes, I happen to grow an interest for an ever growing number of different Tek scope models... and then there are the non-Tek scopes as well !

The 2215 is a very, very basic scope, only 60 MHz and apart from a delayed time base and a cool 10 turn aluminium vernier to go with it, it doesn't have much going for it to be honest !  But I gradually find an interest in more and more of the 2000 series scopes, many of them have a little something that sets them apart in some way.

As for the 2215 particularly, I recently learned on TekScope that it might well be the only Tek scope to feature an automatic brightness control circuitry. I always wondered why such a thing was not standard on just about every analog scope, since it's a real pain to have to constantly manually adjust the brightness knob every time you fiddle with the time base knob/sweep speed. So, I was curious to see for myself how this thing would work in practice, study the schematics/service manual for it, etc.  From what I understand this feature was, in practice, more of a pain than an asset... probably why Tek dropped the idea... then a few years later, the first TDS all digital scopes arrived on the market, rendering this feature irrelevant anyway.

So, when fate put a dirt cheap 2215 on my way... I just could not resist, bought it !
Found one locally advertised at only 15 Euros ! Usually any 22XX scope here, goes for anywhere from 150 to 300... "as is", take it or leave it, no questions asked...  So when I saw this 2215 for only 15 bucks, with a front panel dirty but other wise in excellent shape, with no scratches or dents or damaged or missing knobs... I e-mailed the seller the same day. Obviously the seller replied that he got a million calls and that he intended to pull the ad, the time to rethink pricing...
He had absolutely zero clue what it was worth, nor how to operate it, hence why he sold it real cheap "untested, for parts or repair".

I told him all about the scope, and in the end he agreed to sell it it to me rather than re-listing the item and sell it to someone else who no doubt would have been able to offer more money than I was willing to pay for such a humble and untested scope. We agreed on 40 Euros, much better than the original 15 he wanted for it, and muuuuuuuuuuuuuch better than 150/200/250 that I would normally have to pay ! So it was win-win...

I received the scope yesterday,  was so impatient, I started working on it just as soon. I plugged it to see what it would do, and not do...

Pretty simple to describe the symptoms : "No Power" ... not a sausage ! LOL
Not even a spurious attempt at coming to life, not even a split second blink of an LED, not a hissing from the SMPS, nothing. Just...... plain... cold... dead.

"Great" I thought, probably a power supply issue, gonna be an easy one ! Well, relatively easy at least. Last summer I had an in depth experience repairing the SMPS of another 2200 scope, my 2232, here :

https://www.eevblog.com/forum/repair/tektronix-2232-scope-in-need-of-a-doctor/msg1226100/#msg1226100

I assumed the 2215 might therefore have at least some similarity with the 2232 SMPS, so I cracked the think open to check the basics without the service manual just yet, just with my memories from the 2232 repair. Bad luck... when I pulled off the PSU metallic shield to expose its guts, it didn't look anything like my 2232     So I looked the PSU schematics up in the manual... the circuitry looked familiar bar an important bit : the pre-regulator section was of a completely different design. The 2232 pre-regulator is implemented with a classical PWM chip driving a MOSFET... but that 2215 used a TRIAC !!! I thought these things were only good for domestic appliances and motor speed control for portable electric tools ! Never thought one would design a PSU for a precision/sensitive lab grade professional instrument, based on a TRIAC ! You learn something every day.  The TRIAC was triggered by an SCR, itself driven by an opto-coupler.
Then there is a little daughter board to handle current limiting.

"Great" I thought... something new to play with ! But ... no. When I opened the PSU, first thing I saw was a large board, horizontally mounted, covering all the area... board A18... and there is no such thing in the schematics... Then I pulled that board to look at the little current limit board that was supposed to be mounted underneath it. That's board A19.... but nowhere to be found on my scope ! Also, half of the stuff was missing from the main board/ A10... eh ?! Lots of unpopulated foot prints in the PSU area of the main board... hmmm...

The manual can be downloaded from the Tek Wiki , around 22MB :

http://w140.com/tek_2215_service.pdf

For those with a slow connection, I put together a mini-PDF gathering the various pages pertaining to the PSU, see attached below.  1,8MB, more manageable I guess.  Oh no, it didn't go through, file size is limited to 1MB ... well sorry, I tried ! 

Turns out I found all the answers to my dismay, when scrolling at the very bottom of the manual... turns out Tek made some heavy modifications/revision to the PSU at some point.  Basically they dumped all that fancy TRIAC/SCR/opto-coupler/current limit board altogether, and retrofitted the previous design with the PWM circuitry. Of course that would mean re-spinning the main board, because the PSU is part of the main board... out of the question. So instead they put the PWM pre-regulator, and most of the PSU circuitry actually, on a daughter board, laying flat on top of the existing implementation. That's the one you can see on the first pictures, upside down, with the plastic transparent shield/guard.  They simply unpopulated the main board, ran two wires to tap the mains pretty much straight from the mains switch, then at the output of the regulator, just ran two wires to the main board, going straight to the inverter.

The beauty of it is that the access to components is now 10 times better/easier than it was in my 2232 ! On the 2215 that regulator board is a the very top of the chassis, every thing is readily accessible for you to closely look at , and probe. Plus, the wires are long enough that you can flip the board upside down and access the component side as well, while the scope is running !  What a luxury.  This will no doubt make trouble-shooting so much more pleasurable...
Also, it is now dead easy to break the link between the regulator and the inverter, so you can bypass the former and feed the latter directly, to figure out quickly if the fault comes from the regulator, or the inverter/downstream parts of the scope : the wires are already there, just unplug the regulator board and tap those wires !

So I did just that. First I measured the output voltage of the regulator, while loaded... got around 1 Volt , weird.  I would expect either nominal voltage, or zero if the PWM chip cut power due to a short/overload situation...  Having "some voltage" just didn't make a lot of sense to me.

Still, to make sure it was not the inverter/scope loading abnormally the regulator, I disconnected the regulator output wires and measured again its output voltage.... now 13 Volts or so !  Instead of around 40V.   30% there, getting better by the minute !    Still didn't quite make sense. But at least the regulator was not loaded, so had no excuse, hence I declared the regulator faulty, not the scope itself. As a quick sanity check, I powered the inverter directly with the bench power supply, set to slightly around 40V.

Success !  Houston, we have a trace ! Winner winner chicken dinner !  

I then spent 5 minutes checking the basic features of the scope to see what works and what does not...  behaves quite weirdly. I think 30% of it is definitely due to really dirty switches and knobs that need a good dose of contact cleaner. 30% might be due to that auto brightness feature playing up and interacting in some way with the trace. 30% of it might come from a possible slight under voltage condition throughout the scope, as I set the bench power supply only roughly, as at the time I did the experiment I had not yet found out that the service manual did cover this regulator board... so I was only doing it from the memories of working on my 2232 SMPS.

So that's where I am at at the moment !   

1) Narrowed it down to the regulator board, now need to trouble shoot this board.

2) Possibility of secondary/minor faults further down the line, but maybe not. Might just fix itself once the regulator is back up to specs.


Well actually I already started to do some basic trouble shooting on the regulator board last evening, was real tired and late for bed, but did some quick tests none the less (without the schematics at the time):

- Visual inspection revealed nothing. No damaged components, no burn marks on the PCB, no fried components... it all looks good.
Only thing that caught my eye, are the leads of a couple small Zener diodes : they are completely black, like they were painted. But it does not look like they overheated... their bodies and the surroundings show no sign of overload/overheat what so ever. It basically looks like the leads of these Zener diodes were painted black at the factory ! Can't figure out why though... anyone ever seen this before ?? I am curious.

- Checked all the input protection and filtering section : all good, mains goes through. Full wave bridge rectifier good too. Mains filter cap good (enough, don't have an ESR meter just yet) : DMM shows a steady 320 DC Volts across it and no AC component on top of it. That rectified mains does make it to the drain of the FET.  Did a quick diode test on the FET : it is not shorted... I kinda expected that, otherwise the regulator board would be in a pretty sad shape ! LOL  I should know  because this is precisely what happened to my 2232 !  Not a pretty sight indeed..... no such thing on this 2215.

So maybe the FET is faulty but opened (is this is a possible failure mode  for a FET ? .... ), or the surrounding circuitry fails to drive its gate.

I need to investigate this board closer now, and fire up a scope to look at some waveforms here and there... 11h30  PM here, getting tired.. I will start work on it seriously tomorrow....

Anyway, that was for the introduction ! 

Now let's work this baby until it behaves !   

[tautech]

Cool Vince, and great write up thus far.
Looking forward to following this.

Off to study the schematics........

Be careful with this one Vince, the pre-reg calls for use of an isolation transformer but if you haven't got one a differential probe will be even safer.
Luckily you have some listed DC values to check against first in some effort to zero in on the faulty part.
Next I'd be checking the all the smaller caps then focus on active components if the resistors test near enough in circuit.

[Vince]

Hi Tautech, thanks for stopping by  ! 

Don't have a differential probe just yet, too expensive (might go for the "EEVBLog" rebranded probe, since I understand it's good quality and excellent value for money).

However I do have an isolation transformer, rated 100VA, will be plenty enough for this simple/low power scope. Worked just fine when I troubleshot the SMPS of my 2232... yeah you better watch what you are doing when hooking another scope to the primary side of the SMPS, as it's not referenced to the chassis like the secondary side/scope is.

I have not yet started work on that regulator board, however some additional thoughts came to me :

- As I said the rectified mains does make it to the drain of the FET... there is an important and very happy conclusion that one can draw from this : it means that the only painful/proprietary part on that board, the switching coil/transformer, is NOT blown (like it was on my 2232 due to the FET shorting). So now I already know that whatever is wrong with this regulator board, it will be a cheap and readily available part ! So this thing CAN be repaired easily I would think ! 


- I might have a lead/clue about the 13+ Volts I measured at the output of the regulator, while not loaded. If you read the manual and look at the schematics, 13/14 Volts is the voltage that the circuitry creates at power up, to supply power to the PWM chip. Once this voltage is reached, the PWM chip is able to operate, ans starts acting on the FET's gate to regulate the output voltage. So it's possible that even though the FET is constantly open/not regulating anything, the circuitry, the way it is designed, all by itself, maintains a 13+ Volts level in there. So this probably means that  :

A) The circuitry that powers the PWM chip, is fine
B) The FET is just fine.. just not being driven.

The fact that the output drops to next to zero, 1 Volt or so, when it is loaded by the inverter, could be explained because the circuitry that produces the 13+ Volts to power the PWM chip.. well is only designed to spit out a very low current, just enough to power the chip !  To start with, that circuitry is fed by the rectified mains, 320Vdc via a 150K resistor ! So that's 2mA max, not much at all. So if you hook it up to the inverter, which draws around 750mA, there is no way it can keep up ! So the voltages I measured suddenly make perfect sense in the end...

So from all this, my first guess is that the problem is either the PWM chip not working, or it works but it fails to drive the gate because the gate driver (external/glue circuitry) is not working. 

That will be my starting point...   

Also, it is now very probable that I had indeed a slight under voltage condition in the scope when I tested it and witnessed weird effects here and there : as can be seen in the picture, I set the bench power supply to about 40,5V.   Now  after the fact, I read the manual and it states that the PWM chip regulates at 43V, so I was 2,5 Volts short.   Given that the inverter does not regulate anything (the reference -8,6 Volts can be adjusted, but it is not regulated. So for a given setting of that trimmer, I think that a lower input voltage/regulator output, will automatically lower the voltage of all the power rails... ).
So here again there is hope that a properly working regulator will fix some of the issues I noticed during my brief test ride. 

[Vince]

Oh, just got a message from the seller. Says the scope belonged to his dad who recently passed away. He is delighted that I managed to fix it and that the instrument will enjoy a second chance/new life instead of being dismantled for parts, or just scrapped altogether.

That's heart warming  I now have a duty to finish the repair of this scope, give it a good clean, and keep it with me rather than reselling it to make some profit out of it, which had never been my intention to begin with anyhow...

[oldway]

https://www.eevblog.com/forum/repair/help-with-tektronix-2213-power-supply/msg1530746/#msg1530746

Your posts are too long for those who are not native English speaking. 

[Vince]

Hi Oldway,

Thanks for the link. Looks like the OP, unlike me, has indeed the TRIAC version of the regulator, and is resorting to using brute force to get it working, replacing every component 'til it works, rather than trying to diagnose/locate the fault. I guess we can't always afford the "luxury" of taking the time to properly trouble shoot a board. As for me the trouble shooting / detective work is precisely what I am after, the successful repair is only a secondary goal, the icing on the cake so to speak. What matters most to me, is the journey, not the destination per se.

I am really sorry if you don't like my long messages, though I understand/remember from previous contributions of yours on my past topics, that you are fluent yourself. I guess you speak for others then ?

I would argue the following, if you allow me :

1) I am hardly a native English speaker myself, being French, which are universally recognized as probably the people with the least capabilities when it comes to grasping the English language, that the earth has ever carried. So if I can (about....) do it, anyone can do it, for sure 

2) How did I learn English ? By reading poorly written on liners on forums, comments on Youtube or Twitter or such atrocities ?! Definitely not.  The only way to learn English is to read hundreds and hundreds of pages written by actual  native English people, who master their own language decently. Read technical papers, read the news papers, listen to some BBC podcast programs in the UK, read, read read..  The more you read, the better you get, just like one learns his mother tongue effortlessly simply by listening to it all day long. It progressively but surely makes its way into your brain.

3) Learning English is much more enjoyable/easier if the material relates to something you are interested in : electronics here.   This forum is an opportunity to learn English the easy way.  I would not want to learn English by digesting piles of big books filled with law articles, for example...

4) Reasonably, if one comes to an English speaking forum, he has to be prepared to make the effort to keep up, rather than expect the whole world to accommodate for the lowest denominator. If I had to go to a Russian forum to say, seek info on an old nixie tube or whatever, I would hardly expect our Russian friends to post on liners just because some French chap decided to get on board.

5) Don't get me wrong : this post is long because it originated from me, and was meant to detail/introduce this repair, tell the "story". Subsequent messages don't necessarily have to be as long... 
However If I was REPLYING to someone, who clearly was not fluent and was clearly looking for a quick answer to fix his problem at hand, then yes of course I would make the effort to use simple words, short sentences, and be as synthetic as can be, to help him. But here on this topic, I was not replying to anyone, I was creating my own stuff.. so I dare to think that one is free to express himself the way he sees fit at a given time.. just like anyone is free not to read me, or not reply to me if he gets bored and falls asleep before he could even make it to the end of the post. Every one is different,  people just pick what they like and discard the rest, it's only natural...

6) In my repair topics, the whole point is to share my thoughts/trouble shooting process, since this is what I enjoy, and this is where I think the value is. If all I wanted was to fix this scope, I would not even bother posting on this forum. I would just fix it and not post anything about it... but that would suck. I like and enjoy reading about others repair stories, seeing the guts of a piece of gear I have not seen before... so logically, I like to contribute myself, give a little something back, share repairs when I have the opportunity to do so.
What would be the point of me posting a one liner like: "Hi fixed 2215 woz preregulator replaced a few random caps and that deed it, cool bro ! ". How useful would that be ? Both in terms of learning something on the technical side, or in terms of improving one's English skills, for the many foreign people like me ?

7) Experience on countless forums shows that I am the only one on earth to bother writing such long/detailed messages on forums. So really, not reading my posts will hardly harm anyone, not a big loss... 0.00000000000001 % of the total amount of messages on here.

8 ) What about Dave's EEVBlog ? He speaks at the speed of light, and seemingly without ever taking the time to breathe, for sometimes more than one hour. Should he make 30 seconds videos, speak like he would to a 2 year old, and use only 10 words of vocabulary, just because non native might (and do) watch his YT channel ? No, I surely don't expect him to do that, nor would I ever want him to ! Just the opposite : The more he speaks, the better I get at understanding him, the more new expressions I get to learn. The richest/more varied the vocabulary, the better ! Same goes for any other video blog of course. The brain is a machine that needs to be trained : you need to feed it with material or it will starve and progressively but surely....perish.


9) ... no there is no #9 ! 

Somehow I felt the need to justify myself, God knows why...  maybe because you used a big "thumbs down" icon, that made me think you were really extremely upset, which I felt was ungrounded.

Next time someone "complains" about my long messages... I will just link to the present post , will save me the hassle of explaining the same things over and over again.


I apologize to every reader (ie, the very few who can cope with my writing !  ) for this off-topic mini-essai... back to the repair now, we got a scope to fix ! 

[oldway]

Sorry, too long reply for me....
I'am interested in solving technical problems, not to learn English ....I already speak and write 3 other languages, I think it's enough.

[Simon]

Sorry, too long reply for me....
I'am interested in solving technical problems, not to learn English ....I already speak and write 3 other languages, I think it's enough.

Then go away and don't come back to this thread. No one has to meet YOUR expectations!

I am reopening this thread and any more off topic crap will get the offenders banned. We seem to have far too many people here that think they can make their own rules up.

[floobydust]

We had dozens of 2215's in the lab as workhorses when they first came out in the 1980's and at first I loved them then it turned to swearing and cussing.
With even a few years age, the trigger section degraded somehow and the scopes became finicky, they would not trigger reliably on simple waveforms.

Note the Tek 2215 Service Manual has many ECO updates (last pages), including to the trigger circuit.
I wonder if Tek rushed to get these out to market. It could be Trigger adjustment trimpots R482, or R511 or I faintly recall it was varactor diodes aging but it doesn't make sense they (varactor) was used for this purpose.

How do you find the scope performs compared to others in your collection? I was going to get one but not sure if a 475 kicks butt, to go back a generation.

[Vince]

Alright.  Spent some more time on the pre-reg board. I did all the checking I could do with the DMM, before I resort to firing up a scope to check waveform.
Checked all the passives, lifting a leg here and there when needed, to get a reliable measurement.

- All small signal transistors check OK (I pulled them to be 100% sure...)

- All 3 electrolytic caps measure fine and don't leak, but no idea about ESR since I don't an ESR meter just yet. Soon hopefully, though.
However for the kind of symptoms I see, I don't see how a high ESR could explain it. So I will replace them for good measure and preventative maintenance, because they are bound to go bad later on (scope is 36 year old already...), but I don't expect it would make any difference for the problem at hand.

- The 4 tiny caps around the PWM chip, come in a funky packages. 3 of them come in dipped bright green axial packages, and two come in a transparent glass package with  red terminals, making them look disturbingly like an old Germanium diode !      However I was not too shocked, because I learned about their existence of such packages for caps, from a gray beard who designed crystal oscillator prototypes in the 70's. These glass dielectric caps had the advantage of very high temperature stability, IIRC, which was critical in getting the oscillator to be as accurate/stable as possible, which was his goal back then.

Anyway one of these green caps broke in two when I lifted one of its legs... was gentle I swear. So I guess this cap was just waiting to fail anyway. That was C929.   Then the two remaining green caps (C925 and C920) don't instill a lot of confidence either : I lifted a leg on both them, yet the DMM wouldn't register anything, display was stuck to 00.00nF, didn't budge. That means these caps are.. open. So all of these green caps will be replaced.

Trap : in the pictures, you can see not 3 but 4 of these small green packages... one of them is NOT a cap, but a precision/ 1% resistor ! Go figure... never seen a resistor packages like this. Also, the accompanying resistor (other leg of the voltage divided that sets the  output voltage, R921/R922), comes in a "normal package. So it's not consistent...

- R937 reads a bit high at 8.75K for a nominal 8.3K 5%. However I do'nt think it matters at all, given it's purpose on the board. It's in parallel with the filter cap on the output of the regulator. Not even sure what it's for...  The output being at 43V, is above the usual "safety" level, 24V IIRC (my meter gives a warning above 24V anyway).  So maybe the resistor is meant as a bleed resistor for the capacitor. However I don't see the point of it since the regulator board is supposed to be permanently wired to the inverter, which draws 750mA or so... which would bleed that capacitor real quick anyway...
However if Tek really cared about safety and bleeding capacitors... I would have found it much more important and logical, to bleed the high voltage cap that rectifies the mains : 320Vdc  !  But they is no bleeder there... I timed it, takes quite some time for the cap to discharge, over 30s.

Other hypothesis : that resistor is there to ensure a minimal load on the regulator, for it to work properly. That wouldn't be uncommon. However, again... the inverter it supplies is plenty enough of a load. So, this dummy load would be of any use only when a service tech disconnects the inverter to trouble shoot the SMPS... just like what I am doing right now. IF so.. how thoughtful from you , Tek !  Thank you. 

- R911 reads really too high : 180K for a nominal 150K. Sure it's not a precision item, just the common 5% type. However even at 5%, the maximum value would be 157,5K ... and I measured like 180K.   I am not too sure what kind of impact that would have on my symptoms though, but at least it's worth keep it in mind...  Oh, this resistor as you can see is the beefy 1Watt thing that feeds the 320Vdc to the circuitry that powers the PWM chip.

- Checked the switching coil/transformer / T933. As I suspected, and hoped, it is just fine, hooray. All 4 coils in it, are neither open nor shorted. They all read plausible values, in tune with what I measured on the similar tranny in my 2232 scope. So that's a big relied indeed !


After I had checked all that, I resoldered all the legs I had lifted, reflowed a few solder joints for good measure, then powered the thing back up to see if it somehow made any difference. Did not. As I expected, all the green caps, which I removed because either broken or reading open, didn't impact the symptoms the slightest bit : we still have 13+ on the output, and only 2,8V for the PWM chip power supply instead of 14V...

So I then removed the PWM chip itself (and soldered a chip socket in its place while I was at it), just to see if it might be defective and load it's power supply too much, to the point of dropping it from 14V down to the 2,8V I was seeing. Without the chip, the supply goes back to 10,7 V or so, much better, but still not the expected 14V ! But this time the PWM chip can't be blamed.. so the problem must lie elsewhere... in the circuitry that's in charge of generating that 14V to begin with.   I highlighted that circuitry in the schematics below.  Also attached, the 3 pages of text that describe the operation of the pre-reg, including this little circuitry. 
Basically you have C913, 100uf electrolytic, which powers up the PWM chip. At power up, it progressively gets charged by the rectified mains, through R911, that 1W 150K resistor that I just discussed above, which reads 180K instead of 150K...
Once the voltage across that cap reached 14V, the circuitry then regulates it at this value (by means of transistor Q917 and a Zener diode VR917) and sends this juice to the PWM chip power pin (#12) using transistor Q915.

So, I measured the voltage drop across the Zener diode, which is a 6,2V one... got 2+ volts IIRC, don't quote me on this, which didn't make sense to me.
The Zener tested fine with the DMM, in the forward direction. Obviously I could not test it in reverse bias.
So since I was still looking for leads on this repair... I bit the bullet and pulled that Zener so I can test/characterize it properly, on a bread board.
10K in series, reversed biased of course, then ramped up the voltage progressively with the bench power supply. What I witnessed was most strange.. the Zener is definitely bad, and sure enough the circuitry can't work properly with this things in it !
Now I have no particular knowledge or possible failure modes for Zener diodes, other that that common to any PN junction : either shorted or open...

So this is what happened on the bread board... two problems at once :

1) It's a 6,2V Zener, but somehow the "knee" drifted dramatically : it starts to conduct at a much lower voltage : 2Volts or there about ! 
2) Once it starts conducting, hence we have passed the knee... there is another problem : it still does not regulate anything : the voltage across it keeps rising and rising, somewhat linearly. In other words, it's dynamic impedance has increased dramatically ! It does not "regulate" anything anymore.

So that was quite interesting to me, this scope repair was worth doing for seeing/learning these failure modes in Zener diodes...

I had a few Zener diodes in my drawers (sadly no 6,2V nor any values that I could combine/put in series to add up to 6,2V), new, but about the same age as the scope or almost. 25 years at the very least, or shy of 30.
So I was curious to test them all, see if I just got unlucky on the Tek scope, or if there was a pattern there...
I was not disappointed : had of old new Zener diodes showed the exact same issues !!! 
So there you go : it's definitely not uncommon, it was not just bad luck.. this a definite failure mode for these things !
Not sure how relevant it is, but I also noticed something : out of all my new Zener diodes, only the low voltage ones were affected. Those rated at under 10V.
I had 3,3V ones, and 4,7V.  Then I had 12V, 15V and 25V.
Might just be a coincidence.. but maybe not, because there is a pattern : I had around half a dozen diodes of each value. I tested 100% of them. I noticed that 100% of the low voltage showed this issue, and absolutely NONE of the "high" voltage ones did. 

To finish up with the Zener : I figured the black coating on their leads : it's a fine oxidization layer, easily scrapped off. Only the Zener diodes on this board have this... I guess their leads must be made from a different material than all the other components...

So with a 6,2V Zener that starts conducting at 2V or about, and which is unable to regulate anything, no wonder it doesn't work.

There is still a bit of mystery as to how I can measure 13+ Volts at the output of the regulator, when the PWM circuitry  puts out on ly 10+ V ....
Maybe the tow voltage levels are unrelated then. The only path I see for juice to get to the output while bypassing the PWM circuitry, is via the FET.. so maybe it conducted a little bit even though it is not driven. So might still be faulty in some way, just maybe. I will pull it to see if the output drops to zero.  Ought to.


OK so that's it for today.

I will now order a few components to be able to go further :

- Zener diode, hoping it would cure the problem

- Replacements for all the 3 tiny green caps that either open bad or broken in two...

- New electrolytic caps, for preventative maintenance

- New R911, 150K 1W high voltage resistor in case it reading 20% too high maybe part of the problem, along with the Zener.

- And just in case : a new PWM chip and FET, so I don't have to pay shipping twice just to get hose if I do happen to need them.
Sadly the parts list, maybe because it's related to a revision change and not the main/original parts list... does not give enough technical detail to be able to order a replacement part : just says "N-Channel MOSFET" .. great, how helpful ! So I can't check whether the FET in place is the correct one (might have been replaced and be part of the problem..). But well, I will just go buy the info in my 2232 part list. IIRC the FET is not critical, back then we found several possible modern replacement. Hell I might still have some spare/new FET's from that 2232 repair, that I could re-use in that 2215... need to check my "archive" box...

- Might also recap the main board, to be consistent : the inverter and the secondaries.


Oh ! Forgot a nice thing : the non-working delayed time base I noticed during my quick test ride... works just fine in fact : it's just that always.... I forgot that to for ti ti work/display the trace, you need to get to the "B Trigger level" knob in the upper right corner of the front panel, and cranck it all the way  CW, in the "run only after daily" position... Works much better now ! Operator error, hum hum..


Some attachments and off to bed :

- Principle of operation of that regulator board
- Board layout of the board
- Overall schematic for the SMPS, which of course includes the regulator board, which occupies half of the schematic just by itself...
- Annotated Schematic of the regulator, highlighting my main focus : the circuitry responsible for providing the 14V to the PWM chip.. which we need to get working.

[Vince]

...

[Vince]

We had dozens of 2215's in the lab as workhorses when they first came out in the 1980's and at first I loved them then it turned to swearing and cussing.
With even a few years age, the trigger section degraded somehow and the scopes became finicky, they would not trigger reliably on simple waveforms.

Note the Tek 2215 Service Manual has many ECO updates (last pages), including to the trigger circuit.
I wonder if Tek rushed to get these out to market. It could be Trigger adjustment trimpots R482, or R511 or I faintly recall it was varactor diodes aging but it doesn't make sense they (varactor) was used for this purpose.

How do you find the scope performs compared to others in your collection? I was going to get one but not sure if a 475 kicks butt, to go back a generation.

Hi FloobyDust,

Thanks for sharing your experience with these things... not too promising it seems !  LOL
I will keep your remarks in mind when I get to test it properly...

For now it's still in pieces so it's early to give any opinion on the thing. That said, even assuming I could get it to perform superbly, then the cold fact remain that it is a very basic, very low end scope indeed... you can't change that.  Based on this alone, I can safely that my 2232 is vastly superior to this 2215, in every department...  100MHz BW vs 60MHz for the 2215, storage mode, cursors, XY plotter output, GPIB connectivity.. My 2232 is my main scope. Then if I compare that 2215 to my TDS 544A of course it's not fair for the 2215 ;-)

However I could make a more meaningful comparison, that would be with my old Hameg HM604.  Again an old, basic analog scope, same 60MHz.
In this case, I would choose the 2215 over the Hameg any day. The Tek is just in another league.  Everything looks and feels more refined, and it just looks 10 times better as well, though this is obviously of no technical relevance, and entirely subjective ;-)

Now compared to the 475 you envision.. well I have no experience of this scope, so not much qualified. However from what I understand here and there, they were very dependable scopes, well built, everybody loves them... Meant as a service scope so it's very compact, however they crammed as much performance and features as could into these scopes. So unlike the 2215 which was meant to be, from the get go, an extremely basic and low spec scope... the 475 was just the opposite : a quality and high spec scope. The BW alone speaks for itself : 200MHz and 250 for the 'A' revision.. vs 60 for the 2215. Right there, they are not at all in the same league !

So I would say... yeah, get a 475 any day over a 2215 ! I will repair my 2215, but sure as hell I will get a 475 as well as soon as money permits ! 

You could ask all the gray beards for confirmation on the TekScope forum, but I don't think I am too far off....

[tautech]

WRT the FET, a Tek cross ref will save your bacon, maybe this one:
http://www.sphere.bc.ca/test/tekequiv.html

At the bottom right of the page is another more comprehensive one Sphere ask not to be linked.

Good progress Vince, I'll review it in detail today so get some sleep buddy.

R911 drifted high mightn't let the controller kick off as SMPS IC's usually have UVLO.

[floobydust]

For the mains SMPS MOSFET, Tek changed to a different part in the 2213,2215,2235,2236. Depending on the scope's serial number, see parts replacement kit 050-2242-03
151-1152-00 replaced by 151-1245-00 using the parts kit. Also the main DC-DC and OVP SCR were updated, bunch of resistors added.

151-1152-00 IRF820/MTP475                    500V     3.0 ohm  2.5A 360pF
151-1245-00 MTP6N60E (was MTP6N55)  600V     1.2 ohm  6.0A 1,500pF

The MTP mosfets are obsolete and what I could find in stock at Digikey replacing 151-1245-00 looks like:

IRFBC40 600V 1.2 ohm 6.5A 1,300pF 50W
STP9NM60N 650V 0.75 ohm 6.5A 452pF 70W
STB6NK60Z 600V 1.2 ohm 6A 905pF (has internal protection zeners) but obsolete part

Some relevance: https://www.eevblog.com/forum/repair/suddenly-unable-to-start-tektronix-2235-(analog)-oscilloscope/msg1285884/#msg1285884

[Vince]

Thanks for all the good info Floobydust ! That should sort me out 

IRF820 ? This is what's on my regulator board.... so it was original then. Boy 3ohm Rdson, that's truly horrible... yes will definitely replace it with a better part. I might have some left from my 2232 SMPS repair, need to check...

The SCR is fine on my 2215, I mean it's not shorted...
Yes they definitely modified the SCR part of the schematic... on my 2232 the SCR is shorting the rail solid... but on the 2215 I see they added a beefy 3ohm resistor in series with the SCR.

I Paid a visit to my local electronics shop after work, picked a few components... he didn't have the tiny dipped ceramic caps (too "exotic".. will need to order them from Farnell or whatever), so I won't be able to get the PWM chip to work.
However of course he had a 6,2V Zener diode and a 150K 1W resistor, and the two electrolytic filter caps for the PWM chip and regulator output.... so hopefully I will be able to fix the 14V circuitry that powers the PWM chip, that's the most important ! 

But it's late and I am tired.... so will play with that tomorrow hopefully ! 

[Vince]

SUCCESS ! 

Just back from spending a couple hours on the thing. I first replaced the Zener. Bought two of them just in case.  Now being a tad paranoid about Zener diodes, I first proceeded to test them both on the breadboard... and well, again was interesting ! They were twin sisters, coming from the same reel, adjacent to each other. Yet... somehow one was distinctly performing better than the other one ! One was working just fine, with a sharp "knee", quite accurate at 6,2V despite the 5% tolerance which would allow it to vary quite wildly, from 5,9V to 6,5V roughly. Also, the dynamic impedance was very low, it "regulated" really well. OTOH, the other one had a rather "fuzzy" knee, and had a noticeably higher dynamic impedance. No need for a curve tracer to figure it out, it's easy to spot just by looking at the DMM...

So needless to say, I picked the best of the two.

As I hoped, it fixed it !        I now do get the expected 6,2V drop across the Zener, and the circuitry now produces 15V, much better !    Supposed to be 14 but hey, still happy.

Then, just in case it might bring it down to 14V, or make any difference, I proceeded to replace R911, the beefy 1W 150K resistor that measured 180K....
I bought a 2W instead (I like safety margins in high voltage stuff.... ).. and turned out it has the exact same dimensions ! Twice the power, but same size, hum...
The original resistor is old tech as you can see. Brown body and perfectly cylindrical shape, the kind you found routinely in old tube based gear... so maybe this type was not capable of dissipating as much power as the more modern kind, don't know.

Anyway, not sure why just yet, but replacing that resistor DID make a difference ! However not in the direction I would have liked : instead of lowering the voltage to bring it back closer to the desired 14V.... it threw it even farther : now reading 16,8Volts ! 

Still... not too fussed about it : the PWM chip doesn't really care, and it's got its own 5V reference anyway...

Then I replaced the two LV decoupling caps : C913 (PWM supply) and C934 (regulator output). Somehow that too made a (slight) difference : PWM voltage dropped by 0.2V. Go figure.   Notice just how much smaller the new caps are !  Yeah I know, modern caps are always much smaller than their old counterparts  but... these are really, really much smaller, had to measure them twice with the DMM just to convince myself... They are about a third of the volume of the old ones... even though I increased the voltage rating on both of them !
Yes... Tek really didn't give them much head room !  C913 is for the PWM chip, hence sees 14V. They used a 25V cap. OK, that's about reasonable on this one, just.  Still, I bought a 35V one, because price is so ridiculously low on these, and it would still be smaller than the old one, so no problem fitting it to the board.
Then there is C934. It sees 43V, the output voltage of the regulator.. .and they used a 50V rated cap ! Eh ?! 
I bought a 63V one...

Both OEM caps were Nichicon.. killed me to replace them with crappy brands... but well, that was just for test purposes. I will later place an order on-line from Farnell or whatever, and therefore be able to choose a good brand.  Oh, and Tek used 85°C caps as well ! And the scope does NOT even have a cooling fan ! Grrr...

Was looking good, so I thought what the heck, let's plug the PWM chip into its socket, and see what we get ! Of course with a missing/broken ceramic cap, and 2 dubious ones that look like they are open circuit... the chip should not be able to operate. However... I could at least, probably, see if its internal 5V reference was  working... and it was : reading 4,98V , a good start, it's alive ! 

Then looking more closely at the schematic, you realize that the broken/missing cap, is C929 : it's across the two inputs of the current limit comparator. I thought that must be to reduce signal noise. Useful, but not vital... chip should still be able to operate without it.  You bet it does !!!

When I powered it, I heard the typical "wizzz" noise coming from the board ! Hummm I thought, this thing sounds like it's alive ! 

I rushed to check the output voltage : 43,3Volts (for a nominal 43V)  : VICTORY !!! 

Then checked the PWM chip power supply : down to 13V now, woohoo !!! So the 16,5V I got initially, was nothing to worry about... as soon as it's loaded when then PWM chip is in place, it goes back down !    13.0V is not the 13,9V indicated in the schematic.. but again I think it's just fine...

So... the two ceramic caps that I thought were open circuit.... maybe they aren't after all ? Driving me mad... the DMM was not registering anything, even though it had enough resolution to measure them !  Worse case, may be C925 could indeed be defective and it would not show because its role is not vital : from the schematic it looks like its purpose is to reduce noise on the output, by lowering the output voltage/gain for high-frequency signals.
However C920 can't possibly be faulty as it determines the oscillator frequency !

Had a quick check with the DMM : PWM output is about 35kHz, manual states 40kHz, ball park...

So... seeing as the regulator board was working.. at least well enough for test purposes, I then hooked it to the inverter/scope.  Would the scope fly with its own wings now ? Sure it does ! And now the SMPS makes this horrible, high pitched pulsating noise that would make many suicide... and which is unfortunately typical of any and all 2200 scope, arf...  My 2232 does it too, takes 15+ minutes for it to progressively quiet down. On TekScope this very subject has been discussed at length.. many theories have been proposed as to this noise, many experiments were done... in the end, as far as I remember, there is no definitive answer, never mind a sure fix. Sadly. I guess as I get older/deaf, I won't even be able to hear such a high pitched noise, so I won't care anymore !     But until then.. it kinda drives me nuts.

So, yes, the scope now powers up and runs fine... I am a happy camper 

With the inverter hooked up, I measured again the output voltage of the regulator : yep, when loaded, it drops a bit, from 43,3V down to 43,15V .
Pretty much spot on.

So I call this a fix ! That was a quick one, I got lucky !

All fixed for the price of a 10 cents Zener diode, how cool is that.

Now I will order decent brand caps, and recap the whole scope : secondary side of the SMPS, and also I spotted a few electrolytics on the sweep board as well. Will inspect all the boards to find them all, there will be no mercy ! 

Once this is done, and once I have cleaned all the dirty switches... I think we can then test it thoroughly and see what issues might remain.

I can already see one... which we can probably spot on the pictures too : when I enable the delayed time base, there is a problem with the brightness of the trace : right after the part of the trace which is highlighted, there is "dark" spot where the trace becomes invisible. It's about the same width as the width of the highlighted part. Must be a lead as to what is wrong. Will need to check how the delayed time base works. Maybe a dodgy cap is creating some lag/ghost effect that creates this trailing effect.   So might cure itself once I have recapped the whole scope... hence why I won't bother investigate this issue for the time being...

Anyway, that's my second 2200 SMPS repair... I am starting to get the hang of these things, might buy another one if I can find a cheap one again (unlikely...) ! 

So, next : buy a bunch of caps to recap the whole scope, clean switches, and test the thing again.

See you in a bit then ! 

Bed time again...

[tautech]

Cool Vince, well done and thanks as always for writing it up like you do. 

Those old carbon resistor can drift like hell and any that are suffering any load, it's wise to replace them.
WRT the caps, both ceramic and polarized in the SMPS you can be damn sure Tek knew what they were doing to include them and probably for a little bit of trouble you might be wise to replace all in the PSU maybe with the exception of the primary side bulk cap if measured capacitance is still close to spec'ed value.
There may be different ESR specs for modern caps as opposed to what Tek used and little things like that could be responsible for PSU whine as it's essentially operating with parts outside the original design. Food for thought maybe.
I'm sure the whine could be addressed with some rework of the SMPS values to get the transformer operating in a happier place. You'd probably need a good current probe to do so and do some tweaks to the MOSFET ON time (less). 

Did the display show any problems when run on external PSU ?
If you use Horz Pos, does the dim region move ?

[Vince]

[Just proof read my previous message, corrected a million and one typos. Should make for a smoother ride now... ]


Hi Tautech,

WRT the caps, both ceramic and polarized in the SMPS you can be damn sure Tek knew what they were doing to include them and probably for a little bit of trouble you might be wise to replace all in the PSU maybe

Yep sure, that was on the cards. I am not taking any chances...
I notice that on the secondary side of the SMPS, the caps are really old "Mallory" branded caps, half a dozen of them, and that two of them have already been replaced from what I can see !  No doubt about that. They can be seen in my first round of pictures in my original post. So if only two of them were changed, it means they must have been bad to the point of leaving no choice, so they must have been reallllly  really bad ! So, all the other Mallory still in place are logically ticking time bombs... will replace all of these of course.

with the exception of the primary side bulk cap if measured capacitance is still close to spec'ed value.

Yes I thought likewise. This mains filter cap is kinda special... from what I understand and witnessed on my old Tektronix 317 tube scope, and many reports on TekScope, these caps Tek used were astonishingly reliable/long lasting. My 317 is from 1959 and the LV power supplies are still dead on, and ripple free. I just could not believe it.
So if I were to replace this special purpose cap with a generic 450V low ESR electrolytic, I might well be doing more harm than good, trading a proven and purpose designed cap, with an unproven one... Not a good move I feel.   That cap measures at 91uF for a nominal 75uF and +50% tolerance !  So in other words, it's still welllll within tolerance...
Will only replace it if there is a definite evidence that it's faulty and causing trouble.

There may be different ESR specs for modern caps as opposed to what Tek used and little things like that could be responsible for PSU whine as it's essentially operating with parts outside the original design. Food for thought maybe.

Yes. That said my 2232 made that noise even before I repaired it... I am fairly sure ?!  So if ESR related it is, then even non-modified scopes don't have the appropriate ESR. New caps because they are designed this way, and the old original caps because their ESR increased of the years... different cause, same end result...

I'm sure the whine could be addressed with some rework of the SMPS values to get the transformer operating in a happier place.

I will try to dig out that topic from TekScope. The idea back then was similar to what you suggest. IIRC the OP was suggested to add C and R trimmers to the PWM oscillator, so as to be able to fine tune it, in a hope that one could find a sweet spot where the transformer would not resonate with anything, and just be silent.

You'd probably need a good current probe to do so and do some tweaks to the MOSFET ON time (less). 

Yeah a current probe is on my list of things to get... I recently found a YT video about a nice one at a decent price, 450 bucks or something, going right down to DC, and up to 100kHz of bandwidth.
Now I think of it ! This could be an interesting and useful way to use this little 2215 scope ! I mean, let's be honest, I have zero use for it, my 2232 is my main scope and is much better. HOWEVER, the SMPS in the 2215 is WAY easier to work on, compared to the one in the 2232. Also, since I don't need the 2215, I don't have any problem having it non working, in pieces, for months and months... I just don't care.
So... I think I could use this 2215 as a test bed to work on this whine issue !    When I power the inverter from the bench power supply, the scope is perfectly silent. So the whine must indeed be coming from the regulator board/PWM tranny. And since this board on the 2215 is standalone, very low density and can be accessed and probed with so much ease.... it is the ideal candidate !   Then, once/if I can manage to make the regulator quiet, I could then apply the remedy to my 2232 quickly, meaning no significant down time at all !   
And then I could apply the recipe to any and all 2200 scope I might purchase in the future....

Yes I like the idea ! Playing with a current probe, watching waveforms, and I could even record the audio "track" from the transformer, as there is always a distinct pattern to it, which evolves as the scope warms up. It also reflects on the signal trace on the CRT, which "dances"/wobbles according to the pitch of the whine. The lower the pitch, the higher the amplitude of the wobble.

I also have a spare tranny from my 2232 repair. I could also conduct tests on it : feed it with a range of frequencies and wave forms/shapes to try and see what "triggers" it.

Yeah could make for an interesting little project, and hopefully a very useful one if it eventually succeeds and leads to a fix we can apply to all those 2200 scope, which are still plentiful on the market ! 

Did the display show any problems when run on external PSU ?

Hmmm, good point. Just tried it. Running the inverter from the bench power supply doesn't happen to make the slightest of a difference. 100% similar outcome.

If you use Horz Pos, does the dim region move ?

Yes it does. It "follows" the highlighted portion of the trace invariably, sticks to it no matter what.
Same thing if I fiddle with the delay vernier : if I move the highlighted portion of the trace from left to right, again the dark/dim spot on the right of it, follows suit.

Also worth noting : the dim area does not affect the entire height of the screen, so I guess it's safe to say the issue is not related to the auto-brightness feature possibly playing up, but rather located solely to the delayed time base section of the scope. Whatever circuitry is responsible for handling the brightness of the delayed portion of the trace, is having trouble restoring the nominal brightness of the trace, to the point of "undershooting" its target value, before progressively settling to the nominal brightness. Capacitive effect at play sounds plausible.... 
Even before I recap the scope, I might just out of curiosity, have a peek at the manual/principle of operation,  and schematics to see if I can figure something out...

[tautech]

WRT the whine, I suspect and in part you confirm that component values around the SMPS controller have drifted but heat after a short period of operation lets it drift back into happy regions where the noise abates.
But which components ? Could it be the IC ? Or even unrelated directly to SMPS timing as components in the switched current path will impact on the total circuit and therefore peak switched current. Check the 0.2R current sense resistor. You can see there's quite a few components that could potentially affect the IC's operation, from the VCC supply, current sense and timing. It all needs be carefully checked, particularly component values.
Let's say 5% parts are spec'ed but they measure nearly 5% from value. I think that would have me replace them and check new parts like you did with the zener.

When you start fiddling with switched magnetics you see some very interesting effects as the primary current gets high enough to cause primary saturation and this is what I suspect is happening. Just reducing the FET ON time a few ns can drop you below the start of core saturation and extinguish all whine.

I think in your position before you adjust anything in the scope to attempt to fix the whine, build the same circuit deadbug or breadboard and experiment with some timing adjustments to get a clear indication of just what values need tweaking. All you're after is ON time adjustment 'indication' and you don't even need to switch any 'real' power.
What is the SMPS IC ? Can you find a datasheet for it ? Is it still available ?

Fun times ahead and you can learn heaps if you want to want to properly investigate this one. 
Just tread carefully. Little steps. 

[Vince]

Thanks for all the good info, a good starting point !

The PWM chip is a classic TL494 from texas instruments. My 2232 had a TL594... not sure how they differ... internal block diagram is strictly the same. Whatever the difference, is gonna be very minor I think...

They are still readily available, as is the datasheet on Google.

Well, haven't checked availability of the 494 to be honest, but the 594 I had no trouble buying when I fixed my 2232 a year ago.

[floobydust]

Old thread about Tektronix PSU whine: https://www.eevblog.com/forum/testgear/tektronix-2235-prblems/

[tautech]

Old thread about Tektronix PSU whine: https://www.eevblog.com/forum/testgear/tektronix-2235-prblems/

Yes some pertinent info there but nothing that conclusively gives answers to this whine issue.
Probably the best hint is here:

I already had three 2235 oscilloscopes with this problem.
I have not disassembled and uncoiled the defective T906 inductors to find which winding presented the default because I did not see the usefulness of this information.
The inductance is not repairable anyway and it seems quite difficult to make one yourself because it is very compact.
NB; The ferrite core must have a gap.

The best way to verify if T906 is defective or not is to feed the scope through an isolating transformer and to look with another scope if the waveform of current on R907 (0.2R) is correct or not. (see waveform 39 - 150mV peak)

However to surmise that T906 could be the root of many problems without either insulation checks or dissection is just guessing.
That Vince's one quietens down after a period of operation points strongly to temp related component 'real value' issues in the SMPS controller circuitry IMHO and not a faulty T906.

[Vince]

Thanks for the help guys.

Will read that 2235 thread when I get some time.

As a starting point I think I will follow Tautech's idea, for the reasons he gives. Second reason is that my experience with my 2232 suggests that the tranny is not at fault :  that scope was noisy before I worked on the SMPS (with all original parts in there) and it was still noisy after I rebuilt the regulator. The old tranny was cooked (FET went short circuit, the coil/winding in the tranny burnt open), and the ferrite core had a big crack in it, and the two halves of the core were misaligned.

So, I assumed the horrible condition of the tranny was responsible for the whine.... but, not at all : the replacement tranny was in perfect shape, and even had the drop of white gunk / "silactic" in the center of the core (the old one did NOT even have that), that some trannies have, to keep them from vibrating/making noise.
Yet.... even that new tranny was making noise ! Still is.

I think what I would do as a starting point, would be to get a real cheap and dirty USB scope, so I can record a few signals for a few minutes after power up from cold. One of these signals would be the audio coming from the tranny, so I can then correlated the other (electrical) signals to the sounds they gets produced. I could then search for correlations. I could look at the oscillator of the PWM chip, at the voltage pulses from the FET, also the current in the FET/tranny.
I can't do that with my digital scope (TDS 544A) as it has only 50K points of memory, when I reckon I would need at least 100M points.
A USB scope isn't good for anything but... but this particular task : don't care about accuray, performance, bandwidth, or XYZ feature... just need  up to 1MHz B/W or there abouts, plenty enough, and stream the data live to the PC, where ample amounts of RAM will make it easy tro record many minutes of time correlated signals, which I can then "play back" / study...
Getting a current probe is gonna be the most expensive part by far... but then I always wanted one, so this mini-project is a great excuse/opportunity to finally get on, isn't it... 


The 2235 looks interesting.. from what I can see of it (front panel in pictures and videos here and there), it's basically a 2215 with a higher bandwidth, 100MHz vs 60.

Anyway, the whine is a project for later, possibly much later, no idea, will see... have other projects...
The immediate concern is to recap the scope and try to address this weird brightness issue in the delayed time base...
Let's have a close look at the various boards in the scope, and list all the caps I will need to order....

[Vince]

Spent my Saturday on the thing.... some good, some bad. Overall it seems like a very sick puppy, not sure I will ever be able to get it really spot on... but well for now I keep going.  I think Floobydust must be right... the 2215 is a troublesome beast that was rushed to market... ahem...
If really it appears hopeless, I would at least keep it as a test bed to investigate the SMPS whine issue.  Also, regardless of the outcome, I enjoy the hands on experience it provides me with repairing stuff. I always learn a little bit here and there. Also gives me a deeper knowledge of the 2200 series scopes, which can only be a good thing as I intend on getting more of these, and if anything it will help me get better at maintaining my current 2232 which is my main scope.
So, overall I am still quite happy that I bought this little 2215 scope.


OK, so what I did today :

- Went to examine the scope closer, to try and track down every electrolytic cap I could find, simply by visual inspection.


 Then tried to correlate that with the parts list. Doing so was interesting : I removed the sweep bard from the scope, so I could get a closer look at what I thought were 4 electrolytic caps... which turned out to be  in fact a set of 4 matched caps, Tektronix branded, precision timing caps... make more sense indeed, but boy do they look like old electrolytic caps... from a distance.
BUT wait.... there WAS indeed some electrolytic caps on this board, 2 of them, according to the parts list anyway ! For the life of me I could spot them at first glance. So looked at the component designator in the part list, searched for that on the silk screen underneath the board... OK found them ! Gee no wonder I had a hard time spotting them : they do NOT look like electrolytic caps ...  see picture. They are tiny, axial things, silver looking. They bear the same dimensions as a 1/4 W resistor, and look like a precision film cap. Learned something new then. Granted, these electrolytic are indeed "precision" (by electrolytic standards ie...) caps : +/- 10% , a far cry from the usual -10/+50% of the usual electrolytic. Heck, there is even on in the scope that +75%  !!! 
So in short : what I  thought were electrolytic caps.. where precision film caps. And what I thought were precision film caps... turned about to be electrolytic caps ! Lessons learned.

That was for the (main) sweep board. So in the end I have nothing to replace on this board. As for A11, the front panel board : one small electrolytic, not going to bother with that one I think. Don't think it can affect the scope that much (?), and is way to hard to get too : would need to remove the board, which I am MORE than reluctant doing, because when I tried to that on my 2232 last year, just about each and every of the small "position" knobs, the kind that just snaps onto their shaft... broke ! Plastic is way too brittle. They are hard to find and cost a fortune... still have not replaced them to date, a year later almost ! 
So I hardly fancy adding another 4 or 5 of these buttons to my shopping list, if I can avoid it !  Luckily, removing the sweep board didn't involve any of these snap-on knobs, so I wasn't at risk there.

Then let's look at the B/delayed time base board, A13.   Just like the front panel board, there is only one small electrolytic cap on there, that's C659.  However according to the schematics this one is not your regular dumb power supply decoupling cap. Nope, that one is part of the signal path for something of great interest to me : the 'B' delay vernier . So I will replace that one. Gonna be fun though, as the Alternate/B Sweep board , which is the one mounted vertically right next to the SMPS metal work... is not socketed ! Why, ohhh... why. It's a single row of no less than 27 pins, soldered directly to the main board : "lucky me"....
I think I can manage to desolder it cleanly, given some time... but chances are hight that I will have to do most of my trouble shooting to come, on that very board... so I will solder a socket on the main board, so I can pull that daughter board at will. Gonna add a 32th of an inch of extra pin inductance... probably not the end of the world on a 60MHz scope... I shall see for myself anyway.

Then the last board to look at is... the main board, A10. Other than obvious SMPS rear of the board, I could not see a single electrolytic.... the "bad"/usual type I mean. But I did see a (very) few of these small precision axial electrolytic caps, same kind that I found on the main sweep board, A12. Again, I am not touching these.

So... to summarize the recap job ahead :

- A small one on the 'B' sweep board.
- A handful in the SMPS section, but not that many actually. 10 it appears, a nice round number.


- Time base knob issue....

I didn't mention it so far I think, but the delayed sweep knob was a bit of a pain : I had to pull it so much to get it to engage, that invariably it would pop the little red x10 Mag button with it !     Also, it had an incredible amount of play/slack in it. sometimes the spring action would vanish, letting the know slide freely on its shaft... I feared that the big timing switch on the A12 board would be damaged, the end of the world... in the end it turned out the switch is perfectly fine, and all that was needed was to tighten that bloody knob, it was just a little loose. Problem : to tighten it, you need to removed the red knob, then we can access a nut in inside the knob, concentric to the shaft.  Being an old USA bit of kit, the nut is an imperial size : 1/4" . I do have a set of Allen/hex male keys in Imperial sizes, but that's about it. No female hex socket to go get that screw well recessed into the knob...... or maybe I do ?!  Hmmm... 1/4" happens to be the size of all screw driver bits ! With every power drill or 1/4" ratchet wrench tool set, usually they also give you a little adapter that let's you fit the bit to the drill or wrench. Have a few of them. One was slim enough (outside diameter I mean) to be able to get inside the knob... but hell, the adapter was not DEEP enough to reach the nut ! It could barely touch it but not grab it well enough to turn it. Never mind... I reworked  the adapter to make it deeper, and hey presto, now I can tighten that nut/knob at last !  And sure enough... it now works like a charm, no play, no nothing, it just works, like I bought one ! 
Problem : I will actually HAVE to buy one !  When I did the test, the sweep board was outside the scope... when I refitted to the scope and therefore had to tighten again that nut...  yeah, the thread of the insert that grabs the (splined) shaft... sheared off !  The insert is actually not part of the knob, it could be replaced... but go find someone selling this tiny piece of metal ! So, I will have to go find en entire knob assembly just so I can have a new insert... which obviously is just as likely to break. Why did Tek use crap metal for this part ? Beyond me.

But the good remains that I now know that the this knob can be made to work just like new, no worries, no matter how old that scope is (36 years) ! 


- Main board rework

Once I had removed the sweep board, I now had a clear view on the main board. Did a visual inspection. Found a chip with what looked to me like 3 corroded pins ! Grrr.... See pictures. I cleaned that with IPA and an ESD safe "toothbrush". Then a chip next to this one... looked like it had a cracked solder joint on one of its pins ! Grrr again...  so I bit the bullet, turned the scope upside down and proceeded to reflow (while also adding a bit of fresh solder) each and every pic of each and every chip I could spot ! Shy of 20 ICs IIRC.  Can't harm...
The eagle eyed will notice that next to the chip with corroded pins, there is a Zener diode... which appears to suffer the exact same disease as it's Zener friends on the pre-regulator board : their leads are completely covered with a black oxide of some kind. So there is definitely a pattern here....


- Cleaning

Seeing as I had the sweep board out, I flooded the rotary switches (2 vertical amplifiers + timing switch) with contact cleaner and exercised the switches a few dozen times.  Did the same to all the switches on the front panel. Many are not easy to reach, but with the hepl of the straw, and turing the scope in every imaginable position... one can managed to do all of them bar one IIRC. The small pots for the position knobs and trigger knobs, however, are sealed, so no hope there. They seemed to be working fine though, so it's not a big deal...

Then I proceeded to clean the front panel. Doesn't show no my pictures so far, but up close, in the flesh... the whole things was pretty off-putting. 36 years of use, well it does show....
Again I didn't want to risk breaking all these little snap-on knobs, so I cleaned all of the knobs and switches in situ. Spending a few minutes on each and every knob, until I was satisfied with it level of cleanliness... so to speak.  The CRT bezel and color filter were easy to removed (just a couple Torx screws from the front of the instrument), so I removed them, made cleaning theses parts, and the area, easier.
A gallon of cleaning product, and a tired toothbrush later... the front panel now looks, up close, as good as new, I dare to say ! It can withstand macro shots no worries. I might still come back to it for a couple minor improvements :

- The clear/transparent plastic skirts for the sweep knob and attenuator knobs. Up close, they aren't 100% transparent.  Really nitpicking here, I admit, but I think with a bit of polishing they might come up really great, really glass-like clear. So worth a shot.

- The aluminium vernier looks great, except for the tip of it : the small part that you use to actually grab it with your fingers. It has seen better days I think... I could remove it (just a screw) so I can work it properly, on a polishing wheel or something.

Apart from that, front of the scope looks really pleasing now, quite happy    Just need a bloody replacement time base knob to finish it off ! 




Then I also spent quite some time looking at the schematics to try to figure out the circuitry responsible for highlighting the portion of the 'A' trace, that the the 'B' sweep displays/zooms into, with regard to the dark spot that trails behind the intensified portion. But it's a whole new can of worms as it happens, so will leave that for another post... this one is long enough, I admit...


So for now :

- Order all the caps so I can recap the thing
- Order some kind of socket/in-line connector so I can socket the alternate sweep board to assist in trouble-shooting
- try to find a replacement time base know (Sphere, Qservice or whoever on ebay...) and while I am at it try to find snap-on knobs for my 2232.



A few pics and straight to bed... again late and tired. Sorry for the inevitable typos, will fix them tomorrow when I get back home. In the meantime, be forgiving and try to make sense of stuff where ever it looks a little weird.

[Vince]

OK, not much time during the week to work on it... but every little helps :

-  Desoldered all the caps that will need to be rfeplaced, and double checked on the parts list that the few that were replaced... half of them actually (5 out of 10 ! ) where the appropriate value, capacitance and voltage rating. Somehow, they were all spot on.

The guy who replaced these 5 caps, for once, didn't buy the cheapest/crappiest he found from his local store : he put Nichicon everywhere but one which was Nippon-Chemicon, which is good also IIRC.  However.... he used 85°C caps for all of them... when even the prehistoric Mallory branded original caps... were all 105°C rated ! I means, the old/original FET (which is still in place) is very inefficient/heats up a lot, and the scope does not have a fan to top it all !

Also interesting : even though the scope does not have the TRIAC based regulator on the main board, but rather the PWM daughter board... the big filter cap of the regulator output did not move from the main board of the daughter board... no, they put one on both ! So a 470uF cap on the PWM regulator board, and the 470uF cap that came from the old TRIAC design. So that's almost 1000uF in all.  My only assumption here is that there the two wires aht connect the output of the PWM board to the input of the inverter on the main board, are quite long, 6 inches or so, 12 total... maybe that was too much parasitic inductance, and they wanted to still maintain some "local" decoupling right at the input of the inverter ? Probably...

I also see that one of the replaced caps... was probably replaced under warranty ! It's one of the many (5 of them) 840uF 12V Mallory caps.
I am petty sure of this because the replacement cap bears a part number on it and, most importantly, has the exact odd 840uF / 12V ratings ! Now if the average joe goes on-line to fine such a cap, I wish him good luck !  820uF 10 or 16V yeah sure, that's standard, but 840uF ? Nope. 12V ? Neither...

So that goes to show that these Mallory caps didn't last all that long to start with.
Now... with only 12V rating, what the hell...

I will up the voltage rating quite a bit I think. Dimension wise, the only constraint in this SMSP is the diameter, because 3 of these 840uF caps are touching each other, so I can't afford the replacements to be even one single millimeter bigger than this.
Capacitance wise, the parts list give the tolerance for these 840uF caps as being... fasten your seat belt....  +20% ? + 50% ? ...... nope... they state + 100% !

Just had a look on-line to see what's available,  Think I will put 1000uF as it's a nice round value with a lot more choice and low prices. Luckily, it appears that diameter wise, 12,5mm is a very popular package. Just under the 13mm mark I had to obey, perfect ! And even in thise diameter, they can withstand 25 or 35V, much better than the original 12V.  Price wise, I searched for Nichicon and Panasonic, 105°C, and even then it's only 50 cents or so a piece !

So no worries.

Apart from that... I am getting really tired of printed schematics from these Tek manuals... They are usually 3 page foldouts, so even much larger thanA3. Whenever I want/need to print a schematic, I need to select a small portion of it so that it's  not too small/readable on my A4 paper/printer. But even this is a pain, because the PDF reader (at least the one that comes in Linux/Gnome) doesn't allow to select a rectangular area on page, and print that so it fits an A4 page. So I have to full screen the PDF reader window, and and zoom and pan the foldout to select some part of it, then do a screen capture, then past it into an image editing software, then rework the image a bit from there, and print it... but obviously the result is not always readable because the screen capture is 90dpi or so, when a print would need 300dpi or so to look "clean".

So for example today I would have liked to print a board view of the main board, to highlight the caps that I removed from the board, to assist in putting them back later. But I couldn't be bothered because of all the process I just described... so instead I just drew sketch by hand ! Ugly, but super quick and does the job...

So, I think I am now so tired of doing this over and over again, every time I get a new old piece of gear in the lab... that I might bit the bullet and get myself an A3 size printer. Think I saw one not too long ago, for less than 100 bucks new. Ink jet of course, so the cost per page is high, but I don't care because I am not gonna prints millions of foldout a year, and I only need huge black cartridge, not expensive color cartridges.

Then I ought to try and see if there is a PDF reader that does allow to easily make a rectangular selection on a scanned page like these old manual, and print just that, at full res/high dpi. Maybe Adobe's reader can do that ? Would have to run it in a virtual machine though... but still, I would do it.

But ultimately even A3 printers don't cut it for our typical 3 page foldouts... so i am now even considering getting a plotter ! With a roll of paper, one could print foldouts of any length, no worries !    I mean yeah, these things are expensive, but maybe I can find a small one (with a width equal to the longest side of an A3 sheet), and of course a used one would be just fine....
Plus, it's doesn't have to be state of the art, it's just to print black and white schematics. Doesn't have to be super fast, nor do color, nor have silly high resolution,
nor be top quality... yeah I think I will start investigating this path seriously soon !  I am really soooo tired of the current situation !   

Midnight, late again... goodbye...