Philips PM3231 repair/restoration

[Ero-Shan]

After watching too many YT repair and restoration videos, I felt a sudden urge to do a hopefully easy repair of an early 70's simple scope. Although my repair queue reaches almost the next town (OK, that is slightly exaggerated), not having bought something from the bay for several months (there is a terrible drought), I again went looking and actually found a Philips PM3231 CRO in "not working" condition. Starting price was a bit higher than a working 15 MHz scope should go for, but prices have gone through the roof with everything (shipping added more than half of that again). Being the only bidder, it wasn't too bad (after all, it's a real dual beam scope!). A few days later, the thing arrived. The box was in good condition, yet a clatter could be heard when it was moved.





The noise came from the CRT's protective cap, graticule and parts from the mains switch that had disintegrated. The cap and graticule can easily be reattached, but the switch is gone. Philips used these a lot in those days, and they disintegrate without exception now: I had the same problem with my PM3230 and all my PE1514s.



The side panels came off and revealed the absolute filthiest device ever to sit on my bench! I have had quite a lot of dusty, dirty and grimy things before, but none with that disgusting filth. Yuck!

[ Specified attachment is not available ]

That's it for the first episode. The whole saga might go on for a while. There's at least the mains switch I have to order, which will inevitably take some time. Nor can I promise we'll end with a good-looking, working scope.

Ero-Shan

[Ero-Shan]

Episode 2: More filth and a surprise

As I haven't even shown a full frontal, let us start by doing that:



I have a habit of taking pics from all 6 sides of any new gear before I work on it, and after I'm finished. But I tend to forget to photographically document steps in between ...

First off, the disintegrated mains switch needed to be replaced with my patented DPNT switch (dual pole, no throw - always on ).



Finally it was time to give it a tentative try. Used the Variac and my new dim-bulb tester. No explosions, no smoke. Power lamp lights up, current consumption is in the right ballpark.
Graticule illumination doesn't work. But there are traces! So I attached a probe and checked the CAL output - seems to work alright, however, it doesn't trigger. Wait, if I shift the trace to the right (X POS) it does trigger.
And here's where I just forgot to make any pictures. Sigh.

Thing is, the 'scope basically working surprised me, as I had bought it mainly because of my urge to repair it. Yes, there's the trigger problem, but what then? In order to save my face I might try to do a restoration (not to the exact original state) instead of a plain repair. I might end up with a beautifully clean, perfectly working 15 MHz CRO that will be used for absolutely nothing at all. It'll just take up valuable space. Maybe I should have thought this through before clicking on the dreaded 'bid' button.

The graticule illumination is achieved with 4 small lamps at the screen's edges, wired in series. 4 * 6 V, 0.6 W powered from the 24 V intermediary power. And of course, one is dead. These little buggers are not so easy to find anymore. At least my favorite dealer doesn't have any kind of light bulb anymore.

[ Specified attachment is not available ]

I think I'm going to replace them with some kind of LED lighting. I do not have museum, nor do I work for one.

Ero-Shan

[Edit: I don't know why the first pic is rotated - it is perfectly fine on my computer. Sorry.]
[Edit 2: double -> dual]

[factory]

I've got some lamps of that type in storage, no idea what voltage they are, or the base type. Are they not still available for vintage car use? back when 6V electrics were more common.

Edit: Possibly BA5S or BA7S, can't tell the diameter from the picture, might help finding some, the BA7S is still available in 6V 100mA 0.6W.

David

[Ero-Shan]

I'm not familiar with lamp socket designations. And I've already decided to change the illumination to LEDs (one of the plastic holders broke when I removed it - it's Philips after all ). I Might regret that when there are only diagonally lit graticules, but having already fabricated (bodged together) the lamp + socket/holder replacements, that is my chance to shine.

If I ever will get to the end of this project, it will be demonstrated.

Thanks!

[Ero-Shan]

Episode 3: The inevitable mishap

I guess that the only reasonable thing to do is to go forward and make this CRO shiny and perfectly functional.

No doubt absolutely every switch and pot will need cleaning. Why not start with the Y attenuators? There's just a few connections at their backs and one nut at the front.
Yes, behind the knob is another big chunk of filth. Yuck!



The idiot that I am unsolders the wires from the variable gain potentiometers instead of cutting them. And as one of them just won't come loose, de does a bit too much yanking and pulling. This pull at the pots lug has fatal consequences:



Sigh. So I need to order a new pot also.

Those attenuators are really quite nice and look more like they were designed for much higher frequencies. Massive shielding, very easily accessible trim caps for tuning.

Here's a few shots of them before cleaning.




[ Specified attachment is not available ]

Oh dear.

Ero-Shan

[Ero-Shan]

Episode 4: Disassembly starts and more surprises

Now that I have committed myself to be serious about the restoration, the path is clear: put that thing to pieces!

Out comes the CRT, not without a little hassle, but it finally budges. What a monster!

[ Specified attachment is not available ]



But did I not hear something while moving it? A slight clutter?

Indeed, there are two small, loose pieces in the tube. Looks like glass to me. This is probably not so good ...



I can only hope they will lay passively somewhere on the bottom and don't interfere with the tube's operation.

While I looked closer at the mains transformer and how to get it out, I also discovered that the PE from the mains cable was broken right above the solder joint. No big deal, as I had been using my isolation transformer, but certainly something that must be fixed.
After the restoration is done (if), I will do a safety check anyway, so I couldn't have missed it.

Ero-Shan

[Ero-Shan]

Episode 5: It's a Philips after all

There's (at least) two things that come to my mind when I think of Philips equipment of that era. 1) the PCBs are mounted with those strange gray standoffs/clips which easily hold the top position in any brittleness competition and 2) Philips had apparently a hearty dislike for pcb connectors, instead making beautiful wire harnesses and tons of solder connections on card's edges while reusing the same colors for different signals.
Both of these don't make disassembly (and re-assembly later on!) much easier.





I gratuitously throw in some more filthy impressions just to make you envious:





The time base caps, rather beautiful after the filth is cleaned off. The big one is still in spec, while the 3 smaller ones are about 0.7 % low, according to my LCR meter.

[ Specified attachment is not available ]

That's it for this episode. Hope I can write up another before I'll be away for a few days.

Ero-Shan

PS: Does anybody know why my 'Inline expandable thumbnails' Are all at the end? What am I doing wrong?

[factory]

Episode 5: It's a Philips after all

There's (at least) two things that come to my mind when I think of Philips equipment of that era. 1) the PCBs are mounted with those strange gray standoffs/clips which easily hold the top position in any brittleness competition and 2) Philips had apparently a hearty dislike for pcb connectors, instead making beautiful wire harnesses and tons of solder connections on card's edges while reusing the same colors for different signals.
Both of these don't make disassembly (and re-assembly later on!) much easier.

(Attachment Link)

(Attachment Link)

I gratuitously throw in some more filthy impressions just to make you envious:

(Attachment Link)

(Attachment Link)

The time base caps, rather beautiful after the filth is cleaned off. The big one is still in spec, while the 3 smaller ones are about 0.7 % low, according to my LCR meter.

(Attachment Link)

That's it for this episode. Hope I can write up another before I'll be away for a few days.

Ero-Shan

PS: Does anybody know why my 'Inline expandable thumbnails' Are all at the end? What am I doing wrong?

It's a long standing forum problem, there is sometimes a problem with the first attachment (start with the second attachment to avoid this), I think they disabled the normal inline images, but they didn't want to break the forum trying to figure out the cause.

To get the images inline, first right click on the thumbnail image at the end, then select "copy link", paste that in the edit as shown below, add the IMG code, you could also add the width=300 or height=300 code to limit the inline picture size, the numbers are of course up to you.

Code: [Select]

[img width=300]https://www.eevblog.com/forum/repair/philips-pm3231-repairrestoration/?action=dlattach;attach=1854934;image[/img]

The above code gives this;



Re: The glass in the CRT, try to avoid the pieces from falling on the phosphor, as it will cause damage to the phosphor surface, yes I have a CRT with the same problem, it does still work OK.

David

[Ero-Shan]

PS: Does anybody know why my 'Inline expandable thumbnails' Are all at the end? What am I doing wrong?

It's a long standing forum problem, there is sometimes a problem with the first attachment (start with the second attachment to avoid this), I think they disabled the normal inline images, but they didn't want to break the forum trying to figure out the cause.

That I know and work around.

To get the images inline, first right click on the thumbnail image at the end, then select "copy link", paste that in the edit as shown below, add the IMG code, you could also add the width=300 or height=300 code to limit the inline picture size, the numbers are of course up to you.

Code: [Select]

[img width=300]https://www.eevblog.com/forum/repair/philips-pm3231-repairrestoration/?action=dlattach;attach=1854934;image[/img]

The above code gives this;

So this must be done 'post posting' in an extra edit step? Even during preview, I don't see any images. It takes me already ages to write a posting, this won't make it any quicker, I guess.

Re: The glass in the CRT, try to avoid the pieces from falling on the phosphor, as it will cause damage to the phosphor surface, yes I have a CRT with the same problem, it does still work OK.

David

This is the exact same conclusion I came to. Thanks for confirming and the help concerning the pics. In the old days I used Imgur, which allowed for lots of hi-res images in a single posting. That seems to be frowned upon mow (and finding the upload page on Imgur also gets harder every second day), so I try to make do with the forum software. A bit sad that these problems don't get fixed after months or even years.

Ero-Shan

[Ero-Shan]

Episode 6: Me am dumb!

(Did I watch too much Sir Sic lately?)

The switch and the potentiometer arrived. The pot is rather small, but the mounting diameter fits. That will look strange on the attenuator assembly, but should work okay.



The undersized pot is made up with an oversized switch: not only have I ignored the 0.2 mm mismatch of the hole's diameter (I was so sure I'd replaced the exact same switch on exact same mounting holes in my PE1514s!). It is also too wide to fit beside the power indicator lamp.



So I find myself filing the thread on the switch down a tiny bit and enlarging the hole just a little. That problem solved.

As I'm going to replace the graticule illumination with LEDs, I might as well do that with the power lamp. Alas, the lamp holder is not mounted with screws, but with a clamp on the light diffuser (don't know how to call it) that goes only one way. And with three sides closed up, I just can't get at the clamp to loosen the prongs and get it out. Raw force is the obvious answer. The end result, of course, is an utter failure: That diffuser thingy is pulled through the hole and the mangled user-facing surface should never sit on a front again. The LED will have to shine without that gimmick. As a slightly positive note I find that an old plastic mounting ring for a 5 mm LED fits perfectly into the hole.

On the mechanical side, the PM3230 has quite a bit in common with the PM3231. If I give up on the former (which would be a reasonable thing to do, anyway), I could use it as a donor and virtually undo some of my damage. But then again, I'm not aiming at a totally correct restoration. Who will care?

Ero-Shan

[Ero-Shan]

Episode 7: Alles muss raus!

Complete disassembly begins. Everything fastened to the front plate must let go. Of course, Philips did not just rely on hex nuts.



I first tried with circlip pliers, but that didn't work well. So I used the very last piece from my metal construction kit that I got as a kid (while looking for a translation of Metallbaukasten, I finally learned that I had a Distler Gigant) to make a special tool.



It's not particularly pretty, but it works very well!

I did try to minimize the number of wires to unsolder from the boards, so the two power supplies and the time base board stayed together, along with switches and pots. What an unwieldy mess.



The handle bar that runs on top has two plastic bezels that hide the screws with which it is fastened to the rear and front plate. Need I say that this is Philips plastic? Not much of a surprise that they do not let go gracefully. The other one disintegrated totally.



Only after I had reduced them to tiny pieces I thought of maybe using some warm air would have helped to soften the plastic. Hopefully I can get them off the PM3230 that way and use them for this one.

Ero-Shan

[Ero-Shan]

Episode 8: Just a heap of metal sheets

After sliding off the chrome U-frames from the front panel, I get another reminder of just how pervasive this fluffy gunk is.



The bare chassis sans front, only the delay line is still in its compartment.



The front panel is actually a sandwich of the die cast aluminum mounting plate and the actual front plate. They're held together by the mounted knobs and jacks, as well as the chrome frames. Amazingly, the Philips logo is on a small separate plate that's attached to the mounting plate by unknown means. I cannot fathom what made the designers do this.



Finally, the deed is done and all the parts that make up the housing get a good wash and are left to dry in the warm summer sun.



A bit of anxiety befalls me when I think about putting it all together again: That might be just a little harder than taking it apart ...

Ero-shan

[Ero-Shan]

Episode 9: Let there be light!

Now is the perfect time to get the new illumination working. Access to the front panel will never be better.
With the LEDs on narrow strips of Veroboard, and those stuck through holes on small patches of board that act as mounting frames, this looks admittedly somewhat crappy.



My main aim is to check whether this gives an adequate illumination and to find out good values for the limiting resistors. With a somewhat dusty graticule and filter, the result is this:



Frankly, that blew me away. This is so much better than I had envisioned even in my wildest dreams. That alone is reason enough to get that scope back in working condition!

The final 'light distribution board', with a connector for the power-on LED also.



Initially, I wanted this little board to sit right behind the intensity switch, but that would have hidden the mounting screws. Instead I found a place below the CRT opening, where I bored a hole and added a M3 thread.Much cleaner this way.



Since Philips' color scheme doesn't include black (except for lettering) on the front, the LED mounting ring must oblige.



Ero-Shan

[Ero-Shan]

Episode 10: Attenuator B fix

Cut the axle of the new potentiometer to the right length and added a slot. Great thanks go to Philips for their CAL position latch design. A spring ring (for lack of a better word) and a rather massive notched brass piece. Extremely easy to adjust, with no need to have the slot at a specific angle.



Alas, I couldn't convince my cheap camera to focus on the coupling, but you might get a glimpse of it.



That should have fixed my early blunder. While I was at it, I also replaced the crusty, disintegrating old tape with new Kapton tape.



Now all the easy things are done. I'm quite apprehensive of what lies in front of me ...

Ero-Shan

[Ero-Shan]

Episode 11: Protracting

Work on the scope is mostly stalled. I'm still undecided with regards to the mains lead/mains transformer. Philips soldered the mains lead directly at the mains transformer:



This is where I already had fixed the PE connection. Which was an absolute PITA. Maybe I shouldn't have done it with the transformer still in its place. The stiff PE screen screen around the mains cable doesn't help at all.

Philips loved those wire "springs" to solder wire ends. They work beautifully, but they don't survive unsoldering. The transformer begs to be cleaned.



Oops. Cleaned the type shield a bit too much, I guess. The schematics don't show it, but this thing clearly has an temperature fuse (not so clear to see on the image, sorry). This is also a common Philips feature.



Maybe I go for hard-wiring it for 220 volts and forget about the voltage switching. Then only the fuse would need to be incorporated. With a terminal strip on the back plate it should then be easy to get a clear separation. Mains -> fuse -> strip [just 3 screws here] -> power switch -> transformer. Something like that:



I also haven't shown the delay lines yet (still fluffy), which even got their own compartment!




Ero-Shan

[Ero-Shan]

Episode 12: CRT

It seemed so easy. Except it wasn't. Where does that rubber band go to? How exactly are those brackets attached? Damn, I had taken so many photographs. But not of the details that I needed. Or they were not in focus. It also seemed so straightforward when I disassembled it. But that was such a long time ago, and my brain is only a shadow of its former don't-forget-anything glory. Yes there's only so many possibilities to get it back together, and only one really makes sense in the end. Then you have to accept that the screws are located under the SMPS. Off it goes, not without a slight »plink« telling you that the HV cable to the CRT has snapped off (again). Another centimeter lost. The pins for the deflection plates at the side of the tube prove to be quite weak when you try to get the cables onto them - the lower ones in the back are especially awkward. I really don't want the tube to take its last breath after all that work.

Hours later, the canned vacuum is back in its place, and I'm done for the day.



Ever so slowly, the parts heap, put together, more and more resembles a real oscilloscope.

Ero-Shan

[Ero-Shan]

Episode 13: Timebase Switch

That's what it looked like fresh out of its work place:



This switch has wires going to everywhere. Another few hours (and some cussing) later it is back in its place. Once more, some points to solder proved a bit challenging.



It feels like slowly turning onto the home stretch.

Ero-Shan

[Ero-Shan]

Episode 14: alea iacta est

The point is settled: No more voltage switching 110/220 volts. The transformer no longer doubles as a terminal strip. The back panel's wiring gets trivial:



From the terminal strip on the back panel, only 3 leads connect to the instrument: two to the mains switch, one to the PE lug on the chassis. I inserted a small connector into the voltage selector's slot, making it impossible to change it to the 110 volts position. Fits perfectly, and it can even be held in place by its leads.



The cleaned up transformer anticipating its short trip home ...



... and finally back where it belongs.



This mostly concludes the reassembly. I'll call it a day. Powering up will be done in the next episode.

Ero-Shan

[Ero-Shan]

Episode 15: Power On!

I think I can see the finish line in the not-so-far distance. All operational parts are back in their places. Time to take a deep breath and plug the scope into the variac. At 165 volts, the power LED and the graticule illumination are already shining nicely. I increase the voltage slowly and wait for a trace to appear on the screen. But there is no trace. Instead, a slight stream of magical smoke emanates from the smps (I just keep calling it smps, even though it is not really one, just a chopper). I have a feeling that this is probably not good. It doesn't take too long to find that one of the bazillion voltage rails, -16J, has a short to ground. R856, a 47 Ω resistor, did not like the higher current flowing through it.



So the smoke did not come from the smps. It startet below it and went right through. So I disconnected -16J on all the boards, but none of them plead guilty. However, one white wire from the supply unit was obviously shorted. But where does it go to? Alas, the service manual does not include a wiring diagram, which would have helped a lot here. In the schematics I finally find it: R1, the 'STAB' potentiometer at the front, is directly connected to -16J. Sure enough, I measure 0.3 Ω here. And from here another white wire goes to ... it disappears somewhere behind the switches. More searching brings up another connection: R655 at the time base switch.



Indeed, there's no white wire connected to it. My notes clearly show that there should be one - I obviously missed this connection when I put the time base switch back in. Because there was no white wire dangling around. The white wire going to wherever does not yield when I pull on it. It is hard to see, but there is a tiny gap between the front mounting plate and the vertical chassis plate. Oh dear. When I screwed the front plate to the chassis, the wire wedged itself between without me noticing it, as the spot is hidden behind all the things mounted there.
After loosening the screws, the wire can be pulled out, and the short is gone. The resistor is replaced as well.

Next try! I'm sure It'll work! Yes, two green traces appear on the screen! I play around with the controls, and it really looks promising. After maybe a minute, the traces get darker and darker ... And then there's a loud hiss, followed by a plume of smoke, this time definitely from the upper left quadrant of the smps! I switch it off and let this sink in for a while. Smoke. Twice. And this is a non-smoking household! Where's that finish line, again? I can't see it through all that smoke! There's no way around it, the smps has to get out once more (and the HV cable will break once more).

C809, 32 µF/100 V is still slightly warm to the touch. It has really boiled its electrolyte, even the yellow holder was melted somewhat:



Amazingly, it's capacity is still fine at 38 µF, only the ESR has gone up to 2 Ω.
From the way this incident played out, I don't think that was a spontaneous suicide of the cap. As it sits on a 95 V rail, I rather suspect that the voltage suddenly rose above its rating (which might also explain the traces getting darker).
As I have no 33 µF caps rated higher than 100 V, I replace those on the +95 V and -95 V rails with them. Not ideal, but this scope will not see much use. If they work for another 50 years, that will suffice. In 50 years, I will not even be a memory any more.
Now why did the voltage rise that much? It can be seen in the first picture: The -20 V is the feedback for the linear regulator. And this was not soldered to the lug. So the regulator tried to get that voltage up to 20 volts by raising the input of the smps to whatever it could (my guesstimate: about 23 volts instead of about 16 volts).
I do not remember desoldering the brown -20 V wire from the lug, but it really looks like I have. Shame is on me.

Third time lucky? Now, all the major voltage rails are fine, and even after half an hour, there's no sign of any smoker.

'Checking And Adjusting' runs over 10 pages in the manual. I will leave that for another day. And I'm afraid that some of the checks might show further problems.

Ero-Shan

[Ero-Shan]

Episode 16: Reign of Frustration

Suddenly, everything goes wrong. The adjustment instructions are not too clear. And channel B is acting up: in the x10 gain setting, it is extremely unstable. The output of my HP33120A is too high.

Frustrated, I decide to take a break. This repair is already taking way too long. That's not fun anymore. And a hobby should be fun.

Yesterday, I regained some motivation and looked at the scope again. It didn't take long to find the channel B problem: Another not-connected wire. This one was caused by taking out the channel B attenuator assembly once more to fix something that was 'behind' it (what exactly that was has long been forgotten) and then forgetting to reattach all of the connections. The wire touched the pin of the attenuator and looked perfectly okay, just to deceive me I guess. With this confidence boost, I decided to redo all my adjustment steps (hours!), but not right away.

First I wanted to write a short episode here to keep the few of you reading this informed.

After half an hour of writing (I'm slow, and tend to rewrite each sentence several times) I click 'Preview' only to get an error after waiting endlessly. In the end, I was unable to get my text back, the forum software simply moved my elaborate novel to /dev/null. Back to frustration.

Hopefully, the next episode will be here sooner than this one. (This time I save the text locally before submitting it!)

Ero-Shan

[p.larner]

Dont give up your day job!.

[dmulligan]

Dont give up your day job!.

I, for one, am enjoying this thread.

[Ero-Shan]

Dont give up your day job!.

Of course I don't. I'm being a grumpy old man, as any retired engineer has to be!

That darn thing is still acting up. At Step 5., "Bandwidth", in the x10 position at 5 MHz, channel A is not less than 8 div, but actually up to more than 9! And I don't really understand their circuit ...
So this will keep me busy for even longer.



Sorry 'bout that.

Ero-Shan

[Runco990]

Man, you have the patience of a Saint.  I'd have saved the CRT and tossed the rest (properly smashed to make me feel better AND keep ebay flippers from it) into the landfill long ago.  My own life is getting too short for this much frustration.

But I'm looking forward to the conclusion of this adventure!   

[Ero-Shan]

Man, you have the patience of a Saint.  I'd have saved the CRT and tossed the rest (properly smashed to make me feel better AND keep ebay flippers from it) into the landfill long ago.  My own life is getting too short for this much frustration.

Not quite. That's why it takes so long - just a few hours every other day. I also had to learn that my workbench is at a pessimal place: When the sun is shining during the morning hours (when I'm usually working there), it does so directly into my eyes now that he's sitting low in the sky. Alas, due to other constraints, this was the only feasible place.
The fate you mentioned will most likely happen to my PM3230. It was in pretty bad shape when I got it, and (according to some guys here) it is a cantankerous beast even when in good shape.

But I'm looking forward to the conclusion of this adventure!   

Me too, oh yes, me too!

Ero-Shan