Repairing PM3230 oscilloscope

Hello

Sorry, I don't write Dutch, and I cannot read it well! So I will try in english... but you can answer in german, french or italian if you like!

I have an old Philips PM3230 oscilloscope not working well any more (the traces make funny things, changing with temperature).

I have all schematics and I think it would be a good idea to clean everything and to begin with changing all electrolytic capacitors.

8)7 PROBLEM: how can I access to the backside of the PCBs, to measure something or to desolder capacitors?

Do I have so desolder all the wires on the borders of the PCB and remove them? Any easier solution I didn't see?

Thank you for helping!
Patrick
Bern, Switzerland

fred101

Golden Member

I worked on 2 of these scopes (I think, I can not remember the number) If so, these scopes are a pain to repair. The PSU rail boards where plugins, to the front was a big transformer and the other boards were only accasible after desoldering a lot of wires.

If you have a second scope you can measure the ripple on the powerrails. If that is good your problem is not the caps.
But besides that, it is always the best way to measure in the scope instead of starting with replacing random parts.

It is a Philips so first clean all switches ;-)

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch

Thank you. You are right, it is always better to measure and then replace only if necessary. But with those 50 years old electrolytics, I thought it would be a good idea to change all of them anyways. That was before I looked at the way the PCBs are fixed and wired!

I have a second scope (cheap battery thing, but OK) but the problem with measuring is that there is absolutely no legend on the visible part of the PCB, so I have to look for a component with visible wires to a) identify and b) measure the signal!

I have joined two pictures. Even after removing the plastic things holding the PCB, you can still not move them because on all sides short hard wires are going away!

The inside of the scope next to the transformer looks awful, it was humid in the basement and there is something like mold. But I don't think this should cause electrical problems.

fred101

Golden Member

This scope is totally different as the one I had. And I think a lot older because mine had normal FR4-like PCBs and not the brown cardboard. And it had a lot more PCBs. The PSU alone had something like 5 PCBs

Compared to that one yours has almost no wires. And to be real, for me this one does not have a lot of wires. I often have to desolder a lot more to get boards out. If you tell your brain it is not that much work and just do it you will see it is not much work. Make pictures and notes about what wire goes where.

I would replace caps too in this one but I would first measure. That you measure the powerrails for ripple is because you want to find the fault. If it is OK you can still replace them but then you know there is still something wrong. And If the ripple is to high you know that there is a chance replacing the caps will solve the problem.
If you first change caps and it helps, it is possible you solved the real problem partial or temporary (example: by bending some wire with a very bad solder joint that no makes just enough contact)

Replacing the caps and making new mounts for them is a lot more work. I have an old HP here on the bench as restoration project and I had to make a new PCB to mount the caps and it was a lot of work to extend the wires and solder them on the board (I had to solder through the old holes of the chassis mount caps while blocking my view with my hands

You have schematics and a manual, that tells normally you where to measure and yes, fixing a scope can be a lot of work. And then I am not even talking about the alignment.

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch

this thing had a history of storage with high humidity in the air. its seems tot be a combo build of electron tubes and germanium transistors.

i would not be spanding so much time on it, the possibility is quite big when you get it to work en then not be using it for 6 months or so, its likely to fail again.

the humidity never did any good, caps are mostly irrelayable ( scope is approx 55 years old )

multiturn switches and other ones with open contacts have contact problems due to lang not beeing used.

shortly, a lot of work to get a relayable scope.

waar rook was, werkt nu iets niet meer

You're right, testman but... it would be fun if it would work again more or less correctly, especially because it is an oldtimer :-)

Like a hobby...

Fortunately I have a cheap micro-oscilloscope for real measurements!

So, after cleaning the switches and potentiometers and vacuum cleaning... it works less than before.

Now there is no horizontal scanning any more...

fred101

Golden Member

If you first change caps and it helps, it is possible you solved the real problem partial or temporary (example: by bending some wire with a very bad solder joint that now makes just enough contact)

This is the main reason why I advised to first measure in the scope (you can solve the fault temporary but also make it worse)

It could be you have a dodgy solder contact somewhere. Use a strong loupe to check them all. For most the ones that have moved while you cleaned things.

Check you power rails. If it was not used a long time it is possible one of the electrolytics was leaky thanks to a bad oxidelayer and now shorted out. They are big so they can cool them self enough to do that for some time (a small modern cap will overheat and then blow up much faster)

What did you use for cleaning and how did you apply it ? If the stuff is conductive and you spray it all over the switch and pcb, chassis etc you can get leak currents.
Be aware that you have contact cleaners that need to be removed after doing there job and some do not clean but only lubricate contacts, some do both. I always use K61 but there are others. If that does not help I use K60, but only if removing the switch is a hell of a job and I can remove the K60 after cleaning. If possible I remove the switch and clean it with K60 and after that it goes in the ultrasone cleaner.

This can be hard to find. I had an measurements-amplifier here that gave errors sometimes. It was already returned to the factory but they did not solve it. I was checking every solderjoint and removed all boards to check (but still connected so I could move wires, bend things, knock on things etc to see the response)

There was one tiny board left on the frontpanel with a header for a piece of flatcable going to the LCD pcb a few cm away that I also left in place because that had no problem.
I wanted to make a note but could not find my pen. I lifted one corner of the amp cabinet to look under it and the amp gave its error message. Every time if I lifted the cabinet on that same corner the error appeared. Turned out there was one pin of the header on that small pcb not soldered at all. I could not believe that such a small force on such a heavy frontpanel could influence the contact but it did. I had the amp 2 years later again on the bench (but now something completely different)
And the customer told me it worked perfect all the time.

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch

I used only Kontakt WL, so it should not be a problem.

Thanks for alle the advice. I will try to find out...

Regards
Patrick

fred101

Golden Member

No experience with WL but accoording the spec sheet it is made to clean PCBs and parts from dirt. Not oxides from switches. And it can be harmful for some plastics used in capacitors.

K60 cleans oxidations but you can not leave it. K61 cleans a bit less agressive but for me this was in 95% of the time enough (in measurement and calibration stuff) but more important is that it also lubricates.

In high voltage area of the scope you must be carefull. I once saw blue plasma like streamers go over a switch deck that was cleaned with some unknown cleaner (not by me) (at something like 1500V)

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch

WL should not be very agressive and evaporates completely. It worked well in the sense that after it, all switches worked perfectly again.

K61 is a little aggressive and should be cleaned thoroughly, but how do clean it out of an oscilloscope? Take it with you in the bathtub?

What do you think about isopropanol (IPA)? It is supposed to clean a little, not being to aggressive and it evaporates completely.

fred101

Golden Member

Sorry, you are wrong. You know you can find exact what these are made for easy online? Maybe you believe the spec-sheet from Kontakt :-)
Kontakt 61

You can also find one for WL

From the WL datasheet:

Most currently used
materials in the production of electronic parts, housings and equipment, are compatible with KONTAKT WL. However, plastics sensitive to solvents (e.g. polystyrene, polycarbonate) or rubber should be tested on forehand.

There is a good video from applied science about cleaners

Tektronix used to clean their scopes in a washing machine. If I have industrial boards here for repair I clean them in the sink using water and soap. Then dry them in an oven. I remove parts that hate water. Sometimes I toss them in the US

[Bericht gewijzigd door fred101 op 31 oktober 2019 22:33:17 (10%)]

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch

I believe everything :)

What do you think about IPA? Isn't it a good compromise between cleaning without long-term problems?

fred101

Golden Member

I use among other stuff IPA and PCB cleaner (to remove flux residu)
the most. Watch the video, he demonstrates IPA also

[Bericht gewijzigd door fred101 op 31 oktober 2019 22:57:28 (13%)]

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch

I have had much success with isopropanol, all switches are now working perfectly.

I just poured a drop of IPA on the slider side (I have no good access from behind) and moved the switch a few times.

The question is now if I should add a drop of Kontakt 61 at the same place to create a protective "oil" film? The case against it is that the "oil" I could create a mess on the front plate of the oscilloscope (as I have to put it on this side).

Is the only consequence of using only IPA that I will have to clean the switches again in 50 years? :)

fred101

Golden Member

Op 31 oktober 2019 22:02:54 schreef pafff:
WL should not be very agressive and evaporates completely. It worked well in the sense that after it, all switches worked perfectly again.

Op 6 november 2019 08:18:10 schreef pafff:
I have had much success with isopropanol, all switches are now working perfectly.

?????? :-)

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch

I must be drunken because of the isopropanol vapours 8)7

Explanation: at the beginning I cleaned a few switches with WL (and all of them worked), but then I read the comments (that it could be aggressive) and stopped.

They yesterday I cleaned ALL switches with IPA.

The question remains: use K61 after IPA or not?

... and the second question is: should I try to clean the electronics of the thin layer of sticky dust, or is it better to leave it alone?

Last week I fixed a rotary encoder with WD-40 and break cleaner and to my suprise it actually worked! Would not recommend this procedure for other equipement other then the mouse I was 'repairing' for a friend. His shack didn't have any other liquids in aerosols :D

This is the world we know best, the world of madness
maartenbakker

Golden Member

If the switch was originally lubricated, I'd use K61.

"The mind is a funny thing. Sometimes it needs a good whack on the side of the head to jar things loose."
fred101

Golden Member

The last 10 years, on all my own gear (that is, the ones I restored/repairs, so about 80%) I used K61 after recommendations from Henry S (who is also working in the repair business). Upto now I have not had any potentiometer or switches problem.

Opposite to Maarten, The only time I do not use it, is when a service manual states the use of some specified lubricant like vaseline (petrojelly)
But I do not always follow the manufacturer, Fluke uses some grease in the rangeswitch of a current clamps and there it caused problems.

Cleaning a PCB ? If possible it is always better as leaving the dirt on, they leave the factory pretty clean ;-) Dirt has no use, no advantage but it can cause problems with leakage, flash-overs, damaging parts or temperature.

Ýou do not want to know what I find inside industrial gear or how dirty some PCBs from machines are. But in that case I clean them for most to keep my bench clean and be able to read part numbers.

http://schneiderelectronicsrepair.nl/wp-content/uploads/2017/08/dust.jpg

http://schneiderelectronicsrepair.nl/wp-content/uploads/2016/01/vna1.jpg

www.pa4tim.nl, Reparatie van meet- en calibratie apparatuur, ook oud en exotisch