Philips PM3212 Oscilloscope channel A problem

[Basileus]

Current Status: SLOW  (Im having exams at the university)

Channel A trace is still way too high and cant be lowered. Amplitude seems wrong beyond calibration too. The problem doesnt seem stable. After fiddling inside the scope, the trace is even higher than before. I can see something only when i try to see the CAL signal and set amplitude ratio to 50 mV/div. By lowering the position as much as i can, i can see the bottom part of the square wave. <-This happens because of poor connection of the y pos pot. I reseated the connector and the situation improved. Still, trace too high and amplitude too low.

I'm waiting for another scope to arrive, a Tek 2215, to keep troubleshooting the Philips PM3212. Project is currently on hold.

Got the new scope! Love it.

Original post

Hi!

A recently got my first scope, a Philips 3212. The conditions werent specified by the ebay seller i got it from, there was only written "switches and potentiometers need cleaning".

From milking the front panel i found out:

The B channel works fine, amplitude works fine, focus and intensity controls work and the trace is nice and stable.

OUTDATED
The problem is with channel A. I had a trace on first power up, but some time after it totally disappeared. It appeared a second time after a lot of fruitless troubleshooting and the y position looked like out of cal. I mean, i wasnt able to move the trace down enough: with the y pos control at minimum the trace was still in the upper half of the screen. I tried to display the 1.2Vpp cal signal and the trace amplitude displayed was wrong: 1.1v. Tried with channel B, it works perfectly. After a short period of time, there was no trace on channel A. Again.

PLAYLIST OF ALL VIDEOS I MADE ABOUT THIS OSCILLOSCOPE FAULTS:

https://www.youtube.com/playlist?list=PLIgTdpoYMz1MnNDh-Nt9bgHrnhoELMqr6


UPDATE 1:

Right after writing this post i turned on the scope and.. BANG! there is a trace. Cant lower it much though. Still, i got it on camera and i'm uploading a short video of the fault on youtube. Coming soon.

UPDATE 2:

-Tried to adjust DC balance trimmer pot, the situation improved a bit. Still way too high though. Maybe the trimmer pot is faulty.  wrong. Trimmer pot values are within tolerances.

UPDATE 3:

After cleaning with lots of contact cleaner, channel B trace is unstable too. 

Probably oldway is right about weary contacts. You can tell how clicky-er is the mtb range switch compared to the v/div switches of the channels.

Seems like the faults on both channels have similar effects, but different causes.

Channel B is influenced by the ground coupling. Sometimes, the trace is perfectly fine; After you select ground coupling there is no flat trace, but a distorted, lower in amplitude square wave that goes away with a bit of wiggling of the switch.
Update -> Same thing happens when toggling ac/dc coupling

Channel A is similar, but gnd coupling seems to work fine.

Seems like i've to disassemble the front panel. Again.

UPDATE 4:

Cleaned rotary switches with pencil rubber and lots of patience. The coupling switches made the trace unstable when wiggled. Hopefully, spraying lots of cc in them and pushing the buttons a couple million times did the trick.

UPDATE 5:

Trace on A is unstable again. Probably wear of the contacts. The cc magic has worn off. Hopefully, if the switch is kept centered, the trace is stable. Still a bit dodgy though.

UPDATE 6:

Found the channel A problem: transistor V509. I cant tell what is going on, it seems fine. Swapped it temporary with V511. Roughly adjusted channal A balance. Now channel A is fine.

Other lesser problems

- Invert channel B doesnt work properly, some times it works some times it doesnt. From playing with the switch i think its a mechanical problem. Since its a low priority problem, i'll deal with it in another moment. SOLVED, YAY! (the plastic cylinder inside the switch was cracked, i glued it to the pot shaft and reassembled it. In 2 hours. Those flying springs are a pain.)
Update Now the switch x10 mag feels sticky too.. I'll probably have to repair it too  After i replace the reed contact, i'll know if i have to.    UPDATE I replaced the reed contact. Took me 4 tries but im satisfied of how i soldered it. Yes, the switch doesnt do anything.. Time to pull out the front panel.. AGAIN..

- The x10 switch for mag on the time base doesnt seems to work. Low priority fault. FAULT LOCALIZED, ITS THE RELAY THAT DOESNT SWITCH. Replacement is coming. In 2 months.

-One illumination bulb has failed. I'll probably look for a replacement or replace both bulbs with LEDs. LED SYSTEM INCOMING! when i have time... Done. Now there are 4 red leds (2 over and two under the crt) that illumi ate the graticule. Sexy.

Some pics of the new illumination system:

http://imgur.com/a/0FYc3

[LaserSteve]

I had one for 15 years.  Take the time to fix it, you could not ask for a better beginner's analog scope.
No promises, but I think I might still have the whole manual with troubleshooting guide.   I'll look.

Steve

[dan3460]

Welcome to the board, first thing here is a copy of the manual https://doc.xdevs.com/doc/PHILIPS/PHILIPS%20PM3212%20Instruction.pdf
I'm guessing that you do not have another scope. If you look in page 66 there is a block diagram of vertical system. You are loosing the ability to move the trace somewhere between the Pre-Amp and the switching amplifier. On page 68 there is a note saying that the balance can be adjusted by messing with R674. Maybe that can move the trace down.
As per not getting a full signal display, you need to fix the positioning part first, then it should be a calibration procedure somewhere in the manual.
You didn't said if you have a way to generate signals, you will need this to calibrate the scope. If you do not have anything like that you can use a know source of voltage (such a battery) to get a close enough calibration.
Good luck.

[Basileus]

Welcome to the board, first thing here is a copy of the manual https://doc.xdevs.com/doc/PHILIPS/PHILIPS%20PM3212%20Instruction.pdf
I'm guessing that you do not have another scope. If you look in page 66 there is a block diagram of vertical system. You are loosing the ability to move the trace somewhere between the Pre-Amp and the switching amplifier. On page 68 there is a note saying that the balance can be adjusted by messing with R674. Maybe that can move the trace down.
As per not getting a full signal display, you need to fix the positioning part first, then it should be a calibration procedure somewhere in the manual.
You didn't said if you have a way to generate signals, you will need this to calibrate the scope. If you do not have anything like that you can use a know source of voltage (such a battery) to get a close enough calibration.
Good luck.

R674 is related to channel B "Shift balance channel B". The problem is on channel A, do you really mean R674 or R574, his channel A counterpart?

No, unfortunately i dont have a function generator i can use to cal the scope.

PS I adjusted R574, things improved a bit but its not enough. In the lowest position the trace sits half division over the middle line.

[dan3460]

You are correct, I was looking at channel B, both channels are virtually identical. I hope you have DMM. You need to start checking voltages and resistor values. I have never worked on a Phillips before, but the schematics have a lot of voltage references that can be tested.
For example look for R573 you should have -12V on one side and -5.8V on the other. Check the resistance across R2 (position knob) if is not easily accessible, you can find R571 and r572 and check it there. Check the voltages around V518 and V519, varying the knob position. The voltage at the base of both transistors should be -9V (if I read the schematics correctly).
It is possible that one of those transistors around that circuit, V513, V514, V524, V526, V518 and V519 is bad, the good news is that all of them are available and cheap.
You are going to have to measure parts, measure more parts until you find the one that is faulting. Also, by using the diode function you can check the transistors, in most cases, sometimes you can get a bad reading but that tis because is in circuit. The only way to really check is by lifting one leg of the part out of the circuit. Look in google how to use the diode setting to check transistors.

[Basileus]

You are correct, I was looking at channel B, both channels are virtually identical. I hope you have DMM. You need to start checking voltages and resistor values. I have never worked on a Phillips before, but the schematics have a lot of voltage references that can be tested.
For example look for R573 you should have -12V on one side and -5.8V on the other. Check the resistance across R2 (position knob) if is not easily accessible, you can find R571 and r572 and check it there. Check the voltages around V518 and V519, varying the knob position. The voltage at the base of both transistors should be -9V (if I read the schematics correctly).
It is possible that one of those transistors around that circuit, V513, V514, V524, V526, V518 and V519 is bad, the good news is that all of them are available and cheap.
You are going to have to measure parts, measure more parts until you find the one that is faulting. Also, by using the diode function you can check the transistors, in most cases, sometimes you can get a bad reading but that tis because is in circuit. The only way to really check is by lifting one leg of the part out of the circuit. Look in google how to use the diode setting to check transistors.

Yes i do have a DMM.

Here are my readings:

R573 voltages
Side 1: -11.90v
Side 2: Ranging from -5.48v in the lowest and highest positions of R2, -5.87v with R2 in his center position.

R2 resistance

I unplugged his connector and measured his resistance directly. Checking both paths from wipers to the other two terminals, resistance varies from about 50 ohms to 1050 ohms. Seems good.

Voltage between V518 and V519 bases is -2.98v, doesnt vary at all by manipulating R2.

I checked V513, V514, V524, V526, V518 and V519. They all give reasonable readings.

[dan3460]

ok, here is something to investigate. The voltage at the bases of V518 and V519 should not change by moving the knob, that voltage is set by a voltage divider R569 and R568 and if my numbers are right (R569 59k, R568 178k) it should be -9V. Check the junction of R669 and R668 and see the voltage there. If different check the resistance of R569 and R568.

[Basileus]

ok, here is something to investigate. The voltage at the bases of V518 and V519 should not change by moving the knob, that voltage is set by a voltage divider R569 and R568 and if my numbers are right (R569 59k, R568 178k) it should be -9V. Check the junction of R669 and R668 and see the voltage there. If different check the resistance of R569 and R568.

Voltage between V518 and V519 bases changed to -4.77v, stable. Similar voltage reading between V618 and V619 bases on channel B.

I just realized that right after turning the scope on, the A channel trace starts to shake and slowly climbs up by a bit. I'll leave the scope off for a bit and see if i can catch it on video.

Update

Turned it on again. voltages are back to -2.98v on both channels, exact same values. I think that the supposed dc voltage on the manual is -3v (not sure, the pdf is pretty low resolution and its difficult to tell). And yes, the trace climbs up a bit on warm up. Cant say if its relevant though.

[dan3460]

I made a quick simulation of the circuit, bases of V518 and 519 should be around -3V so your -2.98 maybe is correct. Does the trace starts in the middle of the screen and then goes up? You didn't say if you have checked the resistors around V518-519 - 513-514 - 524-526.

[Basileus]

I made a quick simulation of the circuit, bases of V518 and 519 should be around -3V so your -2.98 maybe is correct. Does the trace starts in the middle of the screen and then goes up? You didn't say if you have checked the resistors around V518-519 - 513-514 - 524-526.

Yes i did test them with the diode mode on my DMM, havent desoldered any (V524 and V526 are socketed). They seem fine.

I'll record the next power up and post it.

Update

Recorded another power up, this time the trace didnt climb up nearly at all, 1/4 of div at best.

Still, i'm uploading another video of the oscilloscope displaying the cal signal. Trace on A is quite distorted and his amplitude is unstable when using the v/div knob.

[Basileus]

Here.

https://youtu.be/wjVrUPN72Dc

[oldway]

Your oscilloscope appears to have 2 different failures on channel A

- wear of the range switch ... it is indeed wear and not a simple oxidation of the contacts because the oxidation occurs on the contacts that are not used.

In a low bandwidth oscilloscope that is normally used for hobby, diagnosis and repair, channel A only is used in 80% of cases and channels A and B simultaneously in 20% of cases.
The range switch of channel A is therefore always much more worn than that of channel B
The contact cleaner spray will not solve this problem, at most it will improve the contacts temporarily
The real solution is to replace the A channel range switch with another in good condition.
What I have often done is exchange the ranges switches between channels A and B, but it's a heavy job

- There is also another fault in channel A: the symmetrical preamplifier does not seem to be balanced, so that the position potentiometer does not allow the trace to be centered.

All this seems to me a repair which is not worth it.

I recomend you to use only channel B.

[Basileus]

Your oscilloscope appears to have 2 different failures on channel A

- wear of the range switch ... it is indeed wear and not a simple oxidation of the contacts because the oxidation occurs on the contacts that are not used.

In a low bandwidth oscilloscope that is normally used for hobby, diagnosis and repair, channel A only is used in 80% of cases and channels A and B simultaneously in 20% of cases.
The range switch of channel A is therefore always much more worn than that of channel B
The contact cleaner spray will not solve this problem, at most it will improve the contacts temporarily
The real solution is to replace the A channel range switch with another in good condition.
What I have often done is exchange the ranges switches between channels A and B, but it's a heavy job

- There is also another fault in channel A: the symmetrical preamplifier does not seem to be balanced, so that the position potentiometer does not allow the trace to be centered.

All this seems to me a repair which is not worth it.

I recomend you to use only channel B.

I guess i'll repair what i can and leave channel A be.
It's really sad. I guess that if someday i'll stumble in another broken oscilloscope i'll get another range switch.

I realized there is another problem. The pull for x10 mag for the time base doesnt seem to do anything at all. I checked the switch behind the x pos pot but it seems fine. Maybe the fault is in the main board connection or in the related circuit.

[oldway]

The pull for x10 mag for the time base doesnt seem to do anything at all.

Check relay K140? (late digit not readable on schematics) for bad contact or open spool.
Check resistance of x10 gain pot. (R1419) ...must be lower than 100R

[dan3460]

That video gives a better idea of what the problem is. Did you use deoxit or similar product to clean the switches? I would give that a try.

[Basileus]

That video gives a better idea of what the problem is. Did you use deoxit or similar product to clean the switches? I would give that a try.

I currently have the front panel completely disassembled, it took me 2 hours of work to repair the invert channel B switch. The plastic cilinder inside the switch was cracked and slided on the pot shaft. Luckily it seems like its a common problem and following these instructions did the trick:

http://siber-sonic.com/electronics/PM3214fix.html

I used dry contact cleaner on all switches and pots (without dissasembling them). Now i'm currently mounting back the front panel controls. After this is done i'll check if i messed something up, if channel A trace stability improved and ill do what oldway suggested to troubleshoot the x10 mag switch.

I'm also planning to replace the illumination lightbulbs with 4 red LEDs copying the circuit shown in this repair video:

[Basileus]

Update

Now the invert channel B switch works like a charm.

But after cleaning with lots of contact cleaner, channel B trace is unstable too. 

Probably oldway is right. You can tell how clicky-er is the mtb range switch compared to the v/div switches of the channels.


Wait...
Seems like the faults on both channels have similar effects, but different causes.

Channel B is influenced by the ground coupling. Sometimes, the trace is perfectly fine; After you select ground coupling there is no flat trace, but a distorted, lower in amplitude square wave that goes away with a bit of wiggling of the switch.
Update -> Same thing happens when toggling ac/dc coupling


Channel A is similar, but gnd coupling seems to work fine.



Posting a video about it.

https://youtu.be/YN70fz6rqN4

[Basileus]

The pull for x10 mag for the time base doesnt seem to do anything at all.

Check relay K140? (late digit not readable on schematics) for bad contact or open spool.
Check resistance of x10 gain pot. (R1419) ...must be lower than 100R

I don't know how its inside that relay, but i compared it to the other relays, and i noticed that the leads going in the center of the coil are open circuit, while the others arent.

[dan3460]

For what I can see on the video all your problems are mechanical. I have never worked on a Phillips before so I don't know how the mechanical parts work, but here is what I will do. Try to dissemble as mucha as possible of all the controls at the front panel, the switches at the bottom of the vol/div, etc. carefully try to clean, sometimes lubrication is useful, I use wd40. Clean and check that the contacts are touching.
The open circuit on that relay can be due to settings of the scope when it was turned off.

[Basileus]

For what I can see on the video all your problems are mechanical. I have never worked on a Phillips before so I don't know how the mechanical parts work, but here is what I will do. Try to dissemble as mucha as possible of all the controls at the front panel, the switches at the bottom of the vol/div, etc. carefully try to clean, sometimes lubrication is useful, I use wd40. Clean and check that the contacts are touching.
The open circuit on that relay can be due to settings of the scope when it was turned off.

What about the a trace pos not being centered? Adjusting the dc balance pot didnt solve it.
Edit: maybe the dc balance trimmer pot resistance is off, maybe it should be replaced. Still, the trace on channel A is deformed no matter what..

I did a continuity check on the relay with the scope is on, it conducts no matter the position of the x10 mag switch

What about adding a touch of solder on the worn conctacts? maybe it could be enough the ensure a good contact..

I already cleaned all pots and switches with dry contact cleaner, should i do it again, maybe opening switches and pots?
should i use dry or lubrificating products?

[Twoflower]

As far as I can see all these relais are the same type. And should behave the same. From your picture it looks like a coil around a reed relais. From the schematic and I'm not aware a normally closed reed relay exists it should be open if no current running through the coil. If not the reed relais contacts might stuck for whatever reason.

By design the reed contacts close if a magnet gets close. It should work even through the coil around the reed contact. So you might try if it closes without power your scope by bringing a small magnet close to it. And as the K1401 is driven from a switch you might check if one side of the coil gets connected to GND if 10x is enabled.

As you say the others are closed if you measure them. This sounds odd to me. Sometimes reed relais can stick so slight tapping a closed and un-powered relais might open the contacts (if not fused together).

P.S.: I found a different schematic for a Pm3212: https://archive.org/stream/philips_pm3212 better readable but the schematics are not identical.

[Basileus]

As far as I can see all these relais are the same type. And should behave the same. From your picture it looks like a coil around a reed relais. From the schematic and I'm not aware a normally closed reed relay exists it should be open if no current running through the coil. If not the reed relais contacts might stuck for whatever reason.

By design the reed contacts close if a magnet gets close. It should work even through the coil around the reed contact. So you might try if it closes without power your scope by bringing a small magnet close to it. And as the K1401 is driven from a switch you might check if one side of the coil gets connected to GND if 10x is enabled.

As you say the others are closed if you measure them. This sounds odd to me. Sometimes reed relais can stick so slight tapping a closed and un-powered relais might open the contacts (if not fused together).

P.S.: I found a different schematic for a Pm3212: https://archive.org/stream/philips_pm3212 better readable but the schematics are not identical.

FAULTY RELAY CONFIRMED!

if i short the... reed contacts together x10 mag is enabled...!

...Now what relay should i get...? I've never seen a relay like this.

Should i look just for a new reed contact of similar dimensions? Should i desolder the faulty one?

Update: i tried with a magnet too, i can hear it switching but nothing happens

[Twoflower]

First you should measure the size of the glass ferule to order a replacement. Search fore reed contact/reed relais at your preferred supplier. They're cheap, probably less than 1€.

Can you check if you might be able to get the glass part out or is it glued in? That's the contact part of the relais. If you're lucky you can cut the two wires from the glass part and solder in a new one (careful they break easily).

If they're glued in I think the best is to unsolder the whole relais and see how to remove the contact.

If it's easy to replace you might think about to replace all reed contacts. Otherwise you should double-check the other ones about reliability.

[Basileus]

First you should measure the size of the glass ferule to order a replacement. Search fore reed contact/reed relais at your preferred supplier. They're cheap, probably less than 1€.

Can you check if you might be able to get the glass part out or is it glued in? That's the contact part of the relais. If you're lucky you can cut the two wires from the glass part and solder in a new one (careful they break easily).

If they're glued in I think the best is to unsolder the whole relais and see how to remove the contact.

If it's easy to replace you might think about to replace all reed contacts. Otherwise you should double-check the other ones about reliability.

I dont think its glued, there is a bit of slop inside despite being soldered on both ends. Now i'll desolder it and slide it out.

PS i love you.

[Twoflower]

If you replace them with the new ones: Handle them very careful as the glass cracks easily (order one or two extra might be a good investment). Probably it's a good idea to solder in two pieces of silver wire into the PCB and solder the reed contact to the wires without bending the contacts of the contact. This will minimize the stress on the device.

If you want to solder them directly into the PCB. Please make sure that the wires are bend exactly so they slide into the PCB holes without additional bending that apply a constant stress to the glass.

I might be a bit over careful here. But cracking them without noticing it might result in the same problem you have now: A unreliable piece of equipment ;-)