Philips PM5131 function generator repair

[Refrigerator]

Got lucky and scored three Philips PM5131 function generators for 20€ each.
They were sold as working but "defective".
I called up the seller and he said that the signal was unstable and noisy.
Hmm do i smell crusty switches? Perhaps scratchy pots?
The seller also said that he couldn't be bothered to fix them, which sounded weird to me but to each their own i guess. 
He was selling three of these function generators for 20€ each and at first i thought about only buying one but the deal was too good to let go so i got all three.
I just brought them from the parcel machine and am waiting for them to warm up before i look any deeper into them.
They're actually so cold that all pots except one are very hard to move because the viscous grease is all frozen.
Anyways, while they're warming up i started taking a look at the outside and it looks like they've been banging around for a while.
There's dents on the sheetmetal case and even the BNC sockets are bent.
Found a cal sticker and looks like these came from Germany, last cal date shows 10th month of 2012, so it's been a while.

I don't think there's much to go wrong inside these. I've looked through the manual and it says that the outputs are "short-circuit-proof" so i don't think the will be much to replace inside other than old electrolytics perhaps. I hope there's no RIFA menace inside.
Also took a look at the schematic and read about the working principle of the device. It's completely analog and everything is controlled through the switches and potentiometers so any bad contact will manifest itself in the output, which i'm pretty sure is what's going on with these.

What do you guys think? Will this be an easy fix?
I'm super keen to get these fixed. 

[Gyro]

I'd say they would be an easy fix, just a little care and attention. Nice analogue [Edit:sweep] function generators without digital artefacts [and decent output drive and attenuators].

The circuitry is pretty much discrete and the Instruction manual includes full service information and schematics... https://elektrotanya.com/philips_pm5131_sm.pdf/download.html

Definitely worth the money, Philips made some nice stuff in that era.

[Refrigerator]

One of the generators has some missing caps for the switches and the seller said that ironically that's the best one out of the bunch.
And now that i have them i can see why. The other two have very wobbly frequency adjustment knobs, while the one with the missing switch caps is much tighter.
But all three share the same wobbly DC offset potentiometers.
So it looks like the frequency adjustment potentiometers might be worn out or might have taken a hit even.
The frequency potentiometers feel like they're wirewound so i hope they're not completely busted otherwise finding replacements might be tricky.

[Gyro]

Looking at the manual, there doesn't look anything special about the pots, they're just panel mount single gang carbon ones (I didn't notice a wirewound at first glance), connected to the PCB by flying leads (same with the BNCs I think). It might even just be loose mounting nuts. At least you have no worries about PCB footprints!

Worst case, you'll probably end up with a good pair and a parts mule.


Edit: Ah yes, the frequency one might be wirewound for linearity but those aren't hard to find either - it's only when you get to multi-gang ones that things get tricky.

[jdragoset]

The cases on these slide off the chassis much like the Tektronix 22xx scopes,
If severe dents prevent removal, the rivets can be drilled to allow removal, dents repaired and aluminum rivets replaced.

[Refrigerator]

Fired the first unit up and now i see what the seller meant by "noisy output".
There seems to be a ~1.7MHz wave superimposed on the output waveform and the shapes of the waveforms seem a bit funky at times.
The 1.7MHz noise is present at all times and isn't affected by any of the knobs or switches.
Bumping the top of the case with my fist fixes the noise issue so there might be something loose inside (percussive maintenance FTW).
I can also hear something rattling inside, which might be my clue.
Frequency is very dicky. The knob works but there's spots where frequency suddenly just jumps or even goes backward for a period.
If i select square and fiddle the switch i can see glitching on the output.

So far:
* DC offset works.
* Amplitude adjustment works.
* Sweep seems to work also.
* Frequency offset (vernier) knob works fine.

The unit has yet to have had the time to warm up but should be well within the operational range.

[Refrigerator]

Bumping seems to also affect the shape of the waveform.
Both included pics are supposed to be a triangle waveform.

[Refrigerator]

....as opposed to what it's supposed to look more like.

[Refrigerator]

Sometimes the frequency starts modulating and a bump in the FFT appears.

[Refrigerator]

Took the cover off and the loose thing inside turned out to be a plastic support for the main board.
And it's not exactly loose but it slides inside the aluminum extrusion.

Ps: also noticed that the insides were quite warm.

[Refrigerator]

First thing first i wanted to check for previous repair or bad solder joints.
Noticed a few joints that were shinier than the rest.
Checked what they were holding and uhmmm, i don't think Philips used soviet electrolytic caps in their test gear?

[Refrigerator]

Two trimpots have a bunch of rust in them, but the rest of the board looks perfectly clean otherwise.

[tautech]

Had one of these for years and to set them correctly you need connect a scope first. 
No longer in use other than for gathering dust. 

But really open frame trimpots are the scourge of the earth and apart from the normal PSU checks that's where I would focus your efforts.
Good luck.

[HighVoltage]

I also had a couple of these a very long time ago, must have been late 80s.
They had to be repaired from time to time and it was always easy.
If I remember right, one problem was the output stage that liked to fail.

[Refrigerator]

University work got in the way of repairs but also served as an opportunity to use the function gens, despite their condition.
We were doing MEMS microphone lab work (i work from home because i have my own lab  ) and i used the function gen to drive a speaker.
Anyways, the function gen was working at first but got quite terrible so i tested all three to see if any are good enough to operate a speaker but they were all equally terrible.
So i sacrificed a few hours of my crunch time to see if i can sort one of them out.
And after opening another one i immediately noticed a terribly leaky capacitor, as well as another capacitor that looked out of place.
I didn't have any new replacements so i salvaged some from boards of a lesser vintage.
The situation did not improve so i started poking around inside with my scope and quickly found that there was a whopping 740mV 1.5MHz ripple on the output of the -22V regulator.
So i replaced the 22µF cap with a 47µF nichicon. Value is not the same but it's just a decoupling cap for a linear regulator, it'll be fine.
And upon powerup the output was finally stable and clean.
Except that the sine was looking a little deformed so i fiddled the four potentiometers next to the sine shaper IC while looking at my FFT and managed to tune the shape quite well.
The pots allow to tune second order and third order harmonics as well as adjust the DC offset of the sine. 

I think all three of these gens just need a full recap and they'll be good as new since they all had the exact same symptoms.
Also let's not forget the scratchy potentiometers, i still need to replace some of those.

[Gyro]

Looking good, you've got the hang of them now. 

[Refrigerator]

Looking good, you've got the hang of them now. 

Yep, fortunately i was able to find out what most of the potentiometers do in case i need to tune them after the repair.
I think the hardest part now will be straightening the dents in the sheetmetal case. 
One function gen is so dented that the metal case doesn't even slide off and gets hung up on the dent.
I'm also thinking about modifying the function gens slightly by adding a power indicator.
The power switch has a cool indicator that opens literally like an eye but it's not illuminated and is hard to see sometimes so a power indicator light would be nice.

[me getting ahead of myself]
I think i might modify the "eye" indicator and maybe add an LED inside the "eye" part to glow when power is on.
I have a fresh roll of white PLA so i can maybe print a replacement part with space for a 3mm LED inside.
I wonder what color LED to choose. Green like the magic eye lamps? Red because terminator? Amber for that vintage incandescent look? Hmmmmm 
[/me getting ahead of myself]

[Refrigerator]

I tried adding an LED to the frequency pointer but lost the part because it slipped and fell into a black hole - aka the floor.
So i designed and 3d printed a new indicator assembly.
Yes an assembly, there are two parts. An indicator and an LED holder+retainer.
The indicator comes in from the front and the LED holder presses on from behind and holds the indicator on.
A 3mm amber LED was used to illuminate the indicator from behind.

[Nchannelmosfet]

Hello!
I'm glad that you have managed to fix them. I also got one in not working condition. A visual inspection of the inside showed a pretty clean circuit board beside a resistor that had probably caused a smoke release. The resistor is connected to the 20dB attenuation switch and must be replaced. However, the bigger problem is similar to what you had, aka a noisy waveform or whatever your seller described. Would you also recommend me to recap the function generator or is there any other solution that might help?

[minifloat]

Recap will most likely do.

I had a very similar 5132 and after replacement of the PSU caps (near the 4pin TO220 voltage regulators) the superimposed 1.7MHz went away.

The small electrolytics (22uF/40V?) at the regs are the main culprit, being in the feedback loop of the regs explains why. The bigger ones after the rectifier also had lost quite a lot.

br, mf

[George Edmonds]

Hi

Replace all blue Philips electrolytic capacitors on sight, by now they will all be faulty.

G Edmonds

[Nchannelmosfet]

Interesting. I could also see a very high frequency ringing at the outputs of both the 22V and -22V rails. I pulled out the two 22uF caps and checked them with my transistor tester. They both showed a capacitance close to 22uF so  I don't know what has caused the ripple. Btw, I have noticed that the dual op-amp IC(NE5532) becomes very hot, is it normal?

[George Edmonds]

Hi

But what was the ESR?

G Edmonds

[Nchannelmosfet]

C1 = 20 uF, ESR = 56 ohm, Vloss 9%
C2 = 21 uF, ESR = 40 ohm, Vloss 8%
The picture shows the voltage at the output of the -22V voltage regulator. It has a peak-to-peak voltage of 1.77V at 1.258 MHz.

[Nchannelmosfet]

Does anybody know what kind of capacitors are these two marked with red boxes?