Philips PM5107 teardown and reverse engineering attempt (WIP)

[Mad_Hank]

My cousin, who works at philips, found a PM 5107 function generator and gave it to me. As far as I know it is in working condition, but I didn't dare to plug it in without letting her warm up a little bit. So while I'm waiting for a variac which I can borrow I decided to tear her open and inspect her guts.

Front


This is a very basic function generator, so basic that it can only generate sine waves or square waves.
The other specs are:
10 hz - 100khz
0-2 volt amplitude + a -20 dB attenuation button
TTL out and of course your normal output
Oh and a lighted power button.


While stripping her down I noticed that philips didn't use philips screws on a philips device


Metalized enclosure


Hand drawn PCB...   


Do you know how simple this function generator is? That 6 pin DIP IC, that is an OPAMP. No specialized chip or something, no just a simple lonely OPAMP. http://g-pb.de/datasheets/TAA761.pdf


Powersupply


The brains


TLL circuitry


Output buffer

Since I couldn't find a (free) service manual I decided to reverse engineer the schematic myself. The results are not yet presentable but if someone happens to have the PM 5107 service manual I would love to have the actual documents. I may do a video teardown with a friend of mine but it all depends of how much time we have on our hands.

Before someone asks: the little bag contains merely a 115 volt sticker.

Maybe upcoming: A philips scope (sorry forgot the name) and a huge 0-120-250 volt 0-10-5A DC philips powersupply (appears to be something custom because there is no information available)

[redtails]

We briefly spoke about this device on IM.

What this function generator is likely doing is using some sort of variable signal generation, coupled into the single-channel opamp. The opamp is set up to have a variable gain until 2v RMS. Seeing as it's a single-channel opamp, the selection between square-wave and sinus-wave will have to be done via a hardware switch that physically changes the opamp circuitry.

It's not a complicated machine, though due to its age it might have set you back quite a pretty penny back in the day. I'm looking forward to the reverse-engineered circuit and the powerup.

[Vgkid]

It looks like it uses the bulb in the classic wien bridge oscillator.

[Mad_Hank]

It looks like it uses the bulb in the classic wien bridge oscillator.

It certainly looks like it; The rest of the circuit can be found on the right side (pin 4-5-6) of the OPAMP consisting of a 115 ohm resistor and capacitors (plus the variable capacitor).

[jancumps]

Aaah, memories.
I used this instrument in my school's lab to document the bandwidth and tone control characteristics of my final year project.

The hard way without sweep; changing frequency and drawing the graphs with a pen and logarithmic paper. I must still have my paper lying around somewhere here -but no idea where exactly.

[Mad_Hank]

Aaah, memories.
I used this instrument in my school's lab to document the bandwidth and tone control characteristics of my final year project.

The hard way without sweep; changing frequency and drawing the graphs with a pen and logarithmic paper. I must still have my paper lying around somewhere here -but no idea where exactly.

So you made an amplifier or some sort of audio tone control?

[jancumps]

pre-amp and amp

[rjeberhardt]

I just came across this post.  Very surprised as I did the initial design work on this before passing it on to one of my engineers to finalise.  I think that must have been in the early 1980s.  Hence the lack of sophisticated chips.

It was designed as a low distortion audio signal generator for testing and servicing audio amplifiers rather than a function generator.  We designed and built it in our Cambridge factory under contract to Philips.

Unfortunately I no longer have the circuit diagram but can say that, yes,  it does use a Wein bridge oscillator.  The thing in the glass bulb is a very small thermistor in a vacuum used to stabilise the amplitude of the oscillator output and maintain low levels of distortion.

Russell.

[tautech]

Welcome aboard Russell.

Have you kept anything special from your Phillips days you can share with us?
Maybe some photos of unusal models or rare manuals?

[tautech]

The OP's model has the same form factor as my first FG, a PM 5131 however the 5107 is in a smaller case.
It's a lot busier inside, but the same sort of beautiful old artwork. 

[dom0]

I recently acquired a PM5132 but must admit that I'm not really impressed with the design. Like in my PM5193 everything is soldered together and one needs to remove quite a few parts and unsolder at least one cable to get real access to the two main PCBs... regarding the circuitry it's a solid and classic design using lots and lots of discrete amplifiers and quite a nicely done current source, although I dislike the use of an ASIC (OQ0011) containing just a sine shaping network.

Another thing I dislike about that series of generators is that they have a fixed power cord (ugh) and the fuse holder is inside the device (ugh). I never got why companies do this... even Tektronix did that on 7000 series mainframes, wtf?!

(The PM5193 has a much smarter mechanical construction, though, with the top PCB mounted on hinges that swivel out of the case - very nice)

[Tac Eht Xilef]

While stripping her down I noticed that philips didn't use philips screws on a philips device

Not surprising - different Phillips (note the 2 'l's).

[rjeberhardt]

Have you kept anything special from your Phillips days you can share with us?
Maybe some photos of unusal models or rare manuals?

Sorry, only memories!  I had to have a big clear-out when I retired and moved from the UK to France.  My biggest regret there is that I gave away an HP digital storage scope that I had won in a competition

Russell.

[astro400]

The oscillator diagram is exactly the same as the  great brother, PM 5109, I own :

I like these old philips instruments, congratulation Rjeberhardt for the design!
Michel

[CJay]

And they spawned many a knock off or 'coincidentally' similar design in their time.

I love this forum, oddball things like a 6 pin DIL op-amp and then, even better, the original designer...

[iStealth]

I can buy one of those very cheap but I was wondering what exactly the "Low Distortion" level for that generator is. I cannot find any manual online. I found in the manual that for Philips PM 5131 the THD in the range of 10Hz-100kHz is about 0.5%. Nowadays Chinese signal generators, for example FeelTech FY3200S, can do 0.18% THD at 1V RMS input at 1kHz, which would be ok for my tests of audio amps. So, is that old Philips PM 5107 better than modern cheap alternatives? and what is approx the THD of that unit?

[free_electron]

I can buy one of those very cheap but I was wondering what exactly the "Low Distortion" level for that generator is. I cannot find any manual online. I found in the manual that for Philips PM 5131 the THD in the range of 10Hz-100kHz is about 0.5%. Nowadays Chinese signal generators, for example FeelTech FY3200S, can do 0.18% THD at 1V RMS input at 1kHz, which would be ok for my tests of audio amps. So, is that old Philips PM 5107 better than modern cheap alternatives? and what is approx the THD of that unit?

that's chinese THD ... conversion formula :

real thd = (chinese thd ^sqrt(pi)) + (rnd[5..100]/10). 

rnd[5..100] is a random number between 5 and 100

[edavid]

I can buy one of those very cheap but I was wondering what exactly the "Low Distortion" level for that generator is. I cannot find any manual online. I found in the manual that for Philips PM 5131 the THD in the range of 10Hz-100kHz is about 0.5%. Nowadays Chinese signal generators, for example FeelTech FY3200S, can do 0.18% THD at 1V RMS input at 1kHz, which would be ok for my tests of audio amps. So, is that old Philips PM 5107 better than modern cheap alternatives? and what is approx the THD of that unit?

The PM5109 has an 0.03% THD spec.  From the posts above, it looks like the PM5107 should be similar.

If you are just doing audio tests, you should be able to generate similar low distortion signals from a sound card or CD player.

P.S. Does anyone have a PM5109 manual or schematic PDF to share?

[iStealth]

that's chinese THD ... conversion formula :

real thd = (chinese thd ^sqrt(pi)) + (rnd[5..100]/10). 

rnd[5..100] is a random number between 5 and 100

Indeed that formula works in practice, but this time that generator was actually tested
here on this forum and it does have THD less than 0.2% for 1V output. The specification says 0.8% but I think is it for the max 10V output. It also means that those THD of 0.5% for Philips are for the max output voltage, and for the reasonable audio range of 0-2Vrms the THD should be about 0.1% or less.

The PM5109 has an 0.03% THD spec.  From the posts above, it looks like the PM5107 should be similar.

If you are just doing audio tests, you should be able to generate similar low distortion signals from a sound card or CD player.

P.S. Does anyone have a PM5109 manual or schematic PDF to share?

I also tried ARTA and RMAA software using my old Dell desktop with build-in sound card and the results were not bad. I did not know if I could trust those tests, but I could play with the input-ouptut levels and see the THD measurements of even less than 0.1% for my tube headphone amp. Also the 2nd and 3rd harmonics were popping up as they should.... so I have doubts if I should get any external generator for those purposes. I was also planning to get a HP 333a Distortion Analyzer which I can also get reasonable cheap... but again it looks like my PC can do pretty much the same job, it could be worse, but to observe some trends and fix a few things it might be enough. 

[jannis53]

Hi, is there anybody who needs a Manual for the PM5107? I just got one, if there is a requirement for a copy, I could scan it and mail a PDF.
I have also some other Manuals - a lot of. Who knews, perhaps there´s the right one in the box...

Greetings from Graz, Austria
Johann