PM5519 (and PM5324) back in the land of the living

[KeithsTV]

I recently acquired a Philips PM5519 TV pattern generator along with a PM5324 signal generator from a test equipment clear out at work. Both had faults, the sig gen had more of a square wave mod waveform that the specified sinewave and the pattern generator had every test pattern except the circle. The sig gen had a leaky electrolytic cap which was replaced and a quick tweak of the pot in the oscillator circuit brought the mod waveform back to spec.

Flushed with success I opened up the pattern generator to investigate the lack of circle. Armed with the circuit and circuit description it took a little while and a lot of thinking to understand how the circle is generated. In its simplest terms it generates two parabolic signals, one at line rate and one at field rate, adds them and feeds the resulting signal into a comparator generating pulses which, when added to the video waveform, produce a circle. It's the electronic equivalent of making a parabolic cone and slicing though it at right angles to its central axis. The resulting slice is a circle. I'm sure I could prove it mathematically once but I've forgotten most of my A level maths many years ago.

I checked the input to the comparator, not helped by the fact that the mains transfomer lies directly over part of the circle generation circuit, and found only a line parabola. The parabolas are generated by double integrating the line and field drive signals each integrator comprising a resistor and transistor with a capacitor between its base and collector. The line was fine showing a sawtooth after the first stage and the parabola after the second stage but no field signals were present as the first stage capacitor was short circuit.

This appears to be a special as it's 130nF 1% and are probably rarer than hens teeth so I improvised a 100nf and parallel 22nF which although not exactly the correct value got the circle back.

So after 2 nights and 2 pieces of test equipment returned to full working order it's time for a rest, probably writing Christmas cards tomorrow.

Keith

[Sideband]

Makes a change to read a success in something other than radio or TV. Good write-up Keith...done in true 'Television mag style'!

[Hybrid tellies]

A good and interesting write up. Thanks Keith

[Phil G4SPZ]

I like restoring test equipment myself. It's a bit more challenging than the average vintage superhet, so well done and thanks for writing it up.

Phil

[KeithsTV]

One of the problems with repairing test equipment is getting service information. It would appear to be easier to get information on vintage radios, TVs etc than old test gear. Perhaps there was less call for it as the test gear was built to a higher standard than domestic equipment.

An internet search for the service manuals for both the PM5519 and PM5324 brings up a lot of responses but they all seem to be exactly the same scan of the manual which lacks detail in the vital areas. However that's all part of the challenge and fortunately the faults were relatively easy to find. One thing that didn't help was the sheer quantity of components on the boards and no identifying numbers on the boards for the ICs and other components. (The components themselves were marked with their type numbers)

The pattern generator did have an interesting fault at first although it's the PAL I variant it appeared to be producing a 60Hz field rate signal as the TV I was using wouldn't lock to it correctly. The manual shows 5 solder bridges that control the type of colour signal. These took ages to find as they are contained in the subcarrier oscillator module. They were however set correctly which lead me to believe it was a problem with the connector the module plugged into. On powering it up, after checking the module set up, it was working correctly so I removed the module again, cleaned the connector contacts and refitted it. There are two other modules so I did the same with those as a precaution.

Keith

[threeseven]

With test equipment manuals I think it's a case of test equipment being made in much smaller quantities and the manuals only being circulated in the trade.

I can't even find a user manual for an Iso-Tech clamp meter I bought online minus it's documentation, and it's a current model. It's usable without but some functions seem to have a mysterious way to activate them!

Well done with the diagnoses, all very interesting.

[KeithsTV]

A quick postscript to the circle restoration. I had noticed that the circle was not centred on the centre of the displayed screen and suspected that the value of the 130nF cap was critical. If it wasn't then they would not have used a 130nF 1% cap. As the value I'd improvised was approx 122nF, well outside the specified value, I just tacked a 10nF cap in parallel and lo and behold the circle changed size and was now almost central to the displayed screen. So the moral is, if its a 1% cap specified the designer meant it to be 1%. OK so the new value is 132nF but it's a lot closer to the original value than before approx 2% high rather than 6% low.

Keith

[ljones0]

I have a PM5519 here as well btw, though mine has two annoying faults! I've not yet looked at it but the first is with the circle -- sometimes you can see it getting drawn, but it then promptly vanishes. A Much worse fault though is after a few mins' of usage the output becomes unstable and breaks up -- almost as if there was a bad contact someplace.

It might - or might not - also be drifting off tune as well though at this stage I'm not sure. Maybe it is all down to bad contacts or a bad power supply? I'm guessing these devices date from the 70s?

ljones

[ljones0]

Update:
I have found that the circle can sometimes just end up as a bright white line at the top of the screen. I have had it appear briefly but normal response seems to be this.

As for the unstableness I found it happened if I moved the handle or laid the device down on its front as opposed to standing upwards though that could have also been a quirk or a fluke (am thinking it is, as I did have the unit on for a time).

ljones

[KeithsTV]

If I remember correctly The circle would appear then appear to expand and disappear. The way it generates the circle is described in the manual but basically it generated two parabolic waveforms at line and field rates. These are then combined and fed into a comparator, the output of which triggers a short pulse to create the circle.

Check the two 130nF caps C518 & C519 in the field circuit and the two 680p caps C516 & C517 in the line circuit. These are the integrators that generate the parabolic waveforms. In mine one of the 130n caps was short circuit.

Not sure what the stability issue is, could be a dry joint or bad connection somewhere

Keith

[dazzlevision]

Hello,

I had exactly the same problems with the circle on my PM5519 (bought new from Pye Unicam Ltd in the early 1980s). I obtained the correct 130nF capacitors from their spares department.

I also have erratic tuning drift, which appears to be due to the preset tuning controls unit (the same sort of thing was used in varicap diode tuned colour TVs). I did remove the modulator unit and check for/rework any suspect soldered joints, but that didn't improve this problem. It would be worth checking the stabilised dc supply that the preset tuning unit requires.

The only other issue I have had is dry joints developing on the small power supply printed circuit board at the rear of the unit.

Regards,

Dazzlevision

[John Caswell]

Hi Keith,
I have the full manual on the PM5324 which I would be happy to copy for you when time permits, but I did lend it to another member of this forum to copy so he may have it available electronically.

John

[KeithsTV]

Thanks for the offer John but I have a copy of the service manual and circuit diagram obtained from a forum member just after I got the sig gen.

Keith