Philips PM5324-RF Oscillator instability.

[djsb]

I'm trying to trace the cause of the RF Oscillator instability in my Philips PM5324 RF signal generator. The instability consists of distortion (sine wave goes out of shape the shape of the distortion changes with frequency) and the amplitude reduces gradually until around 32 MHz (most instability centres on 31.8 MHz or thereabouts) and then the amplitude rises again. The distortion is still present right up to around 110MHz. Up to 10MHz the sine wave signal is stable in amplitude and shape.

Some questions that someone may be able to help me with.

1/ In the snapshot of the schematic below showing the HF Oscillator section what is the topology of the oscillator (i.e Wien Bridge, Colpitts etc)?
2/ What influence would noise on the power supply have on the stability of the Oscillator. In particular could failing electrolytic capacitors introduce HF noise.
3/ The power supply regulator is 741 op amp and Darlington transistor based. Could a fault here be a cause?

I'm going to do some voltage checks tomorrow, so I'll report back. I thought at first that the switches are the likely cause, but they seem OK. Then I had a look at the output amp (no voltage checks yet though). I found that the distortion is present at test point 10 in the schematic so it originates at the Oscillator itself.

More checks in the morning. Just wondering if anyone has any additional things I should be looking for? Thanks.


P.S Reading through the description of how the Oscillator works (attached) I have noticed a 47uF capacitor (Ref 530). It seems that this is involved in the amplitude stabilisation. I wonder if this is the faulty component? I shall investigate in the morning.

[djsb]

I'm doing a complete electrolytic cap replacement job. The components are all accessible and I've just ordered the full set of 18 caps for around £24 off Farnell. Lot cheaper than eBay. Hopefully get most of the job done by Saturday lunchtime.

Stand buy for the exciting aftermath.

[SeanB]

If that does not solve it look at the BAX13 diodes and the BFW10 Jfet, they tend to be leaky with time.

At least this does not contain unobtanium UJT's.

[djsb]

I'm still struggling trying to trace where the waveform distortion is coming from. I've hooked the output to my signal hound spectrum analyser and I can see a few strange harmonics here and there which will explain the distortion. What should I be looking for in particular?
I'm wondering if the oscillator has too much gain above 10 MHz (this is where the distortion begins).
Also, I've got the circuit in LTSPICE, so I'm playing around with that.
Does anyone know how the current mirror, oscillator and amplitude stabilization section (the FET and 2 diodes with cap) work. There is clearly a feedback capacitor (the 100pf one) and transistors 301 and 302 form a 2 stage amplifier. Can anyone explain how the gain is controlled and why the gain would vary with frequency (which may be the problem?).Any help would be appreciated.


PS LTSPICE file attached. Thanks.

[Grandchuck]

What happens when you change pot. 612?

[djsb]

When I turn this pot the amplitude either goes up or down. The amplitude at the output is a little too high (it should be 50mV peak to peak) but the pot can't reduce the amplitude enough when turned to one extreme. I'm going to replace the 4.7k existing pot with a 10k cermet pot. The original 2.8v Stabistor (that is what this component is actually called-It's several stacked PN junctions in a glass envelope) is broken and I've replaced it with a BZV87 2V6 Stabistor in a SOD package mounted on a piece of stripboard. The Oscillator now works but the amplitude needs some adjustment and there is some distortion of the waveform above 10MHz (but it improves around 100MHz).
There some noise on the -12v rail as well which I need to track down. Some RF noise is probably coming from the power supply, so I may experiment with a few ferrite beads to see what they can do.

[Grandchuck]

The AGC control is via the current source (Q305 - 306).  Have you checked around D407?  Looks as if the signal is rectified and then used to control gain.

[djsb]

Here is some more history of the work so far.
1/ All PN junctions (transistors/diodes) check out OK.
2/ All the capacitors have been changed (apart from capacitors associated with tuning). I've also replaced ALL the Electrolytic caps.
3/ All resistors checked and OK.
4/ The 2 diodes on the schematic that are inline with pot 612 have been replaced with a thermistor in parallel with a resistor. This is a later revision done by Philips not me.
5/ HF output amp checked as well (PN junctions, voltages, resistors).