Trio Lafayette KT-320 Question - Page 2 - UK Vintage Radio Repair and Restoration Discussion Forum

cathoderay57 · 22nd Nov 2019, 12:46 pm

I changed C10, C11 & C12 but the BFO would still not start. However, I refitted the 10nF (not shown in the circuit diagram, but found fitted on receipt) across cathode resistor R10 and hey presto, this time the BFO started and appears (so far) to be doing so reliably. Clearly the cap is providing a cathode AC bypass but not sure why that is encouraging the BFO to start as presumably the osc feedback path is from anode to grid? I'm sure there will be someone out there who is much more knowledgeable on oscillator theory than me! Cheers, Jerry

Skywave · 22nd Nov 2019, 1:33 pm

Quote: "knowledgeable on oscillator theory". It's tempting to get into the complications of oscillator design (e.g. Nyquist, feedback equations, calculating the required phase of feedback paths, etc.), but here we are not designing an oscillator, just fixing it, so a 'why analysis' is appropriate instead.

By fitting that cap. across the cathode resistor, you have increased the stage gain. (Without it, the stage has negative feedback). For any oscillator to work, two essential conditions must be met: appropriate feedback that achieves the necessary phase change - usually termed 'positive feedback' - and sufficient gain in the stage when viewed as an amplifier.

Look at the 'mirror situation'. Suppose you've built an amplifier for 450 kHz, but it (unwantedly) oscillates. The fix? Reduce the gain or reduce the feedback - which may be difficult since the unwanted feedback may not be due to any discrete components. So you reduce the gain, the stage is now stable, but now it has very little gain, if any. (There is the alternative approach of neutralization, but that tends substantially to diminish bandwidth and is critical to get right, and that is taking the 'design path' anyway).

In your case, that cap. was not originally fitted - and, presumably, the BFO then worked without it. That simply suggests that the stage gain now is a lot less that when the set was new. And that could be due to many reasons, a tired, low gm valve being one possibility. I would just fit that cap., make a note of this 'modification' for future possible reference, and live with it.

Al.

cathoderay57 · 22nd Nov 2019, 3:40 pm

Thanks Al, that makes a lot of sense! Cheers, Jerry

Nuvistor · 22nd Nov 2019, 4:34 pm

I know you tried the valve and various item, it make me think of something with the coil losing Q, if it is that well the cap across the cathode resistor is probably the easier cure.

Skywave · 23rd Nov 2019, 2:17 pm

"BFO coil losing Q": yes, a possibility, but note that Jerry (the OP) did state in post 16 that when he adjusted the core of the BFO coil, the tuning was very sharp. That suggests to me that the LC circuit that constitutes the heart of the oscillator is O.K.; hence the coil is O.K.

OTOH, Jerry did state that he tested the valve: a 6AV6 and also tried a replacement. In the past, when tackling similar faults in other comms. radios - "Is this valve faulty?" - I have found more that two replacement valves failed to fix the problem - it wasn't until a third (or possible a fourth) valve was fitted that the problem was fixed. Having said that, though, those valves were not new ones.
Consequently, I don't think that here we have unquestionably eliminated a faulty 6AV6 valve as being the culprit.

Hence, one Q. remains. Jerry: you stated that you "tested the 6AV6 valve(s) and they were O.K". What did you use to make that assessment?

Al.

Jon_G4MDC · 23rd Nov 2019, 3:16 pm

As this is a dual function circuit between Qmultipler and BFO, and seeing that the switch between the two (connected to Mixer Anode for Qmult, coupled by stay capacity for BFO) looks to be ganged to the pot controlling bandwidth, this could be down to the pot developing "end resistance".

At BFO setting resistance is minimum so the cathode should have only the 3.3k fixed resistor and that is supposed to give gain enough that with only minimal loading it will oscillate. If the pot track is not going to zero Ohms at that end of the travel, gain will be reduced and it could fail to oscillate. It might be worth trying shorting wiper tag to end tag at the appropriate end just to see if it will start.

cathoderay57 · 23rd Nov 2019, 9:52 pm

Hello all, thanks for the interest and the responses. BTW the BFO is still working..... As far as the 6AV6 valve tests are concerned the results (as tested on a VCM163) were as follows, (Avo manual spec values in brackets): Currently in-use valve Ia=0.8mA (1.2); gm=1.1 (1.6). The spare Ia=0.7mA (1.2); gm=1.2 (1.6). Jon is correct, the Q-Mult switch and pot are indeed ganged; it's a "backwards" pot so for BFO operation the control is turned fully clockwise with the switch off. In this position the resistance measured from cathode to chassis is 3.3k so there isn't any "end float" (for want of a better term) on the pot track. The only other factor that might have a bearing is that the BFO sub-chassis is connected by fairly long leads to the controls on the front panel, including a coax wire some 5 inches long from the HT feed resistor R8 to the grid side of the BFO coil so presumably there will be some losses in that. Maybe a factor.

Skywave · 24th Nov 2019, 12:34 am

I don't think I can now add anything further to this thread which may be of interest or helpful. However, here's a final post from me by way of an 'exit'.

I've owned two of these sets: now, do not own either.

The first one was bought second-hand in 1968. (Previous user had owned it for one year). It was factory built, and at the time - being a novice SWL etc. - I was initially quite happy with it. A couple of years later I became aware of some of its short-comings, so I replaced it with something else.

The second one I acquired cheaply during 2000 as a 'needs a full restoration' project. It had been substantially 'got at', including the removal of the entire BFO sub-ass'y. It went on to the 'round-to-it' shelf for a few years and was eventually passed on to someone else. To be quite candid, I felt that the work required to restore it fully would not be worth the effort, since at that point in time I was even more aware of its short-comings - as far as I was concerned. That conclusion was clearly coloured by my then ownership of a Racal RA-17. (Which I still own).

The one thing I always intended to do to this Trio / Lafayette receiver was to modify the design so that the BFO and the Q-multiplier were distinctly separate stages - which would involve another valve. I never got round to that - and I now know I never will; something I don't regret. But that idea might inspire others: Jerry here, perhaps.

All said: I hope you have fun with your radio, Jerry. I'm glad I could help.
TTFN,

Al.

22nd Nov 2019, 12:46 pm	#21
cathoderay57 Nonode Join Date: Oct 2017 Location: Bristol, UK. Posts: 2,382	Re: Trio Lafayette KT-320 Question I changed C10, C11 & C12 but the BFO would still not start. However, I refitted the 10nF (not shown in the circuit diagram, but found fitted on receipt) across cathode resistor R10 and hey presto, this time the BFO started and appears (so far) to be doing so reliably. Clearly the cap is providing a cathode AC bypass but not sure why that is encouraging the BFO to start as presumably the osc feedback path is from anode to grid? I'm sure there will be someone out there who is much more knowledgeable on oscillator theory than me! Cheers, Jerry

22nd Nov 2019, 1:33 pm	#22
Skywave Rest in Peace Join Date: Jun 2006 Location: Chard, South Somerset, UK. Posts: 7,457	Re: Trio Lafayette KT-320 Question Quote: "knowledgeable on oscillator theory". It's tempting to get into the complications of oscillator design (e.g. Nyquist, feedback equations, calculating the required phase of feedback paths, etc.), but here we are not designing an oscillator, just fixing it, so a 'why analysis' is appropriate instead. By fitting that cap. across the cathode resistor, you have increased the stage gain. (Without it, the stage has negative feedback). For any oscillator to work, two essential conditions must be met: appropriate feedback that achieves the necessary phase change - usually termed 'positive feedback' - and sufficient gain in the stage when viewed as an amplifier. Look at the 'mirror situation'. Suppose you've built an amplifier for 450 kHz, but it (unwantedly) oscillates. The fix? Reduce the gain or reduce the feedback - which may be difficult since the unwanted feedback may not be due to any discrete components. So you reduce the gain, the stage is now stable, but now it has very little gain, if any. (There is the alternative approach of neutralization, but that tends substantially to diminish bandwidth and is critical to get right, and that is taking the 'design path' anyway). In your case, that cap. was not originally fitted - and, presumably, the BFO then worked without it. That simply suggests that the stage gain now is a lot less that when the set was new. And that could be due to many reasons, a tired, low gm valve being one possibility. I would just fit that cap., make a note of this 'modification' for future possible reference, and live with it. Al.

22nd Nov 2019, 3:40 pm	#23
cathoderay57 Nonode Join Date: Oct 2017 Location: Bristol, UK. Posts: 2,382	Re: Trio Lafayette KT-320 Question Thanks Al, that makes a lot of sense! Cheers, Jerry

22nd Nov 2019, 4:34 pm	#24
Nuvistor Dekatron Join Date: Aug 2013 Location: Wigan, Greater Manchester, UK. Posts: 9,431	Re: Trio Lafayette KT-320 Question I know you tried the valve and various item, it make me think of something with the coil losing Q, if it is that well the cap across the cathode resistor is probably the easier cure. __________________ Frank

23rd Nov 2019, 2:17 pm	#25
Skywave Rest in Peace Join Date: Jun 2006 Location: Chard, South Somerset, UK. Posts: 7,457	Re: Trio Lafayette KT-320 Question "BFO coil losing Q": yes, a possibility, but note that Jerry (the OP) did state in post 16 that when he adjusted the core of the BFO coil, the tuning was very sharp. That suggests to me that the LC circuit that constitutes the heart of the oscillator is O.K.; hence the coil is O.K. OTOH, Jerry did state that he tested the valve: a 6AV6 and also tried a replacement. In the past, when tackling similar faults in other comms. radios - "Is this valve faulty?" - I have found more that two replacement valves failed to fix the problem - it wasn't until a third (or possible a fourth) valve was fitted that the problem was fixed. Having said that, though, those valves were not new ones. Consequently, I don't think that here we have unquestionably eliminated a faulty 6AV6 valve as being the culprit. Hence, one Q. remains. Jerry: you stated that you "tested the 6AV6 valve(s) and they were O.K". What did you use to make that assessment? Al.

23rd Nov 2019, 3:16 pm	#26
Jon_G4MDC Nonode Join Date: Nov 2015 Location: Welwyn Garden City, Hertfordshire, UK. Posts: 2,015	Re: Trio Lafayette KT-320 Question As this is a dual function circuit between Qmultipler and BFO, and seeing that the switch between the two (connected to Mixer Anode for Qmult, coupled by stay capacity for BFO) looks to be ganged to the pot controlling bandwidth, this could be down to the pot developing "end resistance". At BFO setting resistance is minimum so the cathode should have only the 3.3k fixed resistor and that is supposed to give gain enough that with only minimal loading it will oscillate. If the pot track is not going to zero Ohms at that end of the travel, gain will be reduced and it could fail to oscillate. It might be worth trying shorting wiper tag to end tag at the appropriate end just to see if it will start. Last edited by Jon_G4MDC; 23rd Nov 2019 at 3:22 pm.

23rd Nov 2019, 9:52 pm	#27
cathoderay57 Nonode Join Date: Oct 2017 Location: Bristol, UK. Posts: 2,382	Re: Trio Lafayette KT-320 Question Hello all, thanks for the interest and the responses. BTW the BFO is still working..... As far as the 6AV6 valve tests are concerned the results (as tested on a VCM163) were as follows, (Avo manual spec values in brackets): Currently in-use valve Ia=0.8mA (1.2); gm=1.1 (1.6). The spare Ia=0.7mA (1.2); gm=1.2 (1.6). Jon is correct, the Q-Mult switch and pot are indeed ganged; it's a "backwards" pot so for BFO operation the control is turned fully clockwise with the switch off. In this position the resistance measured from cathode to chassis is 3.3k so there isn't any "end float" (for want of a better term) on the pot track. The only other factor that might have a bearing is that the BFO sub-chassis is connected by fairly long leads to the controls on the front panel, including a coax wire some 5 inches long from the HT feed resistor R8 to the grid side of the BFO coil so presumably there will be some losses in that. Maybe a factor.

24th Nov 2019, 12:34 am	#28
Skywave Rest in Peace Join Date: Jun 2006 Location: Chard, South Somerset, UK. Posts: 7,457	Re: Trio Lafayette KT-320 Question I don't think I can now add anything further to this thread which may be of interest or helpful. However, here's a final post from me by way of an 'exit'. I've owned two of these sets: now, do not own either. The first one was bought second-hand in 1968. (Previous user had owned it for one year). It was factory built, and at the time - being a novice SWL etc. - I was initially quite happy with it. A couple of years later I became aware of some of its short-comings, so I replaced it with something else. The second one I acquired cheaply during 2000 as a 'needs a full restoration' project. It had been substantially 'got at', including the removal of the entire BFO sub-ass'y. It went on to the 'round-to-it' shelf for a few years and was eventually passed on to someone else. To be quite candid, I felt that the work required to restore it fully would not be worth the effort, since at that point in time I was even more aware of its short-comings - as far as I was concerned. That conclusion was clearly coloured by my then ownership of a Racal RA-17. (Which I still own). The one thing I always intended to do to this Trio / Lafayette receiver was to modify the design so that the BFO and the Q-multiplier were distinctly separate stages - which would involve another valve. I never got round to that - and I now know I never will; something I don't regret. But that idea might inspire others: Jerry here, perhaps. All said: I hope you have fun with your radio, Jerry. I'm glad I could help. TTFN, Al.