Ferranti Lancastria - strange fault

[Ian B]

Although I went to the last Punnetts Town event determined not to buy any more projects, I felt sorry for a rather paint-splattered Lancastria which had the potential to be a very attractive set. It was sound, complete and (importantly) cheap! So I brought it back to play with during the cold weather.
Some of the original caps and one resistor had been changed many years ago,
but the quantity of dust on the top of the chassis testified to a lack of recent work. I changed all the waxies and electrolytics and fitted a new mains lead (the original was about as flexible as dry spaghetti and still had a bayonet connector).
Switch-on produced only a slight hum, but I soon identified an o/c cathode bypass resistor on the output valve. Voltage checks now appeared normal . However, I could only just pick up very strong signals like 5 Live on MW. I also found that maximum volume was at about half way; increasing to maximum reduced the output! The control was rather scratchy, so I fitted a new old stock replacement. Result - no scratch, but still volume reduces past half way position!
I've checked every resistor, rechecked my wiring, tested the valves and also substituted known good ones - still the same. Oh, and although the original volume pot did have a minimum at midway, the replacement measured correctly.
I've been working from "Trader" sheet no 106 which I've been staring at without inspiration; anyone got any ideas please?
In desperation,
Ian Blackbourn

[Station X]

Have you tried feeding an audio signal into the pick up (PU) socket?

[cathoderay57]

Check that the connections to the slider and the "top" (i.e. non-earthy end) of the vol cont pot have not been accidentally reversed. Cheers, Jerry

[PJL]

Grid leakage in the PT4D would do this. Try monitoring the PT4D grid voltage with a DMM as you increase the volume.

[Ian B]

Reversing the end connections of the pot results in max volume at either end of travel.
Feeding an audio signal to the pick up sockets gives a similar result - except that max volume is at around mid-position with connections reversed (as above) and at either end with the connections as original!
DVM attached to top cap of output valve (actually an AC Pen 2DD) reads 76.6 volts at minimum setting of volume control falling to 46.6 at max. It's not a reading I've ever measured before, so I don't know what to expect.

Thank you for your suggestions so far gentlemen, I'm sure the problem has a simple cause, I just can't fathom what it is!
Ian

[cathoderay57]

Hi Ian, I've now had chance to have a look at the circuit. Before I make any further comment I note that there are several variants of the Ferranti Lancastria. The original vol cont pot connections seem OK so put them back as they were. You are using Trader Sheet #106 but are you sure the circuit matches your set? Although the PT4D and AC2/PenDD seem to be equivalents there is an alternative circuit in Trader Sheet (#519) which maybe has some differences in the AF output area (they are drawn differently which makes comparison tricky). Using Trader #106 check the cathode bypass electrolytic C16 (4uF 12v) also C15 (0.25uF) connected between one end of vol cont and cathode of AC2/PenDD. Check C14 (0.02uF), and the resistance across the ends of the vol cont pot R15 (1M) plus R13, R14, R16, R17 and R18. When doing resistance checks disconnect one end of the resistor otherwise the in-circuit measured value will be different. Hope that helps. Cheers, Jerry

[PJL]

The grid voltage should not change with volume (assuming the trader 106 is correct) and is far too high. My guess is a problem with the PT4D cathode resistances R17 & R18.
Some quick resistance checks would be:
PT4D Cathode to Chassis = R17+R18 = 740 ohms
PT4D grid to chassis (min. volume) = R15+R16+R18 = 100600 ohms
PT4D grid to chassis (max. volume) = R15+R16+R18 = 1100600 ohms
Anything within 25% would be OK.

[Ian B]

I'm pretty sure sheet 106 is the correct one, the top and bottom chassis views are correct and the component values and positions agree, whereas 519 looks different.
All the resistances measure well within 25% of marked value. The caps are all new, but I've disconnected them to measure them and they're all ok.

The volume pot in the set as found was 500k rather than the specified 1meg. I fitted a new 500k (I can't find a good 1meg!) assuming it must have worked ok before like that. I've tried a 2meg I discovered - same result.

Tomorrow I'll try to trace every part of the circuit in case I've made a wrong connection somewhere.

wish me luck!
Ian

[cathoderay57]

Hi Ian, OK, more head scratching required then. PJL is on the money I think - 46-76v (presume DC?) on grid 1 of the AC2/PenDD seems very wrong. I could see how it might happen from the circuit in the #519 Trader because there is a potential divider between smoothed HT and chassis comprising R17, R18, R19, R21 and R22. The positive voltage on R22 is tapped off and connected to the bottom of the volume control and also the AC2/PenDD cathode. However R22 is small value compared to the rest of the potential divider so unless other resistors had gone short circuit the positive potential on R22 should be small-ish and is provided for the purpose of delayed AGC. If the cathode circuit on your radio is the same as Trader #106 there is no such potential divider and the cathode is simply connected to chassis via R17 and R18. So the question remains where is the high voltage on the AC2/PenDD grid1 coming from? I would be inclined to substitute a replacement AC2/PenDD if you have one, or get it tested. Internal valve leakage could be the issue here. Cheers, Jerry

[Ian B]

I've tried two spare used AC2Pen/DDs and a brand new AC5Pen/DD. They all test ok on my 2 panel AVO valve tester, and all give similar results; the grid voltage going from about 45 at minimum to 86 maximum.
Should I try increasing or decreasing the cathode bias?

Ian

[cathoderay57]

Hi Ian, what voltage are you measuring between the top cap connector and chassis when you remove it from the valve? Also what voltage are you measuring between the AC2/PenDD anode and chassis, and grid 2 and chassis? Cheers, Jerry

[Goldieoldie]

Have you checked the earthy side of the volume pot is connected to the chassis ?

[cathoderay57]

For Goldie Oldie - Ian says the circuit is the same as Trader #106 in which the earthy end of the pot goes to chassis but via R16 (100k) and R18 (600R). Doing the sums on cathode bias: R17+R18=740R; V3 anode current + screen current = cathode current =43.6mA. Therefore the voltage between cathode and chassis ought to be in the order of 32v (less if the valve has aged and reduced emission as is likely). Can you also measure the cathode voltage to compare with this? R17 and R18 act as a potential divider and so the DC voltage measured between the bottom of the volume control and chassis should be no more than about 26v. The grid current should be very small so I'd expect approx. the same 26v DC to appear at grid1 (top cap). I you can report back on the actual voltages measured at these points we should be able to get to the answer. There is no grid stopper resistor so it is also possible the valve is going into HF oscillation which could be upsetting things. The IF bypass capacitors C12 & C13 could possibly be giving trouble but that would be extremely rare. Cheers, Jerry

[cathoderay57]

Just for completeness as I said above 26v across R18 and 32v across R17+R18 means the grid should be about -6v with respect to the cathode and the grid bias spec for an AC2/PenDD is -5.3v so that would all be about right.

[PJL]

Ian, did you take the resistance measurements I suggested in post #7:
Cathode to chassis
Grid to chassis

My guess is still that something is wrong with the cathode path.

[cathoderay57]

Further thought - what voltage rating electrolytic did you use to replace cathode bypass electrolytic (4uF)? In post #6 I indicated that 12v would be OK but I was wrong. From the above calcs the cathode to chassis voltage should be in the order of 32v so that cap needs to be rated at 50v or 63v. Cheers, Jerry

[Ian B]

Measured the voltages again:-
Anode to chassis 258V
G2 to chassis 294V
Grid to cathode 68V
Grid to chassis 49V Connected to TC
Grid to chassis 54V disconnected from TC

All taken with volume control set to max noise (not max setting). At this setting, grid to chassis resistance about 240K, cathode to chassis 800R (I'm sure it was about 780R yesterday, but it's much colder in the workroom now as the heating hasn't come on yet; perhaps my DVM doesn't like the cold!)
The cathode bypass I used is marked 150V.

Now the thing that really surprised me - disconnecting the top cap connection to the AC2Pen/DD made no change to the volume!
Somebody much cleverer than me will say "That's because ……………"

[PJL]

A fault with the cathode bypass would lower the grid and cathode voltages so it won't be that.

According to this the cathode voltage=49-68V = -19V.
Can you measure the cathode voltage to chassis.
My calculations based on the trader sheet of cathode current = 37.0+6.6mA gives the cathode voltage = 32V and the grid voltage = 24V and the grid bias = -6.1V.

My new guess is instability and the RF signal is upsetting your meter.

[ms660]

No chassis to cathode voltage yet? Just to complete the picture.

EDIT: Blimey, post crossed...psychic.

Lawrence

[cathoderay57]

Hi Ian. Well this one's proving to be a bit of a brain teaser. G2 volts 294 (Trader 305) - OK. Anode 258v (Trader 295) - OK. Grid to cathode 68v not OK - should be about -6v as calculated by me and PJL. It's also odd that you are measuring a higher voltage Grid1 to cathode than Grid1 to chassis - should be the other way around for Grid1 to be negatively biased wrt cathode, as it should be. There's something odd there because with +68v on Grid1 with respect to cathode I would expect the anode to be glowing red hot and the valve well on its way to self destruction. Since we presume that isn't happening, or else you'd have mentioned it, PJL is most likely correct and it's a measurement anomaly. If the valve is going into oscillation normally you'd hear it unless it is oscillating at RF. If you have a capacitance meter check C12 and C13 (IF bypass caps). If they are within 25% of the values shown in the Trader Sheet, and not showing DC leakage, then leave them as-is. The other thing you can try is inserting a Grid stopper resistor of somewhere between 10k and 100k between the Grid1 wire and the TC of the valve, as close to the TC as poss. The McMichael 362 also has an AC2/PenDD AF output valve and has a 100k Grid stopper. Then remeasure the Grid 1 to chassis and cathode to chassis volts, and see if the volume control is functioning normally. The set shouldn't need a Grid stopper otherwise there would have been one fitted by design but tacking one in might help to diagnose what's going on. You can get sound with the TC disconnected via AF coupling internally in the valve from its detector diode I presume, although somebody might know better than me. Top marks for persevering, don't give up! Cheers, Jerry