UK Vintage Radio Repair and Restoration Discussion Forum

UK Vintage Radio Repair and Restoration Discussion Forum (https://www.vintage-radio.net/forum/index.php)
-   Vintage Radio (domestic) (https://www.vintage-radio.net/forum/forumdisplay.php?f=20)
-   -   Oh No! Not another howler. (https://www.vintage-radio.net/forum/showthread.php?t=141099)

Boater Sam 5th Nov 2017 7:22 pm

Oh No! Not another howler.
 
I seem to be plagued by howling Bush radios laterly.
The VHF41 is fixed now that I have found the poorly connected wire.
This one is only an AC41, possibly the simplest Bush chassis of the period.
Once the inevitable rotten wiring was replaced it works but when tuned off station it howls and rumbles.

Replacing the screen decoupler cap, AVC cap, oscillator anode isolator cap and audio grid cap leaves only the tone corrector caps that are waxies.

Its just the same. The twiddler has not been here, its a virgin, as it left the factory.

Valve voltages are about right, resistors are close enough to correct values.
Electrolytics are good, no HT hum, shunting all 3 does nothing.

It is fine in the gram position, so the triode in the ECH42 which is used as 1st audio amp in gram is OK and pulling the ECH42 doesn't stop the racket.

Pull the EBF80 and there is silence.

So it seems to be the EBF80 IF amp/diodes that is taking off.
Swapping V1 and V2 valves for known good ones also has no effect.
Reducing the volume to near to nothing stops it so it has to be before the EL41.
There are so few components in these sets that it is driving me mad.
My only thought is to split the combined V1 and V2 screen feed and decoupler and put a separate one in for V2 and see if that works, but this should not be necessary.

Please, any ideas?

G6Tanuki 5th Nov 2017 7:51 pm

Re: Oh No! Not another howler.
 
"Reducing the volume to near to nothing stops it so it has to be before the EL41."

This makes me think you've got IF signal leaking into the audio-chain, where it will be amplified [an EL41 is good for a couple of MHz] and fed back into the HT line.

Remember that the HT line is only decoupled by an electrolytic or two - which look like inductors at anything more than a few tens of KHz. So your HT line can be 'alive' with hundreds of millivolts of IF!

Try 500pF from the wiper of the volume-control pot to earth. And 100pF from the grid of the EL41 to its cathode. That should keep unwanted IF signals from building-up along the audio chain.

Though some will disagree, I like the "tone corrector" on the output-stage to be from anode-to-earth - that way it shunts any stray IF signals safely to earth. Putting it from output-stage-anode to the HT side of the output transformer feeds any stray IF signals into the HT line, just where you don't want them!

Boater Sam 5th Nov 2017 8:03 pm

Re: Oh No! Not another howler.
 
There are 2 100pf caps marked as signal diode filters in a pi network with a 47K resistor before the volume control, these are OK, shunting them does nothing.
I'll try grounding the IF, your suggestion is a good one, but the damn thing should work without modifications.
I agree with the remarks about the top cut tone caps, I don't like them from anode to HT either. I may just disconnect them and see what happens.

Sideband 5th Nov 2017 8:06 pm

Re: Oh No! Not another howler.
 
It might be a simple Bush....maybe too simple! They've probably cut the circuit to the bone in order to reduce costs so it may have been marginal in the first place. The idea of separating the screen grids is a good one and using separate decoupling on both. You could also try a grid stopper between 47k and 100k in the grid of the EL41. Recently I had a Murphy set that howled like a Banshee above a certain volume.....it was the volume control itself...just something else worth considering.

turretslug 5th Nov 2017 8:07 pm

Re: Oh No! Not another howler.
 
Is this one of the many sets where the output valve screen grid is connected directly to the HT line? That possibility of a mucky-at-HF HT line mentioned could be getting onto the screen grid which is perfectly capable of acting as a control electrode in its own right with gain between it and anode. Not having good HF decoupling on an HT line always strikes me as "busking it"- the smoothing electrolytic may have been "sufficient" when new but less good now. Maybe try a decent 100nF or so between HT and ground, at the screen grid pin if so connected- but after G6T's tips mentioned have been tried and also after trying a series grid stopper of 10-22k or so.

Boater Sam 5th Nov 2017 8:16 pm

Re: Oh No! Not another howler.
 
I suspect that it is not the audio taking off on its own as it is stable at all volume settings when in "gram"
If shunting stray IF to ground fails I may have to think about a few mods.
True about pared to the bone, only 4 valves, common chassis punching with other models, lots of unused holes, and the cabinet, scale pan, speaker/output transformer and knobs are of course common to many models produced at this time.
Even the valve line up is an economic measure, three B8A and one B9A.
The mains transformer is also thinner in the core than some others.

Boater Sam 5th Nov 2017 8:24 pm

Re: Oh No! Not another howler.
 
I have this dislike of pentode screen grids at higher HT than the anode by being direct onto the HT line too. I think a screen feed resistor with a 100nf decoupler is a better bet.
I was surprised that the reservoir and smoothing caps were so good, they check out very well against new parts. Course its a resistor smoothing, no choke.

Boater Sam 7th Nov 2017 1:29 pm

Re: Oh No! Not another howler.
 
Had another quick go at this howler, its more like fast motorboating
A few hundred pfs from EL41 g1 to ground does nothing. 4.7nf anode to ground similarly.
Separated the V1 and V2 screens and gave V2 its own 100k and 0.1uf off the HT line, zilch.
I am going to have to get serious with this set.
Measuring HT voltages today, they are +30v to +40v over but the mains is over 250v AC today. Filaments are at 6.8v AC. Tapping is 230/250v so no leeway there.
But the V1 and V2 screens are at 107v instead of 60v, there is something up with that.
I think I will replace the EL41 when I find a good one in the stock just to satisfy myself that it is OK.

Sideband 7th Nov 2017 3:09 pm

Re: Oh No! Not another howler.
 
Quote:

Originally Posted by Boater Sam (Post 989375)
But the V1 and V2 screens are at 107v instead of 60v, there is something up with that.

Not necessarily. You're probably using a digital meter so the impedance is higher...voltages will read higher.

It could also mean that something is oscillating.....!

Boater Sam 7th Nov 2017 7:39 pm

Re: Oh No! Not another howler.
 
The mains voltage dropped to a more sensible 243 volts this evening, the milking machines are on. With the AVO 8, voltages are now more in line with the sheet.

Tried shunting the IF to ground after the detector diode, no joy, and on the volume control, same racket.
Disconnected both top cut tone caps, no joy, still howling.

Sat and thought about it, decided to see if I could get a clean audio output both on the EL41 grid and on the triode grid. On gram its clean, on any waveband putting a signal into the EL41 it is still obvious that there is oscillation as well. The EL41 has a 47k grid stopper already.

Considered that it had to be before the volume control, as I said before pulling the IF valve V2 killed it.
Looking at the circuit I notice that there is no grid stopper on the g1 of the IF amp. I wonder if.........

Cut the wire from the IF transformer to g1, insert 8k2 resistor, success! No howling, clean reception on all bands. Tried smaller values, but it seems to need 8k2.
Removed all other extra components including the extra screen feed components for V2 and it is still working fine.

Now I don't really want to redesign the set but unless anyone can tell me what else I could do, its staying like this.

Comments most welcome please.

Skywave 7th Nov 2017 7:45 pm

Re: Oh No! Not another howler.
 
Many years ago, a design engineer with whom I used to work, told me one of his favourite mottos: "You can't have too much decoupling". O.K., you could take that to a ridiculous extreme case :-), but the general idea is sound.

After-thought: for this set, there is always to possibility of poor earthing, such as a soldered joint to chassis that looks "O.K.", but . . . . .

Al.

Boater Sam 7th Nov 2017 8:06 pm

Re: Oh No! Not another howler.
 
All chassis earths are on solder tags nut and bolted on, I've been over them with a spanner without finding any fails. I did find the bolts on the tuning cap were loose, the cap was twisted out of line, I was dreading trying to get to replace the grommets, but having tightened them its fine now.

ms660 7th Nov 2017 8:21 pm

Re: Oh No! Not another howler.
 
It won't take much in an IF amplifier stage, the grid is tuned and the anode is tuned, both are tuned with relatively high Q coils, enough regen and off it will go (oscillate)

Lawrence.

Skywave 7th Nov 2017 8:24 pm

Re: Oh No! Not another howler.
 
I know you said that you want to keep the design of the set as original - which is fine, of course. But I'm also sure that you'd like to find the cause of the trouble, even if you then choose not to implement the 'fix'. (Least, I would :-) ).

So, with that thought in mind, is the cause due to unwanted feed-back within the wiring configuration / cct. design? I would liberally distribute 0.01 µF caps. around the set in the many appropriate and various places. Then, hopefully, the instability ceases. At that point, disconnect a total of half on those caps. and see if the instab. remains. You can continue like that, successively removing half of the remaining caps. until you find the critical one (or more! :-) ).

Time-consuming, I realize: just depends on your 'need to know'; availability of time - plus patience and determination.

Al.

Boater Sam 7th Nov 2017 8:37 pm

Re: Oh No! Not another howler.
 
Thanks Al.
Thats more or less what I did, decoupled every HT feed, shunted IF filters etc.
It was suggested that this was due to IF getting into the HT line but if the grid stopper on the IF amp stops the oscillation I can't see that being the case.
I can't see how I could have reaction across the IF valve either, the set was in as built condition, no rewiring or caps.
Using the grid stopper brought about an instant fix, but why?
Its not valves either, all have been subbed.

turretslug 7th Nov 2017 8:50 pm

Re: Oh No! Not another howler.
 
Quote:

Originally Posted by Skywave (Post 989499)
Many years ago, a design engineer with whom I used to work, told me one of his favourite mottos: "You can't have too much decoupling". O.K., you could take that to a ridiculous extreme case :-), but the general idea is sound.

Al.

I've sometimes wondered if commercial designers build a prototype with plentiful decoupling (i.e. the amount that someone inclined to do the job "properly" would use) and that decoupling is then incrementally snipped out until the functioning is significantly impaired. (Keeping bean-counters happy).

"Right- put that last capacitor back, and sign it off for production".

I wouldn't be surprised....

ms660 7th Nov 2017 8:54 pm

Re: Oh No! Not another howler.
 
Out of interest what's the AGC voltage (R8/C13) using a DMM with the grid stopper fitted and tuned to a decent signal?

Lawrence.

turretslug 7th Nov 2017 8:57 pm

Re: Oh No! Not another howler.
 
Perhaps a leaky cap across the high impedance AGC line, stopping the controlled valves biasing back sufficiently?

Boater Sam 7th Nov 2017 9:05 pm

Re: Oh No! Not another howler.
 
I'll measure the AGC, remiss of me not have done but it was freezing out in the farm tonight, dark and no moon. Flushed with success I quit and went for dinner. I know I was getting around -3v on the AGC side of the coupling cap to the output G1, and +0.004v on the other side, aerial is not good, 4m inside the floor above.
The AGC cap is new.

kalee20 7th Nov 2017 10:12 pm

Re: Oh No! Not another howler.
 
Reading down, I'd have said, Ah! Residual IF being amplified by the AF amplifier and getting radiated by the wiring back to the IF amplifier input! But, that seems to have been covered.

I'd also have said, Bad chassis connection somewhere, putting a common impedance to true ground in a high-level stage and a low-level stage! But that seems to have been covered too.

When you tried extra capacitors stuck on here, there, and everywhere, did you take the earthy lead to chassis, or to an existing, possibly dodgy, solder tag which looked as though it was a solid chassis connection but may not have been?

All I can think of otherwise, is, without the EL41 in, does it still oscillate? Of course, how would you know... But maybe you could monitor the AVC line with the EL41 in, see if it goes more negative as you approach howling, remove the EL41 and see if the same happens. It's just barely possible that IF is present on the AVC line, and winding up the volume control is affecting this, with the EL41 just being a red herring!


All times are GMT +1. The time now is 12:31 am.

Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.