Mullard 3-3 voltages query
 

I'd be grateful for advice on my newly built 3-3.

I recently bought an old SE output transformer harvested from a radiogram running a single EL84. I decided to assemble a Mullard 3-3 to test it and I now have that amplifier working. Although it sounds good, there are some problem voltages. I'm measuring voltages with a Fluke 87 DMM with good batteries. Specified voltages in brackets below. I have a working AVO8 to hand.

I've studied all the threads on the 3-3 here and know that correct voltages are especially important for correct operation of this circuit. I have checked that the primaries of the output transformer are correct and checked and re-checked my assembly and solder joints.

My power transformer has a separate 5V winding, so I'm using a GZ30 rather than EZ80 for the rectifier. After a 150R dropper resistor R15, I get 297V (310V) at the first 50uF capacitor (C9). I guess I could use a lower value resistor to raise it a bit. However, the voltage at C6 is significantly lower than it should be at 253V (290V).

I'm puzzled by one symptom: when I measure the voltage at the EL84 anode (pin 7), the sound goes off and I get 250V. But when I touch the grid of the EF86 with DMM probe, the sound returns.

The other problem is higher/lower voltages on both valves:
EF86 anode: 48V (20V)
EF86 G2: 49V (28V)
EF86 cathode: hardly any voltage and nothing specified in the circuit.
EL84 G2: 259V (290V)
EL84 cathode: 50V (28V)

Is the disappearing sound a useful diagnostic symptom?

Re: Mullard 3-3 voltages query
 

That sounds like it might be turning into an oscillator.

Lawrence.

Re: Mullard 3-3 voltages query
 

Yes....it's a common thing with these amplifiers since the wiring layout is quite critical. It's probably going into supersonic (you can't hear it) oscillation. The EF86 is running in 'starvation' conditions and the gain is very high. This can result in supersonic oscillation through poor layout or long runs of wiring.

These really are excellent little amps (I've built three in the past) but you really need to be careful with layout and follow the Mullard wiring as closely as possible.

Re: Mullard 3-3 voltages query
 

Is there a smoothing choke? If so, could it be dropping too much voltage?

Aub

Re: Mullard 3-3 voltages query
 

You didn’t forget to fit the 10MΩ grid-leak resistor on the EF86, did you? I’ve done that myself with this amplifier - easy to miss because in many/most input circuits you might be expecting the grid to get a DC path to ground through the volume control. But in this circuit the v.c. is AC-coupled to the grid, so the valve really needs that 10MΩ to ground.

Mike

Re: Mullard 3-3 voltages query
 

I used a chassis I had to hand, but have followed the layout in the Sterns construction manual carefully.

The 10M grid leak resistor is there.

The OPT connections are temporarily quite long. Could that be an issue? I also need to get a 560 K cathode resistor for the EL 84.

Can I view the oscillation on my scope?

Re: Mullard 3-3 voltages query
 

Here is a pic of the layout.

Re: Mullard 3-3 voltages query
 

Long connections to the output transformer could easily be the problem. I presume you mean a 560 ohm cathode resistor? what value are you using now as 50V on the EL84 cathode equates to 90mA and will burn out the EL84...

Re: Mullard 3-3 voltages query
 

Sorry, yes 560R. I’m using 470+100 in series at present. Should I try increasing it? I see the Radio Constructor article of April 1956 uses 680R.

Re: Mullard 3-3 voltages query
 

Sorry to go against the grain, but I do not like the 3-3 design.
For every one you have working in spec, there are 2 failures...
They are so dependant, amongst several factors, on the EF86 and EL84 characteristics being the same as the originals used in the 'model' amp. First step would be try several other EL84s first, then other EF86s.
You do not say what value cathode resistor you have at the moment?
Also a random output transformer might need feedback changes from the Mullard values. Have you tried it with the feedback disconnected?

If all else fails, do what the BBC did, and AC couple the EF86 to the EL84 as in their AM8/2A circuit...
Alan

Re: Mullard 3-3 voltages query
 

The EF86 anode load looks like 1.5M not the 1M in the Mullard design?

The EF86 anode voltage reading in the Mullard documentation is lowered by the 2M AVO 8 input resistance. Even the 10M of your DMM will have an impact.

You need to make sure it is not oscillating as this will affect the DC measurements. If it is stable then the EF86 is probably weak.

Re: Mullard 3-3 voltages query
 

It could be worth checking again, your photo seems to show it connected on the wrong side of the coupling capacitor rather than directly to the grid...?

Pete

Re: Mullard 3-3 voltages query
 

Yup, that would do it!

Re: Mullard 3-3 voltages query
 

Well spotted!

Re: Mullard 3-3 voltages query
 

Pete, I owe you a beer! :beer:

It was the one thing I didn't worry about, even though Mike (Boulevardier) drew my attention to it yesterday. Thanks to all. What a pleasure to see all the correct voltages slot into place. And it does sound good with the large OPT. I attach pic.

Since last night, I had already tried:
1. raising and lowering the value of the EL84 cathode resistor;
2. swapping the OPT secondaries, following Morgan Jones' advice on oscillation;
3. shortening the leads to the OPT (necessary anyway);
4. tried a number of different EF86s and EL84s.

This a bit of a milestone for me. With much help from members here, this non-engineer has now built all the Mullard amps in the book (5-20, 5-10, 7W stereo and now the 3-3).

David

Re: Mullard 3-3 voltages query
 

Congratulations, David! Well worth the perseverance. I think, for all their simplicity, these are jolly good little amplifiers. I've often wondered if, if you put one beside a 5-10 and equalised the gains, it would be difficult to tell them apart on an A-B blinded test. And, using the sensitive loudspeakers of its period, 3W is quite a surprising level of noise in a domestic environment.

Mike

Re: Mullard 3-3 voltages query
 

Wow - that's a massive OPTX for a single-ended amp like this!

Re: Mullard 3-3 voltages query
 

Hi David, you can't stop now. There is a Philips Tech library book by Rodenhuis
"Hi Fi Amplifier circuits" with yet more of them for you to build

Ed

Re: Mullard 3-3 voltages query
 

Yes Edward, I'm afraid that's what drew me in to this side project! Reading Morgan Jones' Building Valve Amplifiers yesterday, I realise I should learn to dial in optimal feedback values. I just don't know how to connect up a dual gang variable capacitor!

Ed D, I downloaded the Rodenhuis book some years ago and do consult it, but I still turn to the Mullard book first as it's a print copy. I'm contemplating a Maplin Millennium type project next. I have the OPTs and a huge Leak PT.

Re: Mullard 3-3 voltages query
 

I use a 10k 10 turn pot to find the right R FB value; connect between IP & OPT, set at 10k, start lowering until the amp starts to get unstable, then back off a bit. To find the C, get a capacitance decade box,IP a 1khz square wave to the amp, monitor OP on a scope, play with values till you get a nice sq wave, then do the same for 10khz. This is the simple procedure.

To sort an amp properly you need to do lots of sweeps from 20hz (lower if you can) - 100khz at various OP levels with a resistive load, then a capacitive load.This reveals it's instability points, what your looking for is where the OP peaks or where THD shoots up. When you look at your results you'll see something like "peaks at 48.3khz & at 58khz" or peaks at 10hz & 42khz. You then select your FB RC values to form a HPF at the problem frequencies using 1/(6.28 * f * r or c). It's also a good idea to pop a Zobel or Bucheret network on the OPT secondary using the same data & method. What your doing is putting RC filters in various places of the amp to stomp on any instability or silly behaviour. This is why you'll see an RC in parallel with the IP stage anode resistor or small value (picofarad) caps in parallel with other parts of the circuit in an amp,broadly speaking that is.

This results in a stable amp, stable with all loads. It can be very time consuming but it's worth it.

Andy.