Mullard 3 watt amp - gobsmacked !

[yestertech]

Amazing ! who would have thought I could be bowled over by a 3 watt valve amp !
This concerns an unloved Mullard 3-3 which I bought in a job lot recently. Probably not fired up for decades and valve-less ,it looked rather unloved.
Like most of these, it has been well made from a kit of parts right down to the bus-bar earthing. ( the 'right' way to do things )
Just by way of a change, I thought I would try it out. But first a few minor repairs....
The dual 50 +50 electrolytic reformed fine from the bench HT PSU, and having found some valves I fired it up on t'variac as usual.
After a while the HT began to appear. There is no "that capacitor" to worry about as the circuit is DC coupled from Ef86 to El84. Most of the voltages appeared to be about right, perhaps a bit low. Those around the Ef86 were very low and I mean VERY, as this stage is configured at 'starvation' volts to gather maximum gain and to enable DC coupling.
A quick prod of the input tag produce a healthy "barp". I then set about measuring the output and distortion. It seemed happy to produce 3 watts into 15 ohms, but the distortion figures were way out of spec. A quick frequency squeak was none too impressive either, with both LF and HF ends drooping within the normally expected 20 Hz to 20 Khz.
On revisiting the schematic, I could see that the HT feed to the EF86 was low, as was the screen grid supply to the EL84. On measuring the relevant resistors, they were miles out. There was also a 220nF decoupler on the anode load feeder to the EF86, and, yes you guessed it, it was a waxy - and leaking like a sieve. Having checked most of the resistors and found them out of spec, they were all replaced except the green w/wounds which were OK.
Surprisingly, the 2 rather elderly 25 micro electrolytics checked out fine for ESR and value so these were left alone. For good measure, a NOS EL84 was fitted and the whole test sequence re-run.
Result ? A complete transformation ! Distortion now comes in at the quoted 1% at 3 watts and the response is virtually flat from 15 Hz to over 20 KHz. ( 1 watt )
Much of this is a result of the heavy feedback used - A "raw" EL84 single ended generates nearly 10% THD at 3 watts !
So onto a listening test, via my now veritably antique Rogers Exports ( well, one )
Now, these are 8 ohms and this amp only offers 3.75 or 15 , so there was a degree of mismatch...oh and they're not terribly efficient, due to mountains of iron in the crossovers
BUT ....Breathtakingly good ! More than adequate lounge-filling volume until the sound just begins to break up on the peaks.
I'm amazed it sounded so good. Just goes to show how less is sometimes more.

I'm off to look for my PX4 single ended transformer and my chassis cutters now !

Andy

[ppppenguin]

This shows that Mullard knew what they were doing when they designed circuits to use their valves. A very good marketing exercise too.

[murphyv310]

These are wonderful amps and with an efficient speaker loud. I have one up in the loft and fed it from my Quad QC22 preamp and was quite amazing.
I think I will build a matching pair and use another spare Quad as a preamp!
Yet another project for the list

[percival007]

These are fantastic little Amps.

I have built a couple (see other threads on here) and use one daily as my main Stereo Amp. It sounds wonderful.

James.

[Hybrid tellies]

That does look nice and I bet it sounds good as well.

[yestertech]

Wow that 's a pretty looking amp. I'll wager the O/P trannies are a cut above average too !

Andy

[amp_mangler]

no wonder it sounds good look at the size of the output transformer !

Joe

[flyingtech55]

Don't forget that these are 'proper' watts. Not PMP or PMOP or any other of the marketing bo****ks that the marketing pillocks dream up. These watts date back to the days of honesty. If they were marketed these days they would probably be rated at 150Watts Peak Music Power.

TimR

[raditechman]

I have one of these amplifiers from years back sitting on a shelf in the garage. Since reading your post I think tomorrow think I shall get it down re-valve it and bring it back to life again.
I remember when I first used it back in the 1960's I was very impressed with it, I only stopped using it as I went on to the Mullard 5-10 amplifiers.

John

[Valvepower]

Hi,

Yes, they are a rather nice amplifier. I’ve a pair of the Sterns Kit Mullard 3-3 amplifiers with the Parmeko output transformers.

One thing I observed is the EF86 has to be good to get the circuit working at it’s best.

The DC coupling from cathode of the EL84 to the screen grid of the EF86 & anode of the EF86 to the grid of the EL84 is very intresting, along with using the EF86 in starvation mode.

Terry.

[Sideband]

I built one of these back in the 70's to replace an amplifier in a record player. From memory, the amp I replaced used a couple of UCL83's in push-pull on a PCB and you would think would sound better. However the PCB was tracking and causing the UCL's to go into grid current.

So I took it out, built the 3-3 and the results were astounding. I changed the cartridge to a Sonotone 9TAHC and fitted a better speaker (dual cone type) and it was so good, I was asked to fit a line in socket so a tuner could be connected. It was the main source of music and entertainment in my brother's flat for many years. He used to connect it to a larger external speaker for parties and there was more than enough volume.


SB

[Mike Phelan]

The 3-3 has always been one of my favourites since I made one in the 1960s and used it for years. Drove one of those 13" x 8" EMI speakers with two tweeters, available from most of the surplus places.

[johnn]

What a thoroughly absorbing write up Andy!
Impressed by your component testing, before and after tests followed by the listening tests.
John.

[Andrew2]

Teriffic!
Back in the late 80's, I was called out to an elderly gent who had a pair of those he built from kits in the early 60's. They drove a pair of the 'small' Leak Sandwiches.
The reported fault was 'one channel dead', but even after fixing the fault (heaters out on one EL84), the system sounded tired and in need of some loving care.
Although the owner was 'short of brass', I convinced him to let me take both amps back to the workshop where I could do a proper job. I told him I'd go easy when I got to the invoicing bit.
At this point the story becomes almost a carbon copy of yours, Andy - including that decoupler.
I cleaned the pots and re-tensioned the valve-bases and re-valved both amps, and 'forgot' the 3 hours on the bench, just charging him for the original call-out and the parts.
He was such a nice bloke and over the moon to have his beloved Mullards sounding so good again, and we enjoyed a cup of tea while we listened to Gershwin's Cuban Overture from his immaculate vinyl collection. Ah.
Nice one, Andy.

[yestertech]

Great story that - Thanks
Field service for valve audio gear - now that would be a job made in heaven !
Although my field service days are behind me, I still miss the customer contact and that need to think on your feet.
I was mainframe and peripherals so when the system was "down" the pressure was enormous.
But the feeling you got when you left ,a job well done - immeasurable.

All laid to waste by "n+1 everything" That's progress

Andy

[Valvepower]

Hi,

If it’s of interest I’ve uploaded a scan of the Radio Constructor article from 1956 describing the Mullard 3-3.

You’ll notice there are subtle differences between the circuits in the Mullard publications and the Radio Constructor. The differences are in the EF86 anode load and the screen grid supply.

The Mullard circuit can be found here:

http://www.r-type.org/static/3-3.htm

Terry.

[Aerodyne]

I've just completed a 3-3 stereo amp and whilst still not entirely satisfied that I've got it to its peak, the results are so far very impressive and fully justify my memories - back in the 1950s, a friend built a kit version which, in modern parlance, blew me away.

In the later 50s when stereo came about, I built my first stereo amp around twin 3-3s. This amp worked superbly for many years until I passed it on in the late 60s (or possibly early 70s) to a colleague at a firm I worked for. He had the amp for many years before he too passed it on to another. It gave sterling service and I can recommend anyone thinking of building a valve amp but unsure of performance to go ahead, just get the best transformers you can find. These do not come cheap, however, in these days of plastic and microchip.

Great thread, this. Thanks to all and especially Terry for the Radio Constructor files. These throw an interesting light on the design, as the differences he mentions are in fact more than subtle! I'll get around to trying them out.
-Tony

[Sideband]

I'm tempted to build a couple of these again just for old-times sake. One point...anyone building this (or two for stereo) should try and use the best quality resistors available for the EF86 stage. Of course standard resistors will work but hi-stability types will help to further reduce noise levels in this very sensitive stage.


SB

[Valvepower]

Hi,

Tony, no problem I’m glad it was of interest as I felt it was different enough to the Mullard publication to be intresting, plus The Radio Constructor articles make an intresting read.

One of the advantages of DC coupling there is one less phase shift to worry about.

I feel this circuit is really innovative with the DC coupling, and is more than just an EF86 driving an EL84 as used elsewhere.

When replacing the valves one should consider the seesaw action in the DC coupling between the anode of the EF86 and the grid of the EL84 and between the cathode of the EL84 and the screen of the EF86 to get the best performance.

Terry.

[Valvepower]

Hi,

Whilst looking for some Mullard Outlook's for another post I unearthed two more articles on the 3-3 circuit.

Phil_EL41_3 is from the 1954 Phillips book Valves For AF Amplifiers and Phil_3_3 is from the later 1960 Phillips book Hi-fi Amplifier Circuits.

The Phil_EL41_3 is the first version using EL41 and no DC coupling, and Phil_3_3 is a Phillips article of the Mullard 3-3 circuit.

Terry.