12E1 SE Amplifier

[The Shadow]

Hi All,

I've been working on and off on this little project for the last few months since Steve S brought me a box of valves to experiment with.

I've posted the cct below.

I have a couple of questions.
I'm trying to pull back from unnecessary complication so:

1. The 6SL7 cathode follower buffer input stage: Is it strictly necessary?
I put it in to try and avoid any untoward effects from what would otherwise be a passive volume control input, and to provide a low impedance coupling to the grid of the ML4 driver valve.

I could use a 6F8G (lovely looking valve with a top cap) to complement the appearance of the 12E1 output stage instead of the 6SL7

2. Could the output of the cathode follower be direct coupled to the ML4 thereby removing a capacitor from the circuit. or would it cause more problems than it would solve?

The ML4 has a very high transconductance, a reasonably low output impedance and an amplification factor of 12. The 12E1 has a gain of 5. I reckon the ML4 is man enough to drive the 12E1 and I like the idea of using an old Brit valve as a driver. There's not an awful lot of data about for the ML4 so I don't know if there would be any issues with Miller Capacitance. I could put in a C3g in its place if need be.

This will be housed in my newest chassis with the cage

Should be up and running in about 2 weeks if I get on with it.
Steve

[NickG]

IMHO, I would loose the cathode follower. It won't work as drawn anyway, you will need to raise the voltage on the grid somehow, and you can't do that if its connected to the input pot.

I think you may be a little short of gain, maybe another 6-12db somewhere would be good. If it was me, I would throw it together on a board and see how it works.

If the 12e1's are anything like the ones I have, I would have a big bag of grid stoppers to hand, I found they like to behave as oscillators in the Mhz range. If when you turn it on, you hear a odd hissing noise, and nearby DVM's start going mad, keep your hands away as there could be 20W or so of RF being generated.

[The Shadow]

Originally Posted by NickG

IMHO, I would loose the cathode follower. It won't work as drawn anyway, you will need to raise the voltage on the grid somehow, and you can't do that if its connected to the input pot.

I think you may be a little short of gain, maybe another 6-12db somewhere would be good. If it was me, I would throw it together on a board and see how it works.

If the 12e1's are anything like the ones I have, I would have a big bag of grid stoppers to hand, I found they like to behave as oscillators in the Mhz range. If when you turn it on, you hear a odd hissing noise, and nearby DVM's start going mad, keep your hands away as there could be 20W or so of RF being generated.

Hi Nick
OOPS Forgot the input coupling cap after the volume pot which would be needed if a cathode follower was going to be the very first thing the signal would meet on entering the amp - stupid boy!

If it would be better to lose it then that's good. I was dubious about the bloody thing in the first place.

Using the loadlines I had worked out for the 12E1 and the ML4, gain looked marginal no matter how much I fiddled about. Thanks for confirming my fears Nick. Despite having done a few amps now I'm still a bit afraid to go ahead with anything new without asking a few questions first.

Oscillating cathode followers coupled with an oscillating output valve yikes! I'll definitely heed your advice, ditch the followers and get the grid stoppers in there.

A first gain stage will be needed too methinks. I'll go back and have another look

What about a 76 as a first stage? I have the UX5 sockets. Just need to get hold of the valves

Steve

[NickG]

I like 76's, especially when choke loaded, the little hammond 150H chokes are good for that.

[The Shadow]

Thanks Nick,

I obviously have not tried choke loading yet seeing as how I have not been in the valve audio game for that long.

I'll have a read up on it.

Picking a random resistor load say 50K on a valve how does a choke behave as a plate load in comparison to the resistor for setting the operating point?

For instance the Hammond 156C 150H choke has a DCR of 3K7 but the reactance of the thing must also have some influence over how the choke loads the plate.

I don't really know what I'm asking so if you could give me a basic idea of what is involved and how to set the op point for the valve I'd be extremely grateful. I'd then know where to look for more gen on the idea.

Ian mentioned that grid chokes could be a good thing too in a PM ages ago but I would need considerably more inductance for those.

Steve.

[The Shadow]

OK to recap

Using a choke loaded C3g as a driver, the amp gave a good account of itself at Steve's last Saturday and drove Ed's Vofos, Nick's open baffles and Steve's Lowthers nice and smooth.
Bass was tight fast and clean but lacked ultimate extension. Normally I use a sub driven from the preamp so I had not noticed this at home but on big full- range speakers the deficiencies at the bottom end were all too evident.
Also there was a slight hardness to vocals through Ed's Vofos on some recordings especially The Beatles "Let it Be (Naked)" album that had Paul and I wincing.

The cause of this turned out to be too much current drive from the triode strapped C3gs to the power valves. Dropping the quiescent current from 17mA to 8mA cured the problem. However the new o/p dropped the loadline so low we were nearly scraping the bottom of the graph, right where the grid lines were curving. A waste of the C3g's capabilities IMO.

Choke loaded 76s are what I am currently wiring in after experiments on a breadboard proved them to be far more appropriate in this application than the hard driving C3g.

Experimenting with a 10K:10K isolator/balancer transformer on the input proved inconclusive. The transformer coupled input smoothed out the already smooth sound a bit too much for my liking but the soundstage was enormous. Swings and roundabouts there then.

I had a thought that the primary of the IPTX might have been forming a low pass filter with the 470nF output capacitance of the preceding preamp.But since I did not know the inductance of the IPTX and had no means of measuring it there was no point messing about. A change of output cap in the pre might have restored the sparkle and still given the huge soundstage but I was now allowing myself to get sidetracked into complication as is my wont if not restrained, so I shelved the IPTX idea for another day.

I finally settled on the circuit below and am currently building it into a case. The power amp case will then be used for the preamp rebuild as the shiny aluminium top surface is starting to look decidedly grotty and the pre has had so much tweaking about that it looks like a rats nest, something I can't stand.
The output caps will be changed to 100nF Auricaps to suit the traditional 1M input on the power amp. Might have to watch out for hum with that input impedance though, so we'll see what happens.

I have two weeks off as the schools have broken up for Easter so I have a bit of time to get things sorted properly.

Steve

[Paul Barker]

That looks good.

Really the preamp would have to be like the output stage of the wad preamp the transformers belong to to make best use of them. I don't remember, maybe it was an ecc88 common cathode stage. In your case as the bass is not too important you could use it with a nicer sounding 5687 (the "Audionote Valve"). That mightn't improve the preamp, depends what it consists of. There was a fairly long old thread on adapting that pre to CF output, those who did so prefered it, which implies that the transformers add distortion, as we would expect with transformers. They benefit us but bring distortion. Using the same one with the same failing in two stages of an amp compounds the problem. Best to blend methods throughout the amp and not repeat.

The 1meg grid leak resistor on the 76 is quite a high value, although 9 pin valves are OK with this I don't recall the 76 details, but reducing it to 500k wouldn't harm anything else, but would suit the valve better.

[petercom]

Originally Posted by Paul Barker

There was a fairly long old thread on adapting that pre to CF output, those who did so prefered it, which implies that the transformers add distortion, as we would expect with transformers. They benefit us but bring distortion.

That wasn't the reason, Paul, just to correct any misunderstanding.

The WAD Pre used high ratio output transformers which were prone to picking up hum from any field in the vicinity, including the psu power transformer.

Thus the benefits from rewiring to CF output were to lose the hum. Those, like Dr. John, who arranged the preamp so that hum was not picked up actually enjoyed better results from the transformer output.

IME the 'distortion' produced by correctly designed transformers is minimal and benign compared to that produced by amplifier stages, especially CF outputs!

[The Shadow]

Originally Posted by Paul Barker

The 1meg grid leak resistor on the 76 is quite a high value, although 9 pin valves are OK with this I don't recall the 76 details, but reducing it to 500k wouldn't harm anything else, but would suit the valve better.

Cheers Paul
I'll reduce the impedance at the input as you suggest.

Steve

[slowmotion]

Nice simple amp schematic.
I usually end up with something like that, usually sounds very good...