New SE With Triode Mode KT120/150

[Cressy Snr]

Just a little housekeeping.
I thought, for the sake of completeness I'd put up the amp schematic, with the 807s,
and measured voltages, now that the two toroids are in place.

It's a nice little amp. About 6Wpc.
Makes me smile every time I put it on and listen, so it's doing something right

[Cressy Snr]

Have made it self-contained again, so it’s back to being an integrated design. It’s one cable less in the signal path.

Rather than return the controls to the front, where they are a long way from the rear panel, where the sockets are, I let them into the top plate, to the side of the left hand output valve. IMO it makes for a better appearance than having the controls sticking out of the front panel - well it will do when I’ve done something about those toroids.
The rear control is the selector switch whilst the front one is the volume pot.
Whilst you can’t see on this pic, I’ve reduced the anode voltages to the MH4s: down from 200 to 183V using a 0B2 and a 75C1 as the series, glow tube stack instead of the previous pair of 0B2. I’ve decided to treat these 85 year old globe meshplates a bit more gently than having them at maximum plate voltage all the time.

[Ray P]

all well Steve?

[Cressy Snr]

all well Steve?

Yes.

[Ray P]

[steve s]

I’ve decided to treat these 85 year old globe meshplates a bit more gently than having them at maximum plate voltage all the time.

That's wise steve, but I'm guilty of thrashing them unfortunately.
They are playing only 2/3 nights a week driving the px4s.
I do need to swap the mofo back in, but I cant be arsed !!

[Cressy Snr]

Forgive me for blathering on, and I know this is boring, but the so called single-ended triode mode, KT whatever idea I started with, has evolved into something, that is nothing at all like what I intended building. And contrary to the past habit of spiralling into chaos and mediocre sound, ending with the dismantling of the project, this particular item is currently surpassing the sound quality of its original triode mode pentode iteration, by a pretty decent margin. Realistically, it's not a 'magic' sounding SET but it's pretty darned good, nonetheless.

So this is the state of play at the moment: I've been having a bit of an intermittent, loud, left channel crackling problem which has been going on for a few days now. After much cursing and head scratching, I traced the problem to one of the cheap Hammond 156C 'plate' chokes, which seems to have a break in the wire somewhere. Whatever the reason, the choke is fit only for the bin.

I had been looking for a dry joint, reflowing and re-terminating everything on the offending channel, but with no success, until in desperation, I took the choke out of the offending plate circuit and replaced it with a temporary load resistor. It was the last option I had because I'd eliminated everything else including the MH4s, so however improbable, it might have appeared, the choke had to be the culprit, and so it turned out to be. So now, all I had to do was find another one in my box of transformers, as I know for a fact, I have another two of the little blighters, but could I find either of them for love nor money? Could I heckers like......FFS! I spent half a day of my life, half a day I'll never get back, grumbling and furtling about, in every box, in every place, before eventually giving up in disgust, at the utterly slack organisation of the storage and labelling of parts. God only knows where I've put the b'stard things!

More later.

[izzy wizzy]

I just got some Hammond mains transformers and chokes for the fixed bias in the GM70 amps. One didn't go and then was intermittent when I flexed it. Bumbling about located the issue was a not so great connection on one of the primary tabs so I reflowed it and bingo, it was OK. The wire is very fine and hard to see. Could yours have similar or is flying lead?

I now have to hope that it stays that way coz if it fails, no bias on the GM70s could get exciting

[Cressy Snr]

Hi Stephen, it’s flying leads on the choke, and both leads feel firm. Tapping the thing with a toffee hammer whilst measuring the resistance, it goes infinite when each tap lands, then reconnects itself, so I would guess we have a heat expansion issue, making the problem show up when the amp is warm.

I’ve reconfigured the front end with different valves at the moment, but the results I’ve got show that all is certainly not lost, which TBH caused a definite sigh of relief.

[Cressy Snr]

The circuit has been changed to the config below, which was done over the weekend.

We are back with the parallel 6SL7 input and driver stage for the time being. It’s a bit ‘gainy’ for the amp, but some excess gain has been burned off by the lack of a cathode bypass cap, plus the steepened load-line obtained by halving the anode resistor value. The lack of bypass cap linearises the input stage as is usual.
The use of the parallel 6SL7 is not as bad as it was, because it now driving an ultralinear output stage, so Miller capacitance is quite a bit lower.

Power supplies:

The diode between the end of the series resistor and the two 0A2 VR tubes in the input stage power supply, is there because Dick Olsher and Kara Chaffee say it’s a good idea to put one there, so don’t pick on me for it’s presence

Inside is not as neat these days, but I suppose I should be thankful it doesn’t look even worse, given the ragging about it’s endured:

[Cressy Snr]

As it is now:

Sounds quite respectable considering. I mean, the 807 seems not to have much going for it. Most of the stuff I’ve read online, states that it makes a crap triode, it makes a lot of high, odd-order distortion products, and is a dead loss in ultralinear mode because of the poor screen grid voltage limit of 300V. As a result it’s pretty cheap to buy.
There are a few hardy souls who say they’ve made a success of using it and I’ve joined their ranks.

For me, in order to make 807s work, you need ultralinear operation and some other forms of local feedback too.
I have three feedback loops - ultralinear, and cathode feedback around the output valve and degenerative feedback on the input stage. It works.

[Ray P]

There must be something in the air Steve; my Transcendent Masterpiece clone DAC/HPA has acquired a buzz that I have to track down - it disappears when the thing has been powered up for a while so appears to be temperature related. Ho hum.

[Cressy Snr]

Yes these buzzy incidents are very frustrating and time consuming to try to deal with. When I had Linn, Naim, Musical Fidelity, A&R Cambridge, Marantz, Pioneer, and Sansui amplification, I never once had that kind of problem with any of them. They just worked, day in, day out and delivered the signal to the speakers.

This steam engine hi-fi can be an absolute pain in the arse, but it keeps me out of mischief and the missus always knows where I am.

[Cressy Snr]

Pissed off with the rat's nest the inside of this amp had become, yesterday I decided to do a bit of tidying up of the interior:

It's now as tidy as I can make it.

But I decided then to change the front-end again. The 6SL7 parallel input stage wasn't bad but I did feel there was room for a bit of improvement on that score, so a complete rebuild of the front end was carried out. The two large, black cathode bypass caps on the driver stage have been removed since the picture was taken. Now the gain is about right. It was too much before. I suspected it would be, but it's easier to cut bits out than put them in after the fact.

The amp is now a three stager, with a separate input and driver stage, courtesy of a DC coupled 6SN7 cascade.
It is very similar to the one on the Japanese, Sun Audio SV-2A3 amplifier. I think it’s bloody superb. Others may
Circuit is below:

470uF cap decoupling the input from the driver should be 47uF, but I really cannot be arsed to change it.
Who gives a fecking shite anyway.

The power supplies needed a little bit of alteration to suit:

6SL7s should be 6SN7 on the heater diagram, but again I can’t be arsed to change them.
That's everything up to date. Quite a bit of musical and sonic progress
with both the speakers and the amp this week, so I'm a happy bunny.

[IslandPink]

Are you sure that decoupling cap on the driver stage needs to be as big as 470uF ? I reckon you could get away with eg. 47uF !