OK…so I’m on wi’ summat

[Mike H]

Too late now but I was going to suggest RS do some nice can electrolytics that might have filled those holes, smallest value 150uF tho (500V).

[Cressy Snr]

I’m glad I got it going again with the JJ and the Sprague caps. It’s a great little amp and the ultimate quality of the sound doesn’t seem to depend on whether or not electrolytic caps or foil in oil caps are used. Yes, of course it is slightly better in the treble, with foil in oil caps but the difference is not a deal breaker.

The cap holes are covered up with black acrylic and use the same fixing holes as the brass strips, so a set of new motor-runs could be easily installed at a later date by simply removing the cover. However, if I do that, push fit connectors will be used; there will be no soldering irons allowed anywhere near the caps.

[simon]

It's slightly concerning Steve I must say, I use them in the GM70 amp

[Cressy Snr]

I don’t think they pose any danger Simon. My hairy-arsed soldering is probably to blame for the problems with mine.

[Cressy Snr]

A little more experimentation with the power supply today.
I've seen series VR tubes used to feed the input stages of commercial amps such as those from Decware and Icon Audio, so I thought I'd give one a try in my own amp:

EL34SEULSimplifiedPSU.png (89.37 KiB) Viewed 6009 times

22K shunt resistor makes sure around 15mA is flowing to ground through the tube. The sound is bloody marvellous.

IMG_0407.jpeg (106.16 KiB) Viewed 6009 times

[andrew Ivimey]

Weird and wonderful...putting the glorious glow tube with the 22k resistor to ground... I'd have never thought of that and bingo; you get a stabilised h.t. from it. Great!

I like the glow tubes and making them glow - who cares what it sounds like - can't be that bad - warms the cockles.

Onwards!

[Cressy Snr]

The idea came from the fact that the 6SL7s are drawing around 1mA each and I wanted to use a series VR105 tube to knock around 100V off the HT. I wanted that tube simply because unlike the 7pin VR tubes, it looks great when lit and I only have the one.

I had to think of some way to make the tube pass more than 2mA or it would have been a waste of time, so I ended up with the shunt resistor to ground. That shunts about 13mA away and an additional 2mA (approx) gets sent forward to the pair of triodes in the 6SL7, causing the VR tube pass a total of 15mA.

It works a treat and sounds great.

[andrew Ivimey]

Yeah I get it just hadn't seen the possibility.

[Nick]

Though as Steve knows the voltage is not regulated. But the difference between the two b+ will be. That I could argue is just as important.

[Cressy Snr]

Aye,
keeping the space between both HTs constant seems to work beautifully. Guess it’s one of those happy accidents.

[Cressy Snr]

Right then...OK...Er...Here we go again

I've reverted the power supply to choke input, but this time, before doing anything, I made extensive use of PSUD, feeding in the parameters of the chokes I have, looking at all the waveforms, both current and voltage for any sign of trouble. I ended up with a decent power supply, at least it looked good on the software.

So here we are and it works an absolute treat!
In terms of both mechanical and electrical noise, this is the quietest LC supply I've built up to now; ridiculously quiet. Bass is the best I've heard from this amp. It's a noticeable difference. I've never had much success with LC supplies, so this is a turn up for the books.

EL34SEULChokeInput_vers2.0.png (91.99 KiB) Viewed 5707 times

[Cressy Snr]

Potted 4H choke can just about be seen to the rear of the VR tube:

IMG_0409.jpeg (135.25 KiB) Viewed 5682 times

[Paul Barker]

Well you have to find the biggest fook off choke you can find for choke input.

This one is a cube showing the EL34 to show how enormous it is. Because this power transformer is 500mA dc capable. So every other part has to be that capable. Youre case isnt so extreme. But success with chokes isnt an accident, its maths.

[Cressy Snr]

Maths is not my strong point, so I turn to PSUD. I also found a nice book http://www.tubebooks.org/Books/Schure_Rectifiers.pdf which helped me out no end.

The bit on page 46 concerning fitting a bleeder resistor that draws a constant 10% of the full rated DC current output caused things to drop into place sort of. “Improves regulation and ensures the supply does not require excessive inductance to operate correctly.”
It just so happens that the shunt resistance following the VR tube does just about that figure although I didn’t plan it like that, and of course there is a slight delay with it coming into circuit until the tube ignites. But everything comes up nice and steady if an indirect heated rectifier is used.

[Cressy Snr]

I got out the 1950s, grey glass KT66 again, deciding to go fully with old valves. The 6SL7 input/driver is a wartime Sylvania VT 229 and the rectifier is a 1952, ‘fat bottle’ CV378 (GZ37.) They’re all at least 70 years old.

Used in isolation, I’ve not really been able to get the sound I like with these valves, but as a combination, they work well and look fab.

IMG_0417.jpeg (154.06 KiB) Viewed 5314 times