Page 48 of 54

#706 Re: Remember These?

Posted: Sat Oct 05, 2019 11:57 am
by Ray P
I've ordered a couple so I'll let you know if they are as nice in the flesh as in the picture.

#707 Re: Remember These?

Posted: Sun Oct 06, 2019 1:29 pm
by Cressy Snr
Image

I have a question: Ignoring the power supply, which is now not the same and looking at the feedback path, would there be any benefit to be gained from taking it from the 13E1 cathode rather than the speaker side of the output capacitor?

Here I go with the theory as interpreted by an English graduate and no doubt I’m oversimplifying things, ‘cause I know squat about AC theory, but humour me, don’t point and laugh. I’m still trying to get my head around feedback

At the moment there is a high pass filter formed by the 6800uF output capacitor and the 7.2R DC resistance of the speaker; this filter cuts off below 3.3Hz. Taking feedback from the output of this filter and shoving it back up the input is surely causing the amp to try and correct for its effect all the time unless I’m missing something somewhere.

Would taking the feedback from the output valve cathode and blocking the DC with a cap to avoid upsetting the input stage bias work any better. A 100uF blocker after the feedback resistor would form 3.4Hz cutoff with the 470R cathode resistor? and because this filter is now at the input stage instead of the output then the ‘trying to correct’ problem is gone.

Or is it just replacing one problem with another, ie another distortion mechanism in the feedback path (the blocking cap).

Or am I just talking crap?

#708 Re: Remember These?

Posted: Sun Oct 06, 2019 1:35 pm
by IslandPink
Cressy Snr wrote: Sun Oct 06, 2019 1:29 pm
Would taking the feedback from the output valve cathode and blocking the DC with a cap to avoid upsetting the input stage bias work any better. A 100uF blocker after the feedback resistor would form 3.4Hz cutoff with the 470R cathode resistor? and because this filter is now at the input stage instead of the output then the ‘trying to correct’ problem is gone.

Or is it just replacing one problem with another, ie another distortion mechanism in the feedback path (the blocking cap).

Or am I just talking crap?
Doubtful - you're not known for it !
But you're going to have to have a cap somewhere to block DC, unless you run the feedback through a transformer. So there'll be an LF time constant in there. ps. the impedance going into the input valve is likely to be about half what you suggest, since the cathode impedance is in parallel with the 470R resistor.

#709 Re: Remember These?

Posted: Sun Oct 06, 2019 1:52 pm
by Nick
The thing is, you don't care what happens at the cathode of the output valve, what you care about is the voltage across the loudspeaker, so I would leave the feedback as it is.

#710 Re: Remember These?

Posted: Sun Oct 06, 2019 2:01 pm
by Cressy Snr
Thanks chaps.
I’ll leave well alone.

#711 Re: Remember These?

Posted: Sun Oct 06, 2019 5:51 pm
by IslandPink
Nick wrote: Sun Oct 06, 2019 1:52 pm The thing is, you don't care what happens at the cathode of the output valve, what you care about is the voltage across the loudspeaker, so I would leave the feedback as it is.
It took me a little while to fully understand this . Been out on a walk on the hills. The distortion of the output cap is best included in the feedback signal, so it can be reduced by whatever factor the feedback is running at ?

#712 Re: Remember These?

Posted: Sun Oct 06, 2019 6:06 pm
by Nick
Yep, the feedback "sees" the entire amp between the summing point and the point the feedback is taken off as a black box providing gain with distortion. It doesnt care what is doing what, it will act to reduce distortion by burning gain. It also sees the effect the load has on the output of the black box, so the load can be seen to be inside the box. Probably the best point to take the feedback from is the +ve speaker terminal.

#713 Re: Remember These?

Posted: Sun Oct 13, 2019 2:41 pm
by Cressy Snr
Been having an interesting (if geeky) time playing with different tube rectifiers in the driver stage power supply. I started with a solid state bridge followed by a slow start mechanism using a 6U4G TV damper diode. It came after I’d originally set up a hybrid bridge with a 5V tube rectifier. It was one of those things I’d have kicked myself for, if I’d not tried it. I needn’t have worried: it gave the amp a hard, fatiguing, glassy quality and was quickly removed, the hybrid bridge going back in toot sweet.

Seems that the driver stage power supply has a bigger effect on the ultimate sound quality achievable from this amp than the input stage. In the input stage supply; different EZ80 or 81 rectifiers make very little difference, but it’s quite a different story with the driver stage.

I’m restricted to 2A rectifiers for the drivers because that’s the capacity of the 5V rectifier heater winding on the Danbury heater transformer fitted in the chassis.

Mullard GZ34 is too warm and cuddly; top end having a curtailed feeling about it. Brimar 5Z4GY is more open up top and better balanced overall. RCA 5Y3G detailed but a bit flat front to back. Surprisingly, the best sounding rectifier is my 1960s Fivre 5R4GY direct heated jobbie. The 5R4GY is ridiculous overkill, when the driver stage power supply is only drawing a total of 16mA, but it’s lovely. The top end is simply gorgeous, with shimmering cymbals, super inner detail and completely free of grain. Depth is about as good as it gets given the proximity of the speakers to the rear wall. Lovely pianos and double basses...Niiiice.

#714 Re: Remember These?

Posted: Mon Oct 14, 2019 9:59 pm
by Cressy Snr
Now I’m being silly (as usual) but I’m a sucker for, proper old classic Mazda UU5, 4V rectifiers. They are just beautiful looking things. The bulb shape and the anodes are pure Art Deco; the valve being an industrial work of art in my view.
843CE729-A7A5-4B79-AA3C-43AD9AB5C8D6.jpeg
Well I couldn’t resist and this evening, and now bask in the glow of one of these beautiful old rectifiers, in the driver stage supply. They really are lovely old things.

#715 Re: Remember These?

Posted: Mon Oct 14, 2019 11:24 pm
by Mike H
It certainly is fab. 8)

I'm surprised different rectifier tubes are having that much effect on the sound TBH.

#716 Re: Remember These?

Posted: Tue Oct 15, 2019 7:19 am
by Ray P
Cressy Snr wrote: Mon Oct 14, 2019 9:59 pm They are just beautiful looking things. The bulb shape and the anodes are pure Art Deco; the valve being an industrial work of art in my view.
They certainly look the dogs bits...

#717 Re: Remember These?

Posted: Tue Oct 15, 2019 10:52 am
by steve s
Nice Steve.. as usual I have a box full of old 4 volt recs, but I've not ventured further than gz37s& gz32s as i always liked the soft start.

#718 Re: Remember These?

Posted: Tue Oct 15, 2019 11:45 am
by Cressy Snr
steve s wrote: Tue Oct 15, 2019 10:52 am Nice Steve.. as usual I have a box full of old 4 volt recs, but I've not ventured further than gz37s& gz32s as i always liked the soft start.
The UU4/5 types I have, are indirect heated and have a slow start. I do have a few GEC U19 types with the top caps, but as I’m not building any high voltage amps, they are somewhat redundant.

#719 Re: Remember These?

Posted: Tue Oct 15, 2019 12:24 pm
by Cressy Snr
Mike H wrote: Mon Oct 14, 2019 11:24 pm It certainly is fab. 8)

I'm surprised different rectifier tubes are having that much effect on the sound TBH.
I compiled a geeky list of HT voltages produced in my amp by the 2 Amp rectifiers I have:

GZ34 - 160V
5Z4G - 156V
UU4 - 152V
5R4GY - 148V
5Y3G - 141V

The direct heated rectifiers produce the largest voltage drops.

The subjective differences are likely to be down to the different HT voltages, moving the operating points of the driver valves around, plus the differing power supply impedances produced by each of the rectifiers. The differences are quite audible and would be a matter of personal taste as to what presentation is preferable.

It’s nice that the subjective impressions can be tweaked to one’s personal preferences and as Nick said in another thread, these audible differences are about power supplies rather than fancy resistors or expensive caps. I just use Sprague Atom guitar amp axial lead caps and JJ or F&T cans in the driver and input stage power supplies, and high value Nichicon caps in the output stage; nothing fancy.

#720 Re: Remember These?

Posted: Tue Oct 15, 2019 12:37 pm
by Nick
I would be interesting to use a variac to tweek the mains to correct for the HT difference and see if there was still any audible change.