Future plan for 300B XLS amp

What people are working on at the moment
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8867
Joined: Mon May 21, 2007 9:42 pm

#151

Post by Paul Barker »

Great to hear a trusted sound report on the 4p1l.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#152

Post by IslandPink »

You're very kind, Paul ; but will be easier to say more once I get the RF out and later, when I can swap from voltage-reg fil. supplies to DHT supplies - now I'm putting the pressure on Andrew :)

Also I need to get a sketch of the amp done so people can see what it is, but really it's a very very simple circuit . Pentode fed direct from the shunt supply for instance .
Gain is enough for current speaker set-up, btw , but would need prob. 3-4 V input for full power, as I haven't used a by-pass cap on the 4P1L cathode resistor. That is 180R ( RS carbon film 2W ) at present, but may go to 200R ( Kiwame ) soon . The current from the 4P1L plus the OA2 current goes through this resistor - about 45mA .
"Once you find out ... the Circumstances ; then you can go out"
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#153

Post by IslandPink »

Ok, now we're cookin' with Gaz ( thanks again Steve )
Seems stable ... holding my breath .
1.2k grid-stoppers on G1 , ferrets added on the anode pins .

Some goodness noted , from playing a couple of sides :
Treble detail & treble tone better than C3m
More low-level detail ( between the notes )
Low bass seems a touch better .

Upper bass tone and midrange tone is good but maybe a touch short of ideal ( & C3m ) at present .
I have high hopes this will fill out with high-impedance filament supplies.
At present the caps on the end of the V-reg supplies and all the rest of the guff back there is in the signal path .

No evidence of much trouble on microphonics.
The pinging and twanging on warm-up seems to have settled down .
"Once you find out ... the Circumstances ; then you can go out"
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8867
Joined: Mon May 21, 2007 9:42 pm

#154

Post by Paul Barker »

What's the load resistor for the pentode? Current through it? Anode voltage? Screen voltage? Have you any way to measure actual gain?

I go to work now. Looking forward to findings tonight.

I been struggling to decide on next va stage progression. May try this.my only reservation is better this for se output stage,triode for push pull.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#155

Post by IslandPink »

It got better last night as I listened for another hour.
This is definitely a winner . The best I've ever heard the track 'Maria Elena' from Ry Cooder's 'Boomer's Story' . The delicacy in the highs and the microdynamics are excellent .

Paul, data as follows :
B+ : 465V
4P1L runs at about 31mA through anode and 9ma through screen .
Anode load 7.8k ( made up of 11k and 27k Mills )
Feed to screen and VR tube 12.3k ( made from 18k and 39k Mills )
Screen set at 150 by OA2 to cathode.
Cathode resistor 180R , but will trim slightly later to 200R .
Gain will be about 20-22 I think .
It would be 32 with fully by-passed cathode resistor, but the extra VR tube current through the cathode R keeps its value lower than normal .
"Once you find out ... the Circumstances ; then you can go out"
User avatar
Mike H
Amstrad Tower of Power
Posts: 20157
Joined: Sat Oct 04, 2008 5:38 pm
Location: The Fens
Contact:

#156

Post by Mike H »

SteveTheShadow wrote:Atomic batteries to power
Turbines to speed.
ed wrote:gamma rays on
yobbo rays on
boost the meson power slim!
OK I will:
"She canna take much more cap'n"

:lol:
 
"No matter how fast light travels it finds that the darkness has always got there first, and is waiting for it."
JamesD
Old Hand
Posts: 997
Joined: Tue Oct 09, 2007 4:26 pm
Location: North Yorkshire

#157

Post by JamesD »

Interesting data on the 4P1L. I always liked the sound of it but this seems to be a cut above most other options :) Still a bit of a sleeper!

J
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8867
Joined: Mon May 21, 2007 9:42 pm

#158

Post by Paul Barker »

IslandPink wrote: Paul, data as follows :
B+ : 465V
4P1L runs at about 31mA through anode and 9ma through screen .
Anode load 7.8k ( made up of 11k and 27k Mills )
Feed to screen and VR tube 12.3k ( made from 18k and 39k Mills )
Screen set at 150 by OA2 to cathode.
Cathode resistor 180R , but will trim slightly later to 200R .
Gain will be about 20-22 I think .
It would be 32 with fully by-passed cathode resistor, but the extra VR tube current through the cathode R keeps its value lower than normal .
I can't find enough valve data for it, but what I have gleaned is that the transconductance is 6, and I am making an educated guess at your operating point of plate resistance 20k. I have assumed a grid resistor for the next stage of 470kohm. Without a cathode bypass cap your actual gain would be 15 and with bypass 33. If the plate resistance is 15k (quite possible) the figures change to 14 and 30.

The grid resistor doesn't make much difference. 50k changes the figures to 13.75 and 30.

To get a gain (A) of 20 without bypass cap you would need a load resistor of 12k, but I haven't drawn it on the curves to see if it cuts through the bend, most likely it does.

But in any case an A of 15 with unbypassed cathode R is very useful, and very hard to achieve from a triode.
Last edited by Paul Barker on Sat Jun 01, 2013 10:37 pm, edited 1 time in total.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#159

Post by IslandPink »

Yeh, but as I tried to explain, the extra current from the VR tube setting the G2 to cathode voltage ( eg. 15ma ) also goes through the cathode resistor, so the cathode resistor instead of being about 300R is only 180R . Hence the amount of local neg. feedback is reduced. I would sim it on Spice if I could work out a way of modelling the VR tube for this case. Anyway, I think mu is at least 20 .
"Once you find out ... the Circumstances ; then you can go out"
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8867
Joined: Mon May 21, 2007 9:42 pm

#160

Post by Paul Barker »

Yes but I used your figure of 200 ohms and the current through it makes no difference to the formula for pentode gain. Only the resistance counts, which is why I arrived at 15, for your future spec'd resistance. With your present resistance of 180 yes the gain is 21.
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#161

Post by IslandPink »

Ok, sorry .
It does seem surprising the difference between 180 and 200R means going from 21x to 15x :?
I think I can sim it, but could also try & measure next time I do a mod to it, using sig gen and voltmeter .

The treble detail & space on this is just exceptional - just listening to the drums & brushes from Joe Morello on 'Three's to get ready' .
The lower mids and upper bass are lacking a bit of weight . I'm hoping the other fil. supplies might fill that out a bit .
"Once you find out ... the Circumstances ; then you can go out"
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8867
Joined: Mon May 21, 2007 9:42 pm

#162

Post by Paul Barker »

Yes I too was shocked. I changed every other parameter and it made little difference, but changing cathode resistor makes huge difference.

So you must be pulling 16mA through the VR, in my experience a 30mA VR is perfectly happy pulling 30mA this would be a steady state current it is rated at 30mA and doesn't complain.

So your bias now is it 8.5v? and you want a little more?

Anyway on what is known 46mA across resistor gives 8.5v with additional 14mA 8.5v achieved by 141 ohm Gain goes up to 24. A huge single stage gain from an ubypassed resistor.

Only problem now is, what's the point? What I mean is, do we benefit from the de-generative feedback with this VR tube int he mix or do we half benefit? or does the VR issue mess it up alltogether? I am not asking because I know. I don't know, and that is why I am asking.

In my amp I also had a sharing effort with the cathode resistor originally, but when I got nasty RF noise I eliminated everything unusual.

I had forgotten that until now. It could probably go back now that I have discovered the RF was definately in the proximity of the SMPS to the signal section. But there may be insufficient time to change anything, and I have enough gain anyway.

If however your idea lessens the degenerative effect then is it just the same as going for half cathode resistor bias and half fixed bias? Or is there another benefit to the VR going in at the cathode? Or for that matter a downside?
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
User avatar
IslandPink
Amstrad Tower of Power
Posts: 10041
Joined: Tue May 29, 2007 7:01 pm
Location: Denbigh, N.Wales

#163

Post by IslandPink »

OK, there's a few points I can answer.
First of all I'm perfectly happy with 8V or 8.5V bias. It would be good to measure it, but I can't safely get a probe under there while it's running !

I could run the OA2 harder, but I think the white noise is at a minimum somewhere around half its max current, so I've always used around 15mA with this set-up .

As far as I'm aware, connecting the VR to the cathode is the correct way to do it, and it was recommended to me by James. After all, the valve data sheets always assume cathode at zero as reference for everything. This is the way I've done it since moving to the VR tube option.
Previously , I used the voltage divider option ( from B+ to ground) suggested by Thorsten in the 'Legacy' circuit to set the G2, as it was supposed to be more stable with AC than a plain feed resistor from B+ and cap to ground .
The VR tube connected to cathode was more transparent and MUCH more solid in the bass than the voltage divider.
So, the additional bonus you get with the VR tube is more current through the cathode resistor and the reduced resistor value, as we've discussed. This may ( or may not ) make the difference in gain that allows you to avoid a cathode by-pass cap . In my case at the moment, there is just enough gain from a line-level signal from the phono to go 'fairly loud' on 92db speakers .

The RF problems do not appear to be due to the OA2, they went way ( as far as I know ) when I increase the G1 grid-stopper and added a couple more ferrites onto the anodes.
"Once you find out ... the Circumstances ; then you can go out"
User avatar
Nick
Site Admin
Posts: 15707
Joined: Sun May 06, 2007 10:20 am
Location: West Yorkshire

#164

Post by Nick »

Yes I too was shocked. I changed every other parameter and it made little difference, but changing cathode resistor makes huge difference.
I must be getting lost here, how is that being calculated?

I get

Rk = 200, Av = 29.7
Rk= 180, Av = 30.1

With Rp = 20k, Gm=6ma/v, RP=15k, RL=470k
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
User avatar
Paul Barker
Social Sevices have been notified
Posts: 8867
Joined: Mon May 21, 2007 9:42 pm

#165

Post by Paul Barker »

Nick wrote:
Yes I too was shocked. I changed every other parameter and it made little difference, but changing cathode resistor makes huge difference.
I must be getting lost here, how is that being calculated?

I get

Rk = 200, Av = 29.7
Rk= 180, Av = 30.1

With Rp = 20k, Gm=6ma/v, RP=15k, RL=470k
Effective cathode resistance = 1/(s [expressed as 6 requiring conversion to umhos]/1000) + cathode resistance = 366 ohm

Effective plate resistance = 15k//470k//20k//

Gain equals (1/(1/15000+1/470000+1/20000))/366 =22.95 or 24.

Is it your way of calculating effective cathode resistance which is different?

But the figures I calculated above were using Mark's stated anode load resostor of 7.8k. and anode resistance of 15k (which is a guess by me) so the 200 ohm result is 13 and the 180 ohm result 14. without going back which I can't do while I am editing this I realise there was an error in my figures previously. So I must have punched a wrong number in during the calc back last night.

Is it now lining up with your thinking?
"Two things are infinite, the universe and human stupidity, and I am not yet completely sure about the universe." – Albert Einstein
Post Reply