4P1L DHT Line Stage

[andyjevans]

Thanks for that, Charm! That PSE 4P1L amp is almost exactly the same as mine, except I use filament bias.

I was thinking about the 4P1L in a preamplifier. The 4P1L has been used for many, many years in Russia in amplifiers. I can read some Russian and I did look on the Russian forums at one time. I first came across the 4P1L on Klaus's site, and Anatolyi has been using it for many years and probably read schematics of it as a boy or at least teenager.

Klaus and Klausmobile: Do-it-yourself Audio Electronics Resources

http://www.klausmobile.narod.ru/projects/pr_08_mullard_e.htm

http://www.klausmobile.narod.ru/testerfiles/4p1l.htm

 

[Magz]

Ale I believe uses a Fostex unit which is sensitive enough for a 2-stage solution and has recently been experimenting with the 814, mu of 8.3. Other options are 813, mu of 8, or 211 with a very useful mu of 12. These options would give you the bass and power that would take the amp up a level. Once you design around a higher gain output tube (PX25 would be nice if it were not rare and $$$, 4P1L is common and cheap) you start to get some more interesting options for the preceding stage.

If you want a DHT with lots of gain, check out the 833, mu of 35! If you run it with enough voltage (2kV) you won't go into A2 until about 40W output. Even if you use an IDHT to drive it, more than half the gain will be coming from the DHT.

Not for everyone, for sure, but just thought it should be mentioned...

 

[andyjevans]

If you want a DHT with lots of gain, check out the 833, mu of 35! If you run it with enough voltage (2kV) you won't go into A2 until about 40W output. Even if you use an IDHT to drive it, more than half the gain will be coming from the DHT.

Not for everyone, for sure, but just thought it should be mentioned...

2kV not for everyone.........?

If we get on to high gain tubes there must be quite a few out there. There's the EML 20b for a start - very nice and very $$$, mu 20! And also the 812A, mu 29 which is quite affordable. Kind of "super spud" amps with a 4P1L line stage.

 

[enzoastro]

Or even sv811-10, i have a small stock of these and considered them to drive my gm70 amp but they have a higher plate r than a 4p1l! And 4a filliment!

 

[euro21]

There's the EML 20b for a start - very nice and very $$$, mu 20! And also the 812A, mu 29 which is quite affordable.

... and 841.. mu: 30.

 

[mogliaa]

I think I missed an interesting conversation here and I'm glad it didn't go south. Also glad to see that at least my experiments stimulated the exchange of ideas and experiences.

In my view, the listening experience is as valuable as the technical aspects so I decided to do some initial tests, which is the way to do proper comparisons and analysis before listening to the thing itself.

I know that the grid current kicking in early is the challenge I have ahead with a potential 4K output resistance of the 4P1L in 1:2 step up. Other than this, the measurement do look good and is worth considering my experiments I believe.

A good amp with one stage is possible. I built my 814/4-65a with outstanding results (both in A1 and A2) which sounds and measures really well. That is a fact. A 46 driver can do 200Vpp in filament bias with only 0.08% distortion when driving a demanding transmitting valve. I have to say that I haven't seen anyone showing similar metrics around. The input step up sounds really good when driven with a proper preamp.
I invite you to come around and judge it yourself with your ears given that I already posted tons of response measurements in my blog.

@45: Would be good if you can share some measurements when challenging some of the options discussed. It would also help us understand better your points

thanks
Ale

 

[45]

I built my 814/4-65a with outstanding results (both in A1 and A2) which sounds and measures really well. That is a fact. A 46 driver can do 200Vpp in filament bias with only 0.08% distortion when driving a demanding transmitting valve. I have to say that I haven't seen anyone showing similar metrics around. The input step up sounds really good when driven with a proper preamp.
I invite you to come around and judge it yourself with your ears given that I already posted tons of response measurements in my blog.

I see it differently. In the end you need 3 stages. I cannot see a difference in comparison to a 3 stage integrated amplifier except that you have so much heat, and filament bias is one source, that you have no choice but making a pre and power amp. That's not the case of a 300B amp if power isn't wasted....
The THD of your amp is not 0.08% and that matters as well.

@45: Would be good if you can share some measurements when challenging some of the options discussed. It would also help us understand better your points

thanks
Ale

Next time for sure. I don't have a proper lab here like in Italy and maybe I will not take part to such discussions until then. However I don't think measurements are conclusive about the performance and some people want only hear what they like....
Also ask Nigel Pearson about harmonic cancellation and bad influence of 2nd harmonics on sound when too much. Nothing so much better than the 3rd and nothing like the usual flawed reasoning based on stationary signal that has nothing to do with music etc. I think he already has a lot of measurements to show.

 

[mogliaa]

My system has three stages including output stage only because preamplifier is actually more a line stage which helps driving the input transformer, no additional gain is needed.
Yes, my amp generates a lot of heat and despite that is a pleasure to listen to and it was worth the challenge to experiment. I'm learning a lot and if I don't try things out, then how can I improve my designs?

I have to build the 300B stage and see how it performs. So much to try and test, but sad I don't the time I wish to do all this.

I value your feedback a lot given your vast experience so hope to keep reading your posts.

Back to the OP: the 4P1L as line-stage is superb. I'm rebuilding it again shortly!
cheers
ale

 

[Abszero]

62W of heat

My system has three stages including output stage only because preamplifier is actually more a line stage which helps driving the input transformer, no additional gain is needed.
Yes, my amp generates a lot of heat and despite that is a pleasure to listen to and it was worth the challenge to experiment. I'm learning a lot and if I don't try things out, then how can I improve my designs?

I have to build the 300B stage and see how it performs. So much to try and test, but sad I don't the time I wish to do all this.

I value your feedback a lot given your vast experience so hope to keep reading your posts.

Back to the OP: the 4P1L as line-stage is superb. I'm rebuilding it again shortly!
cheers
ale

I agree with Ale. My PSE dissipates about 62W only on filament bias. Nice and warm for the winter season. However, it is well worth the compromise, I cant stop listening to its gorgeous sound. My job is in energy efficiency, and This amp is quite the opposite with a global efficiency of about 2-3%.😢
Thanks Ale for your continuous effort to make progress and keep us busy..
Best,
Radu

 

[enzoastro]

On my 4p1l preamp the bias is 12.6v so i use that to series heat the two mc riaa stage. Two birds with one stone, i get fil bias and a very quiet heater supply for the idh tubes saving at least one power supply.

 

[michaelvv]

Hi mogliaa.

Did you made a LTSPICE schematic over "The Siberian" you want to share ???

Try to learn how to use the program, and I would very much like to build the 4P1L Siberian Pre-amp...

/Thanks Michael

 

[michaelvv]

Hi Michael,
Ask Big Bear Audio or VinylSavor for quotes on LL1660. You can get it gapped to the current you need. If not, 25mA would be great.
Unfortunately haven't had time to document all the variations I tried so far. I'm actually rebuilding mine again with an LL1660, filament bias and Rod's filament boards. Concepts and parts are similar to the Siberian and 26 preamps.

Are u lacking of gain in your system? what is the input stage of your amplifier? are you driving long cables? You need to consider these questions when choosing the best configuration for your system
cheers
Ale

Hi Ale.

I look forward to hear about your impressions when using the LL1660.

I'm reading all the thread regrads the 4P1L through just to ensure I does
not miss anything

/Michael

 

[michaelvv]

Try to draw the schematic in LTSpice.

Hi Ale...

I'm trying to draw your schematic for the 4P1L in LTSpice.

Doing the hard way, but learned a lot.

I'm really far now so I just have one show stopper...

Your current source limiter , think it's a gpoly is my show stopper.

I got this error G§I1 - missing note(s).

My Netlist is
___________________________________________
* G:\Tube_Amps\ltspice\sim_city_web\Working\4p1l_Preamp_new.asc
V1 N001 0 250v
M1 N001 N002 N003 N003 DN2540
M2 N005 N004 N003 N003 DN2540
R1 N002 N005 100
R2 N004 N006 100
R3 N005 N006 130
R4 N006 N007 1K
C1 Out N007 220n
R5 Out 0 47K
XU1 N007 N008 N011 N009 4P1L
R6 N008 In 1K
V2 In 0 SINE(0 1 1K) AC 1
R7 In 0 100K
R8 N011 0 1K
V3 N010 N011 4.2V
G§I1 N010 N009 325mA
.model NMOS NMOS
.model PMOS PMOS
.lib C:\Program Files (x86)\LTC\LTspiceIV\lib\cmp\standard.mos
.ac oct 10 1 100000
.lib tube.lib
.backanno
.end
___________________________________________

My asc is

Version 4
SHEET 1 880 916
WIRE 496 -192 0 -192
WIRE 0 -176 0 -192
WIRE -48 -96 -208 -96
WIRE 496 -64 496 -192
WIRE 0 -16 0 -80
WIRE -208 0 -208 -96
WIRE 128 0 48 0
WIRE 496 48 496 16
WIRE 128 80 128 0
WIRE -208 96 -208 80
WIRE 0 96 0 80
WIRE 0 96 -208 96
WIRE 0 112 0 96
WIRE 0 208 0 192
WIRE 128 208 128 160
WIRE 128 208 0 208
WIRE 0 224 0 208
WIRE 208 304 0 304
WIRE 480 304 272 304
WIRE 480 336 480 304
WIRE 0 352 0 304
WIRE -336 400 -528 400
WIRE -224 400 -336 400
WIRE -48 400 -144 400
WIRE 480 432 480 416
WIRE -528 464 -528 400
WIRE -336 464 -336 400
WIRE 32 464 16 464
WIRE 32 496 32 464
WIRE -528 592 -528 544
WIRE -336 592 -336 544
WIRE 32 608 32 576
WIRE -16 704 -16 464
WIRE 32 704 32 688
WIRE 32 704 -16 704
WIRE -16 720 -16 704
FLAG 496 48 0
FLAG 480 432 0
FLAG -528 592 0
FLAG -336 592 0
FLAG 480 304 Out
FLAG -528 400 In
FLAG -16 800 0
SYMBOL Misc\\battery 496 -80 R0
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName V1
SYMATTR Value 250v
SYMBOL nmos -48 -176 R0
SYMATTR InstName M1
SYMATTR Value DN2540
SYMBOL nmos 48 80 R180
SYMATTR InstName M2
SYMATTR Value DN2540
SYMBOL res -224 -16 R0
SYMATTR InstName R1
SYMATTR Value 100
SYMBOL res 112 64 R0
SYMATTR InstName R2
SYMATTR Value 100
SYMBOL res -16 96 R0
SYMATTR InstName R3
SYMATTR Value 130
SYMBOL res -16 208 R0
SYMATTR InstName R4
SYMATTR Value 1K
SYMBOL cap 272 288 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C1
SYMATTR Value 220n
SYMBOL res 464 320 R0
SYMATTR InstName R5
SYMATTR Value 47K
SYMBOL Tubes\\Triodes\\4p1L 0 400 R0
SYMATTR InstName U1
SYMBOL res -128 384 R90
WINDOW 0 0 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName R6
SYMATTR Value 1K
SYMBOL Misc\\signal -528 448 R0
WINDOW 123 24 132 Left 2
WINDOW 39 0 0 Left 2
SYMATTR Value2 AC 1
SYMATTR InstName V2
SYMBOL res -352 448 R0
SYMATTR InstName R7
SYMATTR Value 100K
SYMBOL res -32 704 R0
SYMATTR InstName R8
SYMATTR Value 1K
SYMBOL Misc\\battery 32 592 R0
WINDOW 123 0 0 Left 2
WINDOW 39 0 0 Left 2
SYMATTR InstName V3
SYMATTR Value 4.2V
SYMBOL Misc\\Gpoly 32 480 R0
SYMATTR InstName I1
SYMATTR Value 325mA
TEXT -560 616 Left 2 !.ac oct 10 1 100000

_________________________________________________

I'm a little lost. Can you point in the right direction

/Thanks Michael.

 

[yagoolar]

GPOLY is voltage controlled current source. Is it what you intended to use?

 

[michaelvv]

Hi yagoolar.

I need the right component from the schematic ,see above.
Should be a current limiter at 325mA.

Can you see it on the schematic...

Michael

 

[yagoolar]

I see a current source. So if you don't want it to be controlled by voltage signal (gpoly) use current source in LTspice.

 

[michaelvv]

As a newbie to LTSpice I saw the same. Just added a current , and now it works.

Thanks anyway.

 

[mogliaa]

Hi Michael,
That is a very old diagram. You ought to run the 4P1L with paralleled filaments and starved at 550mA instead of 650mA to reduce microphonic noise. Also helps to keep the filament resistor value low to provide the negative bias.
I'm not around my workshop so can't access the spice files. You can check on my blog where you will find the Siberian preamp diagrams. If you are planning to load it with a CCS or gyrator you better run it at 20mA. That is where 4P1L performs at its best.
Ale

 

[michaelvv]

Hi Ale...

You said 550mA. You're are using 2.1 Volt then ???.

Think I remember from the datasheet that it required either
2.1v/650mA or 4.2v/325mA.

Was Led Bias very bad on the 4P1L tube ???

/Thanks Michael.

 

[Rod Coleman]

Michael,

Please beware of the reduction in S (gm, mA/V) of the 4P1L, if you use 1K cathode resistor, with no bypass cap. This will reduce the gain to a low value.

You can bypass the cathode resistor (this sounds bad, if electrolytic capacitor is used), or you can use Filament Bias, like Ale does.