• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Phase Splitters

Status
Not open for further replies.
Jump to Latest
teodorom

teodorom

Member
Joined 2004
teodorom.atspace.com/
#1
Usually, phase splitters (in a PP tube amplifier) are in the stage immediately before the output tubes.
Is there anybody that can argue on a different topology: that is a phase splitter (with gain, like a Long-Tail, Schmitt) as first stage, followed by the driver tubes, then by the output tubes ?
I plan for a low impedance driver circuit, like the one on can get with a 6C45P-E loaded by a solid-state constant current source (IRF830), I circuit I have alredy used (with success) in some other my realizations.
Output tubes are to be driven a little bit in class A2.
Thanks
Teodoro Marinucci
 
#3
It depends on the phase splitter and overall gain setup for the amp. For example, the "Crystal Palace" amplifier shown in Morgan Jones's "Valve Amplifiers" has a long tail in the first stage. The Williamson that Tekko mentioned puts a split load in the second stage, between the input amp and the driver. My amp uses an input transformer to do the dirty work.
 
#4
If I've read into this correctly, and you're thinking of using 6S45s in a long-tailed pair to drive a PP output stage into class *2, it may well be best to do phase splitting prior to the driver stage as an LTP relies in part on equal load impedances (assuming a high Z tail) to split phase properly. This will not be the case when one output valve is into grid conduction when the other is nearing cutoff.
 
#6
I just built ...

... a pair of monoblocks with

1) phase splitter by input transformer (cinemag)
2) push-pull of 6H30
3) interstage LL1660 PP
4) push pull of 2A3
5) OPT LL1623

simple enough for a newbie like me and, yes, it works nice.

A few days ago I modified the interstage connections and the amps are now working as SE to PP just to make some experience and hear the differences.

Ciao
Gianluca
 
#7
I prefer to start with the splitter and use PP throughout. It helps especially with minimizing hum. E.G., please see attached schematic of my 16W PP EL34 triode amp.
 

Attachments

  • rwm amp el34 triode p-p gfb.gif
    rwm amp el34 triode p-p gfb.gif
    16.1 KB · Views: 1,205
#8
Interesting point ..

Good point Ray. This the kind of comments I need.
You say that this configuration helps in preventing hum. Good.
It is interesting also you usage of the pentode as Constant Current Source. Usually I do CCS in Solid State (mainly because I have more reliable -than pentode- SPICE models).
Thanks
 
#9
I like to keep stages to a minimum. Preamp direct coupled to a cathodyne or LTP, coupled to output (buffered if necessary; you mentioned grid current, that'll do it). Cathode followers don't really count as stages because they contribute no poles, they're basically invisible (assuming you drive the OP grids directly from the CF cathodes).
If a lot of gain is necessary, I'd rather use a pentode front end (gain circa 400, do that with a 12AX7!) than a second gain stage (balanced or SE). Reason being, a mere two coupling caps in the circuit combined with the inductance of the OPT makes a large phase-shift oscillator.

Tim
 
#10
Agreed, 2 coupling caps and an OPT within a global NFB loop can easily cause oscillation - one of the main criticisms of the Williamson design. I have tried DC coupling between my splitter and driver stages, which would be the ideal arrangement, but the 6SL7 and 6SN7 tubes I have are not sufficiently well balanced and I got almost 20% difference between the 6SN7 driver plate voltages. However, I've found that, using a step network with a local loop around it, as shown in the schematic I posted, gives me sufficient stability margin for global NFB, plus excellent balance.

My choice of a pentode as a current sink is due to the fact that I had some 6AU6 tubes available. My power tranny has a 50v 0.1A winding, so, using a voltage doubler, I was able get 110v @25mA for a negative supply, sufficient for both the 6SL7 LTP/6AU6 CCS combination and the 6SN7 driver.
 
#11
Re: I just built ...

Gluca said:
... a pair of monoblocks with

1) phase splitter by input transformer (cinemag)
2) push-pull of 6H30
3) interstage LL1660 PP
4) push pull of 2A3
5) OPT LL1623

simple enough for a newbie like me and, yes, it works nice.

A few days ago I modified the interstage connections and the amps are now working as SE to PP just to make some experience and hear the differences.

Ciao
Gianluca
Click to expand...

Ciao Gianluca

I'm looking for an input split transformer to do some test with differential topology. I've read about some projects with Lundhal and Sowter input transformers but I've never read reports about Cinemag products. After I've looked at cinemag homepage I have to say that they seem good products indeed. But how much do they cost? Have you purchased them directly from cinemag?
And which model have you used?

I've just read about SY's Heretical Unity Gain Line Stage and so I've also look at Jensen homepage but it seems that they haven't splitter transformers. A centre tap could be obtained with two precision res but I'd like the idea of two matched secondarys.

Hi

Mark
 
#16
Ciao belli,

I did buy them (15/15B) directly from cinemag and they shipped a nice pair to Italy for 50 euro (BOTH!!!!) or so ( can't remember now). As a plus David, the vice president, is a great guy.

Well, they sound quite great but I don't know if a Lundhal, Jensen or S&B sound better ... no comparison as this is my very first amp. Is there any difference with and without? Don't know .... I just dropped them in the amp and never tried without.

😎

I tested with my scope and it gave -.2db at 20HZ and +0.1dB at 20KHz, BTW you impair the 1KHz square wave test unless the secondary is loaded with 10kOhm. With 20KOhm you just get a small overshoot and increasing the load up to 40-50KOhm wont change that oscillation so much. Now the amp is running with 47K and I don't care.

You will need a pair of resistor to split the signal as it comes with no CT.

I (strongly) suggest the PCB mount version as the flying leads versions proved to be uncomfortable to be mounted 😀

For the next project (an hilarious and crazy 2A3/300B (push pull) based preamp) I will use a step-down (2:1) input tran.

Ciao
Gianluca
 
#17
Hi guys, I've looked at Lynn Olson homepage and than I've found some posts on Audioasylum regarding the lundhal 1676:

http://www.lundahl.se/pdfs/datash/1676.pdf

Many people say that this is one of the best split input transformers.
Another solution just as good should be the cinemag CMOL-2x600T2:

http://cinemag.biz/line_input/CMOL-2x600T2.pdf

What do you think about them? Has anyone tried one of these?

Gianluca, your trafo is this one?

http://cinemag.biz/line_input/CMLI-15-15B.pdf

Or maybe the dual Faraday shield version...
Can you post the schematic of your amp, please?

Mark
 
#18
SPICE Models

Interesting points ... even if I started the thread having tube phase plitters in my mind.
I did so mainly because I don't know (it's my fault) how to model interstage transformers in SPICE, while I know well how to model tubes.
Even my output transformers models are very rough.
Is there anybody having SPICE models of power, interstage, output trasformers ?
Thanks
 
#19
hi Teodorom

Microcap has accurate internal model for trafo.
It uses Jiles Atherton hysteresis theory.
but I do not know if it can export it in spice format,
I can try.
surely, it is necessary to fully know the specific material
of the trafo.

Federico
 
#20
@ Mark_Titano

I got the single shield version. D. Geren suggested to use that version. Sorry the price was 50 euro or so for each of the trans ... I checked the invoice ... sorry

:angel:

The amp I built is a classical PP to interstage trans to PP a la Sakuma or if you prefer like the Amity (no CCS BTW).

Ciao
Gianluca
 
Status
Not open for further replies.
Top Bottom