HiFi phase splitter - "best" design option?

[banat]

Merlinb
Than the solution will be ; pentode at input as voltage gain stage DC coupled to grid of standard Concertina PI but powered from bipolar PSU ?

 

[Depanatoru]

The swing problem is not resolved even using 500V supply . A cathodyne can't give you 200Vpp swing , or more , for 6AS7 cathode biased .
And of course you can't push it to the max with high distortion , or what is this , a contest for stinginess ?

 

[jcalvarez]

One thing to remember is that for a cathodyne to supply ~200Vpp the first stage needs to supply 200Vpp.
This is the typical setup used by @kodabmx in his triode-strapped TV tubes amps:

You can get there, but you need a ton of voltage.

 

[banat]

jcalvarez
that may work good in combination with this type of PSU , where +275V will be supply line for 6AS7 PP-OPS , and +550V line to supply input and PI stage .
btw , I wonder what result will be if instead of that 6F12P tube we use 6SL7 (6H9C) tube ?

 

[kodabmx]

Close but this is the actually circuit for driving the big ones 🙂

 

[Depanatoru]

As the output tube 12SN7 is amplifying the strain on the cathodyne and input tube is much lower , 600V is for 12SN7 ... completly different than using cathodyne directly for output

 

[jcalvarez]

To be honest I still believe the LTP is the best option here. Are there other options? Sure, but the LTP is the simplest one, imho.

 

[Lingwendil]

To be honest I still believe the LTP is the best option here. Are there other options? Sure, but the LTP is the simplest one, imho.

Same.

I love the cathodyne and prefer it in every instance where I can use it- but this is not the place for it. Unless you go for a Williamson front end and follow up with higher voltage gain stages it doesn't make sense to use it here. 90% of my amps have used the cathodyne- and of the dozen or so finished push-pull amps on hand that I held onto- all use a cathodyne.

Personally, if this were my project that for whatever reason I was required to make happen- I would grab a filament transformer- run it backwards off of the filament supply, and stack that on the plate supply for the differential and front end. Another ~120-150 volts rectified of headroom will make a huge difference here.

 

[kodabmx]

Or you could use a doubler/quadrupler like I use in the Modular Amp 1?

 

[gerrittube]

Hi Koda,

Thanks for sharing your diagrams. In #85 a line is missing from the two capactors C9 and C10 to the two resistors on the right. On the other diagrams the capacitors are properly loaded.

Regards, Gerrirt

 

[Diabolical Artificer]

Sorry not replied sooner, my PC died.

could create a negative supply from?

Think so, having trouble visualising that. A lot of amps use a tap off the secondary to power a bias supply, see - http://www.valvewizard.co.uk/bias.html I prefer a separate tfmr & bridge rectified supply, mainly because the negative V supply is ripple free & I get more Vout.

Re the CCS, I use the attached. I prefer using a MJE340 or BD139 with or without another transistor. The extra tranny gives a higher impedance but there's not much in it. I use Piher 500r presets with a fixed R for the emitter R and use 6V2 zeners. A 6V2 zener gives more adjustment than a LED. With a 6V2 zener Er is found by dividing 5.4 by what Ik you want. Remember for an LTP thats Ik x 2.

Hope that helps, Andy.

 

[Chris Hornbeck]

Not falling into the main body of the thread, but true to the thread title, sometimes the "best" design option is to accept that a mu of 2, not very linear, output valve is just not a good design choice. Engineering is evolution, mostly by death of the not-fit-est. Good information lives to procreate, both as continuing designs and as cautionary tales.

All good fortune,
Chris

 

[Diabolical Artificer]

Good information lives to procreate

So does bad or incorrect information, EG $1000 power cables made from sudo copper, wood cased coupling capacitors to ensure a woody timbre & all that other audio foollery. If your relatively inexperienced sometimes a circuit or valve choice that gets the job done but may not be 100% optimum is all that's needed.

Kushti bok, Andy.

 

[directdriver]

Another option with split-load/cathodyne/concertina is placed it in the front as the input, acting like a 1:2 input transformer and not inside the feedback loop, and cap coupled the PP signal to your favorite PP or differential/LTP driver circuit.

My allegiance to split-load/cathodyne/concertina is that each of the two phases of PP goes through the same number of active devices. Unreasonable? Maybe...

 

[stocktrader200]

Nothing beats the dn2540 fet in concertina setup. KSC3503 is next best.
I get 60v pp off both legs from 130v B+
Fed from 12AX7.
350v B+ would yield 170v pp per leg

 

[Mission720]

I'm going to breadboard a power supply and one channel so I can play around with various configurations.

The power stage will remain fixed as 6AS7, however I might have to change to bias resistors if the B+ voltage varied a lot. I have octal relay bases which are superb for breadboarding (see attached). There's not a noval valve base relay socket unfortunately, however I'll sort something out.

Breadboarding will allow me to easily change components and configuration of the phase splitter/inverter type. I will buy some DC voltage display meters to wire in so I can easily take notes of the circuit voltage levels.

I've got a digital scope knocking around somewhere so maybe time to dig that out or possibly invest in something newer...

Oh, I've taken delivery of a number of ECF82 valves NOS so I will experiment and compare in LTP with CCS against 6N1P, with CCS. And a measure of negative bias as well.