Phase splitter before the output tube drivers?

Kegger

R.I.P. 1/12/1966 - 6/1/2017
Anyone done or tried this?

(Hammond Type G tone cabinet amps)
I have a pair of hammond mono organ amps that run parralell push pull
with no phase splitter as I assume it's done inside the organ somewhere.

So I was thinking since the 6J5's (1 to each half of output push pull section)
are allready talking to the outputs and have a circuit that works and are
direct to the input could I build the phase splitter ahead of them and send
those signals to each 6J5 (or use the dual triode version 6SN7).

As it stands I was pondering a 6SN7 as input/phase inverter (Long tailed pair)
feeding a 6SN7 driving each half of the push pull outputs.

Or a single triode such as the 6J5 as input/phase inverter (cathodyne) then it
feeding a 6SN7 driving each half of the push pull outputs.

Or a 6SN7 as a driver and phase inverter (cathodyne) then it's feeding a 6SN7
driving each half of the push pull outputs.

I kinda like the third one.

Am I crazy? Will it work? Should I do it?
Thanks for any and all input.
 
It's a great idea and historically a very successful one. The cathodyne has a different Zo at the plate as at the cathode and using dedicated drivers keeps balance while driving the outputs to their limit.

This idea is at the heart of the famous Williamson amp where after the phase splitter, the driver stage devices share a cathode resistor to somewhat cancel 2nd order harmonic distortion, which the 6SN7 produces some of, and produces very little else.

Here is an example page http://www.xs4all.nl/~ideas/amps/chapt14.html
 
Last edited:
Thanks.
Yes I see after more research it is or at least was a fairly common practice.

Cool now I just need to implement the circuit into my amps.
I have a fairly simple version in an amp schematic that I'm looking at.
 
Schematic of organ amp.

Finally found the schem for this thing. :yes:
 

Attachments

  • type_g_sm.JPG
    type_g_sm.JPG
    98.3 KB · Views: 110
The phase schem

Here is the schem I'm contimplating recreating the front portion for the hammond.

Basically from input side of the 6SN7 forward and from the 18K resistor
forward taping voltage from the (6k to plug in connector line) to feed a
new front end with a bit of tweaking to get things right.

Anyone have any thoughts on this?

Major changes need to be done to use a 6SN7 in place of the 6DJ8?
 

Attachments

  • input_sec_PI.JPG
    input_sec_PI.JPG
    38 KB · Views: 76
Something like this.

To where it ends up looking something like this. :D

Yes I still have to deal with the fieldcoils as they create ground for the
center tap P but I found a bulliten from hammond on replacing the field
coil speakers with standard speakers to where they say remove both of
the coils and insert a 250 ohm 20watt resistor for the 250ohm coil. ?
 

Attachments

  • type_g_sm_pi.JPG
    type_g_sm_pi.JPG
    97.4 KB · Views: 62
Last edited:
Kegger said:
Major changes need to be done to use a 6SN7 in place of the 6DJ8?
When I built a Williamson, I did a deal of reading. I formed the opinion that it was the most copied design in history (or may have been once) and that the original was the best. So I decided to find a schematic from (IIRC) Wireless Weekly 1947, and used the specified 6SN7's and the original resistor values where possible.

Since then, I have modded that amp into oblivion but I regret it. I wish I still had it around.
 
If you were to use that schematic with the 6DJ8, you'd end up with way too much gain. The 6SN7 would be slightly better, and then it would look much like a classic Williamson schematic ... but again, more gain than needed with no global feedback and only 6V6's to drive.

Instead, you could just change around a few resistors to convert the existing driver stage to a long tail pair and get reasonably good balance with minimal effort. http://www.aikenamps.com/LongTailPair.htm That would be a two-stage amp, and you'd be able to use the 6J5's with no need to buy 6SN7's.

Just be aware that the output transformers might not be up to modern hifi specs in terms of frequency response, so you may get an old-school sound.
 
Hey!!!,
The speaker's field was the input choke, running in the ground leg. You have a few options. One is the resistor, second would be a choke, and third would be to move the choke to the traditional spot after the rectifier cathode.

It isn't all that critical, a decent PS rebuild will put that amp to rights again...:)
cheers,
Douglas
 
Just got power back on here, been out for around 6hours.

I'll ponder the responses here. Did not think about that in the power supply.
Also we think the 6DJ8 in first stage is to much? How about a 6CG7 in place
of it then, generally considerd a lower gain dropin for a 6DJ8?

Not sure how many 6CG7's I got, but I do have many 6DJ8's and plenty of
6SN7's, not to many 6J5's but a couple there to, these amps came with no
tubes but I got stuff I can fill em with. :D
 
Last edited:
I think there are some decent parts but not looking to make or think it's anything
all that, but mainly just looking to fire them up and see what we got then go from
there as to any mods or changes if things do look promising.

I like the layout and the starting point to maybe do em up pretty good in the future.
Supposably the output trannies have a good chance of sounding pretty decent though.

So will see where this goes.

If someone wants to prapose taking the schem and adding in some front stage to check
out or power supply or whatever, I'm all for it and would take any suggestions someone
may have to offer, what I mean is I'm not adverse to anything at this point.

Still learning quite a bit here. I do like the minimal approach to fire em up though. :D
 
Last edited:
Kegger said:
If someone wants to prapose taking the schem and adding in some front stage to check
out or power supply or whatever, I'm all for it and would take any suggestions someone
may have to offer, what I mean is I'm not adverse to anything at this point.

Kegger, I really don't think you need a front stage, which would make this a much easier project :yes: The LTP or a paraphase splitter come to mind as solutions which, although they require rewiring, would be fairly elegant and cheap.

If you want to just try them with no regard for fidelity, you can probably ground input two and put a 470k resistor between input one and input two... then just drive input one. Lots of 2nd harmonic distortion, probably some hum, but you'll get sound. I think :scratch2:

Here's yet another idea that doesn't require any calculation or significant rewiring: If you want to spend $20, buy a pair of 10k:10k transformers from Edcor and use them as an input transformer phase splitter, which is actually a rather elegant solution and would give you some fun new toys to play with.

Further examples of phase splitters: http://www.bonavolta.ch/hobby/en/audio/split.htm

Soooo many choices. I also have a pair of organ amps (bought from PakProtector) with a fairly lousy phase splitter/driver stage, but I just haven't yet found time to rewire them. But when the time comes, they'll have a simple 6N1P long tail pair. I have a schematic, just no motivation to solder :tears:

Good luck!
 
Last edited:
I like the way you think Jon! Those inexpensive 10k:10k transformers from Edcor
sound promising and like you say another experiment/path I haven't been down.

I'll have to look them up. I'm a tech of both computers and electronics but no engineer!

You another michigander also? :thmbsp: If so you should hit me up if your in the area, we could
probably swap some stories and or other things, anyone who's been at my place is usually
quite taken back by the sheer amount of crap/stuff :D around here. :D

I would assume these are the 10k trannies your talking about:
http://www.edcorusa.com/classx/matching/xsm/XSM10k-10k.htm
 
Last edited:
Kegger said:
You another michigander also? :thmbsp: If so you should hit me up if your in the area, we could
probably swap some stories and or other things, anyone who's been at my place is usually
quite taken back by the sheer amount of crap/stuff :D around here. :D

I would assume these are the 10k trannies your talking about:
http://www.edcorusa.com/classx/matching/xsm/XSM10k-10k.htm

Ha, actually I'm in PA... I bought them from Douglas before he moved. But it definitely seems like all the good tube stuff has been going on in Michigan, rather than here. I keep hearing about some amazing hamfests and the like.

Yup those are the transformers... I cannot make promises on them being "great" but I bet if you do a quick search on diyaudio.com or audioasylum you should find at least a few people who have measured them. They may have some ringing at higher frequencies, but that goes with the territory.
 
AHH Gotcha! Douglas seems to leave a trail wherever he go's!

The transformer idea is intriguing, I'm still leaning towards a tube front end though.
(But will give those serious consideration, for experimenting if nothing else)

Curious your thoughts on the 6CG7 instead of the 6DJ8 in that schematic I praposed.

Any suggested circuit for the easiest/quickest solution to a tube front end on these?
Thanks.

By the way "Further examples of phase splitters: http://www.bonavolta.ch/hobby/en/audio/split.htm"

I've allready been reading that one a few days ago, it is a nice informative piece.
 
Last edited:
So in an LTP you may be thinking 6J5 then a 6SN7?

And in a paraphase just using a single 6SN7?

That could be doable.
 
Now that it's come down to it, the standard Williamson uses (IIRC) 20dB of feedback. I removed this, first partly, then fully. It was OK in my opinion.

This suited me as I have stripped back the output stages from my sources and they need more gain than normal. The point is, as I notice you are also looking at a line stage, I think you'll have plenty of gain to play with if you choose to remove some.

I think I would not choose to use the paraphase phase splitter. The reasons are that one half sees a different load and one half goes through an extra stage, complicating the distortion produced, if this is noticeable. I think it will be if the 'first' half drives the output stage into grid current.
 
Williamson it is!

Ok I believe I have what I'm gunna do, the Heathkit W2 uses the same basic
circuit as the praposed front end I had but uses all 6SN7's with very near the
same values as the 6DJ8 circuit did and the same layout, so this is what I am
going to try, all values have been changed to implement the 6SN7's, I think! :D

Still need to do up the power supply and I'm sure I'll need to fine tune this in,
at least the feedback circuit but it should work, I have all the resistors except
maybe 2, I have all the caps, the tubes, and it will utilize the existing sockets.

So pick up a couple resistors and wire this thing up to see what she does. :scratch2:

Any glaringly obvious mistakes or tweaks?
 

Attachments

  • type_g_sm_pi_3_w2.JPG
    type_g_sm_pi_3_w2.JPG
    100.7 KB · Views: 45
Last edited:
Kegger said:
So in an LTP you may be thinking 6J5 then a 6SN7?

And in a paraphase just using a single 6SN7?

That could be doable.

With an LTP you only need two tubes to do it all, so just the existing pair of 6J5's will work fine after some circuit redesign. I'm sure there is even a 6SN7/6J5 LTP driver schematic somewhere on the web (6SN7 being, simply, two 6J5's in one tube).

The only advantage of using a paraphase splitter is some extra gain, which you won't really need. Sound quality would be worse.

The Williamson option I can't really endorse here. One goal of the Williamson design is having a ton of gain that you can then negate with a big ol' global NFB loop. Here you won't want to use any NFB, and don't need to drive KT66's to full power, so you can get away with a much simpler configuration. My vote: 6J5 LTP :thmbsp: .
 
Back
Top Bottom