12ax7 triode after the phase inverter?

[bruno]

Has anyone tried using circuits with a triode after each side of the phase inverter (between the phase inverter plates and the grids of the power tubes)?

Any interesting circuits that can be tried with this setup?

[eleventeen]

Never tried it. Such additional stages might be "interesting", but what you're really trying to do in a push-pull configuration is to apply mirror-like 180 degrees out of phase signals to the power tube grids. At least in the high-fidelity scheme of things. The cleaner the better. Plus, they need to produce pretty large voltage swings...so, whatever those pre-power tube stages might produce would have to be carefully designed so as not to introduce all manner of amplitude-linearity & phase distortion. It's not clear to me that you would be doing anything but introducing some or several forms of distortion and probably not desirable distortion.

[tubenit]

Look at Dumble SSS on the Amp Garage 

With respect. Tubenit

[sluckey]

Gibson did on many of their big amps such as Atlas, Titan, and Mercury series. Here's one...

      http://www.el34world.com/charts/Schematics/files/gibson/ATLAS_MEDALIST.pdf

Ampeg SVT did also.

[John]

Gibson did on many of their big amps such as Atlas, Titan, and Mercury series. Here's one...

      http://www.el34world.com/charts/Schematics/files/gibson/ATLAS_MEDALIST.pdf

Ampeg SVT did also.

Sharing the cathode resistor on V3 helps the output stay balanced?

[SILVERGUN]

Look at Dumble SSS on the Amp Garage 

+1 all the way.....a very interesting circuit and resulting amp sound

If it's lot's of clean headroom you're after, there are some great clips of this amp and some clones out there....on my bucket list for sure...

Try Quinn Amps - King Tone Consoul
Quinnamp Steel String Singer King Tone Consoul

The "official" term (or so I have heard) for the application is "post phase inverter driver" or PPID

[HotBluePlates]

Has anyone tried using circuits with a triode after each side of the phase inverter (between the phase inverter plates and the grids of the power tubes)?

Any interesting circuits that can be tried with this setup?

The triode after the phase inverter doesn't make the circuit more "interesting".

I have a pair of McIntosh MC-30's; each of those uses a long-tail inverter then a differential gain stage then differential cathode followers, all before the output tube grids.

BUT... McIntosh did that because they were pushing the output tubes close to class B (deep class AB, really) and because their unusual output stage arrangement requires relatively huge drive signals.

So I'd suggest that the only time you'd add a gain stage after the phase inverter is when your output tubes require a drive voltage bigger than what the phase inverter can provide with the existing supply voltage and bias.

[John]

So I'd suggest that the only time you'd add a gain stage after the phase inverter is when your output tubes require a drive voltage bigger than what the phase inverter can provide with the existing supply voltage and bias.

Would you also do that if you wanted to "overdrive" the power tubes easier?

[SILVERGUN]

Awww...come on HBP....

This approach is a little interesting....??

How many schematics have I stared at that all share the same PPI topology?
 
OK....call me easily interested 

[HotBluePlates]

So I'd suggest that the only time you'd add a gain stage after the phase inverter is when your output tubes require a drive voltage bigger than what the phase inverter can provide with the existing supply voltage and bias.

Would you also do that if you wanted to "overdrive" the power tubes easier?

Most every amp design has a lot of excess gain, which is thrown away at the tone stack and volume control (and master volume, if present). Even so,  most amps can distort the output tubes with the volume control set between 1/2 to 3/4.

By way of comparison, an amp that needs the extra gain between the phase inverter and output tube would not distort with the volume at maximum, or a preamp stage would clip before the output tubes would distort at all. Since that's not the case with probably any guitar amp you've encountered, it's just not needed.

For the Gibson amps Sluckey noted, the general approach was to have 2 tone networks, one after each of the preamp gain stages, then a split-load phase inverter (which has no gain of its own). So the gain stage between phase inverter and output tube was probably needed.

That said, you might get the same effect from having a single triode immediately before the split-load inverter. The deciding factor of whether that 1 triode before the phase inverter would work or if you need 2 triodes after the phase inverter is likely the power supply voltage available and the type of phase inverter selected (all considered on the basis of the output tube bias voltage).

[HotBluePlates]

Awww...come on HBP....

This approach is a little interesting....??

Not interesting (which you might see based on my last post). It's a waste of parts unless it's absolutely needed.

Example of when it's absolutely needed:
Your output tubes have a bias voltage well over -50v, and supply voltage is near 300vdc.

Or you have that big bias voltage (which implies you need 50v peak to each side of the output stage, or 100v peak-to-peak) and decide you will only accept a split-load phase inverter. Because the split-load has a gain of less than 1, the input voltage to the phase inverter has to be 100v peak-to-peak or you have to have extra gain after the phase inverter.

[John]

So I'd suggest that the only time you'd add a gain stage after the phase inverter is when your output tubes require a drive voltage bigger than what the phase inverter can provide with the existing supply voltage and bias.

Would you also do that if you wanted to "overdrive" the power tubes easier?

Most every amp design has a lot of excess gain, which is thrown away at the tone stack and volume control (and master volume, if present). Even so,  most amps can distort the output tubes with the volume control set between 1/2 to 3/4.

By way of comparison, an amp that needs the extra gain between the phase inverter and output tube would not distort with the volume at maximum, or a preamp stage would clip before the output tubes would distort at all. Since that's not the case with probably any guitar amp you've encountered, it's just not needed.

For the Gibson amps Sluckey noted, the general approach was to have 2 tone networks, one after each of the preamp gain stages, then a split-load phase inverter (which has no gain of its own). So the gain stage between phase inverter and output tube was probably needed.

That said, you might get the same effect from having a single triode immediately before the split-load inverter. The deciding factor of whether that 1 triode before the phase inverter would work or if you need 2 triodes after the phase inverter is likely the power supply voltage available and the type of phase inverter selected (all considered on the basis of the output tube bias voltage).

Good explanation, thanks!

[SILVERGUN]

As always HBP,,,thanks a million for your insightful and intelligent commentary 

Here's why this idea is interesting to me.......I snatched this schematic and may have broken some unwritten rule by crossing the "border" with it,,,but instead of speaking in hypotheticals,,,,here's a snipet for us to dissect.....
Me, from a guitar player looking for a "secret weapon",,,, and you as the logical, educated one 

Respectfully,,,and with my usual sense of humor...
Why shouldn't this be sparking my interest?

I like to believe that Mr. D'ble knows something that I don't (but should),,,,and it just makes me wonder if there could be something more than  smoke and mirrors here?.......unfortunately for me, I don't have the booksmarts to figure it out

The amp that this P.A. resides in (or at least,,,is believed to reside in) has been claimed to have a beautiful, fat clean sound that doesn't breakup at maximum volume......so much so that SRV fell in love with it..
So this isn't about overdrive for me---this time

[jazbo8]

As HPB already mentioned, it is a waste of tube especially since the 7025/12AX7 was used in the driver position for the SSS, which has dismal current capability, so it was not Class AB2 that Dumble was going after, then that begs the question, why add another stage? In the older, and typically larger Dumbleland designs, the use of PI + CF with 12AU7 + 12BH7 actually make some sense, since they have higher current capability. Just my 2c...

[SILVERGUN]

I woke up at 3am,,,,probably because of the fish and chips I had for dinner (late),,,,,but I'm still interested

SO, you're saying that there is a 0% chance that this "addition" makes any difference in the resulting tone of the amp?

If it was 1%, would that make it interesting to anyone other than me?

why add another stage?

Because it's what we do?....to try and make the amp sound different,,,,,better,,,,,unique

If I add microwaves to my microwave and it microwaves better, was it a waste of time? 

OK, I'll go back to bed now 

[jazbo8]

SO, you're saying that there is a 0% chance that this "addition" makes any difference in the resulting tone of the amp?

Actually it is the opposite, there is definitely going to be a difference if you add a stage, noise for one thing  , tone-wise, dunno, it doesn't hurt to try it on the breadboard I suppose, so go ahead prove us nay-sayers wrong

Jaz

[tubeswell]

The Studio Bass uses a 12AT7 as a post-PI CF to drive 3 x 6L6GC on each side

[SILVERGUN]

As I lay in bed watching the fan blades spin,,,,,my 2 brain cells are trying to convince me that:
One of the greatest amp designer/builders of our time, wasted his time on a design,,,,that one of the greatest guitar players of our time found to be the greatest amp he had ever heard........and that I, as an amp builder, should ignore this design,,,because it is nothing more than a waste of parts.........no wonder I cant sleep 

The more I think about it,,,the more I want to believe it has more to do with the response/feel of the amp than the actual tone

The more I hear sound clips of this sss style amp, and other incarnations,  I hear a bass response that reminds me somewhat of a slap bass....just better bass response

it doesn't hurt to try it on the breadboard I suppose, so go ahead prove us nay-sayers wrong

that sounds like a nice challenge....I've only got 1 ahead of it

The Studio Bass uses a 12AT7 as a post-PI CF to drive 3 x 6L6GC on each side

hmmmmmmm..........interesting 

[jazbo8]

The Studio Bass uses a 12AT7 as a post-PI CF to drive 3 x 6L6GC on each side

hmmmmmmm..........interesting 

The AT7 can handle ~10x the current as compared with the AX7, so it makes sense for Fender to use it to drive 3 tubes in parallel.

Jaz

[HotBluePlates]

... I like to believe that Mr. D'ble knows something that I don't (but should),,,,and it just makes me wonder if there could be something more than  smoke and mirrors here?...

The amp that this P.A. resides in (or at least,,,is believed to reside in) has been claimed to have a beautiful, fat clean sound that doesn't breakup at maximum volume......so much so that SRV fell in love with it.

The cathode followers are a different thing... remember, we were originally talking about a triode gain stage between the phase inverter and output tube.

The Studio Bass uses a 12AT7 as a post-PI CF to drive 3 x 6L6GC on each side

According to the 6L6GC data sheet, the maximum grid circuit resistance when using fixed-bias is 100kΩ per tube. Divide by 3 for 3 6L6's per side and you get the 33kΩ resistors from 6L6 grids to the bias supply.

What is that rating about? There are several mechanisms by which grid current can happen. If that current (small as it is) flows through big enough grid resistance, the resulting voltage drop could counteract the bias voltage. Because cathode bias is naturally self-correcting (more tube current = more bias voltage = less tube current) it has a larger allowable grid resistance.

The phase inverter for the amp tubeswell posted is a long-tail and has 220kΩ plate resistors. The rule of thumb is to make the grid resistors of the following stage at least 4-5 times bigger than the plate load resistors of the previous stage. So obviously 220kΩ -> 33kΩ isn't gonna work and will be loaded down significantly.

While we say "loaded down," what it really means is the phase inverter would try to amplify its input but the small load resistance would require more current to create the same-size voltage across a smaller resistance (ohm's law). Bottom line, our rule of thumb with a lower source resistance driving a larger load resistance insures the source can cope with the load without being "loaded down" and its output being distorted.

So we see the Studio Bass amp is using a cathode follower after the phase inverter to cope with the low load imposed by the small grid resistors required by the six 6L6's.

Dumble probably used a cathode follower for a different reason. We assume the grid by itself is an infinite impedance because it's not connected to anything. If you push an output tube so the grid is driven positive, the grid draws current and is no longer an infinite impedance. This is part of the reason for blocking distortion and "farty bass". The grid current could load down the phase inverter and might cause it to distort at a lower output voltage than it otherwise might.

I think Dumble was trying to minimize the effect of grid current as the amp approached maximum power, and have a low source impedance to drive the output tubes so that they were the defining factor in output power rather than the onset of grid current.

[kagliostro]

Also Hiwatt used a tube between the PI and the Power tubes in the 400W

K

[SILVERGUN]

Great explanation HBP

The cathode followers are a different thing... remember, we were originally talking about a triode gain stage between the phase inverter and output tube.

I'm sorry......I wasn't.....I was only talking about this specific circuit because it stands out to me as sounding incredible,,,,and I have to believe that there is something special going on there so that my drive to build this amp won't diminish

And it got me going to think I was totally off the mark 

I think Dumble was trying to minimize the effect of grid current as the amp approached maximum power, and have a low source impedance to drive the output tubes so that they were the defining factor in output power rather than the onset of grid current.

Yeah....that's what I meant to say 

Interesting yet? 

[jazbo8]

I think Dumble was trying to minimize the effect of grid current as the amp approached maximum power, and have a low source impedance to drive the output tubes so that they were the defining factor in output power rather than the onset of grid current.

We are getting slightly OT, but... The CF has very low output impedance (see below), but if Dumble was trying to minimize the effect of grid current as you said, why not use an even more capable driver tube?  

Output Impedance (Rcf) of cathode followers:
Type        gm      Rcf
12AT7    5,500     182
12BH7    3,100     323
12AU7    2,200     455
12AX7    1,500     667

[SILVERGUN]

Good question jazbo...
I think we are assuming 12AX7 because that's how it was labeled,,,,but I can't tell from the original cave drawings if that is factual 

Like Tubenit suggested...TAG forum goes on and on about this amp....mile long threads........I don't have that kinda time right now

I'm also not gullible enough to believe that all of the magic in this amp comes from that PPICF configuration.......so there's still lots to read before I actually try to build this thing,,,or some form of it

I just won't be spending $4000 on a clone or buying a kit from overseas.....that's no fun 

Bruno,,,,I hope I'm not too far out of bounds sir 

[HotBluePlates]

Also Hiwatt used a tube between the PI and the Power tubes in the 400W

Same reason... cathode followers to drive the low grid resistance needed because there are 6 KT88's (!).

Great explanation HBP

The cathode followers are a different thing... remember, we were originally talking about a triode gain stage between the phase inverter and output tube.

I'm sorry......I wasn't.....

 

Well, I was talking initially about gain stages not cathode followers and had the example of the MC-30 as one such amp. 

The CF has very low output impedance (see below), but if Dumble was trying to minimize the effect of grid current as you said, why not use an even more capable driver tube? 

Output Impedance (Rcf) of cathode followers:
Type        gm      Rcf
12AT7    5,500     182
12BH7    3,100     323
12AU7    2,200     455
12AX7    1,500     667

More "rules of thumb":
- If you want to drive a particular load, care about the best voltage output and not power output, the source should be 1/10th the load impedance.
- Conversely, if you want to drive a load for maximum power output, the source impedance should equal the load impedance.

Why? Because when you look at the whole thing through the lens of an equivalent circuit, you can imagine a perfect signal generator outputting a voltage across a divider made up of the source and load impedances in series. If you do that, you get basically 91% of your source output voltage across the load when the source impedance is 1/10th of the load impedance. When impedance match, you get half power across the load and half power across the source, but you strike the best balance of output voltage and current in the load.

So even 12AX7 would be able to drive a 667Ω * 10 = ~6.7kΩ effective load due to the onset of grid current.

I submit that it's easier not to think in terms of current needed in this case (because grid current is typically under a milliamp or two unless you're really crushing the output tubes) and think in terms of source and load resistance.

Source resistance will either be the internal plate resistance at the operating point for a plate-loaded design, or 1,000,000/Gm for the cathode-follower design.

[jazbo8]

I submit that it's easier not to think in terms of current needed in this case (because grid current is typically under a milliamp or two unless you're really crushing the output tubes) and think in terms of source and load resistance.

I get what you are saying now, focus on the voltage not the power. 

Jaz

[PRR]

> why not use an even more capable driver tube?

How much can you shove into the 6L6 grids before they melt?

It's not clear on the datasheets. With all the many different 6L6es made, and variations in grid-wire processing, it may be best to be conservative.

[jazbo8]

> why not use an even more capable driver tube?

How much can you shove into the 6L6 grids before they melt?

It's not clear on the datasheets. With all the many different 6L6es made, and variations in grid-wire processing, it may be best to be conservative.

When I wrote that, I was fixated on the idea of grid current on the power tubes, somehow I got the idea that the amp operates in AB2, anyway, HPB has set me straight...

But to your point on how much grid current the 6L6 can handle, it turns out quite a bit - definitely way beyond the capability of the poor 12AX7, as shown on the GE graph for 6L6GC below:

[HotBluePlates]

... But to your point on how much grid current the 6L6 can handle, it turns out quite a bit - definitely way beyond the capability of the poor 12AX7, as shown on the GE graph for 6L6GC below ...

Good eye!

Now reiterating a fine point: Dumble did not apparently design the amp to be class AB2. Again, I think he added the cathode follower to help compensate for occasional positive grid voltage, such as what happens when you push the output tube right up to the limit of clean output power. In such cases, a very brief temporary overload is typical.

And G1 = +4v, the grid draws 5mA. That is too much for a 12AX7 to supply, regardless of output impedance of a cathode follower. So I say it's pretty clear then he didn't want to drive the grids positive, but give a bit of cushion to the phase inverter when the output tubes reach overload.

[SILVERGUN]

Great reading guys....thank you!

Just for the record,,,I think the original SSS was designed around 6550s

[kagliostro]

May be interesting also to read PRR in this old tread

http://www.el34world.com/Forum/index.php?topic=6395.0

K

[jazbo8]

Great reading guys....thank you!

Just for the record,,,I think the original SSS was designed around 6550s

Not really, the 6550s were used in the later, larger 150W version, the original 100W version used 4x6L6's. See SSS History

But it appears that you were right that the CF does make a noticeable difference to the sound (from TAG):
"I originally built the amp with a traditional PI and then converted over to the PI/Cathode Follower in order to compare the difference (and to make sure that she worked before moving on to something that I hadn't done before). It was really night and day -- with the post-PI cathode follower, she is fuller, louder, much cleaner and the bass is deeper and has more definition." from here.

Here is another tidbit wrt the post-PI CF (from Sebago Sound):
"The power amp driver circuit was originally found in Ampeg SVT Bass amps and gives this version of the SSS an incredible low end punch and enormous amount of headroom." Although the SVT used 12BH7 to drive 3 6550s per side.

So to sum up, the post-PI CF's main purpose is to 1) provide an easier load for the PI; and 2) to hold the PA bias point steady until Eg1<=0V.

[SILVERGUN]

The power amp driver circuit was originally found in Ampeg SVT Bass amps and gives this version of the SSS an incredible low end punch and enormous amount of headroom

Ya know....before Bruno brought this topic up I thought I was done obsessing about this amp....

BUT, you know I spent my entire lunch hour reading those same TAG posts,,and listening to sound clips,,,,and downloading pictures and any schematics/layouts I could get my hands on........sounds like I'm preparing for something 

I thought this thread was supposed to be about Bruno looking for something to build 

I found myself shopping for transformers by the end of the day..........looking for some quality wood locally,,,,, and since I don't care about clone cosmetics I won't waste any time there

I also found out that if you play the last song on Stevie's last album backwards,,,,it says "build it, build it, build it"   

[SILVERGUN]

Thanks for those links jaz,

Do you guys feel like explaining the bias circuit for this CF driver?
It appears to call for -328vdc,,,,and I'm just really not familiar with how this would work...I've only really studied the standard fixed bias PP circuit......or just point me towards some reading
Would it make any sense to try this in a lower wattage 2 tube output stage where you wanted the punch without all the ear damage?
Please give me both versions ......
1) the technical stuff that I will have to read 10-12 times for it to sink in
2) the simplified version, that myself and other "guitar players" will be able to keep up with 

[DummyLoad]

Here is another tidbit wrt the post-PI CF (from Sebago Sound):
"The power amp driver circuit was originally found in Ampeg SVT Bass amps and gives this version of the SSS an incredible low end punch and enormous amount of headroom." Although the SVT used 12BH7 to drive 3 6550s per side.

i believe that design actually goes further back than that... to the inventors of the  KT series tubes - GEC england. attached is a schematic for a 400W amp published circa 1957. notice the tube they use for CF is 6SN7; a variant of the later 12BH7, or more to the point the 12BH7 is similar to the 6SN7 but in a 9pin package; and that 47K grid-leak is sufficient for 4 x KT88 on a rail.

it is not a post PI CF but rather actually; concertina PI >> gain stage >> CF >> output rail(s). i suspect that ampeg used the used the GEC ckt. as a starting reference for their SVT design, too coincidental to be otherwise.


--pete

[HotBluePlates]

Do you guys feel like explaining the bias circuit for this CF driver?
It appears to call for -328vdc ...

I know I can't, without something we could call an "authoritative" schematic. Meaning, not a hand-scribbled, probably error-laden version of a super-secret voodoo circuit.

If on the other hand we have a schematic that is no-sh!t, as-built (even of a successful clone if not the original), then sure.

[jazbo8]

It appears to call for -328vdc
1) the technical stuff that I will have to read 10-12 times for it to sink in
2) the simplified version, that myself and other "guitar players" will be able to keep up with 

You got to squint harder... that's -32.8V fixed bias on the hand-scrawl schematic, so no special supply requireed  
As usual, a good place to start with the technical stuff is at Merlin's site. The simple layman's version is just to treat the stage as a buffer - what goes out = what goes in, there is no gain. You may ask what's the point then? Well, a cathode follower does have some special properties: 1) very high input impedance, so it is easy to drive and 2) very low output impedance, so it can drive difficult loads. So another way to look at it - it's a impedance "matching" device like a transformer in a sense.

Jaz

[SILVERGUN]

You got to squint harder... that's -32.8V fixed bias on the hand-scrawl schematic, so no special supply requireed 

I don't know jaz,,
Take a look at this:......there's also a confirmed, working layout that shows this same bias supply, but I don't know where the line is for sharing info between sites

I'd love to start a discussion that leads to a true build here......but again, I'm not trying to ruffle any feathers
Why should they have all the fun? 

[SILVERGUN]

If on the other hand we have a schematic that is no-sh!t, as-built (even of a successful clone if not the original), then sure.

Unfortunately......I might have to be the one to draw it,,,,,once I'm done making all my mistakes

I just hope the voodoo amp police don't come to my house........I'd hate for them to find out that my home practice amp is a little Traynor pc boarder that I didn't build and contains no mods   

[jazbo8]

I don't know jaz,,
Take a look at this:......there's also a confirmed, working layout that shows this same bias supply, but I don't know where the line is for sharing info between sites

I'd love to start a discussion that leads to a true build here......but again, I'm not trying to ruffle any feathers
Why should they have all the fun?  

Yikes! I am the one that should get new glasses. Yes, the "bias" node is -328V, but that's a misnomer, it should really be called the negative supply... anyway, the actual bias voltage to the power tube is actually ~-40V as noted (the 820k and 130k form the voltage divider to take the negative supply voltage down), still need to figure out the small discrepancy between the various voltages readings... but close enough for now. Pure speculation on my part, but I think the bias network was done this way because Dumble just used whatever power transformer he had laying around at the shop, so he did not have to order another one with a separate winding on the PT.

[SILVERGUN]

Bruno, jazbo, HBP,,,,,anyone else it may concern

Do you guys think this would be better suited in it's own thread?,,,or should we just continue on here?

Haven't heard anything from Bruno in a while,,,,,hope I didn't bore him to death....
But back to his original question, and Tubenit's original answer:
I think this is interseting enough to look more into, and then possibly use this type of PICF in another design  ,,,,and once we figure out what the correct values should be, try to incorporate it into a non clone amp we can call our own......now that's fun.....I knew I built this breadboard for something other than finishing one amp

Yes, the "bias" node is -328V, but that's a misnomer, it should really be called the negative supply... anyway, the actual bias voltage to the power tube is actually ~-40V as noted (the 820k and 130k form the voltage divider to take the negative supply voltage down)

OK......I see that now  

Here's a hacked up version of a 2 tube application, with an implied standard bias supply voltage (approx -40vdc)......what am I missing?
(this schematic is a cartoon for right now,,,not a known good application, but a starting point)

EDITED---I take it back----my schematic snipet isn't ready

[SILVERGUN]

Here's a better idea.....lets work off of the Fender schematic as supplied by tubeswell, since we know it is proven to work and is already scaled down to 2 6L6
Maybe where Dr. D got the idea 

We also won't upset the Dumble army of cloners either........unless the mere mention of the name is enough to upset 

[tubeswell]

Here's a hacked up version of a 2 tube application, with an implied standard bias supply voltage (approx -40vdc)......what am I missing?
(this schematic is a cartoon for right now,,,not a known good application, but a starting point)

If you're going to use a standard bias supply of around -40V, that won't work because the cathode voltages of the CF will be higher than that (in the order of 180-220V for a 12AX7 - I would guestimate) and you have no coupling caps between the CF and the output stage.

Maybe if your bias supply was -220 to -260, you could make that work as a DC-coupled CF (or not), but that would also probably imply a higher-than-desirable grid leak resistance path for the output tubes (from a high-resistance voltage divider in the bias supply? - I'm only making assumptions here, because you haven't detailed your bias supply)

On the other hand if you had coupling caps, then you would need grid leak resistors on the output tubes, which I think is a more practical solution. YMMV

Edit - Oops I see that while I was typing my reply, you have taken away your schematic that I was commenting on. Never mind.

[SILVERGUN]

Thanks tubeswell,,,,I'm going back to your studio bass schem.......at least we know that works  

I'll try to get some sense from that, and retry my schematic snippet

EDITED----OK,,,I'm gonna just take a time out
Looks like the studio bass circuit has a couple pretty major differences

If you guys have any more suggestions,,,I'd love to hear em.

Does this schematic need coupling caps between the CFs and output stage?

[HotBluePlates]

If on the other hand we have a schematic that is no-sh!t, as-built (even of a successful clone if not the original), then sure.

Unfortunately......I might have to be the one to draw it,,,,,once I'm done making all my mistakes

I hope you didn't misunderstand my comment; I don't mean your drawing, but any speculative scribbles from any source that did not start from an actual, working amp. Since the circuit detail in question is "interesting" it is also highly likely that the little details that make it work properly are overlooked.

Here's what I mean:

... there's also a confirmed, working layout that shows this same bias supply ...

Notice that between these two schematic fragments, the cathode follower's plate is at 379vdc and the cathode is at at least -38.5 or so because the cathode of a 12AX7 will generally settle in to be about a volt or so positive of the grid (and there's -39.5vdc at the grid). That's a total of 379v+39v = 418vdc on a tube rated for 330v max from plate to cathode.

Now, McIntosh cheated in a similar way and had 440vdc on the plate of a 12AX7 and -42vdc at the cathode (482vdc total). These amps seem to hold up fine.

What dissipation does this imply? If the cathode is at -38.5v and the bias supply is -328v then there is 289.5v across the 220k cathode resistor; 289.5v/220k = 1.315mA. 482v (plate to cathode) * 1.315mA = 0.63w. So tube dissipation is good, and the potential threat is arc-over between tube pins.

The problem is switch-on... The SSS has 379vdc at the plate end, and -328vdc at the cathode end (because there is no current at the moment of startup and no voltage drop across the 220kΩ resistor). That's 379v+328v = 707v across the 12AX7 (!!). That's bad design, no matter who did it or how famous they are.

Now look at your output tubes:
The 470kΩ resistors to ground for each output tube should not be there. These are drawn going to ground, but in a fixed-bias amp they would normally run to the fixed bias supply. Since that "bias supply" is the cathode of the 12AX7 cathode followers, those resistors are omitted altogether and you just have the 1.5kΩ grid stoppers running to the 6L6 grids.

Another aside: I get that the 12AX7 cathode follower grids should be more like -45vdc. 328v/(820k + 130k) = 0.345mA, 130k * 0.345mA = 44.85v, but this is ~45v away from ground and toward the -328v bias supply so it's a negative voltage.

The bias supply is not terribly great, especially if you go to the trouble of having cathode followers with a "negative supply" like this. It's relatively high impedance (which the designer was trying to avoid by using cathode followers to start with) and not very clean (maybe not a problem because of the relatively low levels vs signal, but some amount of buzzy hum will be injected right into the output tube grids).

[SILVERGUN]

I hope you didn't misunderstand my comment; I don't mean your drawing, but any speculative scribbles from any source that did not start from an actual, working amp. Since the circuit detail in question is "interesting" it is also highly likely that the little details that make it work properly are overlooked.

No, I totally understood....and I get a lot nervous about building something like this without support from guys like you, so I really do appreciate you taking the time to read over my psyco-babble and try to make sense of it 

I think it is just the idea of a unique design in a PP power section that has me most intrigued....especially one that comes off sounding that good
I'd like to try to come up with a decent application, that maybe even leans on the fender schematic a little more, to just hear a difference.....
Just a solid, safe, idea that will let me add this on to a common amp,,,,like a Deluxe Reverb with a PPICF driver might be cool...
This way I'm not lumped in with the cloners, and I can't hurt anyone's feelings (because it's our "design" implementation),,,,but I might get to hear the benefit of the design for myself.......that's not crazy right?

So how can it work?..,........and would it make any sense at all in a 2 tube power stage?


 
 

[jazbo8]

I guess we will continue the thread hi-jacking until someone tells us to stop  Just did a quick review of the various bias schemes posted on this thread, and not surprisingly, the SSS is the most unconventional, and as HPB pointed out, the 12AX7 CF is pretty stressed. As a cloner on TAG pointed, the PI CF in the SSS made a "night and day" difference on the sound, so obviously it is doing something right, BUT that was for driving 2 tubes in parallel, so if you are asking if it makes sense to use it for a regular single pair push-pull, I am not so sure, and at least I would not copy this particular design, there other ways to provide the headroom and manage the grid current without stressing out the tube.

[HotBluePlates]

... I would not copy this particular design, there other ways to provide the headroom and manage the grid current without stressing out the tube.

... Just a solid, safe, idea that will let me add this on to a common amp,,,,like a Deluxe Reverb with a PPICF driver might be cool...

So how can it work?..,........and would it make any sense at all in a 2 tube power stage?

Make sense? Necessary? Probably not. But I guess if you really want to try it, it doesn't have to make sense... as long as the design fundamentals are good.

I'll volunteer to figure out the details, but you have to define what amp you're gonna try this in. Fixed-bias allows the use of the same basic approach as the SSS. Safety can probably be provided by not using such a big negative supply and/or having a plate voltage source with a lower voltage.

I'd probably drop the 220kΩ cathode load to a smaller value (since the negative supply is less and we won't need to drop as much voltage), which also works better with a lower-mu tube with less resulting output impedance (12AT7 or 12AU7), allows the cathode follower to run at a little higher idle current and probably sink/source grid current much better. In fact, thinking about it now we could probably fine-tune this circuit element to be superior in every way to the Dumble original, at least for the purpose I believe it serves.

[SILVERGUN]

I'll volunteer to figure out the details, but you have to define what amp you're gonna try this in.

In fact, thinking about it now we could probably fine-tune this circuit element to be superior in every way to the Dumble original, at least for the purpose I believe it serves.

OK......I'm taking this very seriously,,,,doing some homework, and I'm leaning towards a single channel AB763 with reverb (no trem), SS rectifier, fixed bias, 2- 6L6, with increased first stage filtering to support increased bass response.......kinda like a '65 Twin Lite
I was also looking at the '63 Blonde Twin, for it's lower voltages,,,but I'd like to have reverb

The goal being that slappy, fat, bouncy, bass response that wont break up as you turn it up.....make Stevie proud 

Can we do that within the voltage range you want to be?

[DummyLoad]

The Studio Bass uses a 12AT7 as a post-PI CF to drive 3 x 6L6GC on each side

so far, sim show 12AT7 has best numeric performance. best PSRR, largest swing driving 47K load and lowest Zout of the common 9pin bottles including 12AU7/6CG7/12BH7. maybe why fender chose it?

--pete