CF-driven phase inverter?

[AdmiralB]

I got a good look inside a Komet Concorde, and one difference between it and the 60 is that it uses the 'left-over' triode as a cathode follower...to drive the PI.

What purpose does this serve?

[Daver]

OK, I'll take a stab at it, and then someone can correct me. The cathode follower would present a lower impedance to the PI as opposed to the standard anode driven set up. Potentially less frequency loss. The CF also has a certain tone to it and a unique sound when clipping. My guess is they did it for tonal reasons. Can you tell us what the coupling cap value is from the CF? Does the 50 also have the series 220K resistor into the PI? Any other differences or info you like to disclose?
Dave

[AdmiralB]

The cap is .022uF.

Keep in mind I've only seen the circuit, I haven't traced it with a meter (lots of under-board jumpers). However the board itself is the same as the 60, and I'm quite familiar with that one (I owned one, and made a clone).

The differences seem to be:

1. The Concorde is SS rectified and has larger at the plate supply.

2. The third triode is not set up as an asymmetrical clipper. Instead of an unbypassed 10K cathode resistor, it is 3.3K, also unbypassed.

3. The third triode's input is not the 220K/330K voltage divider - it is a simple 330K grid reference....I *think*. The 330K's tail could either be grounded, or it could be series resistance into the grid...but since there must be *some* ground reference I think it must be grounded.

4. The third triode drives a CF (standard configuration, 100K cathode resistor), which in turn drives the PI through the .022uF cap and 220K series resistor.

Other than that the differences are in parts selection - there are more Orange Drops in the Concorde, the treble cap is a ceramic disc...there is a big CC resistor in the power supply that was metal oxide in the 60. I didn't have a chance to measure the OT ratio, so I don't know if that is the same as the 60. I think the lam stack is a tiny bit smaller.

[Ray Ivers]

Two questions:

1) Is the CF direct-coupled to the previous gain stage, and

2) Is the PI a long-tailed-pair type?

I saw a Komet up close two weeks ago (don't know what model, but it was brand-new), and the build quality just knocked me out - super nice.

Ray

[AdmiralB]

Yes to both. The PI is identical to that in a typical Marshall.

[Ray Ivers]

OK, well... my guess is that they were looking for that GS/CF interaction that (IMO) can really improve overdrive tone - here's a paper that goes into serious detail on the subject:

www.coastin.com/AESPaper.pdf

BTW, I thought the Komet sounded really good cranked, too.

Ray

[AdmiralB]

There is a 'leftover' triode in the original Express/Komet 60 configuration. In my clone I used it as a CF to drive the tone stack, a la 6G6 Bassman.

When I saw the Concorde I reconfigured mine to match. I didn't like it. It's dirtier, and has more bottom, but it loses a lot of the dynamics and touch-sensitivity of the original.

[Daver]

Admiral B.,
Thanks for the insight. More questions, of course. What type of Orange Drops were used? Did the power supply have the diode in it like the Komet 60? Did you convert your entire amp build to the Concorde preamp or just change the CF position? Thanks!
Dave

[AdmiralB]

Actually I mis-spoke. The only ODs in the Concorde are the PI output caps. Didn't pay attention to type. It does not have the diode in the rail.

I moved the CF in mine to the Concorde location and changed the third triode's cathode resistor to the 3.3K value.

My amp doesn't have the voltage divider in front of the third triode; I have a 500K pot instead. This takes the place of the 'fast/gradual' switch in my build.

[Daver]

Admiral B.,
Thanks for sharing. The sound clips on the Komet sight sound very much the same between the 50 and 60 to me anyway. Interesting how they changed things up a bit. Cool info!
Dave

[Matt T.]

Admiral,
Lemme get this straight: You've tried using a cathode follower in an 'Express' type preamp in two different places; at the end of the preamp just before the PI and also just after the first stage, driving the tone stack (a la 6G6B...which I have built before). Is that correct? I've actually tried the CF after the first stage in an Express-type preamp (I was also 'inspired' by the 6G6B circuit)...but that amp had 'issues' and I never really got to hear it in it's full glory.

You said you didn't care for it at the end of the preamp but how did you like it after the first stage?

Thanks.

[AdmiralB]

Correct. As far as driving the tone stack...I can't hear a difference. I even put it on a switch that bypassed it, and neither I nor any of the three other guys with me could hear *any* difference with it in or out, clean, dirty, whatever. I left it in the circuit because....well, I hate seeing stuff go unused!

When I reconfigured the amp per the Concorde circuit, I didn't like it, but I don't know that that is due to the CF placement. It might also be due to the change in the cathode resistor on the third triode.

[Matt T.]

Thanks, I hate seeing that triode go unused too...does that make me OCD or something?

An idea (maybe a dumb one): wire the socket with the CF 'backwards' (i.e. triodes reversed) and try using a 12DW7...so the 12AU7 half is for the CF. Might make a slight difference...probably a waste of time.

I'm building an Express (with LP power scaling as part of the stock build) right now so that's why I was interested.

[raiken]

OK, well... my guess is that they were looking for that GS/CF interaction that (IMO) can really improve overdrive tone - here's a paper that goes into serious detail on the subject:

The problem is, that cathode follower will never distort in that configuration. It will remain extremely linear throughout it's range. The typical Marshall triode/cathode follower stage distorts the bottom of the waveform because it drives a relatively low impedance, frequency-dependent tonestack (the impedance can go as low as the value of the "slope resistor" for some frequencies if the mid control is zero).

If this one is configured as claimed (I've never seen one in person), it is doing nothing. It isn't even providing a low impedance drive to the PI, because there is a 220K resistor stuck in there, raising the impedance even higher than it would be if there were no cathode-follower at all, so you don't get any extended frequency-response benefit, either. The only slight advantage you might get would be if the output of the cathode follower had to drive a long wire over to the PI. If the 220K resistor was placed over at the PI, you may get a bit less noise/crosstalk pickup because of the low impedance run.

Randall Aiken

[Ray Ivers]

Randall,

Hey, great to see you here!

The way I understood it, the CF doesn't do the distorting itself; instead, it draws grid current when the gain stage swings way positive (I've seen them pull almost 2mA!) and loads down the GS plate before it tops out at cutoff. The Marshall 30th Anniversary amp uses three of these GS/CF pairs in the hi-gain channel, and sounds really great to me - very different from the usual 12AX7 preamp overdrive, almost a push-pull distortion tone.

Ray