Early Model vs Late Model DR-103 Phase Inverter

[W5FH]

Just a quick question for those who have used both types-
    Am looking at a build that will utilize the Hiwatt fixed bias 12AU7 phase inverter. The early version (early 70's) used a triode section of ECC83 as both the fixed bias source (for phase inverter) and a "buffer" for the audio signal, with cathode DC coupled to phase inverter grid. Later versions (sometime in 80's) separated out this circuit by capacitively coupling audio signal to phase inverter grid and using ECC83 triode section strictly as the fixed bias source for phase inverter.
    Wanted to ask folks that are familar with these two versions which is the best sounding circuit? 
Thanks.

[tubeswell]

Hiwatt fixed bias 12AU7 phase inverter. The early version (early 70's) used a triode section of ECC83 as both the fixed bias source (for phase inverter) and a "buffer" for the audio signal, with cathode DC coupled to phase inverter grid.

Let’s look at a schematic


https://el34world.com/charts/Schematics/files/Hiwatt/Hiwatt_100w_dr103.pdf


V3b is a voltage regulator that supplies a fixed bias voltage to the phase inverter

[W5FH]

Thanks Tubeswell and pdf64-
    Yes, the schematic that you posted is for the "later" DR-103; the ECC83 triode section is strictly a voltage regulator supplying fixed bias to the phase invererter grids. The audio signal is capacitively coupled to phase inverter from last amplifier stage.
     I was curious if the "earlier" version DR-103 imparted any sonic qualities to the audio signal being that this voltage regulator triode section was also passing through the audio signal to phase inverter. Let me see if I can attach a link to the "early" version schematic.

[W5FH]

Regarding the voltage source cathode follower, I meant to ask-
    Being the audio signal is passing through this cathode follower is there anything going in this stage as compared to how the cathode follower works in the Bassman 5F6A and other amps using it? I have studied how the DC coupled cathode follower imparts the smooth compression due to drawing grid current all the time. But this Hiwatt cathode follower is not DC coupled to driving stage like in these other amps mentioned, so I am curios what sonic properties it may impart?

[sluckey]

Regarding the voltage source cathode follower, I meant to ask-
    Being the audio signal is passing through this cathode follower is there anything going in this stage as compared to how the cathode follower works in the Bassman 5F6A and other amps using it? I have studied how the DC coupled cathode follower imparts the smooth compression due to drawing grid current all the time. But this Hiwatt cathode follower is not DC coupled to driving stage like in these other amps mentioned, so I am curios what sonic properties it may impart?

There is no audio signal passing through V3B. V3B is not even needed. Same thing could be accomplished with that 1M/220K voltage divider on the grid. I think Hiwatt was just trying to use up an extra triode. How clever!

[66Strat]

Regarding the voltage source cathode follower, I meant to ask-
    Being the audio signal is passing through this cathode follower is there anything going in this stage as compared to how the cathode follower works in the Bassman 5F6A and other amps using it? I have studied how the DC coupled cathode follower imparts the smooth compression due to drawing grid current all the time. But this Hiwatt cathode follower is not DC coupled to driving stage like in these other amps mentioned, so I am curios what sonic properties it may impart?

There is no audio signal passing through V3B. V3B is not even needed. Same thing could be accomplished with that 1M/220K voltage divider on the grid. I think Hiwatt was just trying to use up an extra triode. How clever!

I think that we may be comparing oranges to orangutans. If I'm not mistaken, I believe that sluckey is referencing Schematic A and the OP is referencing Schematic B.

[W5FH]

OK 66Strat-
   Yes, you are correct. I am curious if the early 1970's version (the second schematic, the one I posted) imparts any sonic properties with the triode section serving in both roles - - as a low impedance voltage source for phase inverter bias AND as a cathode follower buffer passing the audio signal from last preamp stage to phase inverter. I understand that this triode stage was not necessary (but still used) in the later DR-103 versions where it served only as a voltage source and the audio signal was capacitively coupled from last preamp stage to phase inverter. In the earlier version it also was passing through the audio signal so that is my question. Being a cathode follower, but not DC coupled to previous stage, I was curious if there was any desirable effects, such as the smooth compression we see in the DC coupled ones. 

[pdf64]

The CF is DC coupled to the preceding stage, so it is a version of DCCF.
The 1M8-1M potential divider just reduces the DC voltage at the CF grid.
The 1M8 is bypassed for audio by the 22nF cap.
The DC potential divider will probably prevent the forward bias effect Merlin describes in his DCCF page, so I suspect that tonally, it’s not doing much.

Personally I’ve got no experience of either circuit, but the designer presumably thought that the later version was an improvement.

I’m aware of a couple of users of other forums who build / have built a few Hiwatts, so you may wish to raise the query at TGP and MEF.

Alternatively, the difference in implementation is minimal, it would be easy try it both ways if you the build the board to accommodate that.

[tubeswell]

Schematic B was produced by M Huss. Mark’s version could be ‘interpretive’ (or not).