12AU7 to 12BH7 to 6CG7 conversion question

Hi,
I have an HK 260 Chorale that I'm finishing restoration on. It's listed in this thread.
http://www.audiokarma.org/forums/showthread.php?t=416045

I've run into a funny kind of snag.
The preamp to driver stage consists of:
2 12AX7s (V1&V2)
2 12AU7s (V3&V4)
2 7247s (V5&V6)
4 6L6GCs (V7-V10).

The two 12AX7s run of DC filaments.
The 7247s and the output 6L6s all run off 6.3VAC filaments.

The 12AU7s? That's a different story.

They are not run in parallel, but cascaded in series with matching triodes by channel. However, the first 12AU7 (V3) runs off the series string DC filament that runs the two 12AX7s (V1&V2). The second 12A7 (V4) runs off the 6.3VAC filaments with the 7247s V5&V6) and the 6L6s V7-V10).

I'd decided I wanted to convert both 12AU7s to 6CG7s.
That was easy with the second one (V4). Break the connection between pins 4&5 and run the wire going to pin 9 to pin 5. Worked fine.
But the first one is wired at 12.6 volts from that DC series string that runs the two 12AX7s also.

So, here's what I did for now.
V3 is using a 12BH7 run at 12.6 VDC.
V4 is using a 6CG7 run at 6.3 VAC.
These two tubes may be alike but seem to have very different functions. V3 feeds V4.

It works fine as is and sounds fine too. But is this optimal? What scheme could I use for converting V3 to a 6.3VDC output from that series string without messing up the voltages at V1&V2?
Alternatively, if I were to convert to the AC filament off the power transformer to V3, how would I compensate for the DC filaments to V1&V2?

I scanned the schematic and hopefully it displays fine. I haven't figured out how to rotate it to landscape format. I can rotate it, but it shrinks the scan to 5 1/2 by 8.

Let's see if this cross-section of an unresized scan is better.

Ya know, I don't even use the tone control filters. There's a switch to bypass them, just like in a Citation I or IV preamp and I keep the tone controls switched out, so this may all be for nothing.

The problem with the series string, is that even though it's 12.6v, the 12BH7 takes TWICE the filament current of the 12AU7 (i.e, it will have HALF the resistance). This will unbalance the string... the 12AX7s will get more voltage than stock, and the 12BH7 won't get enough.

If the power supply will take the load (i.e, it doesn't upset voltages elsewhere in that string), you can parallel a 160-165 ohm 10 or 15 watt resistor (165 ohms is about the impedance of the heaters of two 12AX7s with wired in series), in parallel with the SUM of the two 12AX7s (i.e, so that it's in parallel with the series connection of both of them- connected to the "input" heater terminal of the first one, and the "output" heater terminal of the second one). Then, you can use a 12BH7 in place of the 12AU7. But, unless you disconnect that resistor, you can't put a 12AU7 back in.

If you do this, you might want to try replacing the single diode (half-wave rectifier) in the bias/DC heater supply, with a full-wave rectifier. This will reduce the duty-cycle load factor on that winding of the transformer a bit- which should give you enough "headroom" to draw another 150ma from it.

You will need to re-check the bias current on the output tubes afterward... the voltage in that bias/DC heater circuit may change a tad. Probably not enough to hurt anything, but you will want to make sure you can still bias the output tubes properly.

Regards,
Gordon.

Makes good sense. I'm supposed to show this off today, so I'll just put a 12AU7 back into that one socket for the day. Then I'll give this a shot.

Oh, noted one other thing- that first resistor after the diode, in the DC heater/bias supply circuit, would also need to be replaced, to be able to run a 12BH7. I'd start with half the resistance, and double the wattage value, of the stock one...

Regards,
Gordon.

I just realized that I posted what I did in the other thread.
http://www.audiokarma.org/forums/showpost.php?p=5447243&postcount=25