DR103 preamp - what does the CF do?

[Zozobra]

I'm looking at a schematic for a DR103 and I don't quite understand what is going on with the CF in V3:
http://hiwatt.org/Schematics/DR_Pre2Input.gif

Is it part of the NFB and is it integral to the DR103 tone?
Apologies if this has been done to death. I couldn't find much in the way of useful discussion on this topology.

[Dave H]

It looks like it is providing the bias voltage for the LTP phase inverter.

[Zozobra]

Ah, now I see it. Thanks.

[Zozobra]

Is there any real advantage or tonal difference to using CF for the bias voltage over using a potential divider such as shown Merlin's book? Maybe a MOSFET CF would be a good choice if I didn't want to use another bulb? Will the silicooties police charge me with crimes against Dave Reeves legacy if I do?

I get that using fixed bias like this will reduce blocking so should make for better bass response which is appealing enough reason to try it out.

[Malcolm Irving]

Interesting circuit. The CF has a low output impedance - which means that the bias voltage for the LTP will be better regulated or 'more stable' than it would be with a simple potential divider. This could be part of the reason why Hiwatt's are known for a 'firm response' when driven hard.

[J M Fahey]

Yes, it's just a buffered voltage reference, in this case for the LTP inverter.

Funny there are no filter capacitors.

I'm quite certain that the LTPI would be equally biased with a 1M/220k string, with, say, 10uF in parallel with 220k.

[Gaz]

Yes, it's just a buffered voltage reference, in this case for the LTP inverter.

Funny there are no filter capacitors.

I'm quite certain that the LTPI would be equally biased with a 1M/220k string, with, say, 10uF in parallel with 220k.

Yeah, I've built this with a divider and it functions the same. That's an interesting idea about regulation due to the low impedance, but in practice it's just not audible. The voltage divider is better IMHO because you can set the bias voltage more accurately to control PI headroom. It's been speculated that they designed it the way they did as to not waste a triode. Make sense to me.

[Zozobra]

Thanks guys. With a divider and the 70's 4 hole circuit I can get the design down to just needing 2 ECC83's and 1 ECC81 which is handy as the amp I intend to repurpose only has 3 preamp tubes and I didn't fancy having to jimmy an extra one in there

[mhuss]

In the earlier implementation (http://hiwatt.org/Schematics/DR_Pre4Input_v1a.gif), the signal actually passed through the CF, so a divider with a filter cap to ground would not have worked well. I personally like the way the earlier circuit sounds, but to each their own.

[Zozobra]

Thanks. I see that the Trinity TriWatt uses that CF topology. I must admit that I was scratching my head at what they had done as I was comparing it to the later designs.

[trem]

Pardon me for butting in, but I am just wondering if I am reading this correctly.......so the CF is "just" a quiescent device.?
It sits there at idle and its only job is to provide a constant bias voltage to the grid(s) of the PI.?
Thank You

[mhuss]

...so the CF "just" [...] sits there at idle and its only job is to provide a constant bias voltage to the grid(s) of the PI.?

Yes, in the circuit posted by the OP. I believe this was just a refinement of the older circuit I posted above. Redoing the layout to save a tube, while certainly possible, would have been a fairly major change, while keeping it in place as a voltage reference only involved moving a few wires.

[trem]

Yes, in the circuit posted by the OP. I believe this was just a refinement of the older circuit I posted above. Redoing the layout to save a tube, while certainly possible, would have been a fairly major change, while keeping it in place as a voltage reference only involved moving a few wires.

I suppose you ought to know.....Thank You

[bea]

As this is now adequately answered, may i go a bit into OT:

the circuitry uses an ECC81 as a PI. Sometimes You see an ECC83 instead, notably in the"Mywatt" clones (which otherwise implement the Early 70s Four Input Preamp shown above).

Are there any advantages of using an ECC83 in this place?
Qualitatively, without doing any math i would expect a slightly larger open loop gain of the ciruitry, hence a slightly different behavior of the presence control, but it might be hard to drive the 4 KT88 of the DR201.

[mhuss]

Theoretically, the 12AX7/ECC83 has more gain, but in practice the gain is pretty much determined by the circuit values. IME the 12AT7/ECC81 works better in this position, but for a 50 or 100 it doesn't make much difference until clipping, at which time all bets are off and it's up to what sounds best to you. I don't think I've ever seen a 200 with a 12AX7 in the PI. The 400 actually has 12AT7 pre-drivers for each bank of tubes after the PI.