I was browsing thru the archives, and stumbled across this EQ on gyraf's obscure schematics collection:
http://www.gyraf.dk/schematics/Ear_825_EQ.GIF
I think this circuit is interesting. I have no idea how it sounds (anybody heard one?), or if the circuit topology is a good one at all, but I think it has some really interesting details.
As I understand it, it's a low impedance driver, followed by an opamp-based equalizer (frequency dependent stuff moved between signal input and feedback input with the pots). Opamp built from tubes, diff amp, cascode, output stage.
So far so good.
And here come the questions:
(1) Input stage, built around V1. Feedback via R2 lowers input impedance. And input trafo TX1 has 10k + 10k in series to primary. So is this the (in)famous zero-ohm transformer coupled input stage, where the transformer is used in current mode instead of voltage mode? What does this mean for the transformer? I read in another thread that this configuration can give good performance even from a mediocre transformer. But I guess shielding would be even more important, with the transformer sitting directly on a current summing node?
(2) Transformer based feedback from anode to cathode, T2. Is there a good description for that kind of feedback? I've seen this in output stages, but I have to read up on theory here. Any pointers? And, what winding ratio would you typically choose for that kind of feedback? 1:1 ? Something smaller?
(3) Secondary of T2 feeds the - rather low impedance for tubes - EQ network. Node labelled "C" is where I would expect GND. For some reason, this is not connected directly to GND, but AC coupled via C12. I have no clue why this is. From the output side, DC is already blocked by C11. On the input side, it's floating because of the transformer T2. On the opamp inputs side, the U2 grids are AC coupled with C5 and C6 already. It's clear that nodes "A", "B" and "C" should be on the same DC potential. But why not just make this potential 0V and connect C to GND, making C12 obsolete ??
(4) Yet another variation of the transformer coupled feedback from anode to cathode: V3b, TX3. But wait. The feedback winding of TX3 also "lifts" the foot point of the diff amp's bias resistor R17. Why is this? AFAIK, a signal there would mostly be a common mode signal to the diff amp. So why is the output voltage modulating the amp's common mode? Or is this some compensation, for an otherwise bad CM rejection (small resistors, no current source), to become closer to ideal?
(5) Frequency dependent open loop (forward) gain in both amps. R7, C2, C3 and R22, C8, C9, respectively. I guess this is for stability. Not easy to understand in detail, either.
I hope someone else finds this interesting also. Maybe it's all quite simple for long-time tube experts. For me it contains quite a collection of unusual details.
JH.
http://www.gyraf.dk/schematics/Ear_825_EQ.GIF
I think this circuit is interesting. I have no idea how it sounds (anybody heard one?), or if the circuit topology is a good one at all, but I think it has some really interesting details.
As I understand it, it's a low impedance driver, followed by an opamp-based equalizer (frequency dependent stuff moved between signal input and feedback input with the pots). Opamp built from tubes, diff amp, cascode, output stage.
So far so good.
And here come the questions:
(1) Input stage, built around V1. Feedback via R2 lowers input impedance. And input trafo TX1 has 10k + 10k in series to primary. So is this the (in)famous zero-ohm transformer coupled input stage, where the transformer is used in current mode instead of voltage mode? What does this mean for the transformer? I read in another thread that this configuration can give good performance even from a mediocre transformer. But I guess shielding would be even more important, with the transformer sitting directly on a current summing node?
(2) Transformer based feedback from anode to cathode, T2. Is there a good description for that kind of feedback? I've seen this in output stages, but I have to read up on theory here. Any pointers? And, what winding ratio would you typically choose for that kind of feedback? 1:1 ? Something smaller?
(3) Secondary of T2 feeds the - rather low impedance for tubes - EQ network. Node labelled "C" is where I would expect GND. For some reason, this is not connected directly to GND, but AC coupled via C12. I have no clue why this is. From the output side, DC is already blocked by C11. On the input side, it's floating because of the transformer T2. On the opamp inputs side, the U2 grids are AC coupled with C5 and C6 already. It's clear that nodes "A", "B" and "C" should be on the same DC potential. But why not just make this potential 0V and connect C to GND, making C12 obsolete ??
(4) Yet another variation of the transformer coupled feedback from anode to cathode: V3b, TX3. But wait. The feedback winding of TX3 also "lifts" the foot point of the diff amp's bias resistor R17. Why is this? AFAIK, a signal there would mostly be a common mode signal to the diff amp. So why is the output voltage modulating the amp's common mode? Or is this some compensation, for an otherwise bad CM rejection (small resistors, no current source), to become closer to ideal?
(5) Frequency dependent open loop (forward) gain in both amps. R7, C2, C3 and R22, C8, C9, respectively. I guess this is for stability. Not easy to understand in detail, either.
I hope someone else finds this interesting also. Maybe it's all quite simple for long-time tube experts. For me it contains quite a collection of unusual details.
JH.