I know the inverted Futterman very well, it is true that the effect of the driver crosscoupled to the output tubes make them behave as cathode followers but they are not both cathode followers! the lower tubes is a cathode coupled amplifier so the Miller equation is still valid but as the actual gain is soo low there is little Miller effect. The total input capacitance, (including Miller effect) for each output tube is similar
I belive that in the inverted Futterman, we have a negative feedback in the lower tube thats why it behave as cathode follower, because NFB decrease its output impedance and increase the frequency response, on the other hand in the simple Futterman topology we have positive feedback in the top tube thats why it behave like common cathode stage!
Last edited:
Yes, that is roughly correct, but for Miller effect it only matters what the gain is over the tube, for an OTL with low load impedance the gain will always be lower than 1 so Miller effect is very limited. I have measured this by checking frequency response and I have also simulated the output stage, I found that the poles for both output tubes are very similar and in my case was around 3 - 4MHz for 6C33C.
I have about -50V grid bias, (it differs between tube samples) so I don't have a input voltage to the lower tube of more than about 50V Pk.
I know that you can argue that both Futterman types will give same end result if end gain of the complete amplifier is the same but I have tested both types and found that inverted Futterman gives a bit lower 2nd order distortion for same output power so symmetry is in reality better for the inverted type, at least that is my experience . You can also argue that as the noninverted type would have more global feedback it will be less stable but that can be handled by adjusting the feedback network.
IMO, the OTL needs a lot gain so it has more Gnfb to cut down the distortion. Inverted mode is doing the reverse, somewhat misled about lower output Z, has gain attenuated and need bootstrapped, both are equal sonic penalty too. I apply Pfb so that output Z is adjustable, there shouldn't be stability problem if apply correctly such as embedded Pfb within NFB loop. Good luck to both, hope you can build the non inverted amp one day just to experience it, if you already done so.
All 3 in real world. I only learnt to use simulation in 2013 (I posted a few threads, Smoking Amp is my teacher
. All 3 topology are done before I use simulation. I just glad that the sim agreed with my circuit so I have tried to use it a lot more nowadays, but physical work is mandatory, the sim is to confirm the result, and to do prototype and some small driver PCB productions, but each time I ordered the PCB or prototype, there is new mods coming up, but the audio quality is make progress each time. Until now I finally can populate the remains 40 amps. You can call me "fail or slow to commercialize" but if I get the audio quality, my main target has already been full filled. I also do small quantity guitar amp not yet sold out.
. All 3 topology are done before I use simulation. I just glad that the sim agreed with my circuit so I have tried to use it a lot more nowadays, but physical work is mandatory, the sim is to confirm the result, and to do prototype and some small driver PCB productions, but each time I ordered the PCB or prototype, there is new mods coming up, but the audio quality is make progress each time. Until now I finally can populate the remains 40 amps. You can call me "fail or slow to commercialize" but if I get the audio quality, my main target has already been full filled. I also do small quantity guitar amp not yet sold out.Nice, it's good for you to do all this experiments, I believe that mods and improvements in tube amps and preamps will never stop, as I said before I built 2 OTL amps one Futterman (inverted), which I built it back in 2004, and one Circlotron (built it in 2011), but all those years never stopped from modifying and redesign them, especially the input and driver stages, in preamps it is more easier and inexpensive, all those years I feel like I became more richer in experience and knowledge (the same is for you and everyone who love tubes and experiments) that is the meaning of DIY.
After all those experiments I am convinced that the inverted Futterman have lower distortions and lower output impedance than the Futterman topology, which is different from your opinion, thats all!
After all those experiments I am convinced that the inverted Futterman have lower distortions and lower output impedance than the Futterman topology, which is different from your opinion, thats all!
Not just my options, it is a fact, I know what your amp topology can deliver you can do better. If you still unhappy with what I said consider these follows see it makes you feel better.
My amp distortion is as low as you can get, because there is so much Gnfb and Pfb , from gain of 79 dbs reduced to 32 dbs or 47 dbs, 7 dbs more than Futterman pentode OTL. I recent find another tweet to get even high gain without losing BW. You can get negative impedance if you want to experiment to make the speaker load "disappeared". The distortion and Z, you can figure out. You got to accept your amp topological limitation, IMO.
My amp distortion is as low as you can get, because there is so much Gnfb and Pfb , from gain of 79 dbs reduced to 32 dbs or 47 dbs, 7 dbs more than Futterman pentode OTL. I recent find another tweet to get even high gain without losing BW. You can get negative impedance if you want to experiment to make the speaker load "disappeared". The distortion and Z, you can figure out. You got to accept your amp topological limitation, IMO.
Last edited:
Photos of your amp? or your amps! sorry for this but it seems to me that you are a little bit "deff", Pfb and low distortions (clear sound and detailed) do not match together, even if you apply - 100dB ΝFB , you just want to impress us, one day you will understand your mistake, take it as advice and not as an attack! ciao!
, you just want to impress us, one day you will understand your mistake, take it as advice and not as an attack! ciao!
Last edited:
Mine looks like this at the moment. It will have 4 output tubes, each with it’s individual dc coupled mosfet driver. The input stage will be a differential cascode or a differential pair (based on the gain needed) with a cascoded fet + depletion fet tail CCS set at around 16mA. Anode loads will be resistors or chokes... for the moment it’s all in pieces since I still need to decide on the input stage. Input stage has a salas high voltage reg set at 300V and an unregulated tail at -135V ish. Driver has a separate power supply, unregulated differential +/- 135V ish and power stage has two switching power supplies at 140V each. Coupling caps are 0.1uF russian silver mica or teflon. Speakers are Quad ESL 63. Will keep you updated if anyone is interested in schematics, pictures, tips and tricks, etc
Attachments
Last edited:
If anyone needs some extra information and inspiration you can have a look at this thread:
6C33 Circlotron
6C33 Circlotron