Hello,
This is my first topic in this forum, my name is Jan, I live in Germany and I'm interested in audio circuits building mostly Headphone and Phono Preamps the last years. I often use tubes, my goal here is to use technology from 70 Years ago and build something that performs on a level close to modern circuits.
My last project was an Phono preamp using INA849 instrumentation amps in the input, ECC83/12AX7 tubes for the LF equalization and OPA output stage. The circuit around the ECC83 was “classic”, a feedback loop was limiting the gain to 10dB in the HF range, at 20Hz the Gain was 30dB. The highest harmonics (K2) with 1.55V output was around -85dB, measured with an AP ATS-2. Not bad for tubes.
I recently found a real treasure, 2x EF12 NOS Tubes that have been build by Telefunken before WWII. I really think about using them in a phono hybrid circuit again using OPAs in the input Stage (AD797 instrumentation amp) and do all the equalization in a circuit around the EF12 where noise is not an issue. They will be in a shielded case because unfortunately the cathode is connected to the shielding of the tube which is basically an antenna.
My goal was to improve the performance of the ECC83/12AX7 circuit. The output of my old circuit is limited to around 5V, after that THD is increasing. Typically for such an circuit, THD is rising with higher output levels.
So I was simulating OPA circuits, really good results I got with the HDAM design. (HV+ = 150V, V+/- = +-24V) To my surprise, using tubes and transistors in the differential amps is not really an issue. Sure, the offset is extreme because of the unbalance in the input stage, many Volts. But with capacitor coupling (an using caps in the feadback loop deceasing gain to 1 at DC) this is not really an issue. To my surprise in the simulation this circuit reaches around -96dB distortion (mostly K2) @ 5V output in 1kOhm. Using 2 tubes in the diff. Amp does not really improve things here, THD is around the same. I'm not sure how this will perform in real life, the tube could be better in the simulation than in real life. For sure this circuit has some problems, if you power the circuit, output will rise close to the V+ rail until the tube is heated. Stability is really good, circuit is unity gain stable with phase margin > 60°
Would like to hear your opinion! My feeling tells me I will get trouble because of the Tube/Transistor combination in the diff. Amp, I was surprised how good this is working in the simulation.
Thanks!
Best regards,
Jan
This is my first topic in this forum, my name is Jan, I live in Germany and I'm interested in audio circuits building mostly Headphone and Phono Preamps the last years. I often use tubes, my goal here is to use technology from 70 Years ago and build something that performs on a level close to modern circuits.
My last project was an Phono preamp using INA849 instrumentation amps in the input, ECC83/12AX7 tubes for the LF equalization and OPA output stage. The circuit around the ECC83 was “classic”, a feedback loop was limiting the gain to 10dB in the HF range, at 20Hz the Gain was 30dB. The highest harmonics (K2) with 1.55V output was around -85dB, measured with an AP ATS-2. Not bad for tubes.
I recently found a real treasure, 2x EF12 NOS Tubes that have been build by Telefunken before WWII. I really think about using them in a phono hybrid circuit again using OPAs in the input Stage (AD797 instrumentation amp) and do all the equalization in a circuit around the EF12 where noise is not an issue. They will be in a shielded case because unfortunately the cathode is connected to the shielding of the tube which is basically an antenna.
My goal was to improve the performance of the ECC83/12AX7 circuit. The output of my old circuit is limited to around 5V, after that THD is increasing. Typically for such an circuit, THD is rising with higher output levels.
So I was simulating OPA circuits, really good results I got with the HDAM design. (HV+ = 150V, V+/- = +-24V) To my surprise, using tubes and transistors in the differential amps is not really an issue. Sure, the offset is extreme because of the unbalance in the input stage, many Volts. But with capacitor coupling (an using caps in the feadback loop deceasing gain to 1 at DC) this is not really an issue. To my surprise in the simulation this circuit reaches around -96dB distortion (mostly K2) @ 5V output in 1kOhm. Using 2 tubes in the diff. Amp does not really improve things here, THD is around the same. I'm not sure how this will perform in real life, the tube could be better in the simulation than in real life. For sure this circuit has some problems, if you power the circuit, output will rise close to the V+ rail until the tube is heated. Stability is really good, circuit is unity gain stable with phase margin > 60°
Would like to hear your opinion! My feeling tells me I will get trouble because of the Tube/Transistor combination in the diff. Amp, I was surprised how good this is working in the simulation.
Thanks!
Best regards,
Jan
At a quick glance my only real comment is that R2/R11 seem a pretty large order of magnitude too small- but with your hybrid scheme and overall topology I'm not sure if the traditional rules for plate loads apply. I would have expected to see ~100k or so as a start there at least. Is this some sort of folded cascode with load multiplication going on?
If HV+ is also 150 V for this circuit, Q13 may go in breakthrough. It is rated for 120 V and almost the full voltage appears to be across it.
If your models are reasonable, and they probably are, there's no reason why the sim and real life results would not be reasonbly thye same.
There will be small differences but as a 'proof of concept' it should give you a good feeling.
And there is no real reason why a tube here wouldn't work nicely with the solid state stuff, when you account for the tube specific parameters like the offset you mentioned and the lower gain compared to the transistor.
Jan