I'm not a tube expert but I think this might help: JB biases the 6GM8 with 187 ohms in his 24V Aikido.
An externally hosted image should be here but it was not working when we last tested it.
Be aware that the ECC86 (6GM8) might not be capable of driving power Mosfets high enough in frequency. My attemps to get a similar circuit going failed miserably due to the high input capacitances in the mosfets, even when whey are connected as source followers.
Give it at try, but keep this in mind.
Choosing the smallest possible Mosfet might help.
Give it at try, but keep this in mind.
Choosing the smallest possible Mosfet might help.
Hi Tyimo!
I wonder if E88CC works well at such low voltages.
ÓTOH, if you´re going to build a choke loaded follower you cannot use direct coupling and once you´ve put a coupling cap there you can build a decent driver stage with good tubes operating at whatever voltage they need.
My early experiments with hybrid amps all sounded dreadful because of whimpy driver stages, use a tube with low plate resistance that can handle plenty of current, you´ll need all the current you can get to feed the mosfet´s capacitances.
Power calculation formula, hmm, lemme see...
If you ignore stuff like saturation voltage and losses in the choke, just divide your rail voltage by 1,414, square the result and divide by the load impedance.
Max current? As much as the gear can handle...
I wonder if E88CC works well at such low voltages.
ÓTOH, if you´re going to build a choke loaded follower you cannot use direct coupling and once you´ve put a coupling cap there you can build a decent driver stage with good tubes operating at whatever voltage they need.
My early experiments with hybrid amps all sounded dreadful because of whimpy driver stages, use a tube with low plate resistance that can handle plenty of current, you´ll need all the current you can get to feed the mosfet´s capacitances.
Power calculation formula, hmm, lemme see...
If you ignore stuff like saturation voltage and losses in the choke, just divide your rail voltage by 1,414, square the result and divide by the load impedance.
Max current? As much as the gear can handle...
Hi Fuling!
First: thanks a lot for the output power calculation formula!!
Second: I agree with your oppinions about the hybrids with low voltage whimpy driver stages
I think I would use something like the Pathos TT, but maybe it would be even better to use Mads K's BOSEWA preamp!!
We made a test with my SEWA +BOZ and SEWA+REMUS tube pre and you will not believe, but the BOZ was better!
Please visit the OTA thread for further discourse!
Greets:
Tyimo
First: thanks a lot for the output power calculation formula!!
Second: I agree with your oppinions about the hybrids with low voltage whimpy driver stages
I think I would use something like the Pathos TT, but maybe it would be even better to use Mads K's BOSEWA preamp!!
We made a test with my SEWA +BOZ and SEWA+REMUS tube pre and you will not believe, but the BOZ was better!
Please visit the OTA thread for further discourse!
Greets:
Tyimo
I´ve been thinking a bit about the circuit in the first post.
It appears that a lot of my shortcomings with ECC86 and mosfets was due to low grade test equipment rather than bad circuitry, but I still doubt that an ECC86 can drive a mosfet easily.
The rescue might be an additional follower based on Jfets, more precisely a cascoded White follower. Input capacitanse close to zero and very low output impedance, seems to me like the perfect thing to squeeze in between the tube and the power Fets.
Have a look at this schematic (by wrenchone):
http://www.diyaudio.com/forums/attachment.php?postid=802626&stamp=1135983661
It might look complex, but since direct coupling to the tube would be possible we can remove the bias network and the input/output caps. Suddenly the component count is down to a half.
A three stage CCT might not be as elegant as a two stage CCT, but amps that doesn´t make it to 20kHz ain´t much fun either...
In a while I´ll start building an active crossover based on similar buffer circuits, so there´s no reason for me not to order some extra Jfets and try this one out.
I have friends within reasonable driving distance that has much better test equipments than I do
PS: I have a 30VA toroid transformer with 13 + 13+ 7V secondaries, perfect for a complete 6,3V and 35V PSU!!!
It appears that a lot of my shortcomings with ECC86 and mosfets was due to low grade test equipment rather than bad circuitry, but I still doubt that an ECC86 can drive a mosfet easily.
The rescue might be an additional follower based on Jfets, more precisely a cascoded White follower. Input capacitanse close to zero and very low output impedance, seems to me like the perfect thing to squeeze in between the tube and the power Fets.
Have a look at this schematic (by wrenchone):
http://www.diyaudio.com/forums/attachment.php?postid=802626&stamp=1135983661
It might look complex, but since direct coupling to the tube would be possible we can remove the bias network and the input/output caps. Suddenly the component count is down to a half.
A three stage CCT might not be as elegant as a two stage CCT, but amps that doesn´t make it to 20kHz ain´t much fun either...
In a while I´ll start building an active crossover based on similar buffer circuits, so there´s no reason for me not to order some extra Jfets and try this one out.
I have friends within reasonable driving distance that has much better test equipments than I do
PS: I have a 30VA toroid transformer with 13 + 13+ 7V secondaries, perfect for a complete 6,3V and 35V PSU!!!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.