I've run a couple simulations of the Aikido, using a CCS or CV power supply.
According to what I've seen, a regulated voltage PSU seems to be superior to a constant-current PSU.
When you use a constant-current PSU, it forces current down the second branch when it really doesn't want extra current - this stresses the bottom right triode. Basically, the second branch is forced to take whatever current the first branch doesn't want ... why make the amp act unnatural?
A regulated voltage PSU seems to make a lot more sense, because then the two branches are allowed to take whatever current is required.
Comments?
According to what I've seen, a regulated voltage PSU seems to be superior to a constant-current PSU.
When you use a constant-current PSU, it forces current down the second branch when it really doesn't want extra current - this stresses the bottom right triode. Basically, the second branch is forced to take whatever current the first branch doesn't want ... why make the amp act unnatural?
A regulated voltage PSU seems to make a lot more sense, because then the two branches are allowed to take whatever current is required.
Comments?
Does anyone know how these voltage regulators work? If I use two 0D3, it will provide a regulated 300VDC output.
http://frank.pocnet.net/sheets/127/0/0A3.pdf
What about current?
0D3 operates between 5ma to 40ma. There is 35ma of current swing. Will this range be enough of a cushion to satisfy kashmire's concern (instead of forcing current down the second branch, excess current will simply pass though the regulator tube)?
http://frank.pocnet.net/sheets/127/0/0A3.pdf
What about current?
0D3 operates between 5ma to 40ma. There is 35ma of current swing. Will this range be enough of a cushion to satisfy kashmire's concern (instead of forcing current down the second branch, excess current will simply pass though the regulator tube)?
http://www.tubecad.com/2005/July/blog0051.htm
found it
mr. Bas - maybe better to do a little math before - just to check for the expected results - for Your ECC99 mu is around 22 as for my 6n6p - so for the upper res. of 100k the lower res. will be arround 120k
best regards
edit - i have thought a little - if we optimize those resistors in follower for the best performance we optimize for the whole circuit - that means that we can't simply put 120k (for mr. bas application) because the tubes are not paired - we have to be arround 120k - how close it will depend how close are the mu's of the tubes - right? - that means that we would have to optimize circuit every time we substitute the tubes to get the best performance.........
found it
mr. Bas - maybe better to do a little math before - just to check for the expected results - for Your ECC99 mu is around 22 as for my 6n6p - so for the upper res. of 100k the lower res. will be arround 120k
best regards
edit - i have thought a little - if we optimize those resistors in follower for the best performance we optimize for the whole circuit - that means that we can't simply put 120k (for mr. bas application) because the tubes are not paired - we have to be arround 120k - how close it will depend how close are the mu's of the tubes - right?
- that means that we would have to optimize circuit every time we substitute the tubes to get the best performance.........
just have to tune the resistors to the correct ratio
Bas Horneman said:
I know ..
sorry - this was not a critique - i am the same guy - hate math
but - thought I would help.....
yes - it would be nice to use pot at the bottom res. since the tubes are not paired and everything so You wouldn't end at the calculated value - keep us informed about Your progress - i am very interested to know what happened.....
wow - it allmost give me an heartattack - kidding - o.k. thanks - i will proceed with my 6n6p aikido - it will be interesting to see how this tube will work in srpp and in follower..... 
heater said:
Had a chance to revisit this over the weekend. Spice provides little guidance since the 1 Meg resistors are respectively across cathode-anodes of tubes which are matched perfectly in software. Addition of these resistors therefore has no impact on the sim results but might in reality.
Not much later my notebook motherboard roasted in a BIOS upgrade taking with it access to these sims so I can't offer any help for a while, contingent on the whims of post-Carla Compaq. Might be a long while.
rdf said:
Had a chance to revisit this over the weekend. Spice provides little guidance since the 1 Meg resistors are respectively across cathode-anodes of tubes which are matched perfectly in software. Addition of these resistors therefore has no impact on the sim results but might in reality.
Not much later my notebook motherboard roasted in a BIOS upgrade taking with it access to these sims so I can't offer any help for a while, contingent on the whims of post-Carla Compaq. Might be a long while.
that is bad - sorry about that
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.