Great work generg.
Just want to add my 2 cents here. If you trying to design for a 4Ohm load you are best off not relying on the aleph ccs to get you there. I also have 4Ohm speakers and I believe the correct approach to get the best sonics is to bias it to get your nominal power from the static quiescent current, and only rely on the Aleph CCS to do it's job for the reactive nature of speakers.
As a bare minimum I would be aiming for 4A quiescent current. P=4^2x4 = 32W avg (64W pk).
5A would be even better.
With the semisouth you should be able to dissipate 50W per device no probs, but I would probably limit it to 40W per device (or about 1.6A per device with 24V rails). So three pairs at 1.6A each should get you where you want to be.
This is going to mean monoblocks. I hope this is not a problem.
You might find once you do this that the irfp240 aleph ccs is not so bad after all. Obviously the preferred approach is to use semisouth for aleph current source as well, but you might the difference between the two is smaller once you increase the quiescent current.
Just want to add my 2 cents here. If you trying to design for a 4Ohm load you are best off not relying on the aleph ccs to get you there. I also have 4Ohm speakers and I believe the correct approach to get the best sonics is to bias it to get your nominal power from the static quiescent current, and only rely on the Aleph CCS to do it's job for the reactive nature of speakers.
As a bare minimum I would be aiming for 4A quiescent current. P=4^2x4 = 32W avg (64W pk).
5A would be even better.
With the semisouth you should be able to dissipate 50W per device no probs, but I would probably limit it to 40W per device (or about 1.6A per device with 24V rails). So three pairs at 1.6A each should get you where you want to be.
This is going to mean monoblocks. I hope this is not a problem.
You might find once you do this that the irfp240 aleph ccs is not so bad after all. Obviously the preferred approach is to use semisouth for aleph current source as well, but you might the difference between the two is smaller once you increase the quiescent current.
I thank you all for your friendly and encouraging comments...... its very pleasant.
@flg: I understand the new xistor, thanks to your painting, only the collector of the EF goes to a cap, did you mean the bootstrap cap? This cap must be also connected to the connection of the two resistors coming from the upper rail to the collector of the ZTX.
... and I must look who is Scott Wourcer!
@Melon Head: good advice, I suppose I will end up with this way, maybe I start with two pairs of Semis at 1,6 A at look at my heatsinks. The best would be to ask
@Zen Mod: is it possible to drive Babelshark J
with two sets of Semisouth at 2 x 1,6 or 2 x 1,4 A. or is it not possible to parallel Mu-followers?
@flg: I understand the new xistor, thanks to your painting, only the collector of the EF goes to a cap, did you mean the bootstrap cap? This cap must be also connected to the connection of the two resistors coming from the upper rail to the collector of the ZTX.
... and I must look who is Scott Wourcer!
@Melon Head: good advice, I suppose I will end up with this way, maybe I start with two pairs of Semis at 1,6 A at look at my heatsinks. The best would be to ask
@Zen Mod: is it possible to drive Babelshark J
with two sets of Semisouth at 2 x 1,6 or 2 x 1,4 A. or is it not possible to parallel Mu-followers?
....
@Zen Mod: is it possible to drive Babelshark J
it's completely possible
just double SS-es , along with source/drain and gate resistors.
they must be tightly matched in lower , same as in upper half ....
maybe sound will be darker somewhat , and you can compensate it in two ways - decrease gate resistors of SS-es and increase current through input LTP - to 30mA
input Jfets are cascoded , so dissipation will not be a problem
O.K that are very very good news .....so I can stop here my little walk through the history of Pass amplifiers, back to the future .....
but I will try two or three SK170 as CCS for the input LTP
and I will try ....
you see I get a little bit older in amplifier building and more renitent.....
nothing is worse than a finished amplifier, my Accuphase E 213 rests in my basement...
nothing is worse than a finished amplifier, my Accuphase E 213 rests in my basement...
....... pretty bad kismet for 3150 EUR amp ..... anyway - we can always make it better .... up to one point - melting heatsinks
ah, I see "fate" in German "Schicksal"..
you are right I must sell him, but it was the beginning of my electronic revival..
I destroyed
- four cable connections
- one minifuse
- one little inductivity
and had to resolder the pot, because I tried another
and the clients buying Accuphase will not forgive me, they want the original glamour
so it will not be easy to sell it.
-
you are right I must sell him, but it was the beginning of my electronic revival..
I destroyed
- four cable connections
- one minifuse
- one little inductivity
and had to resolder the pot, because I tried another
and the clients buying Accuphase will not forgive me, they want the original glamour
so it will not be easy to sell it.
-
well thought , but needs actual test confirmation
maybe nice thing implementing follower there , but will it work more depends of type of lower bjt than anything else ....
say that BC546 will work OK , but BC550 will be starved ......
in any case - collector of upper bjt must be connected to bootstrap point , where any residual positive PSU garbage is already well attenuated
Actually 1.2V (expected Vgs) seems enough for proper BJT operation in my opinion. What should be checked is the gate current. When MOSFET is used the gate current is practically inexistent. It is not the case with enhancement type power JFET, tough I can’t tell what would gate current look like with full output voltage swing. EF helps anyway, just consider where to connect its collector (I guess V+).
even if that voltage seems too low to me , it's worth investigating/testing ;
but - I'll not test it - just because to me is relevant enough Papa's claim that J2-like output stage is more suited for SS Jfets than Aleph CCS .
when you need bigger amp - make bigger , don't try to squeeze more from small one
gate current - pretty much predictable - counting on Jfet input/transfer capacitances ; small bjt is already small ..... why choking him also with smaller voltage
naah ..... we are just fooling around ..... and fooling sole worker in this case - Generg
Consider, F1J and F2J use constant current source, not Aleph, as far as a know.
So possibly an ordinary CCS is best here.
I had this low voltage already some months ago, when I made BA-1 outputstage completely with SemiS.
It worked, no failure of the amplifier.... sound was a little bit sharp in the consonants
and it worked yesterday with A3 sound more precise, but less soul, dynamics?
very difficult to say..
when I try the "EF xsistor" flg suggested.... can I measure anything to proof the success?
only hearing is very difficult.......
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.