Hi,
* In the manuals of the Aleph3/5 Nelson states that single ended classA is better because in push pull classA, it's like two guys sawing with one saw, not achieving perfect coordination.
When building a circuit in X configuration, isn't that introducing push pull ?
* The Aleph 3 current source, is it dynamic or not ?
In the manual I can't find anything about that.
It just says 2A.
* Is it sufficient in Aleph to have idle = 2 * output current ?
If the load is very high impedance and the load current is max. 50mA, is it ok to idle at 10mA ?
* Wouldn't a single ended DIY op amp always be better as a chip just because of being single ended classA ?
The chip op amps are always push pull
* In the manuals of the Aleph3/5 Nelson states that single ended classA is better because in push pull classA, it's like two guys sawing with one saw, not achieving perfect coordination.
When building a circuit in X configuration, isn't that introducing push pull ?
* The Aleph 3 current source, is it dynamic or not ?
In the manual I can't find anything about that.
It just says 2A.
* Is it sufficient in Aleph to have idle = 2 * output current ?
If the load is very high impedance and the load current is max. 50mA, is it ok to idle at 10mA ?
* Wouldn't a single ended DIY op amp always be better as a chip just because of being single ended classA ?
The chip op amps are always push pull
To answer part of your question, yes it is introducing an aditional factor but i'd be hesitant to really call it push-pull. Given that both ends are driven by N channel devices each in Single Ended configuration it's likely that we will do far less damage than the comparatively poor linearity of P vs. N channel devices plus when you consider the advantages of a balanced drive system, IMHO you end up quite a way ahead of push-pull amplification driving a single end of the load with the other end grounded.
P channel devices are generally about 1/10 as good as their N channel counterparts and even P channel devices that are meant to be a match to their N channel counterpart are often quite a long way off being a good match for the N channel devices. The input stage is low power and again is single ended so does not incur as much of a penalty as would be the case with having P channels in the outputs. The real problem is with push-pull stages as the linearity of N channel vs. P channel is quite different so apart from the crossover distortion, you get quite different linearity for either side of the wave.
AudioFreak said:
To make my understanding better, I put the Aleph3 on the sim and I am very
confused.Idle per Fet is 0.7A
Input voltage is 1.7Vpp
Output voltage is 36Vpp
The current through one lower 0.47ohm swings between 0.4A and 1.1A
The current through one upper 0.47ohm swings between 0A and 1.5A
Is this single ended
?
When the lower fet is fully on the upper fet is fully off that is true... you could certainly argue that in this instance, it is no longer Single-Ended amplification .... Some members have talked about maybe adding some hysteresis to the current source so it is never fully off however, it would seem hard to do this without adding some component that will adversly effect sound quality. I haven't checked yet but maybe decreasing current gain to something just under 50% will allow the current source to remain on at all times.... just something to ponder. Of course, it's not so quite so simple with the Aleph X circuit (well not if you want the circuit to have a true constant current draw as the original does) but I'll leave that for another time.
In the Aleph, when the upper Fet's current is going up, the lower Fet's current is going down.
For my understanding:
This is push pull, and when it is Aleph-X, it is double push pull.
Single ended is one current is constant, the other one swings.
In the Alephs, both Fet's currents swing in opposite direction at the same time.
When the lower fet is fully on the upper fet is fully off that is just before clipping...
For my understanding:
This is push pull, and when it is Aleph-X, it is double push pull.
Single ended is one current is constant, the other one swings.
In the Alephs, both Fet's currents swing in opposite direction at the same time.
When the lower fet is fully on the upper fet is fully off that is just before clipping...
In push pull the two output transistors works independent on each others work, in their own way, they don´t know each others existance so to speak, and their language differs a little.
The transistor in the Aleph Current Sources is just following the work of the singleended outputransistor wich mainly constitutes the charater of the output, which is only sligthly influenced by the ACS.
I guess this is the difference.
The transistor in the Aleph Current Sources is just following the work of the singleended outputransistor wich mainly constitutes the charater of the output, which is only sligthly influenced by the ACS.
I guess this is the difference.
I think take a push pull design ( I have and it's my own) and idle with that much current through the output, it will sound just as good as any Aleph and then some. Please don't get me wrong, the Aleph is exceptionally good. But my point is 95% of the designs out on the market don't idle enough current through the output, in fact not even close. My two cents
We refer to the Aleph as single-ended because the gain
for the signal comes from only one side. Even though the
current source is active, it is not literally part of the gain path.
We refer to the Aleph X amps as "balanced single-ended" a
term coined by another manufacturer some years ago, referring
to two single-ended amplifiers operating differentially, forming
balanced outputs.
Like the SOZ, technically this would be thought of as push-pull,
although not in the usual sense.
for the signal comes from only one side. Even though the
current source is active, it is not literally part of the gain path.
We refer to the Aleph X amps as "balanced single-ended" a
term coined by another manufacturer some years ago, referring
to two single-ended amplifiers operating differentially, forming
balanced outputs.
Like the SOZ, technically this would be thought of as push-pull,
although not in the usual sense.
I think there is a lot to be said for N-channel-only output stages. You can make them perfectly symmetrical (to the eye at leastAudioFreak said:
) this way: Instead of driving the same end of the load alternate directions, just drive alternate ends of the load the same direction. Works for me. 
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.