time to pull the triger
Well I am about to purchase the toroidal transformers for the Aleph x
low power versionand I am a bit hesitant. The tranys are 300va
24v ct. One tranny for each mono block with one bridge configured for
a dual supply +/-. When doing simulations in PSUD I get about
14 volts which comes to aprox 38 wrms into 8 ohms. Is it feasable to use this trany with the original schematic (+-15 v rails) without much tweaking? or asprin
?
Well I am about to purchase the toroidal transformers for the Aleph x
low power versionand I am a bit hesitant. The tranys are 300va
24v ct. One tranny for each mono block with one bridge configured for
a dual supply +/-. When doing simulations in PSUD I get about
14 volts which comes to aprox 38 wrms into 8 ohms. Is it feasable to use this trany with the original schematic (+-15 v rails) without much tweaking? or asprin
?
There's a difference between his preference and what he builds for the market.
It's clear from the charts in the IRF MOSFET spec sheets that the various capacitances in the MOSFETs decrease fairly rapidly as Vds increases, leveling off somewhere in the vicinity of 25V. Strictly speaking, that would give you an amplifier capable of something like 125W. That's not to say that you can't build an amplifier with any arbitrary rail that you might want, only that if you use a low-ish rail you'll have to accept somewhat higher distortion as a consequence.
Keep in mind that an increase in NFB will pull the distortion back down. I tend to prefer minimal feedback, but others seem to feel otherwise, judging from the number of "the more NFB the better" posts I see.
I posted the Aleph-X as a variant on the Mini-A. I used 15V rails on the Mini-A because it was a reasonable compromise between voltage, current, and power dissipation for a single pair of output devices. The Aleph-X was likewise a minimalist one-pair-per-side design; two Mini-As strapped together. If you want to run higher rails then do so with my blessing. You'll have somewhat lower distortion, but you'll reach a point where you have to start looking at two or more pairs per side in order to handle the power. Like Nelson's Zen series, it was a workable example of the topology, without necessarily being a contender for state of the art.
I have no idea how many have actually been built--probably at least two or three (ahem)--but those folks seem to be satisfied.
Grey
P.S.: For what it's worth, when I look at the Aleph-X schematic these days, I see several things I'd do differently, having had a few extra years to mull it over. But, no, increasing the rail voltage isn't necessarily one of them. That's easy enough to do on your own and the basic principles of the Aleph-X have been flogged to death in a number of threads.
It's clear from the charts in the IRF MOSFET spec sheets that the various capacitances in the MOSFETs decrease fairly rapidly as Vds increases, leveling off somewhere in the vicinity of 25V. Strictly speaking, that would give you an amplifier capable of something like 125W. That's not to say that you can't build an amplifier with any arbitrary rail that you might want, only that if you use a low-ish rail you'll have to accept somewhat higher distortion as a consequence.
Keep in mind that an increase in NFB will pull the distortion back down. I tend to prefer minimal feedback, but others seem to feel otherwise, judging from the number of "the more NFB the better" posts I see.
I posted the Aleph-X as a variant on the Mini-A. I used 15V rails on the Mini-A because it was a reasonable compromise between voltage, current, and power dissipation for a single pair of output devices. The Aleph-X was likewise a minimalist one-pair-per-side design; two Mini-As strapped together. If you want to run higher rails then do so with my blessing. You'll have somewhat lower distortion, but you'll reach a point where you have to start looking at two or more pairs per side in order to handle the power. Like Nelson's Zen series, it was a workable example of the topology, without necessarily being a contender for state of the art.
I have no idea how many have actually been built--probably at least two or three (ahem)--but those folks seem to be satisfied.
Grey
P.S.: For what it's worth, when I look at the Aleph-X schematic these days, I see several things I'd do differently, having had a few extra years to mull it over. But, no, increasing the rail voltage isn't necessarily one of them. That's easy enough to do on your own and the basic principles of the Aleph-X have been flogged to death in a number of threads.
Nelson,
Thanks for your response. I have wondered that for some time... just curious more than anything.
I am personally very pleased with my Aleph-x with the 15 volt rails (and it is not quite complete - if there is such a thing). Thank you Grey and Nelson (and many others) for such a great DIY community!!!
Jeff
Thanks for your response. I have wondered that for some time... just curious more than anything.
I am personally very pleased with my Aleph-x with the 15 volt rails (and it is not quite complete - if there is such a thing). Thank you Grey and Nelson (and many others) for such a great DIY community!!!
Jeff
You can count me in for one of those 3? people
I am very happy with the Aleph-X indeed. I have never heard anything comming close The soundstage is tremendous indeed, and the overall sound is just coherent! At least in my setup I have been listening to a load of fancy amps, but none of them came close to the Aleph-X So at this point, I am looking forward to hear the F4 Mr. Pass himself, claims it might be the best sounding amp, I guess he means in the right setup?Steen
The F4 has very high bandwidth and no global feedback, which is nice (for me at least). But it has unity gain, so unless you have enough output level from your source (which I doubt), you need a good preamp upfront which can swing up to +/- 20V.
Some one did ask at the F4 thread how to add a frontend with gain to the F4. I thought about it once, was not as simple as I thought, and thought that it was definitely a bad idea to mess around with a master piece. So I think it is best to leave that to the Master himself.
Patrick
Some one did ask at the F4 thread how to add a frontend with gain to the F4. I thought about it once, was not as simple as I thought, and thought that it was definitely a bad idea to mess around with a master piece. So I think it is best to leave that to the Master himself.
Patrick
There are many possible variations on the F4 topology. If there is sufficient interest, it might be worth a post or two--possibly even a dedicated thread.
Grey
Hint #1: Cascode the input buffer JFETs to relieve the voltage constraints. How? Reference the cascode devices not to an absolute, but to their output the way I did in the New-Tron. Nuthin' to it.
That alone is good for at least three or four sub-variations.
(And if this leads you think in terms of increasing the rail voltages for more power, then you can express your thanks by sending pieces of pliable paper printed with presidential pictures. [Don't you just love alliteration...])
Grey
Hint #1: Cascode the input buffer JFETs to relieve the voltage constraints. How? Reference the cascode devices not to an absolute, but to their output the way I did in the New-Tron. Nuthin' to it.
That alone is good for at least three or four sub-variations.
(And if this leads you think in terms of increasing the rail voltages for more power, then you can express your thanks by sending pieces of pliable paper printed with presidential pictures. [Don't you just love alliteration...])
my biased opinion
Well after using 13V supply@12 A, I used a 4 ohm load resistor to drain and then to 13v ,source was grounded. I managed to drop 10 v
across the resistor easily(2.5A )however that makes only 3 volts Vds.
This seems to be operating in the resistive region of the curve and not the active region where Id is constant with higher values of Vds and a given Vgs. I think I need more B+ for acurate matching Yes ?
J
Well after using 13V supply@12 A, I used a 4 ohm load resistor to drain and then to 13v ,source was grounded. I managed to drop 10 v
across the resistor easily(2.5A )however that makes only 3 volts Vds.
This seems to be operating in the resistive region of the curve and not the active region where Id is constant with higher values of Vds and a given Vgs. I think I need more B+ for acurate matching Yes ?
J
Unless, of course, you intend to use them at bias currents greater than one amp.
It's not really necessary to get crazy. I have seen some minor shifting in the matching of devices at different currents, but not the sort of thing to make your hair stand on end. That said, I try, when possible, to get as close to actual operating environment as possible, meaning voltage, current, temperature, etc.
Grey
It's not really necessary to get crazy. I have seen some minor shifting in the matching of devices at different currents, but not the sort of thing to make your hair stand on end. That said, I try, when possible, to get as close to actual operating environment as possible, meaning voltage, current, temperature, etc.
Grey