Netlist
I looked at the wuffwaff spreadsheet and would like to verify one of its calculations. When you put in a value for bias say 8A, is the spreadsheet calculating it as only only one half of the amp? IOW, only two of the four mosfets on the hifizen board will carry the burden of these 8A?
I looked at the wuffwaff spreadsheet and would like to verify one of its calculations. When you put in a value for bias say 8A, is the spreadsheet calculating it as only only one half of the amp? IOW, only two of the four mosfets on the hifizen board will carry the burden of these 8A?
No kidding!Netlist said:We hided them a bit too much to be useful but its all there.
BTW, past sense of hide is hid.
I plan on using a regulated supply, so wouldn't say 700 VA be enough?
I'm not sure how much would be gained with a regulated supply. At least regulation with semiconductors. The project would get rather complex. Wouldn't Mr. Pass be using regulation if it was worthwhile? Also, I thought the circuit design took care of the problems associated with small variations in the power supply.
I'm not sure how much would be gained with a regulated supply. At least regulation with semiconductors. The project would get rather complex. Wouldn't Mr. Pass be using regulation if it was worthwhile? Also, I thought the circuit design took care of the problems associated with small variations in the power supply.
Regulation is fine with me, as long as it is followed by a
big steaming pile of capacitor to ground afterword - about
the same amount that an unregulated supply would have.
/pass/: You can never be too rich or have too much capacitance.
Too thin is out, as it is apparently possible.
big steaming pile of capacitor to ground afterword - about
the same amount that an unregulated supply would have.
/pass/: You can never be too rich or have too much capacitance.
Too thin is out, as it is apparently possible.
I have long advocated a regulated supply followed by capacitance, preferably lots of it. Film caps if you can afford it. Certain commercial manufacturers use regulation with no capacitance and sound damned good. But then someone does something rude--like, say, hit a drum--and the sound drifts away on a thin whisp of a breeze. This leads people to say that regulated supplies have no dynamics. This should be recognized as a partial truth. The amended version should say that regulated supplies with no capacitors have no dynamics.
I will freely grant that regulating the main power buss in an amplifier can be an intimidating task, but there's no excuse for not regulating the rails in preamps and such. I attribute the glorious sound I get from my main tube amps to the power supply, not the circuitry (which is really fairly normal). I regulate the first and second stages, and the bias supply. All separately. That's another nifty trick if your circuit and budget allow for it. Keeps one stage from modulating the others via the rail. Marvelous, simply marvelous.
Do not, repeat, do not regulate the front end of an Aleph or Aleph-X separately from the output stage. Regulate the whole thing or not at all. You have to consider each circuit individually as to whether it can be broken up into separate rails.
Grey
I will freely grant that regulating the main power buss in an amplifier can be an intimidating task, but there's no excuse for not regulating the rails in preamps and such. I attribute the glorious sound I get from my main tube amps to the power supply, not the circuitry (which is really fairly normal). I regulate the first and second stages, and the bias supply. All separately. That's another nifty trick if your circuit and budget allow for it. Keeps one stage from modulating the others via the rail. Marvelous, simply marvelous.
Do not, repeat, do not regulate the front end of an Aleph or Aleph-X separately from the output stage. Regulate the whole thing or not at all. You have to consider each circuit individually as to whether it can be broken up into separate rails.
Grey
Hy,
i have an Aleph 4 and would like to use the case for an Aleph-X. It runs normally at about 57 degrees, heatsinks are 21" wide, 14" high and 1,5" thick per side. What max. output power can i drive save in this case. I look more for voltage swing but current, so what values would you recommend.
Thanks in advance
Frank
i have an Aleph 4 and would like to use the case for an Aleph-X. It runs normally at about 57 degrees, heatsinks are 21" wide, 14" high and 1,5" thick per side. What max. output power can i drive save in this case. I look more for voltage swing but current, so what values would you recommend.
Thanks in advance
Frank
I couldn't give you a better answer then looking at the AXE1.2 spreadsheet, assuming you know what K/W your heatsinks have.sippo said:what values would you recommend.
/Hugo
Hello Hugo,
that´s my problem. I choose the heatsinks for my Aleph 4 by guessing and it turned out to be the right choice. Am i right in assuming that an Aleph 4 dissipates about 250 watts per channel? If this is right (according to manual) i can go to the spreadsheet an choose the new values for the Aleph-X.
Frank
that´s my problem. I choose the heatsinks for my Aleph 4 by guessing and it turned out to be the right choice. Am i right in assuming that an Aleph 4 dissipates about 250 watts per channel? If this is right (according to manual) i can go to the spreadsheet an choose the new values for the Aleph-X.
Frank
Regulator
How does this fit with the Zen V4 with its regulator which has little capacitance after the regulator?. Is it acceptable because of the constant draw of class A or because the regulator is modulated or because it is using a constant current source.
Would it be worth adding more capacitance after the regulator?
BDP
Regulation is fine with me, as long as it is followed by a big steaming pile of capacitor to ground afterward
How does this fit with the Zen V4 with its regulator which has little capacitance after the regulator?. Is it acceptable because of the constant draw of class A or because the regulator is modulated or because it is using a constant current source.
Would it be worth adding more capacitance after the regulator?
BDP
Any circuit which draws something close to DC, for example a differential, can do with less capacitance than a circuit that injects more AC into the rail. That doesn't mean that more capacitance isn't welcome.
You can look at power supply capacitance in different ways. First, there's the filter-out-the-power-supply-hum idea. That's not so unusual. Then there's the reservoir-of-power idea which isn't so weird, either.
But let me suggest another point of view: The bypass-the-signal-to-ground idea. Think of the capacitor bank as a filter...not for the AC hum from the transformer/bridge, but for the AC that gets injected into the rail by the circuitry. In other words, the music. Ideally, the power supply will supply DC and only DC. That implies that it has an infinite impedance to ground at DC and zero impedance at any AC frequency. When music gets into the rail, you want it to go away, not hang around and cause trouble by modulating the circuitry. So where does it go? To ground, via the caps in the power supply. Now, let's assume that the circuit has some intrinsic impedance, say 10 ohms, meaning that if you were to insert a 10 ohm resistor that it would draw the same amount of current as the circuit. Treating it as a simple 6 dB/oct filter, how much capacitance would you have to have in order to roll off 20 Hz? Fooey, 1000uF would do it, right? But that was an oversimplified view of things. In the real world, an amp will reduce to the equivalent of a fraction of an ohm on peaks. Capacitance needs increase accordingly.
Note that this decreases the need for the regulator to respond well at high frequencies. The cap bank takes care of them. Since active devices are really much happier delivering DC, anything you can do to make their lives easier will give you better performance in the long run. Run this scenario backwards and you'll see a possible explanation for regulated circuits having a poor reputation regarding high frequencies and dynamics. They're used with little or no capacitance after the regulator.
Incidentally, people who tell you that film cap bypasses in power supplies are useless since all you're filtering is 120Hz hum from the rectifier are silly. You're also trying to provide a path to ground for 20kHz signals coming into the rail from the circuit. Given that electrolytics are terrible at high frequencies, it makes perfect sense to give those high frequencies an alternate path to ground.
One nifty trick is to design a circuit that works like a see-saw so that current drawn by one part of the circuit matches that relased by another part. They sum to DC (or nearly so) and this reduces the AC in the rail. For instance, the Aleph-X tends to even out its current draw in this manner. Sneaky, eh?
Another thing you can try is to regulate each stage individually. That way, no matter how rowdy that stage gets, it won't bother anything else. Yes, it's expensive and annoying to implement, but it's something to think about.
And of course, circuits that tend to be more immune to racket in the rails are always a nice idea. Think in terms of differentials and perhaps current sources, stuff like that.
Nelson's papers are starting points, meant to stimulate interest. You're free to push them further if you want.
Grey
You can look at power supply capacitance in different ways. First, there's the filter-out-the-power-supply-hum idea. That's not so unusual. Then there's the reservoir-of-power idea which isn't so weird, either.
But let me suggest another point of view: The bypass-the-signal-to-ground idea. Think of the capacitor bank as a filter...not for the AC hum from the transformer/bridge, but for the AC that gets injected into the rail by the circuitry. In other words, the music. Ideally, the power supply will supply DC and only DC. That implies that it has an infinite impedance to ground at DC and zero impedance at any AC frequency. When music gets into the rail, you want it to go away, not hang around and cause trouble by modulating the circuitry. So where does it go? To ground, via the caps in the power supply. Now, let's assume that the circuit has some intrinsic impedance, say 10 ohms, meaning that if you were to insert a 10 ohm resistor that it would draw the same amount of current as the circuit. Treating it as a simple 6 dB/oct filter, how much capacitance would you have to have in order to roll off 20 Hz? Fooey, 1000uF would do it, right? But that was an oversimplified view of things. In the real world, an amp will reduce to the equivalent of a fraction of an ohm on peaks. Capacitance needs increase accordingly.
Note that this decreases the need for the regulator to respond well at high frequencies. The cap bank takes care of them. Since active devices are really much happier delivering DC, anything you can do to make their lives easier will give you better performance in the long run. Run this scenario backwards and you'll see a possible explanation for regulated circuits having a poor reputation regarding high frequencies and dynamics. They're used with little or no capacitance after the regulator.
Incidentally, people who tell you that film cap bypasses in power supplies are useless since all you're filtering is 120Hz hum from the rectifier are silly. You're also trying to provide a path to ground for 20kHz signals coming into the rail from the circuit. Given that electrolytics are terrible at high frequencies, it makes perfect sense to give those high frequencies an alternate path to ground.
One nifty trick is to design a circuit that works like a see-saw so that current drawn by one part of the circuit matches that relased by another part. They sum to DC (or nearly so) and this reduces the AC in the rail. For instance, the Aleph-X tends to even out its current draw in this manner. Sneaky, eh?
Another thing you can try is to regulate each stage individually. That way, no matter how rowdy that stage gets, it won't bother anything else. Yes, it's expensive and annoying to implement, but it's something to think about.
And of course, circuits that tend to be more immune to racket in the rails are always a nice idea. Think in terms of differentials and perhaps current sources, stuff like that.
Nelson's papers are starting points, meant to stimulate interest. You're free to push them further if you want.
Grey
Capacitance multipliers float. If your circuit can tolerate a change in the operating point with grace, then they're an option. For my money, if I'm going to go to the trouble to stick an active device in the rail, I'd prefer to nail the voltage down while I'm at it. To my way of thinking, once the pass device is in the rail, the damage is done in terms of sticking another device in the signal path. If you've chosen to go that route, you might as well work it hard enough to earn its keep.
That said, they work well as a kind of intermediate between no regulator and a regulator.
Grey
That said, they work well as a kind of intermediate between no regulator and a regulator.
Grey
I like the simplicity of the circuit. It's elegant because it is so simple. I would think that active regulation makes the project too complex. You could go to the Nth degree adding regulators to regulate the regulators. Sort of like the government?
Now all of a sudden you are building a Mark Levinson instead of a Pass design. Seems like overkill to me. If the thing ever burns up, it's gonna be costly.
Now some passive components like capacitance and inductance would make a good supply that won't be easily depleated.
I still believe Nelson would have used regulation in his production amplifiers if he felt it was necessary.
Now all of a sudden you are building a Mark Levinson instead of a Pass design. Seems like overkill to me. If the thing ever burns up, it's gonna be costly.
Now some passive components like capacitance and inductance would make a good supply that won't be easily depleated.
I still believe Nelson would have used regulation in his production amplifiers if he felt it was necessary.
- Home
- Amplifiers
- Pass Labs
- Aleph-X builder's thread.