Pass DIY Addict
Joined 2000
Paid Member
Wonders Never Cease!
Wow, out of the blue we have two momentus events on the same page of this thread: Nelson gives in and uses high speed diodes and Grey surfaces for some "air"!
What a day for the DIY community! I just can't wait to see what happens next 
Nelson, that Aleph-X looks like a mighty beast! Can you post some size & weight specs on that bad boy?
Wow, out of the blue we have two momentus events on the same page of this thread: Nelson gives in and uses high speed diodes and Grey surfaces for some "air"!
What a day for the DIY community! I just can't wait to see what happens next Nelson, that Aleph-X looks like a mighty beast! Can you post some size & weight specs on that bad boy?
If you decide to leave the compensation capacitor out (the small one in the feedback loop), you will see mild overshoot on a 10Khz squarewave, realworld!
Sim indicates the same. Phase margin suffers.
If you don't mind this, don't tell us you have not been warned when your nice speakers go bust or worse.
Sim indicates the same. Phase margin suffers.
If you don't mind this, don't tell us you have not been warned when your nice speakers go bust or worse.
Real world. I presume then, that you have built this circuit. Which iteration? Did you use the resistors to ground at the output? Any other real world observations that would be useful? Sorry for the 20 questions, but it's good to hear more info from someone who has actually built the circuit. With respect to the feedback cap, is it possible that it was eliminated as a result of a particular board layout?
You could have a look at page 23 of this thread.
page 23 of the alpeh-x thread
There are some pictures on one of the next pages.
I build the first channel in July of this year. Since then I have finished a second channel.
I am still working on the case. I am also finishing a schematic for a
larger version. (12 or 16 output fets).
Yes, I did use the output resistors to ground. Even added the famous resistors from output to input CS.
Matching the input fets is critical, if you want low offset on the output.
page 23 of the alpeh-x thread
There are some pictures on one of the next pages.
I build the first channel in July of this year. Since then I have finished a second channel.
I am still working on the case. I am also finishing a schematic for a
larger version. (12 or 16 output fets).
Yes, I did use the output resistors to ground. Even added the famous resistors from output to input CS.
Matching the input fets is critical, if you want low offset on the output.
I believe that the resistors to ground are in place in the XA200.
They serve at least two purposes.
1. They help stabelize the absolute offset (to ground).
I found they only help a little. The famous resistors have a bigger effect.
2. They make sure the amp sees a load at very high frequencies.
This is a very good thing for the stabilty, and reliability of this amp.
Since I try to build all my amps as stable as possible, I regard these resistors as essential.
I also suspect ( have no proof) they play a little role in preventing the "hall of mirrors" effect, as described in the patent.
They serve at least two purposes.
1. They help stabelize the absolute offset (to ground).
I found they only help a little. The famous resistors have a bigger effect.
2. They make sure the amp sees a load at very high frequencies.
This is a very good thing for the stabilty, and reliability of this amp.
Since I try to build all my amps as stable as possible, I regard these resistors as essential.
I also suspect ( have no proof) they play a little role in preventing the "hall of mirrors" effect, as described in the patent.
Fancy diodes (again)
I was quoted about $8 and 14 weeks delivery on the Infineon SiC Schottky diodes today. I guess they would not be very useful after all in a traditional amplifier with large transformer and cap banks.
Thus we can assume that the fancy diodes used by Pass are of the more traditional type available from various vendors including International Rectifier.
(((Off topic: BTW, I also tried to get quotes on some Infineon C2 type CoolMOS switching FET's but they appeared to be unobtainable.)))
Petter
I was quoted about $8 and 14 weeks delivery on the Infineon SiC Schottky diodes today. I guess they would not be very useful after all in a traditional amplifier with large transformer and cap banks.
Thus we can assume that the fancy diodes used by Pass are of the more traditional type available from various vendors including International Rectifier.
(((Off topic: BTW, I also tried to get quotes on some Infineon C2 type CoolMOS switching FET's but they appeared to be unobtainable.)))
Petter
fcel,
If by story you mean the SF stories I write, there's one on the stands now--"Tower of Babel" in the Nov. '02 issue of <i>Analog</i>. Yes, I've been writing others.
If you mean have I been building things of a stereo nature, the answer is yes (of course). Details will be forthcoming in the proper time and place. I've still got to deploy a preamp to match this amp...amongst other things.
Petter,
I'm betting on IRF diodes. I used them in my big tube amps, but confess that I still haven't gotten around to swapping in regular diodes and comparing sonics. Oh, well.
Grey
If by story you mean the SF stories I write, there's one on the stands now--"Tower of Babel" in the Nov. '02 issue of <i>Analog</i>. Yes, I've been writing others.
If you mean have I been building things of a stereo nature, the answer is yes (of course). Details will be forthcoming in the proper time and place. I've still got to deploy a preamp to match this amp...amongst other things.
Petter,
I'm betting on IRF diodes. I used them in my big tube amps, but confess that I still haven't gotten around to swapping in regular diodes and comparing sonics. Oh, well.
Grey
I was looking at some capacitors for my AlephX PSU. First I had my eye on Nichicon Super Through capacitors until I read a reply from Nelson
Now I understand why they use computer grade capacitors and not special audiophile types as the Nichicons (Ir = 9A).
My bias current will be between 9 and 10A. So the rimple current must be 30A. Now I want to go for BC Components (Philips) serie 114/115 capacitors. The only thing I don't now is the preffered capacity for an AlephX PSU. Is it the same as for an Aleph2 (100kF - 150kF) or can I use more because of the higher bias current?
Now I understand why they use computer grade capacitors and not special audiophile types as the Nichicons (Ir = 9A).
My bias current will be between 9 and 10A. So the rimple current must be 30A. Now I want to go for BC Components (Philips) serie 114/115 capacitors. The only thing I don't now is the preffered capacity for an AlephX PSU. Is it the same as for an Aleph2 (100kF - 150kF) or can I use more because of the higher bias current?
Petter,
The CoolMOS C2 MOSFETs have been discontinued, they are replaced by the C3 series, same part number types, just a change in the suffix. Technically, they have a lower Vth (about 3.3V, instead of 4.5 - 5V), and they have higher transconductance and higher pulse current capability.
Early next year we will be introducing a new 500V series, with some additional types, such as 33A 0.1 ohm device, and 15A device.
For circuits with large capacitors on secondary I would recommend more conventional diodes than the SiC types, because *all* Schottky diodes (including silicon and Sic) have poor surge current capability, because they are unipolar devices- no conductivity modulation through hole carrier injection.
OTOH, I'm using them anyway in my new preamp, DAC, and power amp designs- just need to parallel a few. But you have to watch the surge. Easiest way to due this is to parallel with a "slow" silicon double diode set; during inrush surge, the power will be clamped/handled by the silicon; at normal operating conditions, only the SiC diode conducts.
Regards,
Jon
The CoolMOS C2 MOSFETs have been discontinued, they are replaced by the C3 series, same part number types, just a change in the suffix. Technically, they have a lower Vth (about 3.3V, instead of 4.5 - 5V), and they have higher transconductance and higher pulse current capability.
Early next year we will be introducing a new 500V series, with some additional types, such as 33A 0.1 ohm device, and 15A device.
For circuits with large capacitors on secondary I would recommend more conventional diodes than the SiC types, because *all* Schottky diodes (including silicon and Sic) have poor surge current capability, because they are unipolar devices- no conductivity modulation through hole carrier injection.
OTOH, I'm using them anyway in my new preamp, DAC, and power amp designs- just need to parallel a few. But you have to watch the surge. Easiest way to due this is to parallel with a "slow" silicon double diode set; during inrush surge, the power will be clamped/handled by the silicon; at normal operating conditions, only the SiC diode conducts.
Regards,
Jon
rtirion:
Your simulation is close enough; I provide the 5 pF through
stray capacitance, Muuhahahahahah!
everybody:
Also, the XA200 does use the output resistors to ground,
10 X 1000 ohms at 3 watts on each side.
The fast/soft diodes we are using came from STMicro, and
I don't think it matters a whole hell of a lot. I haven't
subscribed to this approach yet myself, but Wayne is
measuring better RF emissions, and both Wayne and
Joe Sammut insist it sounds better.
By this you may discern that mine are not the only ears
in the place, and this is probably a good thing.
Your simulation is close enough; I provide the 5 pF through
stray capacitance, Muuhahahahahah!
everybody:
Also, the XA200 does use the output resistors to ground,
10 X 1000 ohms at 3 watts on each side.
The fast/soft diodes we are using came from STMicro, and
I don't think it matters a whole hell of a lot. I haven't
subscribed to this approach yet myself, but Wayne is
measuring better RF emissions, and both Wayne and
Joe Sammut insist it sounds better.
By this you may discern that mine are not the only ears
in the place, and this is probably a good thing.
I imagine whether the RF issue is important will often boil down to the mechanical layout of the circuit--whether there's any shielding (or distance) between the power supply and the circuitry, for instance.
On the other hand, RF might spray back up the AC line and influence other components or it might...okay, okay, lemme think about this.
Grey
On the other hand, RF might spray back up the AC line and influence other components or it might...okay, okay, lemme think about this.
Grey