Hello gl
Still...mmm yes still looking for the right speaker, heard it on some kef 104.2 and sounded the best. So now trying to track some of these old speakers down.
But yes great sounding amp, just need to lower bais as running about 60 degrees and these speakers are 4 ohm so will make one half run hotter.
Cheers
Made any more changes to yours?
Still...mmm yes still looking for the right speaker, heard it on some kef 104.2 and sounded the best. So now trying to track some of these old speakers down.
But yes great sounding amp, just need to lower bais as running about 60 degrees and these speakers are 4 ohm so will make one half run hotter.
Cheers
Made any more changes to yours?
Dan,
Glad to hear things are sounding good. I haven't made any recent changes. I'm just listening.
Tad,
I'm very happy with the amps. The bass performance is excellent IMO and I have MG3.6's which are difficult to get good bass out of. The bass performance improved when the JFET front end went on but I couldn't tell you why.
Like Nelson said, Pass Labs made a generational change when moving from the XA line to the XA.5. They went to a follower type output topology among other things. I would expect serious improvement to the sound especially the bass. Still though, the XA line received very favorable reviews and is still super primo gear IMO.
I have only compared the AX100J's to my old Audio Research D115MkII and the AX100J's won hands down. I can't think of any existing DIY project out there that I would judge to be an upgrade. The Maggie's needs lots of power so I would need to build something substantial. The only thing that might do it for me is souping up the balanced version of the F4 to the 200-250 watt level. But there are other projects in the hopper first.
It's interesting to me that this thread continues to receive hits even though it's almost four years old and was started years after the main bubble of interest in the Aleph-X had subsided. The thread is mostly a rehash of stuff from other threads. It has no stars, and has not been selected by anyone as one their favorite threads. Yet the number of views is currently over 90K and growing. There aren't many threads here that get up to that number of views. So something is going on. I think it's a combination of a classic DIY design at a power level that is more than enough for most people's needs. And I think people are continuing to build the Aleph-X. I'm just happy to see that others are benefiting by what's here or being inspired by it.
OK, I'll shut up now before people start to think I'm a GR sock puppet.
Graeme
Glad to hear things are sounding good. I haven't made any recent changes. I'm just listening.
Tad,
I'm very happy with the amps. The bass performance is excellent IMO and I have MG3.6's which are difficult to get good bass out of. The bass performance improved when the JFET front end went on but I couldn't tell you why.
Like Nelson said, Pass Labs made a generational change when moving from the XA line to the XA.5. They went to a follower type output topology among other things. I would expect serious improvement to the sound especially the bass. Still though, the XA line received very favorable reviews and is still super primo gear IMO.
I have only compared the AX100J's to my old Audio Research D115MkII and the AX100J's won hands down. I can't think of any existing DIY project out there that I would judge to be an upgrade. The Maggie's needs lots of power so I would need to build something substantial. The only thing that might do it for me is souping up the balanced version of the F4 to the 200-250 watt level. But there are other projects in the hopper first.
It's interesting to me that this thread continues to receive hits even though it's almost four years old and was started years after the main bubble of interest in the Aleph-X had subsided. The thread is mostly a rehash of stuff from other threads. It has no stars, and has not been selected by anyone as one their favorite threads. Yet the number of views is currently over 90K and growing. There aren't many threads here that get up to that number of views. So something is going on. I think it's a combination of a classic DIY design at a power level that is more than enough for most people's needs. And I think people are continuing to build the Aleph-X. I'm just happy to see that others are benefiting by what's here or being inspired by it.
OK, I'll shut up now before people start to think I'm a GR sock puppet.
Graeme
Last edited:
Graeme,
I believe this threads popularity lies in its content. The OT crowd stayed away and let you dazzle us with your fine project. Mostly meat and potatoes no frill.
Thanks to you and of course Rollins and Nelson.
Tad
P.S maybe when you get time you could elaborate on the newer x.5 follower output scheme. Would be nice to do up to date cloning -- with permission of course.
I believe this threads popularity lies in its content. The OT crowd stayed away and let you dazzle us with your fine project. Mostly meat and potatoes no frill.
Thanks to you and of course Rollins and Nelson.
Tad
P.S maybe when you get time you could elaborate on the newer x.5 follower output scheme. Would be nice to do up to date cloning -- with permission of course.
Hi michaelbehrendt,Here a scematic idea, i don´t know this will work, have to changed the Fet´s to FQP3P20 and FQP4N20 and the bt to 2SC4793 and 2SA1837, who can answer.
Michael
I have been looking at your schematic up3ppse.pdf and like what I see, has this amp been built yet, and is it working. I need to know this before I begin to build my X100
Hi Tad,
I appreciate your comment regarding the OT posts. I would agree. I am not sure if I would use the word dazzle to describe anything about me.
That said though, it appears that my sniveling has resulted the thread gaining five stars at least for the moment!
The XA100.5 output section (each side) appears to be a scaled up version of the F4 without the input JFET's. The 5W of SE bias is provided by adding the appropriate resistive load from the output node to one of the power rails. The XA100.5 received a great deal of speculative attention on the Aleph-X builders thread between Jan 08 and Aug 08 driven primarily by one member. Here is one of the last schematics he posted:
http://www.diyaudio.com/forums/showthread.php?p=1584055
It's a relatively credible effort and includes lots of hints and tips from Nelson. On the down side there are number of confusing things shown and many missing values. And there is at least one thing that IMO will produce oscillation. Nevertheless, I believe that one of the experienced heavy-duty builders on this site could make this work and work well. Myself I'm not willing to lay out the time and money and take the risk.
The design referenced by scatterbrain001 has similar issues. For instance there is no way to adjust the output bias.
I generally don't attach any value to schematics unless they represent something that someone has built, and debugged, and liked the sound of.
Graeme
I appreciate your comment regarding the OT posts. I would agree. I am not sure if I would use the word dazzle to describe anything about me.
That said though, it appears that my sniveling has resulted the thread gaining five stars at least for the moment!
The XA100.5 output section (each side) appears to be a scaled up version of the F4 without the input JFET's. The 5W of SE bias is provided by adding the appropriate resistive load from the output node to one of the power rails. The XA100.5 received a great deal of speculative attention on the Aleph-X builders thread between Jan 08 and Aug 08 driven primarily by one member. Here is one of the last schematics he posted:
http://www.diyaudio.com/forums/showthread.php?p=1584055
It's a relatively credible effort and includes lots of hints and tips from Nelson. On the down side there are number of confusing things shown and many missing values. And there is at least one thing that IMO will produce oscillation. Nevertheless, I believe that one of the experienced heavy-duty builders on this site could make this work and work well. Myself I'm not willing to lay out the time and money and take the risk.
The design referenced by scatterbrain001 has similar issues. For instance there is no way to adjust the output bias.
I generally don't attach any value to schematics unless they represent something that someone has built, and debugged, and liked the sound of.
Graeme
Last edited:
Graeme,
Not sniveling just hard work. From what I have been reading Nelson has no patients for sniveling. He hints and cajoles but does not give it all up. Hell he might even decide to bread crumb us to x.5 and that one elusive resistor value you have wanted.
By the way are you real pleased with your maggies the pair only cost 599.00.
Tad
Not sniveling just hard work. From what I have been reading Nelson has no patients for sniveling. He hints and cajoles but does not give it all up. Hell he might even decide to bread crumb us to x.5 and that one elusive resistor value you have wanted.
By the way are you real pleased with your maggies the pair only cost 599.00.
Tad
Graeme,
I was talking about Magnapan's introduction model. That level is something I can justify. At my age I have become very frugal with money.
This forum allows a lot of us to obtain extremely high quality audio without the excessive out lay of cash. I could never see myself buying audio equipment today like I did 30 years ago.
Anyway I will quit OTing this thread.
Tad
I was talking about Magnapan's introduction model. That level is something I can justify. At my age I have become very frugal with money.
This forum allows a lot of us to obtain extremely high quality audio without the excessive out lay of cash. I could never see myself buying audio equipment today like I did 30 years ago.
Anyway I will quit OTing this thread.
Tad
Tad,
count yourself lucky, consider how much people elsewhere have to pay for Magnepan or any other US gear.
For example, a set of 3.6 Maggies retails for $10k overhere.
If i could get an MG3.6 set at US rate, use it for 15 years and then trade it in for $1k end value, then the annual write-off of my LSP system would be less than what i pay for a pair of shoes.
Money talk Off.
count yourself lucky, consider how much people elsewhere have to pay for Magnepan or any other US gear.
For example, a set of 3.6 Maggies retails for $10k overhere.
If i could get an MG3.6 set at US rate, use it for 15 years and then trade it in for $1k end value, then the annual write-off of my LSP system would be less than what i pay for a pair of shoes.
Money talk Off.
XA100-jfet option
Hi GL it has been a long time since I posted but I have recently retired and I now have more time to spend on Audio projects. I don't have to be on an airplane all the time.
I was thinking that I would like to re bias my XA100 to a new bias point reducing it down from about 5.5amps to 3.5 or so as I don't need this much class A power and it might help quiet down my trannies. I am thinking that a simple adjustment is all that is needed without any component changes. The voltages from the PS would remain the same. I want to use them on my ribbon tweeters which are crossed over at between 2100 to 2500khz at 60db using my DEQX digital crossover. I measured using my scope about 10volts p to p during a normal level listening session which is just a few watts. Let's say with some safety I could get by normally on 20watts class A with maybe peaks of 50watts (hard to imagine in less I got into the wine stock a little too much). I probably would use a 40uf caps on the outputs for tweeter protection (12 5" ribbon tweeters per side at over $100 bucks each is a reason to at least think about protecting them). Builiding a protection relay scheme is not to interesting. What do you think?
I was also considering building an F3. I remember that they may put out about 10 watts and this could be enough. I can't find the schematic on the F3 thread but it must be somewhere. I also thought about trying to build the Jan Didden design in the recent Audio express but it is not a balanced input.
Hope all is well with you and yours in this new year. dave
Hi GL it has been a long time since I posted but I have recently retired and I now have more time to spend on Audio projects. I don't have to be on an airplane all the time.
I was thinking that I would like to re bias my XA100 to a new bias point reducing it down from about 5.5amps to 3.5 or so as I don't need this much class A power and it might help quiet down my trannies. I am thinking that a simple adjustment is all that is needed without any component changes. The voltages from the PS would remain the same. I want to use them on my ribbon tweeters which are crossed over at between 2100 to 2500khz at 60db using my DEQX digital crossover. I measured using my scope about 10volts p to p during a normal level listening session which is just a few watts. Let's say with some safety I could get by normally on 20watts class A with maybe peaks of 50watts (hard to imagine in less I got into the wine stock a little too much). I probably would use a 40uf caps on the outputs for tweeter protection (12 5" ribbon tweeters per side at over $100 bucks each is a reason to at least think about protecting them). Builiding a protection relay scheme is not to interesting. What do you think?
I was also considering building an F3. I remember that they may put out about 10 watts and this could be enough. I can't find the schematic on the F3 thread but it must be somewhere. I also thought about trying to build the Jan Didden design in the recent Audio express but it is not a balanced input.
Hope all is well with you and yours in this new year. dave
Hi Dave,
It's good to hear from you. Congratulations on your retirement. It's a condition I look forward to in another 5 years or so if all goes well. It's also good to hear that your system continues to serve you well.
As far as the bias reduction goes, I would be inclined to remove output transistors rather than turn the bias down on the whole output section. Turning the overall bias down and keeping the current gain percentage where you have it set will require changing a number of resistor values.
The more elegant overall approach would be to build an F3 as you suggest. The schematic is on the Firstwatt site here:
http://www.firstwatt.com/downloads/F3-service-manual-sm.pdf
Things are going well here although 2009 was very brutal to my business. Things are turning up with the start of 2010. I hope my comments are helpful.
Cheers,
Graeme
It's good to hear from you. Congratulations on your retirement. It's a condition I look forward to in another 5 years or so if all goes well. It's also good to hear that your system continues to serve you well.
As far as the bias reduction goes, I would be inclined to remove output transistors rather than turn the bias down on the whole output section. Turning the overall bias down and keeping the current gain percentage where you have it set will require changing a number of resistor values.
The more elegant overall approach would be to build an F3 as you suggest. The schematic is on the Firstwatt site here:
http://www.firstwatt.com/downloads/F3-service-manual-sm.pdf
Things are going well here although 2009 was very brutal to my business. Things are turning up with the start of 2010. I hope my comments are helpful.
Cheers,
Graeme
ax100 revisions
GL thanks for your comments, after reflecting on this you are right and if things are working well leave enough alone. Noisy trannies irritate me and I have put caps across the hexfreds in the PS and added the chokes but still they wine. I found out that the avel lindberg trannies are made in China. This can be hit or miss on quality. I worked in China for a number of years and their quality is all over the place. This is why avel trannies are cheaper.
I even put in a dc blocking circuit in the ac input on the primary side but nothing has stopped it. Talking to Avel they say "oh welll stuff happens".
Thanks for the schematic, I will go back through my audio x press mags and find the old Nelson articles and read up on current source amps.
I should tell you that my dream retirement job is to be an tasting consultant and organizer at one of the big NAPA Valley vineyards. Hey I could ** all day and work on audio at night. The music would definetly sound better.
Or I could volunteer as a listening consultant at Pass Labs and help them all day by listening to their equipment and giving opinions. Just have to sell my house first. dave
GL thanks for your comments, after reflecting on this you are right and if things are working well leave enough alone. Noisy trannies irritate me and I have put caps across the hexfreds in the PS and added the chokes but still they wine. I found out that the avel lindberg trannies are made in China. This can be hit or miss on quality. I worked in China for a number of years and their quality is all over the place. This is why avel trannies are cheaper.
I even put in a dc blocking circuit in the ac input on the primary side but nothing has stopped it. Talking to Avel they say "oh welll stuff happens".
Thanks for the schematic, I will go back through my audio x press mags and find the old Nelson articles and read up on current source amps.
I should tell you that my dream retirement job is to be an tasting consultant and organizer at one of the big NAPA Valley vineyards. Hey I could ** all day and work on audio at night. The music would definetly sound better.
Or I could volunteer as a listening consultant at Pass Labs and help them all day by listening to their equipment and giving opinions. Just have to sell my house first. dave
Back in 2006 when this thread started and the Aleph-X was all the rage, everybody would have known the answer to this question. Some things do get lost in the mists of time.
It's a reasonable question however, so here goes.
AX100 - This is a 100W version of Grey Rollins original 38W Aleph-X project that is documented in the thread of the same name. In general all amplifiers here in this forum whose names start with the letters AX derive in some way from GR's original. In the case of the AX100 numerous changes have been made to bring it closer to the factory version. Most of these changes come from hints dropped by NP himself.
XA100 - This is a Pass commercial product. When referring to the XA series one usually uses that term or the term 'X Aleph'. The owners manual for this product is on the Pass website.
XA100.5 - This product replaced the XA100 and is a very different design. There are a number of threads discussing this amp.
There.
Graeme
It's a reasonable question however, so here goes.
AX100 - This is a 100W version of Grey Rollins original 38W Aleph-X project that is documented in the thread of the same name. In general all amplifiers here in this forum whose names start with the letters AX derive in some way from GR's original. In the case of the AX100 numerous changes have been made to bring it closer to the factory version. Most of these changes come from hints dropped by NP himself.
XA100 - This is a Pass commercial product. When referring to the XA series one usually uses that term or the term 'X Aleph'. The owners manual for this product is on the Pass website.
XA100.5 - This product replaced the XA100 and is a very different design. There are a number of threads discussing this amp.
There.
Graeme
Well for starters the first post in this thread states that the amplifiers run at 21 degrees C above ambient - not 45 to 50.
Second, you state in so many words that the mathematics indicate that the amplifiers shouldn't work. Yet you provide only a couple thermal resistance coefficients without stating where you found them or showing how you derived them.
And finally, the amplifiers work fine and have done so for over four years.
Here's some math: Taking one heat sink plate we have two transistors which input 12 watts of heat each for a total of 24 watts, and a temperature rise of 21 degrees C at the edge of the plate. So we have 21/24 = a T.R.C. of .87 (at the edge of the plate). Not that far away from your number of .7 actually. At the transistor the temperature measures about 7 or 8 degrees C higher.
However, the thermal gradient of the plate is quite poor. The key to success is keeping the heat dissipation on each transistor low, using lots of transistors, and spreading them out physically. Works like glue.
Graeme
Second, you state in so many words that the mathematics indicate that the amplifiers shouldn't work. Yet you provide only a couple thermal resistance coefficients without stating where you found them or showing how you derived them.
And finally, the amplifiers work fine and have done so for over four years.
Here's some math: Taking one heat sink plate we have two transistors which input 12 watts of heat each for a total of 24 watts, and a temperature rise of 21 degrees C at the edge of the plate. So we have 21/24 = a T.R.C. of .87 (at the edge of the plate). Not that far away from your number of .7 actually. At the transistor the temperature measures about 7 or 8 degrees C higher.
However, the thermal gradient of the plate is quite poor. The key to success is keeping the heat dissipation on each transistor low, using lots of transistors, and spreading them out physically. Works like glue.
Graeme
Sorry, I did not mean to offend.
I did the maths using two sides of each plate this time and your results are then very close.
The equation that I normally use is Tha = 50 / (Square root of Area (cm2))
So for a plate that is 17" x 8"; Area = 1745cm2
50 / (Square root 1745)) = 1.2 K/Watt
I'm planning on building an Aleph 4 but the cost of heatsinks is beating me. The Aleph 4 dissipates nearly 300W / Channel. This is spread over 12 devices so 24W / Device.
All 12 devices on one heatsink would require a heatsink of less than 0.15 K/Watt - ENORMOUS.
I did the maths using two sides of each plate this time and your results are then very close.
The equation that I normally use is Tha = 50 / (Square root of Area (cm2))
So for a plate that is 17" x 8"; Area = 1745cm2
50 / (Square root 1745)) = 1.2 K/Watt
I'm planning on building an Aleph 4 but the cost of heatsinks is beating me. The Aleph 4 dissipates nearly 300W / Channel. This is spread over 12 devices so 24W / Device.
All 12 devices on one heatsink would require a heatsink of less than 0.15 K/Watt - ENORMOUS.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.