The new batch won't necessarily match up with the old ones. Consider them as a separate group.
.01V is a good start. See how things work out.
Incidentally, the MOSFETs run at ~10mA each. The 20mA figure is for the current source, but that's divided by the two sides of the differential.
Grey
.01V is a good start. See how things work out.
Incidentally, the MOSFETs run at ~10mA each. The 20mA figure is for the current source, but that's divided by the two sides of the differential.
Grey
aleph 1.2 to XA100
GL and Grey, thanks, after reading much of the current threads relating to dc offset on the xa and x designs, I am trying to keep my trouble shooting to a minimum. My Aleph 1.2's run at about 40mv dc offset cold and that goes down a bit after they warm up.
They never have exceeded this number and since NP advertised the design specs at <100mv for the Aleph series I never tried to mess with it. I think that I went with matched devices at .01 or less but I can't find my notes. I built the aleph's seven years ago.
I am trying to match all the 9610's per mono amp.
My other input design issue is a volume control. I have an Alps stereo volume control on each mono aleph and trimmed the input of the controls so that they matched. I connected the amps in balanced mode. I did this because I was on an assignment in Puerto Rico and my Sony SACD/CD player could drive the amps directly. I left my analog stuff home.
I want to remove this volume control once the amps are tested
and my preamp has an output impedance of <100ohms and drives about 20 feet of silver sonic interconnect to where the amps are going to be located. I am leaving the 10k input gate resistors in but I am removing the 4.7uf coupling cap and 100k resistors that GL installed. I hope I don't run into trouble with the 10k input imedance in balanced mode. More to come-dave
GL and Grey, thanks, after reading much of the current threads relating to dc offset on the xa and x designs, I am trying to keep my trouble shooting to a minimum. My Aleph 1.2's run at about 40mv dc offset cold and that goes down a bit after they warm up.
They never have exceeded this number and since NP advertised the design specs at <100mv for the Aleph series I never tried to mess with it. I think that I went with matched devices at .01 or less but I can't find my notes. I built the aleph's seven years ago.
I am trying to match all the 9610's per mono amp.
My other input design issue is a volume control. I have an Alps stereo volume control on each mono aleph and trimmed the input of the controls so that they matched. I connected the amps in balanced mode. I did this because I was on an assignment in Puerto Rico and my Sony SACD/CD player could drive the amps directly. I left my analog stuff home.
I want to remove this volume control once the amps are tested
and my preamp has an output impedance of <100ohms and drives about 20 feet of silver sonic interconnect to where the amps are going to be located. I am leaving the 10k input gate resistors in but I am removing the 4.7uf coupling cap and 100k resistors that GL installed. I hope I don't run into trouble with the 10k input imedance in balanced mode. More to come-dave
aleph 1.2's to Ax100
GL or Grey, I have been thinking of ways to shorten my wires etc in my Aleph 1.2 to get the cleanest grounds and least noise pickup.
I will take a picture of my current amp layout as soon as my wife gets her digital camera back from repair. I have the traditional horizontal layout (like Krells or Marsh amps). Transforme in front and driver circuit board way in the back (2 feet away) with 8 ps supply caps inbetween and 24 fets (12 per side per mono amp) on the heat sinks mounted on the sides of the chassis.
I got an email from NP 6 years ago giving me a tip on how to connect all the ground leads to the Center tap to minimize noise.
My Chasis is about 7" tall and the caps are set vertically so I don't have a lot of clearance between the top cover and cap connections.
I noticed by looking in the 350.5 through the cover that NP does not use an Inductor. How does he get his commercial amps to sound clean without one?? I would like to remove the inductor if possible. I did notice that he has more of a vertical arrangement with the caps right above the tranny for shorter wiring lenths.
I have 8 - 14k uf sangamo CG PS caps per mono amp (currently in my Aleph 1.2's). This should be enough capacitance to eliminate hum. I think I measured less than .6v ripple without the Inductor years ago. With two 1uf inductors I get virtually no ripple but I still get buzzing at the speakers (low level and not evident at the listening position from my current Alephs's). Any suggestions for a lower noise line? I am going to use standard thick wire for the power supply upgrade for the XA100 conversions. I have some silver wire from other projects to use as connecting wire at the circuit board etc. dave
GL or Grey, I have been thinking of ways to shorten my wires etc in my Aleph 1.2 to get the cleanest grounds and least noise pickup.
I will take a picture of my current amp layout as soon as my wife gets her digital camera back from repair. I have the traditional horizontal layout (like Krells or Marsh amps). Transforme in front and driver circuit board way in the back (2 feet away) with 8 ps supply caps inbetween and 24 fets (12 per side per mono amp) on the heat sinks mounted on the sides of the chassis.
I got an email from NP 6 years ago giving me a tip on how to connect all the ground leads to the Center tap to minimize noise.
My Chasis is about 7" tall and the caps are set vertically so I don't have a lot of clearance between the top cover and cap connections.
I noticed by looking in the 350.5 through the cover that NP does not use an Inductor. How does he get his commercial amps to sound clean without one?? I would like to remove the inductor if possible. I did notice that he has more of a vertical arrangement with the caps right above the tranny for shorter wiring lenths.
I have 8 - 14k uf sangamo CG PS caps per mono amp (currently in my Aleph 1.2's). This should be enough capacitance to eliminate hum. I think I measured less than .6v ripple without the Inductor years ago. With two 1uf inductors I get virtually no ripple but I still get buzzing at the speakers (low level and not evident at the listening position from my current Alephs's). Any suggestions for a lower noise line? I am going to use standard thick wire for the power supply upgrade for the XA100 conversions. I have some silver wire from other projects to use as connecting wire at the circuit board etc. dave
Hi Dave,
Grounding is one of those black art things. Nelson Pass has given out tons of info on this subject. Read through his projects at passdiy again looking for grounding tips and ground loop info. In your case you're building monoblocks which will make life easier than building multiple channels in one chassis.
I have no hum or noise. You could re-read my notes about this on page 1 of this thread and also look at the pictures again.
As for the inductors; you don't really really need them. Most commercial amps don't have them. I personally think that they improve the soundstaging and reduce grain. Most of the value of the inductors can be had by using power resistors of approximately the same DC resistance value - about .2 to .5 ohms I would guess.
Cheers,
Graeme
Grounding is one of those black art things. Nelson Pass has given out tons of info on this subject. Read through his projects at passdiy again looking for grounding tips and ground loop info. In your case you're building monoblocks which will make life easier than building multiple channels in one chassis.
I have no hum or noise. You could re-read my notes about this on page 1 of this thread and also look at the pictures again.
As for the inductors; you don't really really need them. Most commercial amps don't have them. I personally think that they improve the soundstaging and reduce grain. Most of the value of the inductors can be had by using power resistors of approximately the same DC resistance value - about .2 to .5 ohms I would guess.
Cheers,
Graeme
Aleph 1.2 to AX100
GR, I just soldered my diff front fets to the board. I measured the GS voltage with the input fets attached to the same heatsink to stabalize the temperature. I checked it at about 72 degrees.
One set measured 3.901 for both and the other set there is a .003v difference, A change is temperature will be enought to affect this meausrement. Did you measure your with 10ma of current? I did it using a 9volt dc battery and sized the source reistor for 10ma. I got good luck from a group of 16 fets. dave
GR, I just soldered my diff front fets to the board. I measured the GS voltage with the input fets attached to the same heatsink to stabalize the temperature. I checked it at about 72 degrees.
One set measured 3.901 for both and the other set there is a .003v difference, A change is temperature will be enought to affect this meausrement. Did you measure your with 10ma of current? I did it using a 9volt dc battery and sized the source reistor for 10ma. I got good luck from a group of 16 fets. dave
hi guys!
i haven't been visiting this forum for about one month 'cause i've been in vacation.
yeaahhh!
I visited California, New york and so on....
i have noticed than newer versions of the aleph ax or whatever weren't released.
Did i miss something or nelson decided to not up-to-date the "old" Aleph X?
I was excited to see a new version of the aleph XA or i don't know a new version for the preamplifier...or just sooomethingg
i haven't been visiting this forum for about one month 'cause i've been in vacation.
yeaahhh!
I visited California, New york and so on....
i have noticed than newer versions of the aleph ax or whatever weren't released.
Did i miss something or nelson decided to not up-to-date the "old" Aleph X?
I was excited to see a new version of the aleph XA or i don't know a new version for the preamplifier...or just sooomethingg
aleph 1.2's to Ax100
GL, I am just finishing the boards and after I return from a week's vacation I will be ready for some testing. I just need some hints.
I will reread NP's article on the Aleph ccs in one of his zen articles but I have the following situation:
1kva 25volt ct transformers per channel that put out 35 volts per rail dc no load (1.4 x 25). I would expect a bit higher than this due to the xformer regulation. That means after losses due to the 2uh Inductor and diode losses I expect a solid 30-32 volt rail voltage. I also have 6 output devices per quadrant.
If I run 540ma per device with 12 devices that is approximately 6.5amps per rail. What would be a good guess for R11 & R33 assuming that I would have to increase them from the 100k to take account of the increased voltage and extra device? I am getting lazy and taking the board in and out after hooking up the power supply and the output section is a pain. I relooked at the Aleph 1.2's schematic and it uses +/- 60 volt rails and R11&33 are lower.
I guess I could try to calculate it if I knew the ma draw of the input differential pair?
Or I could given the fact the trannies have 20 amps per rail available go to 600ma per device. I have enough heatsinking. What would be your guess on this or would I still have to do the ac test replacing R14 &R32 and measure at 50%?
I could test the boards out of the circuit as I have 4 leftover output devices. I would not use an inductor but could find a resistor to mimic the voltage drop. I have enough old caps lying around and a variac. Obviously I would be putting 6.5 amps through 4 not 12 devices but it would be just for short term testing.
I wouldn't go much over 600ma due to the heat.
I will post some pictures tonight of my Aleph's and the boards under construction. I am doing pt to pt and I will upgrade the chasis front plate with a thicker aluminum. The amps are already
gargantuan. Thanks for the help--dave
GL, I am just finishing the boards and after I return from a week's vacation I will be ready for some testing. I just need some hints.
I will reread NP's article on the Aleph ccs in one of his zen articles but I have the following situation:
1kva 25volt ct transformers per channel that put out 35 volts per rail dc no load (1.4 x 25). I would expect a bit higher than this due to the xformer regulation. That means after losses due to the 2uh Inductor and diode losses I expect a solid 30-32 volt rail voltage. I also have 6 output devices per quadrant.
If I run 540ma per device with 12 devices that is approximately 6.5amps per rail. What would be a good guess for R11 & R33 assuming that I would have to increase them from the 100k to take account of the increased voltage and extra device? I am getting lazy and taking the board in and out after hooking up the power supply and the output section is a pain. I relooked at the Aleph 1.2's schematic and it uses +/- 60 volt rails and R11&33 are lower.
I guess I could try to calculate it if I knew the ma draw of the input differential pair?
Or I could given the fact the trannies have 20 amps per rail available go to 600ma per device. I have enough heatsinking. What would be your guess on this or would I still have to do the ac test replacing R14 &R32 and measure at 50%?
I could test the boards out of the circuit as I have 4 leftover output devices. I would not use an inductor but could find a resistor to mimic the voltage drop. I have enough old caps lying around and a variac. Obviously I would be putting 6.5 amps through 4 not 12 devices but it would be just for short term testing.
I wouldn't go much over 600ma due to the heat.
I will post some pictures tonight of my Aleph's and the boards under construction. I am doing pt to pt and I will upgrade the chasis front plate with a thicker aluminum. The amps are already
gargantuan. Thanks for the help--dave
Hi Dave,
Sounds like you're making excellent progress.
1) I would recommend putting in the original Aleph-X resistor and pot in place of R11 and R33 then adjusting until you get the right bias.
2) Yes, you absolutely do need to do the deal on dialing in the 50% current gain. You've changed too many things from my amp to use that for anything other than a rough guide. For starters your source resistors are a different by a fair chunk IIRC. And you're using 6 transistors instead of my 5. The total transconductance of the CCS is significant and it's not in the equation in the patent. This really is a measure to fit deal here. Ya gotta do it.
3) Take your time and be patient even if it means dis-assembling and re-assembling a few times.
Regards,
Graeme
Sounds like you're making excellent progress.
1) I would recommend putting in the original Aleph-X resistor and pot in place of R11 and R33 then adjusting until you get the right bias.
2) Yes, you absolutely do need to do the deal on dialing in the 50% current gain. You've changed too many things from my amp to use that for anything other than a rough guide. For starters your source resistors are a different by a fair chunk IIRC. And you're using 6 transistors instead of my 5. The total transconductance of the CCS is significant and it's not in the equation in the patent. This really is a measure to fit deal here. Ya gotta do it.
3) Take your time and be patient even if it means dis-assembling and re-assembling a few times.
Regards,
Graeme
aleph 1.2's to Ax100
GL, one last question in a universe full of them. I have 1.5 ohm source resistors on the outputs of my Aleph 1.2. I remember that you said that the CCS may have trouble being set to 540ma with these and you suggested that I parallel them with another 1.5 since I still have 50 of them lying around. Is this still your thinking? 600ma x 1.5 ohms is .9volts across the resistors. Using .75ohms x 600ma =.45volts which is exactly the what I measure on my Alephs with 24 devices (.5volts/1.5ohms=330ma per device).
GL, one last question in a universe full of them. I have 1.5 ohm source resistors on the outputs of my Aleph 1.2. I remember that you said that the CCS may have trouble being set to 540ma with these and you suggested that I parallel them with another 1.5 since I still have 50 of them lying around. Is this still your thinking? 600ma x 1.5 ohms is .9volts across the resistors. Using .75ohms x 600ma =.45volts which is exactly the what I measure on my Alephs with 24 devices (.5volts/1.5ohms=330ma per device).
aleph 1.2's to Ax100
GL and Grey, I am on vacation this week in SF to visit my son. Too bad NP doesn't do factory tours!!!!!
When I return I will be getting ready to test my driver boards for the AX 100 and if you could comment on my test procedure, I would appreciate it.
I have not disassembled by Aleph 1.2's and I thought I would test the boards out of the circuit.
1. I have 4 output mosfets remaining as spares from my aleph projects. I could hook them up as one per quadrant on a heat sink with a large adjustable fan cooling them. I want to run 3.2 amps through each quadrant. I can use my new 1kva trannies and I have some caps and a variac. I wouldn't connect ps inductor but I could if you think I need it.
2. I would attach the boards to the output and the heatsink. The boards have 2200uf of local ps caps and politically appropriate by passing. I have some very expensive minature linear pots at 100k and 50k. I can use GL's method of adjusting the bias. I have installed his 100k resistor for 540ma but that was based on 5 devices, I have 6 so I think the resistor value will need to go up and I want to run 600ma per device times 6 devices is 3.2 amps total per quad. So I will adjust the resistor R11 and R33 on Gl's schematic until I get 3.2 amps using the pot in series with the 100k resistor (if this doesn't work and then I will go down in value).
2. For adjusting the 50% gain I have again some very expensive 1k and 5k linear pots. I would just wire these in the board and adjust using these. In Gl's notes he says that the 1K and 475 ohm resistors were substitued for R12 and R34. These resistors were "disconnected" and I take that to me "this resistor spot was left open". This also means that C1&C6 are out of the circuit. With R12 and R34 OPEN AC gain was measured across the source resistors and noted based on a 100hz sine wave. The resistors were then reinserted and the ac gain adusted until the scope showed a 50% reduction in the previous value. AM I CORRECT? This will be done after the dc bias was set using the 4 fets at 3.2amps and allowed to stabilize? Of course I would have the 100 ohm dc to ground reistors in the circuit
3. Do you think I can ge close in the test jig or should I abandon this approach? Is there any chance I can blow up 24 matched fets if I do the in circuit test? I have a big powersupply that can run a welder. The main reason for testing this way is to make sure that I have eliminated any and all mistakes in constructing the board (pt to pt). I will do continuity tests and visual inspection but one cold solder joint could sent me out drinking.
4. Has anyone ever left the bias and ac gain pots in the circuit after finishing the adjustments? I know this sounds like heresey.
dave
GL and Grey, I am on vacation this week in SF to visit my son. Too bad NP doesn't do factory tours!!!!!
When I return I will be getting ready to test my driver boards for the AX 100 and if you could comment on my test procedure, I would appreciate it.
I have not disassembled by Aleph 1.2's and I thought I would test the boards out of the circuit.
1. I have 4 output mosfets remaining as spares from my aleph projects. I could hook them up as one per quadrant on a heat sink with a large adjustable fan cooling them. I want to run 3.2 amps through each quadrant. I can use my new 1kva trannies and I have some caps and a variac. I wouldn't connect ps inductor but I could if you think I need it.
2. I would attach the boards to the output and the heatsink. The boards have 2200uf of local ps caps and politically appropriate by passing. I have some very expensive minature linear pots at 100k and 50k. I can use GL's method of adjusting the bias. I have installed his 100k resistor for 540ma but that was based on 5 devices, I have 6 so I think the resistor value will need to go up and I want to run 600ma per device times 6 devices is 3.2 amps total per quad. So I will adjust the resistor R11 and R33 on Gl's schematic until I get 3.2 amps using the pot in series with the 100k resistor (if this doesn't work and then I will go down in value).
2. For adjusting the 50% gain I have again some very expensive 1k and 5k linear pots. I would just wire these in the board and adjust using these. In Gl's notes he says that the 1K and 475 ohm resistors were substitued for R12 and R34. These resistors were "disconnected" and I take that to me "this resistor spot was left open". This also means that C1&C6 are out of the circuit. With R12 and R34 OPEN AC gain was measured across the source resistors and noted based on a 100hz sine wave. The resistors were then reinserted and the ac gain adusted until the scope showed a 50% reduction in the previous value. AM I CORRECT? This will be done after the dc bias was set using the 4 fets at 3.2amps and allowed to stabilize? Of course I would have the 100 ohm dc to ground reistors in the circuit
3. Do you think I can ge close in the test jig or should I abandon this approach? Is there any chance I can blow up 24 matched fets if I do the in circuit test? I have a big powersupply that can run a welder. The main reason for testing this way is to make sure that I have eliminated any and all mistakes in constructing the board (pt to pt). I will do continuity tests and visual inspection but one cold solder joint could sent me out drinking.
4. Has anyone ever left the bias and ac gain pots in the circuit after finishing the adjustments? I know this sounds like heresey.
dave
Hi Dave,
It's too bad you aren't visiting your son October 22-23. I'm thinking about taking the AX100's to the burning amp deal in SF.
OK here goes: (I have included 2 twos like you did - for clarity.)
1) I have no informed opinions to offer on the merit of this idea. If it was me I wouldn't bother.
2) I would recommend again to go back to the pot/resistor combination used in the original Aleph-X. The small trimmer pots used by everyone are available from everywhere and are very cheap.
2) Your interpretation of my comments is correct. I didn't write that description very well. The method was given on this forum by NP a long time ago. The circuit was first measured with R12 and R34 removed - and yes the caps were left open. The 1K pot and 475 ohm resistor in series with it were inserted in place of R12 and R34 and the measurement taken again - this time adjusting the pots.
3) For myself - I would just go with amp as built. In fact that's what I did do!
4) I think that's what most people do.
Cheers,
Graeme
It's too bad you aren't visiting your son October 22-23. I'm thinking about taking the AX100's to the burning amp deal in SF.
OK here goes: (I have included 2 twos like you did - for clarity.)
1) I have no informed opinions to offer on the merit of this idea. If it was me I wouldn't bother.
2) I would recommend again to go back to the pot/resistor combination used in the original Aleph-X. The small trimmer pots used by everyone are available from everywhere and are very cheap.
2) Your interpretation of my comments is correct. I didn't write that description very well. The method was given on this forum by NP a long time ago. The circuit was first measured with R12 and R34 removed - and yes the caps were left open. The 1K pot and 475 ohm resistor in series with it were inserted in place of R12 and R34 and the measurement taken again - this time adjusting the pots.
3) For myself - I would just go with amp as built. In fact that's what I did do!
4) I think that's what most people do.
Cheers,
Graeme
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.