I have tryied 2 different options: at the beginning at 0.22 V at the Source resistors having ca. 45 Degrees Celcius.
At the 0.45 Volt i had 54 Celcius and Very low Absolute DC (in the region of 10mV and the Retive witch i am not so sure about. It was 20mV at the one and 118mV at the other.
Should i let it be? What is more critical Excatly the same Voltage on the source resistors everywhere or small deviations but very good Relative DC?
What Watts output is it at 22 Volt rail and 0.45 Volt at the recistors and what at 0.3, 0.35 and 0.4 Volts? I what to see if it is what is enough for me. My speakers are going to have ca. 90 bB.
Thanks for your help and the knowledge you provide at everyone and Especially to Pass and Gray.
At the 0.45 Volt i had 54 Celcius and Very low Absolute DC (in the region of 10mV and the Retive witch i am not so sure about. It was 20mV at the one and 118mV at the other.
Should i let it be? What is more critical Excatly the same Voltage on the source resistors everywhere or small deviations but very good Relative DC?
What Watts output is it at 22 Volt rail and 0.45 Volt at the recistors and what at 0.3, 0.35 and 0.4 Volts? I what to see if it is what is enough for me. My speakers are going to have ca. 90 bB.
Thanks for your help and the knowledge you provide at everyone and Especially to Pass and Gray.
GRollins said:
billman said:
My comment was not directed at you. I never make fun of people who are new to electronics. It was more along the lines of implied criticism of people who make boards and profit from them, but do not support them. I was less than surprised to read that you received no help from the fellow who sold you the stuff.
As noted above, the original thread (which resurfaced recently and shouldn't be too hard to find) and this one have the answer to everything you might think to ask and the answers to some questions you might not have thought of.
All things being equal, 22V rails should give you something on the order of 100W. The quiescent bias doesn't effect wattage, per se, only how low an impedance you can drive.
The DC offset is a function of parts matching and how you set the pots. That's covered extensively in the original thread and probably this one as well.
I am somewhat concerned about the four electrolytics in a row towards the back of the preceding photograph. They appear to be sitting on top of other components. Are you certain that nothing is shorting out?
Grey
I have lowered the Bias on one side in order to have the same Relative DC and under 100 mV. Both channels sound the same and have the same loudness. Problem now is the following: When i connect RCA input i have a Buzz on the loudspeakers. When the RCA is not connected but only the speakers are on the amp i hear almost nothing.
Any ideas? Do i need some sort of a DC - Filter?
This is the inside of it:
Any ideas? Do i need some sort of a DC - Filter?
This is the inside of it:
An externally hosted image should be here but it was not working when we last tested it.
OK, today i bought Balance Cables and connect it with it. The loud Buzz is away and i only have a very low humm. Can i also trimm it away?
I suppose the loud buzz on the RCA is there but due to the balanced cables i dont hear it anymore. But was causing it? sould i check something?
So my target is If i make the RCA Buzz go away i suppose that the Overall humm (that i have alway when the Amp is on even without Preamp) would also go away.
I suppose the loud buzz on the RCA is there but due to the balanced cables i dont hear it anymore. But was causing it? sould i check something?
So my target is If i make the RCA Buzz go away i suppose that the Overall humm (that i have alway when the Amp is on even without Preamp) would also go away.
billman said:
Are you connecting the balanced or the RCA from the same source? If so, and if the noise disappears (or almost) when you use balanced, the cause is ground current between your source and your amp. Often it is just leakage current through the transformers in each equipment to their chassis. Then if you connect the RCA (which uses the chassis-to-chassis screen as ground) you can get this. This is really THE reason to use balanced interconnects.
If you want to get rid of the noise using the RCAs you can try running the source or the amp NOT connected to the mains ground. Note that potentially that may be unsafe.
Another cure can be to isolate the RCA from the chassis and connect the signal cable screen inside the case directly to ground near the signal input, on the pcb, and not to chassis.
As Hugo says, there are other factors involved like the internal equipment grounding setup.
Jan Didden
Hi, thanks for the reply. I understand what you are saying. My RCA-problem is not so serius because my PreAmp has Blanced out.
That what i must get rid off is this humm on the one channel. I heard it as if it has the same frequency (same frequency the Humm with the balanced on one channel with the buzz on both channels with RCA. I also made a Frequency pass of the Amp (like we do with speakers) but i cant read anything (maybe because i dont know on what to look).
That what i must get rid off is this humm on the one channel. I heard it as if it has the same frequency (same frequency the Humm with the balanced on one channel with the buzz on both channels with RCA. I also made a Frequency pass of the Amp (like we do with speakers) but i cant read anything (maybe because i dont know on what to look).
current source connections
Gl and GR--In converting my Aleph 1.2 to the XA100Plus-I noticed one thing that I did with the Aleph that I am intuitively doing with this makeover.
The connection at the node of the Current sourse output fet between the Source (Q10 & Q1) and the 1ohm resistor (R40 &R5) that goes back to the base of the current source sensing transistor (Q3&Q8) is taken from just one output fet. Each output fet is conntected to the output rail after the resistor. I am only sampling one fet and not all 6 ( which are not connected by a rail before the resistor) of each quad. This makes sense to me since you want to sample the 500ma to 600ma bias current using about +/-30volt dc rails and not the total 3.2 amps (in my case of total output bias)
This is the same for the driver fets where I sample from just one output driver fet back to the base of Q1&Q9. This has worked to set the bias current for all the fets.
Let me know how you guys did it. I have finished the boards and did my continuity tests and have rewired the outputs and installed the new trannies and my PS noise reduction circuit. I am still thinking about whether to have a CLCC supply or a CLCLC power supply to reduce ripple to anal levels. dave
Gl and GR--In converting my Aleph 1.2 to the XA100Plus-I noticed one thing that I did with the Aleph that I am intuitively doing with this makeover.
The connection at the node of the Current sourse output fet between the Source (Q10 & Q1) and the 1ohm resistor (R40 &R5) that goes back to the base of the current source sensing transistor (Q3&Q8) is taken from just one output fet. Each output fet is conntected to the output rail after the resistor. I am only sampling one fet and not all 6 ( which are not connected by a rail before the resistor) of each quad. This makes sense to me since you want to sample the 500ma to 600ma bias current using about +/-30volt dc rails and not the total 3.2 amps (in my case of total output bias)
This is the same for the driver fets where I sample from just one output driver fet back to the base of Q1&Q9. This has worked to set the bias current for all the fets.
Let me know how you guys did it. I have finished the boards and did my continuity tests and have rewired the outputs and installed the new trannies and my PS noise reduction circuit. I am still thinking about whether to have a CLCC supply or a CLCLC power supply to reduce ripple to anal levels. dave
Hi Dave,
The feedback is taken from only one output transistor in the case of the Aleph CCS or the gain transistor. This is the way everyone has done it. The others transistors are simply connected in parallel and if they're matched they will follow along. As you say the original Aleph's worked the same way.
My CLC filters produce a dead quiet result.
Cheers,
Graeme
P.S. I still haven't decided if I'm going to the Burning Amp event but I'll let you know.
The feedback is taken from only one output transistor in the case of the Aleph CCS or the gain transistor. This is the way everyone has done it. The others transistors are simply connected in parallel and if they're matched they will follow along. As you say the original Aleph's worked the same way.
My CLC filters produce a dead quiet result.
Cheers,
Graeme
P.S. I still haven't decided if I'm going to the Burning Amp event but I'll let you know.
These amplifiers will be at the Burning Amp event in San Francisco next Sunday. Another member has graciously offered to carpool and the final details are being worked out.
The amps are essentially unchanged and no attempt will be made to spiff them up. I have not swapped out the old junk box phono connectors and binding posts for the Vampire and Cardas units I said I was going to install. I've been too busy listening. This is inexcusable considering that Michael Percy lives within walking distance of my house. I have added Hansen snubbers to the power supply and I've removed the input caps. The jury is still out on these mods.
Graeme
The amps are essentially unchanged and no attempt will be made to spiff them up. I have not swapped out the old junk box phono connectors and binding posts for the Vampire and Cardas units I said I was going to install. I've been too busy listening. This is inexcusable considering that Michael Percy lives within walking distance of my house. I have added Hansen snubbers to the power supply and I've removed the input caps. The jury is still out on these mods.
Graeme
burning amp event
GL-do you have the address or website for this years event which shows the times and places? Perhaps I can talk my son into going there for a few hours to visit. He lives near Mountain View, Ca.
So there really is a Michael Percy? I bought stuff from him back in the 80's. I thought he retired to audiophile heaven and was listening to Radio Shack equipment-lafeyette models. Is he still in
business? His adds have disappeared, maybe due to the fact he is extremely wealthy now? Say hello to him for me. dave
GL-do you have the address or website for this years event which shows the times and places? Perhaps I can talk my son into going there for a few hours to visit. He lives near Mountain View, Ca.
So there really is a Michael Percy? I bought stuff from him back in the 80's. I thought he retired to audiophile heaven and was listening to Radio Shack equipment-lafeyette models. Is he still in
business? His adds have disappeared, maybe due to the fact he is extremely wealthy now? Say hello to him for me. dave
Hi,
I've finished one alephX
I've put a 4 khz square signal ( from a lm555cd with 10 ns rising up).
Here's the alephX out signal :
It seems I have a lineic inductance somewhere ? ( my fet are on external pcb with r5 and r5, r2 and r3 are on the fet pcb too). My power filter is a cLc.
Listening the alephX is ok on my cheap loudspeakers.
I've finished one alephX
An externally hosted image should be here but it was not working when we last tested it.
I've put a 4 khz square signal ( from a lm555cd with 10 ns rising up).
Here's the alephX out signal :
An externally hosted image should be here but it was not working when we last tested it.
It seems I have a lineic inductance somewhere ? ( my fet are on external pcb with r5 and r5, r2 and r3 are on the fet pcb too). My power filter is a cLc.
Listening the alephX is ok on my cheap loudspeakers.
aleph 1.2's to Ax100
Gl & GR--one last design issue for my conversion from Aleph 1.2's to the XA100+. Did you guys put in output fuses or a dc relay etc? In my Aleph 1.2's I did not use output speaker fuses just the AC fuse on the input. With the XA100 designs having a push pull output, I am afraid to do that since 1/2 the amp will be still putting out ac signals should one fuse blow.
I never fuse the power rails near the capcitors without a dc relay protecting the speaker or else you will have a lot of dc on the outputs maybe frying a woofer.
Since I will be using these amps in a bi-amp mode to drive my ribbon tweeters, I was going to put 20uf caps on the input to the +/- connectors on my tweeters since these amps are fully balanced and this should protect the tweeters from any dc getting to them if something should happen.
When used with my older set of speakers with a passive crossover only the woofs would be at risk from any failure in the amp. I had a volume control on my Aleph's which doubled as a mute switch to protect the inputs. I have removed these but I wonder if I should have left them there. They were always set to full on. I have never had output devices fail and I am careful about starting up the amp after all other components are operating. So to fuse or not to fuse the outputs? Thanks in advance for your opinion. I turn on the amps tonight for testing.
dave
Gl & GR--one last design issue for my conversion from Aleph 1.2's to the XA100+. Did you guys put in output fuses or a dc relay etc? In my Aleph 1.2's I did not use output speaker fuses just the AC fuse on the input. With the XA100 designs having a push pull output, I am afraid to do that since 1/2 the amp will be still putting out ac signals should one fuse blow.
I never fuse the power rails near the capcitors without a dc relay protecting the speaker or else you will have a lot of dc on the outputs maybe frying a woofer.
Since I will be using these amps in a bi-amp mode to drive my ribbon tweeters, I was going to put 20uf caps on the input to the +/- connectors on my tweeters since these amps are fully balanced and this should protect the tweeters from any dc getting to them if something should happen.
When used with my older set of speakers with a passive crossover only the woofs would be at risk from any failure in the amp. I had a volume control on my Aleph's which doubled as a mute switch to protect the inputs. I have removed these but I wonder if I should have left them there. They were always set to full on. I have never had output devices fail and I am careful about starting up the amp after all other components are operating. So to fuse or not to fuse the outputs? Thanks in advance for your opinion. I turn on the amps tonight for testing.
dave
Hi Dave,
I don't use output fuses or relays and I don't believe in them - at least not as an amplifier feature. If you want to put an optional fuse on your tweeters then fine.
I think your idea of using a coupling cap is fine. IIRC there was already a discussion on this or another thread on your use of this 20uF cap at this location. It will protect your tweeters against DC but not from transients. There is a lot of good discussion on the MUG forum regarding Magnepan ribbon tweeters and fuses. Lots of opinion!
I don't recall anyone using any fuses in an Aleph-X except for the mains.
Regards,
Graeme
I don't use output fuses or relays and I don't believe in them - at least not as an amplifier feature. If you want to put an optional fuse on your tweeters then fine.
I think your idea of using a coupling cap is fine. IIRC there was already a discussion on this or another thread on your use of this 20uF cap at this location. It will protect your tweeters against DC but not from transients. There is a lot of good discussion on the MUG forum regarding Magnepan ribbon tweeters and fuses. Lots of opinion!
I don't recall anyone using any fuses in an Aleph-X except for the mains.
Regards,
Graeme
aleph 1.2's to Ax100 testing 1st day
GL and GREY, well my first day of testing and here are some initial results.
I measured the voltage across the source resistors 24 in all and 6 fets per quad. After warm up they were between .450 and .601mv divided by 1.2 ohms or about 400ma to 500ma per device x 6 outputs = about 3 amps ( a little less) per rail or 6 amps total with a rail voltage of +/- 29 volts.
differential DC off set when warm is less than 10mv but when cold it is less than 100mv.
I haven't tracked dc offset vs ground across the 100ohm resistors but a quick measurement was around 8 volts after 1/2 hour.
However, my air coiled inductors in the PS got so hot that they melted the plastic covers on my ps caps (where they inadvertantly were touching) and just about put my test bench on fire so I had to remove them. I left the iron core inductors in the power supply and they got hot and started smoking when I did signal testing--full power for a few minutes. My heatsinks got too hot to touch so either I will have to get more heat sink or put in a fan or reduce the bias a bit. Minor problems to fix.
I did some signal testing in the single ended mode and it was perfect with my scope connected to the positive output only and with a 5ohm 250 watt resistor loading the output.
Obviously I have more to do, I have not set the ccs gain and I am using the resistor value that you used for now. The bias resistor is 169k which is quite a bit higher than the 100k you used but I have more devices and a higher rail voltage. I had no ripple with both inductors in the circuit and I haven't measured with just the one. I will do that later after I figure out how to keep this monster cool.
The 1kva trannie groans a little bit but not as loud as the 2kva Aleph trannies. More to come. dave
GL and GREY, well my first day of testing and here are some initial results.
I measured the voltage across the source resistors 24 in all and 6 fets per quad. After warm up they were between .450 and .601mv divided by 1.2 ohms or about 400ma to 500ma per device x 6 outputs = about 3 amps ( a little less) per rail or 6 amps total with a rail voltage of +/- 29 volts.
differential DC off set when warm is less than 10mv but when cold it is less than 100mv.
I haven't tracked dc offset vs ground across the 100ohm resistors but a quick measurement was around 8 volts after 1/2 hour.
However, my air coiled inductors in the PS got so hot that they melted the plastic covers on my ps caps (where they inadvertantly were touching) and just about put my test bench on fire so I had to remove them. I left the iron core inductors in the power supply and they got hot and started smoking when I did signal testing--full power for a few minutes. My heatsinks got too hot to touch so either I will have to get more heat sink or put in a fan or reduce the bias a bit. Minor problems to fix.
I did some signal testing in the single ended mode and it was perfect with my scope connected to the positive output only and with a 5ohm 250 watt resistor loading the output.
Obviously I have more to do, I have not set the ccs gain and I am using the resistor value that you used for now. The bias resistor is 169k which is quite a bit higher than the 100k you used but I have more devices and a higher rail voltage. I had no ripple with both inductors in the circuit and I haven't measured with just the one. I will do that later after I figure out how to keep this monster cool.
The 1kva trannie groans a little bit but not as loud as the 2kva Aleph trannies. More to come. dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.