It sprays in both directions, but the the AC line is the
place where various countries have decided to implement
trade protection via emissions regulations.
I've never seen much issue on the circuit end, as the Aleph
and X mosfet circuits seem immune to RF. We have gone
so far as to place a transmitter on top of the amplifier without
being able to see artifacts.
Great design, or slow circuit?
place where various countries have decided to implement
trade protection via emissions regulations.
I've never seen much issue on the circuit end, as the Aleph
and X mosfet circuits seem immune to RF. We have gone
so far as to place a transmitter on top of the amplifier without
being able to see artifacts.
Great design, or slow circuit?
Great design but problem may be elsewhere
NP,
I would say GREAT.
However, I suspect the problem often surfaces elsewhere. If an amp puts out a lot of noise on the net, lesser components may be harmed. I have a Threshold Stasis 3 which greatly benefited from a cheap RF filter on the CD player ...
The Stax headphone amp did not seem to benefit at all.
I suspect that noise generated by the CD player on the line caused the Stasis 3 to work extra ultrasonic duty, but that the Stax amp had a more substantial power supply (for size) and was either immune or did not care too much. It is also possible that the Stax cabinet was not grounded by 10 ohms or so ...
Anyway, I am pleased that you changed to soft recovery diodes, but suspect your customers have high end anciliarly equipment and will be less affected than us el-cheapo dudes.
Petter
NP,
I would say GREAT.
However, I suspect the problem often surfaces elsewhere. If an amp puts out a lot of noise on the net, lesser components may be harmed. I have a Threshold Stasis 3 which greatly benefited from a cheap RF filter on the CD player ...
The Stax headphone amp did not seem to benefit at all.
I suspect that noise generated by the CD player on the line caused the Stasis 3 to work extra ultrasonic duty, but that the Stax amp had a more substantial power supply (for size) and was either immune or did not care too much. It is also possible that the Stax cabinet was not grounded by 10 ohms or so ...
Anyway, I am pleased that you changed to soft recovery diodes, but suspect your customers have high end anciliarly equipment and will be less affected than us el-cheapo dudes.
Petter
Nelson Pass said:It sprays in both directions, but the the AC line is the
place where various countries have decided to implement
trade protection via emissions regulations.
I've never seen much issue on the circuit end, as the Aleph
and X mosfet circuits seem immune to RF. We have gone
so far as to place a transmitter on top of the amplifier without
being able to see artifacts.
Great design, or slow circuit?
A slow circuit can be potential unstable when subjected to rf! Like dc offset and so on...
Am i nice or what?? It is the same as saying 'Great design'
1000 X 10 = 100 ???
Dear Nelson thanks for giving your art to do the magic recipie!The pie is hot...and I think it will be VERRRY good!But is it possible that the amp needs the 100R value to get really stable DC offset?And not 30 ohms?
Possibly build a pair for extensive listening!Time,always time...and money!!!...
By the way.I see a +/-24V supply for my prototype;a little over 1 A by transistor (the 240s,I have tons of surplus from digikey...only need a match),and total 8 IRFP240 for the bridged version.Total bias will be 4 A, 2 by side.Gain will be 20dB balanced,I think 26dB is too much for me.Should be better for phase margin too.
So I will get a little more than 100W in 8 ohms.But...imagine I want to hook up with speakers at 2 ohms impedance...will the amp do it?And what about the double current source behavior?Do I still have to give to the current source 1/2 the output variation? 
Big thanks dear Nelson...
Dear Nelson thanks for giving your art to do the magic recipie!The pie is hot...and I think it will be VERRRY good!But is it possible that the amp needs the 100R value to get really stable DC offset?And not 30 ohms?
Possibly build a pair for extensive listening!Time,always time...and money!!!...
By the way.I see a +/-24V supply for my prototype;a little over 1 A by transistor (the 240s,I have tons of surplus from digikey...only need a match),and total 8 IRFP240 for the bridged version.Total bias will be 4 A, 2 by side.Gain will be 20dB balanced,I think 26dB is too much for me.Should be better for phase margin too.
So I will get a little more than 100W in 8 ohms.But...imagine I want to hook up with speakers at 2 ohms impedance...will the amp do it?And what about the double current source behavior?Do I still have to give to the current source 1/2 the output variation?
Big thanks dear Nelson...
Whatever you load the outputs with is going to help
stabilize the absolute DC offset. I would say 30 ohms
is about the minimum value, as below this we start
loading the circuit more than we need to.
I have chosen 100 ohms on each side to ground because
it conveniently works with the banks of 3 watt resistors I
can fit and so on. 1000 ohms will help stabilize it as well,
but this is not the only technique. Also, absolute DC
voltage is not particularly important, I just like to keep it
down to a fraction of a volt.
The design is stable at unity gain, which is the figure you
get with no input connection. There is only a bit of feedback
anyway, and we arbitrarily burn off some open loop with
resistance from the inputs to ground. See the Inverting
Op Amp thread over in Solid State.
So set the gain anywhere you like.
To drive 2 ohms, you will need to seriously increase the
bias, but the current source gain is still optimally .50
stabilize the absolute DC offset. I would say 30 ohms
is about the minimum value, as below this we start
loading the circuit more than we need to.
I have chosen 100 ohms on each side to ground because
it conveniently works with the banks of 3 watt resistors I
can fit and so on. 1000 ohms will help stabilize it as well,
but this is not the only technique. Also, absolute DC
voltage is not particularly important, I just like to keep it
down to a fraction of a volt.
The design is stable at unity gain, which is the figure you
get with no input connection. There is only a bit of feedback
anyway, and we arbitrarily burn off some open loop with
resistance from the inputs to ground. See the Inverting
Op Amp thread over in Solid State.
So set the gain anywhere you like.
To drive 2 ohms, you will need to seriously increase the
bias, but the current source gain is still optimally .50
See Nelsons answer to my question wich was the same as yours
http://www.diyaudio.com/forums/showthread.php?s=&threadid=3748&perpage=15&pagenumber=30
http://www.diyaudio.com/forums/showthread.php?s=&threadid=3748&perpage=15&pagenumber=30
nar:
i calculate that you should have at least 5.5A total bias for an 8 ohm load driven by +/- 24V rails, and that's allowing a 2V margin off each rail. 6A would be the total bias for one channel if you calculate rail-rail without the margin. If your speakers have an impedance dip to 4 ohms, as mine do, then accounting for this would give a whopping 12A total bias, or 6A per side for one channel of the Aleph-X with current source gain of 0.5! 8 output devices per channel will certianly be stressed by this, at almost 75W per device!
When Nelson said:
Make sense?
Just wanted to clarify that, since I was a little surprised initially when I started calculating operating values for my Aleph-X. Since my speakers do dip to 4 ohms, and I hadn't fully accounted for this in my preliminary design workup, the idle currents ended up being much higher than I had initially planned on. Not that it mattered much, since I hadn't bought any parts yet, but I did have to rethink and rebudget a little. If you're stuck with a certain heatsink or a certain number of output devices, and a load with impedance dips below 8 ohms, you may want to reduce your rail voltage...
i calculate that you should have at least 5.5A total bias for an 8 ohm load driven by +/- 24V rails, and that's allowing a 2V margin off each rail. 6A would be the total bias for one channel if you calculate rail-rail without the margin. If your speakers have an impedance dip to 4 ohms, as mine do, then accounting for this would give a whopping 12A total bias, or 6A per side for one channel of the Aleph-X with current source gain of 0.5! 8 output devices per channel will certianly be stressed by this, at almost 75W per device!
When Nelson said:
He was referring to peak and idle bias currents for the standard Aleph topology, or at least it isn't clear from this statement what the implications are for the X amps. Just remember that the X topology gives the same output power for half the rail voltages and double the total bias currents of a standard Aleph. So, for an Aleph-X, you'd have 10A total idle current as opposed to only 5A for a regular Aleph... but your voltage rails would be roughly half that of the Aleph.
Make sense?
Just wanted to clarify that, since I was a little surprised initially when I started calculating operating values for my Aleph-X. Since my speakers do dip to 4 ohms, and I hadn't fully accounted for this in my preliminary design workup, the idle currents ended up being much higher than I had initially planned on. Not that it mattered much, since I hadn't bought any parts yet, but I did have to rethink and rebudget a little. If you're stuck with a certain heatsink or a certain number of output devices, and a load with impedance dips below 8 ohms, you may want to reduce your rail voltage...
Whaoww...high values...
I guess an Aleph design will peak
twice the initial current draw,for a
standard Aleph output stage.
Since we mate 2 aleph rear ends,
and the speaker will be linked
differentially,we can assume that
if each of the 2 halves have a 2 amp.
bias,then each side can peak at twice
that amount into the loudspeaker.
Which means peak output current
can be +/- 8 A into any loudspeaker.
On 8 ohms,to get 100W RMS we need to swing
P=(U X U)/R,or for 100W and 8 ohms a +/- 28V on output.
That makes for us also P=(I X I)XR,or a I which is about 3.6A.
The traditional output stage can't do it of course,but since it's an Aleph it can peak 2 times the DC bias,or about 2.4A X 2=4.8A.
(correct me if wrong.)
But as the XA has 2 halves,if I plan a 2 A bias per se,
I believe each half beeing able to peak at 2 times
the DC bias,and as we run bridged then total current
can be (2X2)X2,or 8 amps.
For a 2 ohms speaker,to get 100W we need to
draw 7 Amps(P=IXIXR).
Nelson stated 5 amps total bias to drive 2 Ohms.
I believe it is right,because it means each half beeing
biased at 2.5A;so each half can peak twice its DC bias,so we get
2.5X2X2=10amps.Just what Nelson stated.
But I guess 2 ohms is very low...my
speakers are mainly 8 ohms,
so I don't see why a 2 amps.bias by se
wouldn't do it.Or a 4A total bias.
I also have acces to a small Westlake pair
of speakers,near 2 ohms this time.
This is why I would internally
increase bias to 2.5A per se-
5A total bias- to be in order to
drive them happily
I think the rules for bridged mode
are:twice the voltage gain(compared
to se),the output power is done
by 4 (with same current)
Hope it is clear and that I've not done wrong...
I guess an Aleph design will peak
twice the initial current draw,for a
standard Aleph output stage.
Since we mate 2 aleph rear ends,
and the speaker will be linked
differentially,we can assume that
if each of the 2 halves have a 2 amp.
bias,then each side can peak at twice
that amount into the loudspeaker.
Which means peak output current
can be +/- 8 A into any loudspeaker.
On 8 ohms,to get 100W RMS we need to swing
P=(U X U)/R,or for 100W and 8 ohms a +/- 28V on output.
That makes for us also P=(I X I)XR,or a I which is about 3.6A.
The traditional output stage can't do it of course,but since it's an Aleph it can peak 2 times the DC bias,or about 2.4A X 2=4.8A.
(correct me if wrong.)
But as the XA has 2 halves,if I plan a 2 A bias per se,
I believe each half beeing able to peak at 2 times
the DC bias,and as we run bridged then total current
can be (2X2)X2,or 8 amps.
For a 2 ohms speaker,to get 100W we need to
draw 7 Amps(P=IXIXR).
Nelson stated 5 amps total bias to drive 2 Ohms.
I believe it is right,because it means each half beeing
biased at 2.5A;so each half can peak twice its DC bias,so we get
2.5X2X2=10amps.Just what Nelson stated.
But I guess 2 ohms is very low...my
speakers are mainly 8 ohms,
so I don't see why a 2 amps.bias by se
wouldn't do it.Or a 4A total bias.
I also have acces to a small Westlake pair
of speakers,near 2 ohms this time.
This is why I would internally
increase bias to 2.5A per se-
5A total bias- to be in order to
drive them happily
I think the rules for bridged mode
are:twice the voltage gain(compared
to se),the output power is done
by 4 (with same current)
Hope it is clear and that I've not done wrong...
I'm sorry to not have been sufficiently clear. To
produce 100 watts (rms) into 2 ohms requires
10 amp peaks, and therefore 5 amp bias for each half
of a balanced class A amp, in this case an Aleph X.
This means that each channel of the amplifier would
draw 10 amps continuously, 5 amps for each half.
produce 100 watts (rms) into 2 ohms requires
10 amp peaks, and therefore 5 amp bias for each half
of a balanced class A amp, in this case an Aleph X.
This means that each channel of the amplifier would
draw 10 amps continuously, 5 amps for each half.
High currents indeed...
Hmm.. you seem to have a few things confused, here...
I'm not quite sure how you arrived at this conclusion. In differential mode, one side will source current to the load, and the other side will sink that same current. The currents do not add because each output stage is effectively in series with the load, not in parallel. So, if each side has a 2A idle current, and peaks at 4A, then 4A is the peak current delivered to the load, and 4A is also the total idle current. Kinda nifty that the Aleph-X ends up drawing a constant current from each rail.
Now, if you used 8 ohms as your reference load, then there are a couple of different ways of going about the calculations for lower load impedances, depending on what you want to achieve. They will give different results... let me elaborate:
You have two choices:
1. set up the amp for 8 ohm loads, then calculate what kind of power you will be able to get into a 4 or 2 ohm load.
2. Set up the voltage rails for the desired peak power into 8 ohms, then calculate what current will be required to drive this voltage into a lower impedance load, and set the bias current accordingly.
Method 1 just tells you how much power you can get into a lower load impedance. If you attempt to drive a low value load with an amplifier set up this way, you will not be able to drive peak voltage into the load... your amp will be current limited into loads under 8 ohms.
Method 2 allows you to drive 8 ohms at the desired power level, and ensures that your amp will be voltage-limited into lower loads, down to 4 ohms or whatever you used to calculate the bias current. This method of course results in much higher idle currents, since it will be necessary to drive lower loads to the same voltage (and hence much higher power levels) than an 8 ohm load.
IMHO, method 2 is superior when you will be driving real loudspeakers, since most 8 ohm speakers have impedance dips to 4 or 6 ohms at some frequency range. Since speakers are generally designed for voltage drive, you'll want to be able to give them the full voltage your amp is capable of without being current limited.
So, to continue the design example using method 2:
100W into 8 ohms dictates 22V rails, giving roughly 40V peak into the load. If the loudspeakers have an impedance dip to, say, 4 ohms (you'll have to verify this with your own speakers), then 40V/4R = 10A. This will be the peak current to the load, and each half of the Aleph-X will have to idle at half this current, for a total idle current of 10A.
Hmm.. you seem to have a few things confused, here...
I'm not quite sure how you arrived at this conclusion. In differential mode, one side will source current to the load, and the other side will sink that same current. The currents do not add because each output stage is effectively in series with the load, not in parallel. So, if each side has a 2A idle current, and peaks at 4A, then 4A is the peak current delivered to the load, and 4A is also the total idle current. Kinda nifty that the Aleph-X ends up drawing a constant current from each rail. Don't forget that this is 28V RMS! Peak voltage will need to be 40V (28V * sqrt(2)), so your supply rails should be around +/- 22V to obtain the peak differential voltage of 40V, with a little margin to spare. Same thing goes for your current calculation... this is an RMS value, not the peak current which the amp will have to be able to supply. So, for 100W into 8 ohms, we have Irms=3.54A, and Ipk=5A. To supply this, each side will have to idle at minimum of 2.5A.
Now, if you used 8 ohms as your reference load, then there are a couple of different ways of going about the calculations for lower load impedances, depending on what you want to achieve. They will give different results... let me elaborate:
You have two choices:
1. set up the amp for 8 ohm loads, then calculate what kind of power you will be able to get into a 4 or 2 ohm load.
2. Set up the voltage rails for the desired peak power into 8 ohms, then calculate what current will be required to drive this voltage into a lower impedance load, and set the bias current accordingly.
Method 1 just tells you how much power you can get into a lower load impedance. If you attempt to drive a low value load with an amplifier set up this way, you will not be able to drive peak voltage into the load... your amp will be current limited into loads under 8 ohms.
Method 2 allows you to drive 8 ohms at the desired power level, and ensures that your amp will be voltage-limited into lower loads, down to 4 ohms or whatever you used to calculate the bias current. This method of course results in much higher idle currents, since it will be necessary to drive lower loads to the same voltage (and hence much higher power levels) than an 8 ohm load.
IMHO, method 2 is superior when you will be driving real loudspeakers, since most 8 ohm speakers have impedance dips to 4 or 6 ohms at some frequency range. Since speakers are generally designed for voltage drive, you'll want to be able to give them the full voltage your amp is capable of without being current limited.
So, to continue the design example using method 2:
100W into 8 ohms dictates 22V rails, giving roughly 40V peak into the load. If the loudspeakers have an impedance dip to, say, 4 ohms (you'll have to verify this with your own speakers), then 40V/4R = 10A. This will be the peak current to the load, and each half of the Aleph-X will have to idle at half this current, for a total idle current of 10A.
Making mistakes I learn...
Thanks to all of you.I really misunderstood the thing!!!....
Now I've got it.Thanks a lot HIFIZEN.Thanks a lot to Nelson.Whithout you gurus I wouldn't have a real thing working...
Anyway,this project is gonna take so much time for me.
I really want the amp to make it on 2 ohms.I am deranged,most of you will say.
So I plan to use :10 heatsinks sk109 from fisher.2 mos on each.
Total 20 IRFP240.Bias will be 1 amp.by transistor,+constant current source.I will have 22V rails and total bias for 2 sides 10A.
For the power supply,let's be conservative.I will use 2 toroids at 18V X2/500VA,secondaries in parallel.2 diode bridges at 600V 40A.Capacitors:8X 47.000 uF/25V.The supply will be a C-R-C-R-C-R-C with 24 power thermistors for the Rs.
Chassis will be self designed,in thick AU4G 5mm.I will have to work on it
Final result: 2 huge massive mono-blocks willing to drive all speakers of the planet Earth
Thanks to all of you.I really misunderstood the thing!!!....
Now I've got it.Thanks a lot HIFIZEN.Thanks a lot to Nelson.Whithout you gurus I wouldn't have a real thing working...
Anyway,this project is gonna take so much time for me.
I really want the amp to make it on 2 ohms.I am deranged,most of you will say.
So I plan to use :10 heatsinks sk109 from fisher.2 mos on each.
Total 20 IRFP240.Bias will be 1 amp.by transistor,+constant current source.I will have 22V rails and total bias for 2 sides 10A.
For the power supply,let's be conservative.I will use 2 toroids at 18V X2/500VA,secondaries in parallel.2 diode bridges at 600V 40A.Capacitors:8X 47.000 uF/25V.The supply will be a C-R-C-R-C-R-C with 24 power thermistors for the Rs.
Chassis will be self designed,in thick AU4G 5mm.I will have to work on it
Final result: 2 huge massive mono-blocks willing to drive all speakers of the planet Earth