I went to 2k2 for the famous magical resistors.
I went to 2k2 for the famous magical resistors.
and that would be R12 and R34 on the original shematic?
Thanks
Josip
I went to 2k2 for the famous magical resistors.
and that would be R12 and R34 on the original shematic?
Thanks
Josip
Nar,
if you read the zen variations on Nelson's site (I think it's part 2, the one where he adds the variable current source) there is a description of how to do it with a multimeter and a freq. generator. I believe you set the input to 60 hz, and then measure across the upper transistor's source resistor and the group of resistors in series with the output. There should be approx. a two-one ratio of the ac voltage.
The output group of resistors should measure 2x the added AC voltage of all the current source source resitors ac voltages. So, if you have 4 outputs per side, you would want 1/8 the ac current as what is going through the outputresistors.
-from zen var. part 2:
if you read the zen variations on Nelson's site (I think it's part 2, the one where he adds the variable current source) there is a description of how to do it with a multimeter and a freq. generator. I believe you set the input to 60 hz, and then measure across the upper transistor's source resistor and the group of resistors in series with the output. There should be approx. a two-one ratio of the ac voltage.
The output group of resistors should measure 2x the added AC voltage of all the current source source resitors ac voltages. So, if you have 4 outputs per side, you would want 1/8 the ac current as what is going through the outputresistors.
-from zen var. part 2:
oops... in my first paragraph, I said something wrong- according to the quote below, the ac voltage ratio is 70%, not 50%. Sorry.
The Magical Resistors are implemented between Outputs and Source pins of the differential.
4k7 3k3 2k2....is to try,the optimum value
is to find experimentally against total bias
and number of output devices.
They are not represented in Grey's original
schematic.You have to add them 😉
best regards
Anael
4k7 3k3 2k2....is to try,the optimum value
is to find experimentally against total bias
and number of output devices.
They are not represented in Grey's original
schematic.You have to add them 😉
best regards
Anael
magic resistor
Thanks Anael...
Whell I have in my heands shematic fir AlephX revision 1.0 ,so the magic resistor woulld be R46 and R47 ... am I right???
I 'am going to order last parts that I am missing on monday...and I will use 8 mosfets per chenel (2xQ1,2,10,11)...
I will try to get 50W from this amp ...What volue do you sugest ???
Do I nead to change any other parts if I'am going to use 8 instead 4 mosfets?
any help will do ...
Thanks
Josip
Thanks Anael...
Whell I have in my heands shematic fir AlephX revision 1.0 ,so the magic resistor woulld be R46 and R47 ... am I right???
I 'am going to order last parts that I am missing on monday...and I will use 8 mosfets per chenel (2xQ1,2,10,11)...
I will try to get 50W from this amp ...What volue do you sugest ???
Do I nead to change any other parts if I'am going to use 8 instead 4 mosfets?
any help will do ...
Thanks
Josip
Interested in "optimizing" current source?
If anybody is interested in "optimizing" the current source in Gray's original example, these are the suggested values to obtain the same adjustment range:
R24: 276
R26: 666
V2: 2000
This results in a relatively linear response with minimal current passing throught the adjustable resistor. If you use a higher value for V2, you will get less linearity. If you use a lower value sucvh as 1000 Ohms, you will get more linearity at the expense of slightly higher current. Here are the results from V2=1000.
R24: 296
R26: 568
V2: 1000 (this may well be the sweet spot)
And for V2=500
R24: 361
R26: 461
V2: 500
These "optimizations" have been made targeting the same resistive range for the resulting compound. In Gray's original, the current is shared just about equally between fixed and variable resistance which is likely to result in shorter life span for the variable resistor.
Your mileage may vary.
You can do such triplet optimizations yourself using Excel's solver functionality and theready made fully annotated spreadsheet available from here: http://diyaudio.com/projects/petter/ It is quite useful since the problem is not trivial to work out symbolically given the infinite number of solutions.
Petter
If anybody is interested in "optimizing" the current source in Gray's original example, these are the suggested values to obtain the same adjustment range:
R24: 276
R26: 666
V2: 2000
This results in a relatively linear response with minimal current passing throught the adjustable resistor. If you use a higher value for V2, you will get less linearity. If you use a lower value sucvh as 1000 Ohms, you will get more linearity at the expense of slightly higher current. Here are the results from V2=1000.
R24: 296
R26: 568
V2: 1000 (this may well be the sweet spot)
And for V2=500
R24: 361
R26: 461
V2: 500
These "optimizations" have been made targeting the same resistive range for the resulting compound. In Gray's original, the current is shared just about equally between fixed and variable resistance which is likely to result in shorter life span for the variable resistor.
Your mileage may vary.
You can do such triplet optimizations yourself using Excel's solver functionality and theready made fully annotated spreadsheet available from here: http://diyaudio.com/projects/petter/ It is quite useful since the problem is not trivial to work out symbolically given the infinite number of solutions.
Petter
Hello Joke,
I don't have the revision you talk about.Anyway,as my prototype works very well I am studying a >100W version.
Here are the tips:
12 mosfets per channel.Mono block option.
V rails +/- 22,7V
Bias 6 A (each branch 1 amp,3+3=6 A total bias.
The sinks will be huge
!!!
the advantage of a 6 amp bias is to drive well loads under 8 ohms 😉 Here is a Xcel file to calculate output power vs supply and bias
Have fun !!!
BEST REGARDS
Anael
I don't have the revision you talk about.Anyway,as my prototype works very well I am studying a >100W version.
Here are the tips:
12 mosfets per channel.Mono block option.
V rails +/- 22,7V
Bias 6 A (each branch 1 amp,3+3=6 A total bias.
The sinks will be huge
!!!the advantage of a 6 amp bias is to drive well loads under 8 ohms 😉 Here is a Xcel file to calculate output power vs supply and bias
Have fun !!!
BEST REGARDS
Anael
By the way this is revision 1.0 ...I thik it vas posted by HiFiZen I 'am not shure...
I have AlephX PCB V1.0 from group buy so I 'am gonna build that
Josip
I have AlephX PCB V1.0 from group buy so I 'am gonna build that
Josip
The Aleph X effect
When I listen to the Aleph X instead of regular Aleph,what is the subjective impression?
Well,here is the thing.I have the impression that Mr Pass changed my ears and gave me better ones
The Aleph X magic recipe is really incredible.
I have to force me to go to sleep 😉
Thanks again Mr Pass you are awsome we all love you
For all things you don't make like others,
we greatly appreciate !!!And for all others :
or 
sometimes !
You are the best.No kidding...
I have Pat Metheny "Imaginary Day Live"in DVD.The test couldn't lie....
Best regards
Anael
PS:By the way Petter thanks for the tips.Great work !
When I listen to the Aleph X instead of regular Aleph,what is the subjective impression?
Well,here is the thing.I have the impression that Mr Pass changed my ears and gave me better ones
The Aleph X magic recipe is really incredible.
I have to force me to go to sleep 😉
Thanks again Mr Pass you are awsome we all love you
For all things you don't make like others,
we greatly appreciate !!!And for all others :
or sometimes !You are the best.No kidding...
I have Pat Metheny "Imaginary Day Live"in DVD.The test couldn't lie....
Best regards
Anael
PS:By the way Petter thanks for the tips.Great work !
Has anyone compared the sound off AlephX with SOZ?
My next project will be a high/mid amp (Raven R2/ ATC75-150S) and I need an amp with sounds "better" (more detailed, more exact soundstage and very neutral) than the ZEN V4.
I use no preamp with ZEN V4 and it sounds really good. Could a combination with BZLS and SOZ or Aleph X sound better?
I also want to build an amp for my sub-woofer(<100 Hz, 90 dB/Wm). I have a mono solid state amp for this job. Is an Class A amp (e.g. Aleph X) with 100 Watt better for this hard work.
regards
Günter
My next project will be a high/mid amp (Raven R2/ ATC75-150S) and I need an amp with sounds "better" (more detailed, more exact soundstage and very neutral) than the ZEN V4.
I use no preamp with ZEN V4 and it sounds really good. Could a combination with BZLS and SOZ or Aleph X sound better?
I also want to build an amp for my sub-woofer(<100 Hz, 90 dB/Wm). I have a mono solid state amp for this job. Is an Class A amp (e.g. Aleph X) with 100 Watt better for this hard work.
regards
Günter
I would like to say a few words on Aleph X use.
In my system it is linked to my ZLS (not X-ied yet!!!) balanced .
I have inrush current suppressor on the mains.
The Aleph X is capacitively coupled at the input,4,7uF series for now.Also like Petter Daniel a 100k to ground before each cap.
When I power up the amp I have no turn-on thump at all !!!Offset is 42mV differential and stays still.The input differential 9610 are thermally linked.The parts are matched but not hyper-matched like Petter does
When powering down,I can see a slight turn off thump,but on my multimeter it doesn't go higher than 120mV and vanishes quickly.
Does that mean that no output relay for the amp is required ?
In case of problem...hell I don't know.
Makes me think to a power saving mode,and maintain capacitors charged while amp is standby.
Would it be OK to cut V+ rails only ? Nelson gave this advice to me to implement the thermostatic switch 70 degrees into regular Aleph
Anyone ?
Feedback greatly appreciated
Best regards
Anael
In my system it is linked to my ZLS (not X-ied yet!!!) balanced .
I have inrush current suppressor on the mains.
The Aleph X is capacitively coupled at the input,4,7uF series for now.Also like Petter Daniel a 100k to ground before each cap.
When I power up the amp I have no turn-on thump at all !!!Offset is 42mV differential and stays still.The input differential 9610 are thermally linked.The parts are matched but not hyper-matched like Petter does
When powering down,I can see a slight turn off thump,but on my multimeter it doesn't go higher than 120mV and vanishes quickly.
Does that mean that no output relay for the amp is required ?
In case of problem...hell I don't know.
Makes me think to a power saving mode,and maintain capacitors charged while amp is standby.Would it be OK to cut V+ rails only ? Nelson gave this advice to me to implement the thermostatic switch 70 degrees into regular Aleph
Anyone ?
Feedback greatly appreciated
Best regards
Anael
Attachments
By the way.If input capacitance is an issue,I propose to move R 18 and R 28 to 5K and R 16 and R 30 to 50k.In theory gain is the same,i.e. 20 dB balanced but lower input impedance,did someone make modif and saw a change in the high frequency roll off ?
Should give better THD measures too,although you have to get a preamp capable of driving lower impedances.
Henrick ? Are you there ? ....
Should give better THD measures too,although you have to get a preamp capable of driving lower impedances.
Henrick ? Are you there ? ....
nar said:
Product XA's do not require a relay either and have minimal thump. Interestingly, if the output stage of one side fails, the other side will track any DC as common mode, so it's very difficult to damage a speaker this way. Of course we haven't sent enough XA's into the field to see a failure as yet.
12 volts is a bit low, even if you mean +/- 12 volts. The Mosfets like some extra voltage the same way they like some extra bias, and I would expect only 10 to 15 watts, and not particularly good watts, either. The minimum I work with is +/- 25 volts, and of course you can get higher voltage switching supplies...Bernhard said:
Bernhard said:
Yes, 24V sps are dirt cheap too
I thought you were planning on an ultimate amp.😉