Hi !
Hi everyone!
I will overpunch the new scheme.
In new amplifier has used several ideas thereof forum.
Shift to the right if there is blunders.
I am a beginner amateur , me a great deal not known.
Shall be grateful for responses.
To message join; archive with document from Micro-Cap 7.1.0
Prompt , cost(stand)s try this collect?
Thank you for attention.
Hi everyone!
I will overpunch the new scheme.
In new amplifier has used several ideas thereof forum.
Shift to the right if there is blunders.
I am a beginner amateur , me a great deal not known.
Shall be grateful for responses.
To message join; archive with document from Micro-Cap 7.1.0
Prompt , cost(stand)s try this collect?
Thank you for attention.
Re: Hi !
it looks good. a few questions:
a) do you really need two ccs for the VAS?
b) Q7 and Q10 idles at about 10ma. so the voltage drop over R12 is 200mv. Is that enough to open up those two mosfets?
c) it is probably OK to use a resistor to set bias for the two mosfets. But a Vbe multiplier would be more versatile.
d) you may want to use two source resistors here. They aren't required for the type of the mosfets you use but I always use source resistors to help stabilize bias and provide versatility: you can always jump them with a straight wire.
samod said:
it looks good. a few questions:
a) do you really need two ccs for the VAS?
b) Q7 and Q10 idles at about 10ma. so the voltage drop over R12 is 200mv. Is that enough to open up those two mosfets?
c) it is probably OK to use a resistor to set bias for the two mosfets. But a Vbe multiplier would be more versatile.
d) you may want to use two source resistors here. They aren't required for the type of the mosfets you use but I always use source resistors to help stabilize bias and provide versatility: you can always jump them with a straight wire.
hi
thats a neat design..i am inexperienced too (very in fact!!) so what i say might be garbage, BUT i dont understand why the degeneration resistors for the diff pair are different??...and also, is R5 necessary?...
the design is simple i think..not a huge signal path which is good as far s i understand...maybe someone can enlighten me? it would be much appreciated...excellent attempt though, better than i could have done
thats a neat design..i am inexperienced too (very in fact!!) so what i say might be garbage, BUT i dont understand why the degeneration resistors for the diff pair are different??...and also, is R5 necessary?...
the design is simple i think..not a huge signal path which is good as far s i understand...maybe someone can enlighten me? it would be much appreciated...excellent attempt though, better than i could have done
Thank you for attention to my message
I write using translator .
He not all correct translates.
Answer that questions which realized:
I have enclosed the documents , which explain the choice an element.
One drawing was will take from this forum.
I have considered such scheme justified , distortion less...
( this about modes.Q7 Q10)
The Second drawing emulation dynamic DC Resistors : R18+R19 , R12 possible not constant!
The Choice output transistor is conditioned by their warm-up dependency.
All while has tested on emulator. Shall Tomorrow try to collect , can this and in vain spent time.
Shall try.
-------------------------------------
Can Not enclose again files , does not allow the forum.
See beside me...
http://www.kievsat.com/phpBB/download.php?id=970
I write using translator .
He not all correct translates.
Answer that questions which realized:
I have enclosed the documents , which explain the choice an element.
One drawing was will take from this forum.
I have considered such scheme justified , distortion less...
( this about modes.Q7 Q10)
The Second drawing emulation dynamic DC Resistors : R18+R19 , R12 possible not constant!
The Choice output transistor is conditioned by their warm-up dependency.
All while has tested on emulator. Shall Tomorrow try to collect , can this and in vain spent time.
Shall try.
-------------------------------------
Can Not enclose again files , does not allow the forum.
See beside me...
http://www.kievsat.com/phpBB/download.php?id=970
demons_wing said:
For the degeneration resistors, yes probably for DC offset cancellation. But the main problem is that the DC input impedance at J0 FET is 100k and the DC input impedance at J1 FET is 22K. This creates a high DC imbalance. Normally, the same value should be used for R9 and R13 to minimize the DC offset. If the FET input pair is not well matched then even similar input impedances may still require a DC adjustment but for that use a small trimpot (200 ohms or more for example) because the actual degeneration resistors chosen value will only work in this particular project.
I do not see the purpose of R5 since Q8 is able to sustain a lot higher voltage and power than that.
C3 is probaly a little low to avoid a relatively high phase shift at 20 Hz.
I suppose C8 is used for test purpose only of response to square wave in capacitive load.
What is the resulting mosfet bias current with R12 = 20?
Re: Thank you for attention to my message
at me. The same thing.
If you use software, try to write by yourself. I convinced that you can do it better. Sometimes I use french -> english and I don't understand a thing. when I use english -> german theysamod said:I write using translator .
He not all correct translates.
at me. The same thing.Works!
Greetings to all!
I pleased with result of the work.
This amplifier on a model payment today has soldered.
It has earned! It was necessary to pick up R12, it became 100 ohm. Resistor R13 has put 100 k, R7-2,7 k. Transistor Q10 has replaced on 2SB649, Q7 - 2SD669 / a Feed +-42Â. was not possible to compensate completely drift 0, R18+R19, the tuning resistor, there were 0,014 V.
Likely it is necessary to pick up an entrance pair (I did not select). How to check up key parameters? Can there are programs for it?
Thanks.
Greetings to all!
I pleased with result of the work.
This amplifier on a model payment today has soldered.
It has earned! It was necessary to pick up R12, it became 100 ohm. Resistor R13 has put 100 k, R7-2,7 k. Transistor Q10 has replaced on 2SB649, Q7 - 2SD669 / a Feed +-42Â. was not possible to compensate completely drift 0, R18+R19, the tuning resistor, there were 0,014 V.
Likely it is necessary to pick up an entrance pair (I did not select). How to check up key parameters? Can there are programs for it?
Thanks.
FET inputs pair are hard to keep a constant DC offset since it usually drifts with small internal temperature changes. Matching FET inputs may not be necessary since a DC offset of 14 mv is not so bad considering the use of FET in input pair. For R18+R19, have you used a trim potentiometer? I would try to match FET input transistors only if R18 is very different from R19. Matching FET inputs pair (in DC) could be probably done trying to match IDSS currents for each FET.
Re: Hi !
200 ohms difference between R18 and R19 may be acceptable(personal opinion). Again, with 200 ohms, 14 mv DC offset should be OK (personal opinion).
Using too high R18, R19 values would reduce the voltage gain of jFET. Since the voltage gain of jFET is function of the inverse of the transconductance of the jFET and that it is most probably high compared to actual 200 ohms, there should not be problem. Thus R6 probably does not need to be increased (by the way the jFET ID current would not change since it is determined by the current source of the input pair). If you need (in another realization of this circuit) a higher value than 200 ohms for R18+R19 then it is possible to use low frequency bypass capacitor to preserve the open loop gain of J0-J1 jFETs. See example in post #11 of "JFET preamp recommendation?" thread on this forum.
samod said:Thanks for the response.
Instead of R18+R19 I used an accurate potentiometer 200 ohm. At equality R18=R19, displacement 0 = 0,128 V.
To adjust it is necessary in the end.
200 ohm + 0 ohm turns out...
Can it is necessary to pick up R6?
But then current JFET will change...
While I think.
200 ohms difference between R18 and R19 may be acceptable(personal opinion). Again, with 200 ohms, 14 mv DC offset should be OK (personal opinion).
Using too high R18, R19 values would reduce the voltage gain of jFET. Since the voltage gain of jFET is function of the inverse of the transconductance of the jFET and that it is most probably high compared to actual 200 ohms, there should not be problem. Thus R6 probably does not need to be increased (by the way the jFET ID current would not change since it is determined by the current source of the input pair). If you need (in another realization of this circuit) a higher value than 200 ohms for R18+R19 then it is possible to use low frequency bypass capacitor to preserve the open loop gain of J0-J1 jFETs. See example in post #11 of "JFET preamp recommendation?" thread on this forum.
Hi !
I welcome all at this forum!
Is still vozmozhnosi on completion...
Has decided to apply in the entrance cascade dynamic loading, and has a little changed the second cascade.
It should be better. So it is easier to achieve "0"V on an output.
Still there is an idea umoshchnit the target cascade.
Realized on transistors of one structure (N-MOSFET).
They at work on low power are switched - off and included necessarily.
At such inclusion has disappeared necessity in termostabilizatsii.
The scheme and a file for Micro-cap 7 I put.
http://www.kievsat.com/phpBB/viewtopic.php?p=18403#18403
I welcome all at this forum!
Is still vozmozhnosi on completion...
Has decided to apply in the entrance cascade dynamic loading, and has a little changed the second cascade.
It should be better. So it is easier to achieve "0"V on an output.
Still there is an idea umoshchnit the target cascade.
Realized on transistors of one structure (N-MOSFET).
They at work on low power are switched - off and included necessarily.
At such inclusion has disappeared necessity in termostabilizatsii.
The scheme and a file for Micro-cap 7 I put.
http://www.kievsat.com/phpBB/viewtopic.php?p=18403#18403
1
http://www.diyaudio.com/forums/showthread.php?s=&threadid=31004&highlight=
Thank you for reference!
If gravely obtain justification on current transistor one structures then with two structures variable is added.
I do not love to process the signal in two stages (+ half , - a half ) , their afterwards it is necessary to pack...
Yes and find the transistors to conductivity miscellaneous ( êîìïëåìåíòàðíàÿ pair ) problem.
I seem that ëîãè÷íåå to get the maximum quality on one transistor than pack from two.
Create him such condition ( the dynamic load , complex component...) , under which he will give the maximum!
Thank you for helps.
http://www.diyaudio.com/forums/showthread.php?s=&threadid=31004&highlight=
Thank you for reference!
If gravely obtain justification on current transistor one structures then with two structures variable is added.
I do not love to process the signal in two stages (+ half , - a half ) , their afterwards it is necessary to pack...
Yes and find the transistors to conductivity miscellaneous ( êîìïëåìåíòàðíàÿ pair ) problem.
I seem that ëîãè÷íåå to get the maximum quality on one transistor than pack from two.
Create him such condition ( the dynamic load , complex component...) , under which he will give the maximum!
Thank you for helps.
M1 and M2 power transistors should be low cost "medium" power transistors at lower current. Too much power lost in M1 and M2. Only M3 and M4 provides "audio" power to speakers.
Current mirror loading in input stage: good.
Buffer VAS and cascode: good.
"Negative" open loop gain value is much higher than "positive" one: less good
Current mirror loading in input stage: good.
Buffer VAS and cascode: good.
"Negative" open loop gain value is much higher than "positive" one: less good
1
Thank you for commentary!
This always supports!
From scheme not is obviously seen that power in load will form from power which will give M1 + M2 (25-35 W) and M3 + M4 (rest possible ).
For this reason I did not put the mania powerful (M1 , M2) transistors.
Yes and the price on ,2SJ76-2SJ79 , 2SK213-2SK216 - $8/items , but on 2SJ162 , 2SK1058 - $3/pc .
Not equality of the second cascade I do not disturb , the main to was not a restrictions for M1 + M2 , for level 25 - 35 W.
But cascade on M3 + M4 takes the offset with R27 and R28 ( the fall of the voltage under output power more than 25 - 35 W ) Accordingly high output voltage of the second cascade no need. ( possible I mistaken ).
So I thought and realized the idea...
Shall pleased help!
Thank you for commentary!
This always supports
!From scheme not is obviously seen that power in load will form from power which will give M1 + M2 (25-35 W) and M3 + M4 (rest possible ).
For this reason I did not put the mania powerful (M1 , M2) transistors.
Yes and the price on ,2SJ76-2SJ79 , 2SK213-2SK216 - $8/items , but on 2SJ162 , 2SK1058 - $3/pc .
Not equality of the second cascade I do not disturb , the main to was not a restrictions for M1 + M2 , for level 25 - 35 W.
But cascade on M3 + M4 takes the offset with R27 and R28 ( the fall of the voltage under output power more than 25 - 35 W ) Accordingly high output voltage of the second cascade no need. ( possible I mistaken ).
So I thought and realized the idea...
Shall pleased help!
1
Hi !
Here is test interconnection of the charge...
On charge do not cost(stand) so far:
L1 , C7 , R20 , R22 .
Their it is necessary to place the awning , or hands will reach - shall do the charge completely . :?:
Sizes 147 mm * 40 mm.
C5 , C6 - 3300 Uf 63v
The Resistors R29 , R 30 - solder on the part of track.
Hi !
Here is test interconnection of the charge...
On charge do not cost(stand) so far:
L1 , C7 , R20 , R22 .
Their it is necessary to place the awning , or hands will reach - shall do the charge completely . :?:
Sizes 147 mm * 40 mm.
C5 , C6 - 3300 Uf 63v
The Resistors R29 , R 30 - solder on the part of track.
Hi !
I have built the new amplifier , interesting was as will sound...
Features came in well On emulator.
Here is new variant , without feedback.
Plays aloud , but that that with low frequency , did not yet understand , shall pleased help.
http://www.freewebs.com/samodelkin/no.jpg
http://www.freewebs.com/samodelkin/emul.jpg
I have built the new amplifier , interesting was as will sound...
Features came in well On emulator.
Here is new variant , without feedback.
Plays aloud , but that that with low frequency , did not yet understand , shall pleased help.
http://www.freewebs.com/samodelkin/no.jpg
An externally hosted image should be here but it was not working when we last tested it.
http://www.freewebs.com/samodelkin/emul.jpg
An externally hosted image should be here but it was not working when we last tested it.
Hi Samod
You have changed a lot your design since your first post!
I am not too much experienced with NO overall feedback design (but it is really the case since there is the DC servo?) but having seen some others I can see that yours seem interesting. Similar basic topology as Plantfeve Mosquito 2 but with several enhancements (VAS cascode, buffer before mosfet output stage, DC servo). Is it a current feedback design?
I had not seen before this way of "powering" the supply of the DC servo op-amp, it is nice.
However, I believe that the input of the amp should be lowered a bit since 2 megohm at the DC servo output is quite high.
In the input you could add a small series resistor with small cap to filter high frequency from entering the amp.
R74: is it needed since its low value compared to R72 and R73?
How do you "adjust" the bias current at output?
Fab
You have changed a lot your design since your first post!
I am not too much experienced with NO overall feedback design (but it is really the case since there is the DC servo?) but having seen some others I can see that yours seem interesting. Similar basic topology as Plantfeve Mosquito 2 but with several enhancements (VAS cascode, buffer before mosfet output stage, DC servo). Is it a current feedback design?
I had not seen before this way of "powering" the supply of the DC servo op-amp, it is nice.
However, I believe that the input of the amp should be lowered a bit since 2 megohm at the DC servo output is quite high.
In the input you could add a small series resistor with small cap to filter high frequency from entering the amp.
R74: is it needed since its low value compared to R72 and R73?
How do you "adjust" the bias current at output?
Fab
Hi !
The Regard!
Thank you for interest , which you have to my "test"!
This really completely worker scheme - a sound goes! Scheme simple , will repeat any.
I developed her(its) itself , can and not aptly , do not know else whole.
Bad understand the program ( the shell of the project).
All in this amplifier is fixed automatically! The Turning no need.
Possible, it is necessary to look the reinforcement upper and lower shoulder of the input cascade , probably so much for.
Beside I work this variant , more idle time , with one pair on output.
High power I expected for several vapour(pair)s...
In my variant he works from +/-40v.
If interesting - place;put;lay file for MC7 ...
http://www.freewebs.com/samodelkin/nofeedback.htm
The Regard!
Thank you for interest , which you have to my "test"!
This really completely worker scheme - a sound goes!
Scheme simple , will repeat any.I developed her(its) itself , can and not aptly , do not know else whole.
Bad understand the program ( the shell of the project).
All in this amplifier is fixed automatically! The Turning no need.
Possible, it is necessary to look the reinforcement upper and lower shoulder of the input cascade , probably so much for.
Beside I work this variant , more idle time , with one pair on output.
High power I expected for several vapour(pair)s...
In my variant he works from +/-40v.
If interesting - place;put;lay file for MC7 ...
http://www.freewebs.com/samodelkin/nofeedback.htm
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.