OK I get it now.
So the supplies to the whole input section are tied to follow the output potential.
I predict you will have a lovely special case at unity gain - almost all the distortion mechanism will disappear, as everything will be at a quasi-static point on its load line.
At gain of two, you might expect the distortion to halve, as the bootstrap is moving too far for the input stage.
So for gain of two, try splitting the bootstrap voltage to half the output, so it roughly tracks the input - this will make the input section quasi-static.
So the supplies to the whole input section are tied to follow the output potential.
I predict you will have a lovely special case at unity gain - almost all the distortion mechanism will disappear, as everything will be at a quasi-static point on its load line.
At gain of two, you might expect the distortion to halve, as the bootstrap is moving too far for the input stage.
So for gain of two, try splitting the bootstrap voltage to half the output, so it roughly tracks the input - this will make the input section quasi-static.
Russ White said:
Thanks - I get it now. I predict something special will happen at unity gain.
At higher gains, the boot strapping will always be a compromise between the needs of the input pair and the VAS.
Correct here is unity gain 🙂
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 2.000e+04 9.945e-01 1.000e+00 -0.84° 0.00°
2 4.000e+04 3.370e-08 3.389e-08 -172.27° -171.44°
3 6.000e+04 6.994e-08 7.033e-08 -87.29° -86.45°
4 8.000e+04 1.811e-09 1.821e-09 165.99° 166.82°
5 1.000e+05 3.596e-09 3.616e-09 -53.70° -52.86°
6 1.200e+05 1.128e-09 1.134e-09 155.37° 156.20°
7 1.400e+05 8.294e-10 8.340e-10 -115.91° -115.08°
8 1.600e+05 2.393e-10 2.406e-10 -142.16° -141.32°
9 1.800e+05 6.182e-10 6.216e-10 116.05° 116.89°
Total Harmonic Distortion: 0.000008%
Harmonic Frequency Fourier Normalized Phase Normalized
Number [Hz] Component Component [degree] Phase [deg]
1 2.000e+04 9.945e-01 1.000e+00 -0.84° 0.00°
2 4.000e+04 3.370e-08 3.389e-08 -172.27° -171.44°
3 6.000e+04 6.994e-08 7.033e-08 -87.29° -86.45°
4 8.000e+04 1.811e-09 1.821e-09 165.99° 166.82°
5 1.000e+05 3.596e-09 3.616e-09 -53.70° -52.86°
6 1.200e+05 1.128e-09 1.134e-09 155.37° 156.20°
7 1.400e+05 8.294e-10 8.340e-10 -115.91° -115.08°
8 1.600e+05 2.393e-10 2.406e-10 -142.16° -141.32°
9 1.800e+05 6.182e-10 6.216e-10 116.05° 116.89°
Total Harmonic Distortion: 0.000008%
That really is low distortion! > 140dB down on the fundamental!
I wonder if there will be a comparable special value of bootstrap fraction for other gains?
Unity is a special case, as everything moves together.
For other gain values, you have to trade between keeping the VAS static, and keeping the input LTP static. Your present arrangement (tie supply to output) keeps the VAS centred, but increases (in fact reverses) the common mode swing seen by the input LTP.
If you reduce bootstrap to fraction of output equal to inverse of gain, the input stage is the quasi-static, but VAS only gets modest benefit. Now this might be very worth doing, as it removes most common mode distortion, and VAS is inside FB loop.
Great stuff, best of luck, and I have to go now - I will check in and see how you got on tomorrow.
I wonder if there will be a comparable special value of bootstrap fraction for other gains?
Unity is a special case, as everything moves together.
For other gain values, you have to trade between keeping the VAS static, and keeping the input LTP static. Your present arrangement (tie supply to output) keeps the VAS centred, but increases (in fact reverses) the common mode swing seen by the input LTP.
If you reduce bootstrap to fraction of output equal to inverse of gain, the input stage is the quasi-static, but VAS only gets modest benefit. Now this might be very worth doing, as it removes most common mode distortion, and VAS is inside FB loop.
Great stuff, best of luck, and I have to go now - I will check in and see how you got on tomorrow.
I wonder if we are calling this a bootstrap incorrectly... 🙂
The reason is it seems more to be some sort of follower or cascode.
One down side seems to be that gain is limited by the forward voltage drop of the diodes. At least to way I have done it.
I am still investigating.
The reason is it seems more to be some sort of follower or cascode.
One down side seems to be that gain is limited by the forward voltage drop of the diodes. At least to way I have done it.
I am still investigating.
Ok I got it to work respectably if I separated the Buffer stage from the input/VAS.
What I mean is, I gave them both there own current sources and ran the input at rail voltage, and buffer for the bootstrap/cascode/follower thingy. 🙂
Then I was able to get full swing.
What I mean is, I gave them both there own current sources and ran the input at rail voltage, and buffer for the bootstrap/cascode/follower thingy. 🙂
Then I was able to get full swing.
fierce_freak said:
In simulation hey work great, but it can be more tricky to get the bias easily adjustable.
I have not built one with them yet. Perhaps I will though.
Cheers!
Russ
Russ White said:
Here is a combinatin with MAT02/SSM2210, BC847BS, two BC550C and two BC850C
Here is a plain TO92 and SOT23
Hi Mr White,
i am trying to build your design, and have decided to get the
boards manufactured by a house, instead of trying to do them myself...while they say that the pdfs are okay they would prefer gerbers... are you willing to share these?
if you arent then can you upload the pdf for the solder stop mask
as i would like to have that on the board...
many thanks..
cheers,
mymindinside
i am trying to build your design, and have decided to get the
boards manufactured by a house, instead of trying to do them myself...while they say that the pdfs are okay they would prefer gerbers... are you willing to share these?
if you arent then can you upload the pdf for the solder stop mask
as i would like to have that on the board...
many thanks..
cheers,
mymindinside
I will gladly post gerbers for individual use only. I am not at home right now. When I return later today I will post them.
Cheers!
Russ
Cheers!
Russ
Thanks Mr. White,
with the gerbers i will be more confident of getting a good pcb.
I assure you i will only be using them for my personal, non-commercial use.
another thing i wanted to ask you was if it is ok to put a volume pot before the input of the amp? is that all thats needed to control volume or will i need some other components?what value pot should i use?
sorry for the (multiple) newb questions, just want to get it right the first time
looking forward to hearing how this amp sounds...
love,light and peace
mymindinside
with the gerbers i will be more confident of getting a good pcb.
I assure you i will only be using them for my personal, non-commercial use.
another thing i wanted to ask you was if it is ok to put a volume pot before the input of the amp? is that all thats needed to control volume or will i need some other components?what value pot should i use?
sorry for the (multiple) newb questions, just want to get it right the first time
looking forward to hearing how this amp sounds...
love,light and peace
mymindinside
A 5-25K pot would work just fine. Or a stepped attenuator.
You could use a 50K pot, but I would change the Rz value to 100K if you do.
10K is what I would use.
The Joshua Tree attenuator works very well in front of this too.
Cheers!
Russ
You could use a 50K pot, but I would change the Rz value to 100K if you do.
10K is what I would use.
The Joshua Tree attenuator works very well in front of this too.
Cheers!
Russ
Added drill file
I forgot to include the NC drill file with the gerbers, It is in the zip now. re-download it if necessary.
Cheers!
Russ
I forgot to include the NC drill file with the gerbers, It is in the zip now. re-download it if necessary.
Cheers!
Russ
Hi
I went out to buy the parts today, and saw what might be a discrepancy between the values on the schematic and the ones in the BOM. In the Schematic the value for R16, R17, R22, R23 is 4.99R while the BOM lists these parts as 4.75R. Which one is correct and how critical is the value, will a 4.7R 1% resistor work?
Another thing is that some of the other values like 16.2k, 2.21k 221R are not available here in India. is it okay to put two resistors in series to get the correct value. would the correct way to do this be 15k +1.2k for 16.2k? and 2.2k + 10R for 2.21k? i will have to use two parallel caps for the 60pf C7 positon as that value is not available either. is this okay, would anyone recommend anything else?
Is a cap rated for 35v ok for C1 and C2, the electrolytics? the amp will be running off 29v batteries in series for 18v.
How critical is the FV of the leds in the circuit? the vendors here said that it will work between 1.5 to 3 v,and didnt understand when i asked them about forward votage. is there a way to measure this?
sorry for so many questions but the good side is that the thread has been bumped so more people can see the design...
thanks for all the help
cheers,
mymindinside
I went out to buy the parts today, and saw what might be a discrepancy between the values on the schematic and the ones in the BOM. In the Schematic the value for R16, R17, R22, R23 is 4.99R while the BOM lists these parts as 4.75R. Which one is correct and how critical is the value, will a 4.7R 1% resistor work?
Another thing is that some of the other values like 16.2k, 2.21k 221R are not available here in India. is it okay to put two resistors in series to get the correct value. would the correct way to do this be 15k +1.2k for 16.2k? and 2.2k + 10R for 2.21k? i will have to use two parallel caps for the 60pf C7 positon as that value is not available either. is this okay, would anyone recommend anything else?
Is a cap rated for 35v ok for C1 and C2, the electrolytics? the amp will be running off 29v batteries in series for 18v.
How critical is the FV of the leds in the circuit? the vendors here said that it will work between 1.5 to 3 v,and didnt understand when i asked them about forward votage. is there a way to measure this?
sorry for so many questions but the good side is that the thread has been bumped so more people can see the design...
thanks for all the help
cheers,
mymindinside
mymindinside said:
For R16,R17, R22, and R23 4.7R is just fine.
For R11 with the supplies you are running any value from 12K to 22k will work just fine.
For R18 and R19 any value from 100 to 470R should be just fine. Just be sure those two resistors are the same.
RF and RG do not have to be precisely 2.21K Anything from 1K to 10K or so is fine for either spot. Just keep in mind that those two resistors set the gain of the amp. The values I have on the schematic will give a voltage gain of 2X(6db).
The formula for gain is A = 1 + (RF / RG).
So a 2.2K resistor for each spot is just fine. No need for a series resistor.
The value for C7 is not too critical but it should be >= 33pf so anything from 33pf to 240pf should be just fine. I would try to stay < =100pf.
Just about any LED will work. Most green LEDS have a FVD of 1.75 to 2V. Just use the same type LED in all LED locations.
It is pretty easy to measure forward voltage, just take a 9V battery and a 2K (not very critical just choose a resistor that will not allow too much current through your LED) resistor and put the resistor in series with the LED. Now measure the voltage across the LED. Remember to observe proper LED polarity. Short lead goes to the negative terminal.
For your power supply 35V caps are just fine. Even 25V would be fine.
Don't be afraid to ask questions. I am glad to help.
Cheers!
Russ