I built the CFA-XH V1.3 Mod1 that BV posted. On the pic of the board with modifications that I posted earlier, I missed the two new 10K resistors off of the collectors of Q1 & Q2. There are two more cuts that need to be made. I added the resistor under the board. I am attaching a revised picture. I tested the IPS but didn't have time to hook it up for a listen. Square waves look very clean and square. I will do more in the morning. Adding a pic just showing the trace cuts for clarity.
Blessings, Terry
This is not a "scary" mod. thanks for subjecting one of your boards to this
abuse. I am not DX .. ALL possible changes and suggestions are both welcome and accepted.
This is how we will make the best amps on the forum .
All those who are contrary (with their comments) will really like the
final product.
Sometimes perfection takes a few tries .... (but no "duds" in between)
OS
my slow progress...
i will be modding the cfa-xh according to VB mod and just like terry's board.
i still have to change some resistors in the IPS. i found some resistors with watt rating that is not enough.
i will be modding the cfa-xh according to VB mod and just like terry's board.
i still have to change some resistors in the IPS. i found some resistors with watt rating that is not enough.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Terry ,can you say what compensation has you put there?I built the CFA-XH V1.3 Mod1 that BV posted. On the pic of the board with modifications that I posted earlier, I missed the two new 10K resistors off of the collectors of Q1 & Q2. There are two more cuts that need to be made. I added the resistor under the board. I am attaching a revised picture. I tested the IPS but didn't have time to hook it up for a listen. Square waves look very clean and square. I will do more in the morning. Adding a pic just showing the trace cuts for clarity.
Blessings, Terry
Value and type if it's known
Attachments
Hi OS is this the correct schematic with all modifications?This is not a "scary" mod. thanks for subjecting one of your boards to this
abuse. I am not DX .. ALL possible changes and suggestions are both welcome and accepted.
This is how we will make the best amps on the forum .
All those who are contrary (with their comments) will really like the
final product.
Sometimes perfection takes a few tries .... (but no "duds" in between)
OS
Attachments
Cool , join the "crowd" ....
I will truly hit "the next level" soon.
Even the poor will experience a higher level of audio excellence
than the "halcro crew"
PS - I have a lot more circuits in my "bag" ... I just been "dripping" a
few to the forum. (30+ more IPS's - all unique .. you have not seen these)
This is why I made the "slew".
OS
I will truly hit "the next level" soon.
Even the poor will experience a higher level of audio excellence
than the "halcro crew"
PS - I have a lot more circuits in my "bag" ... I just been "dripping" a
few to the forum. (30+ more IPS's - all unique .. you have not seen these)
This is why I made the "slew".
OS
Hi OS is this the correct schematic with all modifications?
Yes , that will get you single digit PPM + -90db PSSR.
BV's additional small caps can trim out a few ppm ....
BUT ... that is the basic scheme .. to do it.
PS - absolutely no slew or stability degradation involved ...
in fact , things get better !
OS
Terry ,can you say what compensation has you put there?
Value and type if it's known
I think I have 82p there. I will have to check tomorrow when I'm at my computer.
I matched the schematic in post 2801. It is ready to go I just have to plug it into the OPS.
Blessings,Terry
In "raw performance" (pSSR +THD + slew)
Spooky = CFA-XH (mod)
Wolverine and symasui can match 120 V p-p THD
Wolverine IS the PPM "champ" at 3-5.
BUT , PPM is not the sonic "key" ??
The reason for my hawksford "abuse"(R loads) , is to get H2 as dominate ...
even if I am at single digit PPM.
All these amps are "wires with gain" (as they should be) , if one
seems to be more musical , this might be definitely subjective.
OS
Spooky = CFA-XH (mod)
Wolverine and symasui can match 120 V p-p THD
Wolverine IS the PPM "champ" at 3-5.
BUT , PPM is not the sonic "key" ??
The reason for my hawksford "abuse"(R loads) , is to get H2 as dominate ...
even if I am at single digit PPM.
All these amps are "wires with gain" (as they should be) , if one
seems to be more musical , this might be definitely subjective.
OS
Terry, my question is about drivers compensation not pre drivers.I think I have 82p there. I will have to check tomorrow when I'm at my computer.
I matched the schematic in post 2801. It is ready to go I just have to plug it into the OPS.
Blessings,Terry
Please OS can you say something about this?
Last edited:
Please OS can you say something about this?
Thimios , have any amps smoked yet ?
Just make my last schema of the CFA-Xh work , all will be good.
PS - It's not a major change ...
OS
No, but as i know this compensation may have big difference in listening.OK, you know betterThimios , have any amps smoked yet ?
Just make my last schema of the CFA-Xh work , all will be good.
PS - It's not a major change ...
OS
Last edited:
No, but as i know this compensation may have big difference in listening
Yes , I did the SW sim ... it is faster. In fact ... I could get 1.5K V/us with
this config. (op-amp level).
Build and listen ...
PS - VSSA is left in the "dust" ....
OS
Last edited:
I am waiting listening impression for the new compensation. Please Thimios, Terry do it. I will try it next week, may be.
With simple cfa (no diamond buffer), new compensation, and 2ef output, I get THD 0.006295% at 20 kHz, 60W/8Ohm with power supply +-44V.
RC filter on transistor driver power supply can increase PSRR almost flat at 128 Hz to 3kHz but reduce maximum output voltage (I can accept this trade-off). With capacitance multiplier on VAS, it can achieve 126 dB PSRR at 100Hz. Slew rate is 760V/uS. At least on simulation.
I agree with OS. This kind compensation can beat VSSA
With simple cfa (no diamond buffer), new compensation, and 2ef output, I get THD 0.006295% at 20 kHz, 60W/8Ohm with power supply +-44V.
RC filter on transistor driver power supply can increase PSRR almost flat at 128 Hz to 3kHz but reduce maximum output voltage (I can accept this trade-off). With capacitance multiplier on VAS, it can achieve 126 dB PSRR at 100Hz. Slew rate is 760V/uS. At least on simulation.
I agree with OS. This kind compensation can beat VSSA
Last edited:
Member
Joined 2009
Paid Member
PS - I have a lot more circuits in my "bag" ... I just been "dripping" a
few to the forum. (30+ more IPS's - all unique .. you have not seen these)
This is why I made the "slew".
OS
I'll never be able to keep up
I guess this compensation scheme including the input pair answers a question I posted long ago to someone else, Abraxalito, why the feedback loop did not go all the way back to the earliest point where it could be applied. At that point we were speaking of monolithic op-amp designs and his conclusion on why they were not as good as the discrete designs was that you could not include the LTP in the feedback loop of the monolithics. Perhaps this is the answer to both types of amplifier but I wouldn't know if you could implement this compensation in the 3886 type of chip amps.
OS,
With your current comment about this change beating the VSSA amp could the current CFA-XH and a single pair output section be made as small as a vssa if it was on one board? Just to drive a tweeter @2.5Khz up and nothing more. Only needs low power, nothing over 30 watts would really ever be used.
OS,
With your current comment about this change beating the VSSA amp could the current CFA-XH and a single pair output section be made as small as a vssa if it was on one board? Just to drive a tweeter @2.5Khz up and nothing more. Only needs low power, nothing over 30 watts would really ever be used.
Terry, my question is about drivers compensation not pre drivers.
Please OS can you say something about this?
Hi Thimios,
I just checked and I did install 82p there. Maybe it is not needed, I don't know.
I noticed that I had some errors in the last picture I posted. I have reworked to reflect exactly the changes I made to my boards to match BV's last schematic. I will be testing it shortly.
Blessings, Terry
Attachments
The slew master has indeed evolved since the last time I peeked on this thread. Great going Ostripper and members. Thank you for your contributions to audio research.
I humbly request for a third party listening test review between this amplifier and either the 1DIFFQC or the SYMEF or the 2DIFFET. If indeed this amplifier does sound better (in the event it does, the distinction shouldn't be clear as day and night) than the listed amplifiers then that information will be valuable data for my hypothesis.
kind regards,
Harrison.
I humbly request for a third party listening test review between this amplifier and either the 1DIFFQC or the SYMEF or the 2DIFFET. If indeed this amplifier does sound better (in the event it does, the distinction shouldn't be clear as day and night) than the listed amplifiers then that information will be valuable data for my hypothesis.
kind regards,
Harrison.
Member
Joined 2009
Paid Member
Instead of the EF3, how about the Roender output ? - it's also a triple but uses CFP-EF instead. I tried it once (TGM3) with a CFA ips and it sounded just fine. It does require some compensation of course for stability.
Attachments
Last edited:
OS,
With your current comment about this change beating the VSSA amp could the current CFA-XH and a single pair output section be made as small as a vssa if it was on one board? Just to drive a tweeter @2.5Khz up and nothing more. Only needs low power, nothing over 30 watts would really ever be used.
The hawksford CFA will work at low rails. But , being a high Z VAS ... it
needs an EF3 or MOSFET EF2.
Either way , you will lose 6-8V from the rails. So , 40V would be needed
to get that 30-40W.
OS
Last edited:
- Home
- Amplifiers
- Solid State
- Slewmaster - CFA vs. VFA "Rumble"