I think that the schematics michel mentiones are the ones in product datasheets and are only simplified type. You can get the idea of the cicuit. I have seen the original schematics some time and there really isn't anything mysterious. Just the same schematic + output over current protection included to the output stage. The output transistors and the pre output stage transistors are from Sanken.
Ergo
Ergo
Elso,
if you want to experiment with CFA look at the Hiraga on Geoff site. Easy to build, hard to get to work (the PSRR is low to nil). I also have an old Aloia article (1981) about a project that looks very 'Accuphase-esque' there was a commercial version of this amp called, I think, ST-200. This latter design was easier to tame. The author claims a max value of 14dB feedback at any freq. I would have posted already if I had a scanner. I built them both and I love them both.
if you want to experiment with CFA look at the Hiraga on Geoff site. Easy to build, hard to get to work (the PSRR is low to nil). I also have an old Aloia article (1981) about a project that looks very 'Accuphase-esque' there was a commercial version of this amp called, I think, ST-200. This latter design was easier to tame. The author claims a max value of 14dB feedback at any freq. I would have posted already if I had a scanner. I built them both and I love them both.
Check
http://perso.wanadoo.fr/jm.plantefeve/
I would like to hear from anyone who has experimented with current-feedback amplifiers. I am in the process of building one.
Michel
I agree with your comment for the need of a good power supply, do you see any other pitfalls in current-feedback designs. I am aware of their speed and potential for low distortion but that is not usually an indicator of good sound.
Jam
http://perso.wanadoo.fr/jm.plantefeve/
I would like to hear from anyone who has experimented with current-feedback amplifiers. I am in the process of building one.
Michel
I agree with your comment for the need of a good power supply, do you see any other pitfalls in current-feedback designs. I am aware of their speed and potential for low distortion but that is not usually an indicator of good sound.
Jam
The alexander amplifier started it. But i wanted to have a discrete stage instead of the opamp used in the alexander stage.
The inputstage wich is a buffer looks a bit like the output bufferstage in the Jfet/bipolar amplifier i have made.
It can be made a lot simpler. And i did have problems with the inputbuffer ... It worked like a "mixer" you know the ones they user in FM receivers!!!! I have made lowpass filter in the input to get loose of the problem.
The two currentmirrors in the alexander are betaenhanced with an ekstra transistor. The currentmirrors output is then cascode coupled.
The outputstage is a normal darlington stage with 3 pairs of transistors (2SA1302 and 2SC3281).
It is compensated to drive capacitive load like 100uF||8Ohm without oscillation of course only 2 volt out, not to destroy the amplifier. If i remember right it rings only 2 times on a squarewave and with about 20% overshoot.
BW is around 100KHz.
The sound : Alive with good response i the Bass(1KVA + 2 times 47000uF does it all), detailed Mid and high.
A weak powersupply does have a impact on the sound.
I did have the chance to compare it with a Thule (I think it was a IA100?) on some cheap JBL and got a better Bassresponse.
I have also tried it in compare with an old int. (E-202? it is a long time ago!) Accuphase again same chance like the Thule but also better mid and high. They was compared on my "M1 Super".
If have the design?? yes!!! but i will not give it away.. There is something to chance ... It is not DIY friendly!!!!
Sonny
The inputstage wich is a buffer looks a bit like the output bufferstage in the Jfet/bipolar amplifier i have made.
It can be made a lot simpler. And i did have problems with the inputbuffer ... It worked like a "mixer" you know the ones they user in FM receivers!!!! I have made lowpass filter in the input to get loose of the problem.
The two currentmirrors in the alexander are betaenhanced with an ekstra transistor. The currentmirrors output is then cascode coupled.
The outputstage is a normal darlington stage with 3 pairs of transistors (2SA1302 and 2SC3281).
It is compensated to drive capacitive load like 100uF||8Ohm without oscillation of course only 2 volt out, not to destroy the amplifier. If i remember right it rings only 2 times on a squarewave and with about 20% overshoot.
BW is around 100KHz.
The sound : Alive with good response i the Bass(1KVA + 2 times 47000uF does it all), detailed Mid and high.
A weak powersupply does have a impact on the sound.
I did have the chance to compare it with a Thule (I think it was a IA100?) on some cheap JBL and got a better Bassresponse.
I have also tried it in compare with an old int. (E-202? it is a long time ago!) Accuphase again same chance like the Thule but also better mid and high. They was compared on my "M1 Super".
If have the design?? yes!!! but i will not give it away.. There is something to chance ... It is not DIY friendly!!!!
Sonny
Could it not be fun to make a new design from the scratch! ... We could all post our idears .. So that we in the end can end up with a "diyaudio CFB amplifier" ...mmhh Then we need a "project folder" on this webside.
It is only an idear. I would start redraw the inputstage and the current mirror and then post it?
Sonny
It is only an idear. I would start redraw the inputstage and the current mirror and then post it?
Sonny
Sonny,
Great idea!
If you could e-mail me at jsomasundram@msn.com there are a few schematics on current feedback designs I could send you.
I had looked at the Alexander design and thought about replaceing the output stage with mosfets but never got around to building it.
Jam
Great idea!
If you could e-mail me at jsomasundram@msn.com there are a few schematics on current feedback designs I could send you.
I had looked at the Alexander design and thought about replaceing the output stage with mosfets but never got around to building it.
Jam
pitfall
Hi there,
I have build both Hiraga designs as shown on Geof's page and those are to my knowledge current feedback designs. Both sound very good to me but i will not comment on quality in relation to other designs. I have found one pitfal though and that concerns a bandwith limitation cap in the feedback path. I've build the Hiraga 20 w several times and in several forms using the various schematics available on the net. Some of these have no bandwith limitation and some do. This led me to experiment and for now i have to conclude that quality of the cap in the feedback path is very influential. One schematic of the hiraga 20 watt is shown on the bonavolta site and shows a 2200 pF cap in the feedback path. Experiments (by ear) leeds me to conclude that other ways of bandwith limitation seem to be better for this sort of design. A standard Cdom cap was much less intrusive for example. Thus i think that if you want to use current feedback you should keep it away from bandwith limiting arrangements.
greetings Joris Z.
Hi there,
I have build both Hiraga designs as shown on Geof's page and those are to my knowledge current feedback designs. Both sound very good to me but i will not comment on quality in relation to other designs. I have found one pitfal though and that concerns a bandwith limitation cap in the feedback path. I've build the Hiraga 20 w several times and in several forms using the various schematics available on the net. Some of these have no bandwith limitation and some do. This led me to experiment and for now i have to conclude that quality of the cap in the feedback path is very influential. One schematic of the hiraga 20 watt is shown on the bonavolta site and shows a 2200 pF cap in the feedback path. Experiments (by ear) leeds me to conclude that other ways of bandwith limitation seem to be better for this sort of design. A standard Cdom cap was much less intrusive for example. Thus i think that if you want to use current feedback you should keep it away from bandwith limiting arrangements.
greetings Joris Z.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.