My 2-cents on class-G; hope it is of interest to the slightly different class-D crowd:
Entirely by chance, I ended up driving most of my system with 2 stereo amps of the Kenwood Basic family which I understand are Class G. Aside from the advantage of running cool in the summer, I would never given even the briefest consideration to having an ungainly output design of that sort in my high-quality music system. Never.
After about 8 months of listening and system renewal testing, I find no reason to think they sound any way different than other hifi amps. (I believe there are no mysterious and ear-only differences between amps unless there also are measurable differences that distinguish them.)
You can see from my signature below that I use ESL speakers. My Kenwood M2A is driving the mid-range seven octave speakers is frequently running full blast, something quite rare this side of PA applications with cone speakers. In other words, the high-power transistors are coming into play substantially and the transition line is crossed lots.
BTW, for 40 years I have been interested in motional feedback from the voice coil of drivers. The Sigma Drive circuit of the Kenwood Basic M1 and M2 are cousins to that feedback concept... a concept sorely needed to make its way into sound reproduction.
Entirely by chance, I ended up driving most of my system with 2 stereo amps of the Kenwood Basic family which I understand are Class G. Aside from the advantage of running cool in the summer, I would never given even the briefest consideration to having an ungainly output design of that sort in my high-quality music system. Never.
After about 8 months of listening and system renewal testing, I find no reason to think they sound any way different than other hifi amps. (I believe there are no mysterious and ear-only differences between amps unless there also are measurable differences that distinguish them.)
You can see from my signature below that I use ESL speakers. My Kenwood M2A is driving the mid-range seven octave speakers is frequently running full blast, something quite rare this side of PA applications with cone speakers. In other words, the high-power transistors are coming into play substantially and the transition line is crossed lots.
BTW, for 40 years I have been interested in motional feedback from the voice coil of drivers. The Sigma Drive circuit of the Kenwood Basic M1 and M2 are cousins to that feedback concept... a concept sorely needed to make its way into sound reproduction.
Out of interest - and off topic I know - I've taken some passing interest in using feedback to "servo" drivers (played around with some ideas, but never got anywhere).
Have you had any success doing this, and if so what kind of sensing techniques have you used?
Have you had any success doing this, and if so what kind of sensing techniques have you used?
At last, somebody has shown an interest in the final frontier of open loop.
To make a long story short, I think there is no defensible method of feedback except using a dual voice coil or putting the voice coil in a Wheatstone Bridge that balances away the static coil values leaving the motional impedance.
It can only be applied to sealed boxes correctly but the result is fabulous. Ever have a driver push back at you when you shove the cone? Its ALIVE!
i have a theory that it eats room resonances.
Try not to destroy any drivers while experimenting.
Hi..I have a basic M2.. but got it 2nd hand and came with no instructions on how to use the sigma drive, only running like normal amp now, kinda curious on how much of a difference it would make( sound wise) if I could figure out how to hook it up to use the feed back part of the amp...can someone explain or post a link or something
Lucky guy.
There is a person in Colorado who keeps the flame alive. Manuals and service data are out there. See my PM.
The sigma drive is just like remote sensing on a lab power supply: is the device (the speaker terminals) getting the EXACT unflawed signal the amp is sending from the output terminals? So you connect some auxilary light gauge leads (using the sigma drive terminals) from the amp to the speaker terminals to ensure that is the case.
The mathematically adept among us will immediately recognize this drives down the amp's output impedance to a point vastly lower than it needs to be anyway.
In practice and except for leads over 100 feet long (IMHO), no important improvement likely to the ear.
The sigma drive can be made into a sort of motional feedback, I am pretty sure. If any skeptics are interested, you put a little resistor in the speaker lead and what is detected at the speaker terminals becomes the feedback signal which recognizes back-EMF from the VC, obviously.
Ben
A few years down the line, I can point to a couple of instances of these techniques being used comercially:
D&B Audioteknik use a sense-at-the-speaker-terminals system that they call SenseDrive. As bentoronto says, this counteracts the impedance of the speaker cables, presenting a very low impedance at the speaker itself. It's not a big deal for home hifi, but in PA where you've got long runs carrying high current it could become significant. Can't comment myself, but a friend who used to own their systems said it was a lot better with SensDrive turned on.
See for details:
http://www.dbaudio.com/fileadmin/docbase/TI340_1.3EN.PDF
Coda Audio (the sister compeny of driver manufacturers BMS) have gone one further and have sensor (servo) controlled subwoofers. I don't know how the sensing is done, but I suspect that it's a second voice coil - I'm waiting for a driver to get wrecked badly enough that I can pull it apart to find out... Interestingly, they also apply the technique to vented boxes, even though this doesn't make a lot of sense theoretically. The cabinet behaviour is simply corrected for in an open-loop kind of way using EQ filters before the amp.
See:
Coda Audio: White Paper
D&B Audioteknik use a sense-at-the-speaker-terminals system that they call SenseDrive. As bentoronto says, this counteracts the impedance of the speaker cables, presenting a very low impedance at the speaker itself. It's not a big deal for home hifi, but in PA where you've got long runs carrying high current it could become significant. Can't comment myself, but a friend who used to own their systems said it was a lot better with SensDrive turned on.
See for details:
http://www.dbaudio.com/fileadmin/docbase/TI340_1.3EN.PDF
Coda Audio (the sister compeny of driver manufacturers BMS) have gone one further and have sensor (servo) controlled subwoofers. I don't know how the sensing is done, but I suspect that it's a second voice coil - I'm waiting for a driver to get wrecked badly enough that I can pull it apart to find out... Interestingly, they also apply the technique to vented boxes, even though this doesn't make a lot of sense theoretically. The cabinet behaviour is simply corrected for in an open-loop kind of way using EQ filters before the amp.
See:
Coda Audio: White Paper
Tomi - Kudos for mentioning BR makes no sense with motional feedback.
Voice coil (bridge) feedback is possible using nothing more than the driver voice coil, without needing another vc.
Motional feedback is the last frontier of getting the feedback loop around the whole system. Eons ago, there were tube output transformers with purpose-built feedback windings. Pity there are no special drivers today with windings meant for feedback, just dual 4-ohm voice coils (which I suppose are OK for the purpose).
Ben
Voice coil (bridge) feedback is possible using nothing more than the driver voice coil, without needing another vc.
Motional feedback is the last frontier of getting the feedback loop around the whole system. Eons ago, there were tube output transformers with purpose-built feedback windings. Pity there are no special drivers today with windings meant for feedback, just dual 4-ohm voice coils (which I suppose are OK for the purpose).
Ben
Back when Infinity built "Servo-Statik" speakers, the 18" sub used motional feedback. The earliest ones had 10 turns of feedback wound onto the voice coil. The later ones used the voice coil former as a single turn.
To me, that always looked kind of like a transformer, and I wondered how much real motional feedback we really had.
Years later, I remember selling dozens of the little Phillips systems, with accelerometers mounted behind the dust cap. I think that's likely the easiest way to provide real feedback.
To me, that always looked kind of like a transformer, and I wondered how much real motional feedback we really had.
Years later, I remember selling dozens of the little Phillips systems, with accelerometers mounted behind the dust cap. I think that's likely the easiest way to provide real feedback.
- Status
- Not open for further replies.