Well
I don't know if it is their high impedance input, or their inductive nature (or both) that make these AVCs easy on peaking up extraneous signals. But they are more picky on that compared to active (semiconductor) preamplifiers or low resistance volume potentiometers.
I called Magic Bus to ask him do a xtalk listening test on a set of AVCs, this time having both inputs connected to a source and sending music only to one channel.
He called me a few hours later, reporting that there is no interchannel xtalk issue anymore with this arrangement.
Success.
George
I don't know if it is their high impedance input, or their inductive nature (or both) that make these AVCs easy on peaking up extraneous signals. But they are more picky on that compared to active (semiconductor) preamplifiers or low resistance volume potentiometers.
I called Magic Bus to ask him do a xtalk listening test on a set of AVCs, this time having both inputs connected to a source and sending music only to one channel.
He called me a few hours later, reporting that there is no interchannel xtalk issue anymore with this arrangement.
Success.
George
Hi egellings
It is complicated, no doupt!
In theory, AVCs (and TVCs) work as signal attenuators as well as impedance transformers btn source and load.
They multiply the load impedance that reflects back on the source.
I can't tell if it is this or their added distortion profile that makes an impact on the sound but my listening impression is they are detectably different to potentiometers of ladder step attenuators.
Alas, it is difficult and costly to find AVCs for having a listening evaluation yourself...
George
It is complicated, no doupt!
In theory, AVCs (and TVCs) work as signal attenuators as well as impedance transformers btn source and load.
They multiply the load impedance that reflects back on the source.
I can't tell if it is this or their added distortion profile that makes an impact on the sound but my listening impression is they are detectably different to potentiometers of ladder step attenuators.
Alas, it is difficult and costly to find AVCs for having a listening evaluation yourself...
George
Always something new to learn! I would never had thought to test cross talk this way. It explains why there was no impact on the soundstage. Simply, there is no cross talk when it is connected to both channels. These things work different from ordinary attenuators in many aspects making it worth to investigate.
I just realized I haven't posted impressions from George's latest AVC with 3k4 turns of 0,1 mm wire on 80% Ni core. I felt that the added turns restored the missing body of the previous prototype on the same core. Still, it added even more airiness too! There is something magical happening in the recording ambience. Very rich backround. Although there is no doubt it is artificial, it is not at all annoying. No fatigue, no forced presence, just the air. The absolute remedy for dull systems but also welcome to mine, a rather bright one.
I think that for this core material, this is the golden ratio of turns, no more, no less. It would be interesting to see what is the sweet spot for other materials.
I just realized I haven't posted impressions from George's latest AVC with 3k4 turns of 0,1 mm wire on 80% Ni core. I felt that the added turns restored the missing body of the previous prototype on the same core. Still, it added even more airiness too! There is something magical happening in the recording ambience. Very rich backround. Although there is no doubt it is artificial, it is not at all annoying. No fatigue, no forced presence, just the air. The absolute remedy for dull systems but also welcome to mine, a rather bright one.
I think that for this core material, this is the golden ratio of turns, no more, no less. It would be interesting to see what is the sweet spot for other materials.
(again)
Me thinking loud... Why the steel core didn't suffer cross talk even with the clumsy way I tested it? Is it reasonable to think that it doesn't have high enough inductance? But the bass was beefy. What should we expect from more turns? If I was to apply my experience from the two prototypes with the 80% Ni core, I would say it should take less turns but then inductance should reduce moving it too far from the Ni core. And what about the 49% Ni? I was thinking that turns should be reduced from 7k5 to much less if it's going to behave like the 80%, but now I don't know what to suggest...
My impression is that a high-quality rotary switch with 1% metal film resistors will pretty much be optimal if impedances don't present a problem requiring a different solution.Hi egellings
It is complicated, no doupt!
In theory, AVCs (and TVCs) work as signal attenuators as well as impedance transformers btn source and load.
They multiply the load impedance that reflects back on the source.
I can't tell if it is this or their added distortion profile that makes an impact on the sound but my listening impression is they are detectably different to potentiometers of ladder step attenuators.
Alas, it is difficult and costly to find AVCs for having a listening evaluation yourself...
George
egellings - My recent quick and dirty experiment with some Jensen autoformer's I had laying around showed improved low level detail and dynamics, better decay, and more resolution compared to the 50k stepped attenuator currently in my preamp. It was promising enough to make the investment in some Slagle units, which should be a big step up (Permalloy lams, more favorable inductance, etc).
George - Sure. This is my first experience with these things. For decades it's always just been some flavor of resistive attenuation like most people have. Dave sends his out at max inductance (1X1 stack) these days from the looks of it... ~150H. I'm assuming that's because he has no way of knowing what people are going to be putting in front of the thing. I use a single source that has low output impedance so I restacked the lams for less inductance (reads 30H on my LCR meter). It's certainly easy enough to experiment with...
In my experience most economical LCR meters measure at 100Hz though some also measure at 1khz and 10khz. From the size of the Intact Audio
AVC I think they are quoting 150H at 100hz. As George says the measured inductance varies with frequency because the permeability of the core
material varies with frequency. In the case of 80% nickel its roughly inversely proportional to frequency so 150H at 100hz would be around 300H
at 50hz and so on. Other materials like amorphous vary in a different way.
AVC I think they are quoting 150H at 100hz. As George says the measured inductance varies with frequency because the permeability of the core
material varies with frequency. In the case of 80% nickel its roughly inversely proportional to frequency so 150H at 100hz would be around 300H
at 50hz and so on. Other materials like amorphous vary in a different way.
Could it be that input capacitance was causing a HF roll-off when using the 50K pot, while the transformer's much lower source impedance was not affected as much by that? The roll-off caused by the RC pair would produce a dulling of the sound and loss of HF detail. Transformers have distortion causing mechanisms that are not present in resistive devices like potentiometers. The magnetic core materials in the xfmr are responsible for that. An air-core xfmr would not have those problems, but they would not work well at audio, especially in the low frequency range. Also, if there is ringing in the xfmr, that could brighten the high frequency components, producing a false sense of HF definition & detailing. A transformer can work well, but so can a pot. I have Jensen SUTs for my MC cartridge in my turntable, and they do a good job. I have a high-quality ALPS brand pot for my home-brewed line stage's volume control and it does a good job. I measured the line stage's frequency response with the volume set to a low level, where it will have the most attenuating effect on HF response, and it was flat about to about 23 or so kHz, well above human hearing's ability. So yes, the pot does affect performance, but from an audibility standpoint, it's inconsequential. A pot can have higher noise level than a transformer (if the xfmr is not picking up hum from stray AC magnetic fields) but again, in a line stage, it's too little to worry about. From an audibility standpoint either approach can work well.
Last edited:
is well consolidate 10years at last ...that the AVC , the Slage ones , is the best pot we can use...
the new ones is what ZM use...