Very Low Freqs do affect or hearing (& perception)

jkeny said:

Hi Orb
It's called the IQtest & I posted about it here last year

Yes, Jim showed measurements for a DACMAgic DAC compared to a Halide Bridge (asynch USB) - I believe the low freqs timing blips were mostly because the DACMagic was using adaptive USB & not for the reasons Jim gave at the time - that it was the loading on the PC

Click to expand...

That is it well remembered
Then Paul Miller with his digital knowledge and skills in measurement tool development took the investigation a stage further.
In theory the issue would go beyond USB, but scope and scale of effect in those situations are unknown.

Cheers
Orb

DonH50 said:

Just had to note a poster from Portugal citing a French audio magazine about Japanese designers -- talk about multi-cultural!

Allen Bradley? Caddock & Vishay are commonly cited film resistors now, was it one of those? Seems like there were a lot of them back then...

Click to expand...

No, I was just addressing vintage carbon resistors. I will look for the names.

Up to the 80's, the second language of Portuguese people in secondary schools was French, and most of the international press arriving in the country was either in Spanish or in French. French underground high-end received a strong influence from Japan, mostly coming from reports and translations of Jean Hiraga, son of a Japanese father and a French mother. He was exposed to Japanese triode culture in the mid-Sixties - twenty-five years before the American triode revival in the early Nineties - and brought it back to Europe. http://www.nutshellhifi.com/library/europe3.html

I have access to the full collection of the Audiophile, spanning from 1977 to 1995. We can find there many texts about our "magic" subjects, including to tests carried by Japanese manufacturers on capacitors using ultra sensitive devices that would listen to the micro vibrations of passive components subjected to electrical music signals. All of it in French, surely.

BTW, following Gary Koh advice in WBF, I replaced the expensive Vishay thick film power resistors with wire Mundorf resistors in my SoundLabs - it really improved them.

microstrip said:

No, I was just addressing vintage carbon resistors. I will look for the names.

Up to the 80's, the second language of Portuguese people in secondary schools was French, and most of the international press arriving in the country was either in Spanish or in French. French underground high-end received a strong influence from Japan, mostly coming from reports and translations of Jean Hiraga, son of a Japanese father and a French mother. He was exposed to Japanese triode culture in the mid-Sixties - twenty-five years before the American triode revival in the early Nineties - and brought it back to Europe. http://www.nutshellhifi.com/library/europe3.html

I have access to the full collection of the Audiophile, spanning from 1977 to 1995. We can find there many texts about our "magic" subjects, including to tests carried by Japanese manufacturers on capacitors using ultra sensitive devices that would listen to the micro vibrations of passive components subjected to electrical music signals. All of it in French, surely.

BTW, following Gary Koh advice in WBF, I replaced the expensive Vishay thick film power resistors with wire Mundorf resistors in my SoundLabs - it really improved them.

Click to expand...

Interesting history, thanks!

My boxes of Audio magazines were accidentally tossed in a move. I had a complete collection from the first to the very end when they closed and moved me to Stereo Review or whatever (it had a much lower technical content so I did not last long). I was sick when I discovered they had been trashed.

Thanks for the link at the beginning of this thread.

I've felt for a really long time that timing constants in amplification can be critical- and if ignored can result in LF noise. Now we can see a demo that shows that even though the speakers are nowhere near that sort of bandwidth, that 'subaudible' modulations can cause noise effects.

For this reason I have felt for a very long time that direct-coupling from input to output in an amplifier or preamp is a bad idea. This is because the bandwidth of the circuit exceeds that of the power supply, which has some very low timing constant, always greater than zero unless powered by a battery. If the circuit can exceed the LF bandwidth of the power supply, modulation effects occur. 'Motorboating' is a manifestation of this phenomena. Now it appears that if LF modulation is present, noise effects are made more audible.

We avoid this, despite extensive direct-coupling in our amps and preamps, by making sure that there is a timing constant in the circuit that causes it to cut off well above the cutoff of the power supply. The result is less LF noise at the output of the amp or preamp. I've often felt this is one reason why a tube amp can play bass better than solid state, although there are certainly other variables involved. If the amp is limited in the noise it can make in its power supplies, it will play bass better because the power supplies have greater stability. For this reason we run separate power supplies for our driver circuits in our amps, so that any noise that might appear in the output section can't affect the driver. The result is lower IMD. I'm sure this practice could help solid state amps as well, but I think there must be a culture of direct-coupling that eschews coupling caps...

IME, which please note is mostly very high speed data converters for wideband RF systems and not audio designs, one of the most insidious time constants arises from a frequency doublet in the compensation circuit of an amplifier. Pole-zero cancelling is routinely used in op-amps to extend the stable bandwidth. That is, a zero is used to cancel a pole that would otherwise roll off the response and cause instability (ringing, oscillation from poor phase and/or gain margin). There are tricks to get them very close and make them track over process, voltage and temperature (PVT) variations, but it is virtually impossible to get them to exactly cancel (i.e. exactly match pole and zero frequencies). The difference between them (the pole and zero frequencies) is usually very small but introduces a time constant related to the difference between. That small frequency difference leads to very long time constants ("settlers") in the circuit. Miss by a Hertz, and all of a sudden you have a 1 second ('ish) time constant in the circuit...

Another significant source of long-term settlers is thermal time constants. Probably not worth trying to explain here (I probably already lost everyone with the babbling above), but signals inside the device cause the temperature of what might otherwise be perfectly matched devices to vary, leading to slow (thermal) settlers.

Pulse testing is something I have always felt audio could do better at presenting. I think a lot of designers do it, it just doesn't make it to the data sheet.

FWIWFM - Don

Thanks Don, for that. I think you just put your finger on part of why feedback has such a variable reputation in amplifiers. If a designer knows how to use it, he can get good results. If he just goes with the textbook formulae taught in school, then you will have this and other issues cropping up, and causing what might have otherwise been a good design to suddenly sound bright and harsh for no reason.

Why is 250hz considered "very low"?