John Curl's Blowtorch preamplifier part II

[PMA]

EMI/RFI is the key point. All of the possibilities, i.e. mains coupled, air coupled, and on signal and speaker cables. That's why we need to test the audio chain as a whole.

[mikelm]

Having said all that - even though my audio system sounds much better for all my filtration efforts - I still noticed a spike on my scope connected to my amps when I turn off a light - that was a bit depressing !

[Wavebourn]

Here is one more unexplained phenomenon:

[mikelm]

Originally Posted by PMA

EMI/RFI is the key point. All of the possibilities, i.e. mains coupled, air coupled, and on signal and speaker cables. That's why we need to test the audio chain as a whole.

Good point - I'll rethink my low noise strategy with this in mind

thx

[bcarso]

Originally Posted by mikelm

Because switching diodes ( even schottky diodes ) and transformer secondary resonance ( when diodes switch off ) generate their own RFI I usually concentrate my filtration efforts after the bridge.

Although I sympathize with anyone coming in late to this nearly-interminable thread, some of the early schematics dealing with power supplies did show some carefully-positioned common-mode chokes in "unconventional" locations in the power supply.

[Bob Cordell]

Originally Posted by john curl

Kgrlee, I think you are only making 'trouble'. For example, do you have any idea how the Hirata test works, or what might set it off to read something significant? Have you seen examples or done the test, yourself? If not, please read up, first, before demanding satisfaction.
Your PIM argument is more 'reasonable', but again there is no clear mechanism to create PIM. The input capacitance is LOW, and the input is cascoded. Where oh where can we get PIM from this circuit?

Hi John,

PIM most certainly can occur in the absence of negative feedback, and I have measured it and shown it in theory. This has been pointed out by myself and others many, many times, including on this thread. Any situation where the signal will cause phase to change as a function of signal will create PIM. There are numerous such mechanisms in most amplifiers, some stronger than others. Any amplifier in which the bandwidth is a function of signal creates PIM. Video designers have been worried about differential gain and phase for over 60 years, and those amplifiers generally did not employ negative feedback.

Cheers,
Bob

[RNMarsh]

Originally Posted by bcarso

Although I sympathize with anyone coming in late to this nearly-interminable thread, some of the early schematics dealing with power supplies did show some carefully-positioned common-mode chokes in "unconventional" locations in the power supply.

Some of the better audio designers keep on top of everything they can. I am not surprised to see that here -- the thread started in--what? 2006? Those two line noise photo's are from my archive and are dated 4/2006! Other data goes back 10+ years before that. What wasn't wideley known and still isnt appreciated is the source of all this line RFI.... it is not from the incoming ac lines and thus you need each noisy product to have a sepeerate filter on its line cord/power input (output). That means Isolation of each noisy product.... which is just about all products. Now that we have everything powered with switching supplies (and on-line wall-wart battery chargers), the RFI from equipment sharing a common line/outlets is worse than ever..... forget about living in an apartment -- the poor audiophile apartment dweller has everyone's RFI producing equipment to contend with. In their case incoming filtering IS a problem as well as isolating their own noise generators. Circuits with widebandwidth are often an outcome of great designs but have a down side, too. Naturally. never a free lunch.
[note the max number and level of unwanted frequencies are generated within the bandwidth of audio circuits. Above those freqs are filtered as required by UL, FCC et al] Thx-RNMarsh

[mikelm]

thx barsco,

I'm just about to set off on a trip to India and instead of taking a book to read in my spare time I decided to read these blowtorch threads from start to finish - I hope I have enough time !

I have experimented with common mode chokes - not sure if my use was conventional or not - my understanding was that they would help with noise on the earth and I think they did help significantly . . . . . but nowhere near as significantly as full differential chokes on all power lines after the bridge - that was a completely different sonic story and my default arrangement on all my supplies at present.

[Esperado]

Originally Posted by PMA

and on signal and speaker cables.

I was looking yesterday for shielded speakers cables (i always use them). Not easy to find, and more at a reasonable price ! I wonder why.
I thing i will shied my power cords as well.
Mr March: good idea, your "line of ac power conditioners with outlet to outlet isolation". Not to be obliged to modify each part of his equipment. Power supplies are a real nightmare, and, somehow, the key of everything.

Originally Posted by Bob Cordell

Any amplifier in which the bandwidth is a function of signal creates PIM.

Reason for my preference for negative feedback, less sensible to this problem ?

[Joshua_G]

Originally Posted by RNMarsh

This rfi on the ac line may be the real reason why people hear improvements when they use small value caps in shunt on the DC power lines. Or, better, in general with wideband low Z power supplies. A well developed ac line filter with isolated outlets often makes an audible improvment according to listeners. BTW - almost no noise of wide BW and of continuous nature comes in on the ac line. Its coming from within the home, guys. We are our own worst enemy!

RFI on the AC power lines comes both from outside and inside the house.

On my stereo setup there are 4 isolation transformers, with balanced output and Pi capacitors filter. They make marked improvement on the sound quality of my setup.

Also, I have a dedicated in-wall power line to the stereo setup, fed from one of the three power phases which go to circuits which are inactive while I listen to music (bedroom and boiler).