John Curl's Blowtorch preamplifier part II

[Wavebourn]

There's line noise in my apartment now that gets into everything and has rendered low noise measurements almost impossible. It is synchronized to the mains but has obviously a lot of information embedded. I have no idea what it is, but I hate it.

May be it is signal from power counters that forced me to remove from garage wall a floodlight with motion sensor?

 

[31697B]

I found ---

That would be the Richard Gray Power Company.

Stereophile reviewed it some time back and had a sidebar which showed that it wasn't storing any energy to speak of under AC conditions (it's a big *** inductor so natch).

At one of the shows, Gray was demonstrating it's ability to store massive amounts of energy by energizing it with a ***DC*** current, then flipping a switch which shorted a lightbulb across it, the flyback voltage causing the lightbulb to flash.

se

When ever extreame descriptions are used on a product (always seems to come from High-End Audio), I manage to get one and measure it.

I found that the two large (many Henries) coils did not do any filtering and the outlets were bussed together (no isolation).

Being the test junky I am, I put two TEK photometers and a new German PC based photometer system on my plasma screen and ran it thru its paces. There was no measured change what-so-ever in any of the colors or black/white level nor could I see any change at all when the R.Grey unit was used or not used.

It might have a slight regulating affect more like smoothing. But I didnt see any video affect nor measure it on a stable ac line.

Thx - RNM

 

[Joshua_G]

Shocking !

It isn't obvious.
Not always the more expansive items sound better to me.
For instance, I found out that certain power cables which are priced quite reasonably sounded to me (with the proper plugs attached) much better than other much more expansive cables (with the same plugs attached).

Definitely, not always and not necessarily expansive products sound better than equivalent moderately priced ones.
(What I didn't find, so far, is very cheap products that sound very good. I did find fairly cheap Chinese power plugs that will sound very good to those people who prefer a brighter sound to their systems).

 

[Joshua_G]

FYI -- Some of the biggest and most common noise generators are often found in cheap lamp dimmers in the home (without zero crossing switching and rfi filters). And, many switching power supplies built as 'wall wort' supplies are another very common source. In fact, many more products are using switching PS.. some which draw considerable current - like TV's. So, be sure when you inject noise into ac line that there are not already a great number of noise frequencies there already to mask your test. -RNM

Indeed.

I found out that I need no extra source of injected noise to the power lines in order to note which products have beneficial effect to the sound of my setup.

 

[Steve Eddy]

Sounds like he missed part of EE101...

Or perhaps not. Maybe he just knows his audience really well.

When I was visiting my friend that wrote for TAS I was complaining about one of the reviews and he picked up the phone and called Tom Miiller thinking I would be impressed. I proceeded to tell him that the concept of a mechanical diode in the context of the SOTA turntable made no physiical sense, waste of time since a basic intuitive feel for physics is a burden to these folks.

Ah yes, the fabled mechanical diode (read cone) that magically drains away all vibration, never to he heard from again.

se

 

[Steve Eddy]

When ever extreame descriptions are used on a product (always seems to come from High-End Audio), I manage to get one and measure it.

I found that the two large (many Henries) coils did not do any filtering and the outlets were bussed together (no isolation).

Yeah, about the only filtering it would do would be from the parasitic capacitance of the windings.

Being the test junky I am, I put two TEK photometers and a new German PC based photometer system on my plasma screen and ran it thru its paces. There was no measured change what-so-ever in any of the colors or black/white level nor could I see any change at all when the R.Grey unit was used or not used.

It might have a slight regulating affect more like smoothing. But I didnt see any video affect nor measure it on a stable ac line.

Makes sense.

se

 

[gpapag]

For what had been claimed Zobel network would work better: a capacitor with resistor in series.

This approach (multiple such parallel networks, adopted by RNMarsh in his patent US 5,260,862 A-C power line filter) is acknowledged as of a different methodology and discussed in columns 3 & 4 of Gray’s patent.

V = L*dI/dt is always valid, regardless L is linear or not.

Yes.
V= L*dI/dt is the mechanism by which the choke will function by counteracting any sudden change across it. The emf has a negative sign.

Ref. L nonlinearity: I should have been more clear : The L of the choke is a function of the permeability mu of the “iron”. Close to saturation, mu drops very low, thus L drops low , same with emf too. End result, miniscule reaction.

But RNMarsh reports having tested the device here and found no appreciable effects. No use for me to theorize anymore.


George

 

[scott wurcer]

Shocking !
Then again, if I were in the oiling business, I'd have your picture gold-framed on my wall.
I can hear it when my shoelaces are too tight.

(thanks for posting the schematic, Mr Wurcer)

No problem, any questions just send them my way.

 

[john curl]

Demian (audio1) and I are working together on a servo 'latch-up' problem.
I designed a dual servo (common mode and differential mode) as usual, as it would be used in the Parasound JC-2 or the CTC Blowtorch (more or less).
The greatest difference is using a QUAD IC (4134) rather than 2 duals (2 X 2134) as per usual. I am NOT using one of the sections. How would you suggest we tie the pins of the unused op amp, to keep it from effecting the others, especially on turn-on?

 

[SY]

Ground the two inputs, let the output float (or tie it to ground with a 2k-10k resistor).

 

[john curl]

Thank you, SY, for your input. Now what does Scott think?

 

[Wavebourn]

I would connect output to inverting input, and ground non-inverting input. Or may be connected both inputs to the rail, depending on topology.

 

[john curl]

Thank you, Wavebourn.

 

[bcarso]

What Wavebourn said.

 

[abraxalito]

I would connect output to inverting input, and ground non-inverting input.

What Wave said except I'd use a 10k resistor between non-inverting in and gnd.

 

[31697B]

If you are having latching problems -- maybe exceeding common-mode range - do this:

Place output to (-) input, and ->
Place (+) input at zero via center tap of a pair of equal value (10K?) resistors - one 10K from the + supply and one 10K from the - supply. R values not critical. -RNM

 

[PMA]

Now what does Scott think?

Free lunch?

 

[Mooly]

How would you suggest we tie the pins of the unused op amp, to keep it from effecting the others, especially on turn-on?

Parallel them with the used section and keep this opamp away from high impedance sources if possible.

 

[john curl]

Thanks everybody. It appears that what we did, initially, is OK.

 

[jneutron]

NO, I made a mistake, it is probably 50 uH. Sorry about that. You could have HAD me, Scott. Keep sharp! '-)

Darn, I missed a shot also...

About 150 micro Henry for 2 wires, 20 meters. How long wires does Billy Gates have?

Sigh..

You sure, Wave? Rule of thumb, 7nH per cm of conductor in free space, I make it 15uH for that scenario ...

Frank

Sigh..

Let's check.

An externally hosted image should be here but it was not working when we last tested it.

For 20M of 2mm thick copper wire it is 39386.7389298843 nH
2 wires in a cable will give approximately 4 times more. 140 micro Henry.

My rule of dumb gives closer result.

Did I miss anything?

Here is online calculator: Inductance of a Straight Wire: A Calculator

Sigh..

Boy! This is great. We have 15uH, 50uH, 140uH and 150uH. Good thing you all dont do design work for a living.

Yah, but we have plenty of choices...

Join the company. What will be the number if to design for living?

Here ya go..this is the result of applying the Terman equation to a 12 guage wire.

A single wire inductance can only be calculated as the internal inductance of the conductor, which is 15 nH per foot. The external inductance of any wire has to include the return path of the current. Without knowing the return path, one cannot calculate total inductance.

John

BTW, the vertical scale is in NANOHENRIES PER FOOT of conductor pair.. #12 romex is probably about .4 inch spacing.