Audibility of output coils

Status
Not open for further replies.
Jump to Latest
#124
john curl said:
Remember engineers, you have to subtract the real component from the inductive component to get a better estimate.
Click to expand...

John, assuming we're both talking about the DM88, you need to re-check your inductor calculations.

I re-checked mine. 0.1 Ohms in series with 0.78uH works out to 0.14 Ohms magnitude at 20 kHz as reported by Stereophile. I double-checked the original back calculation with a SPICE sim of 0.1 Ohms and 0.78uH.
 
#126
andy_c said:


John, assuming we're both talking about the DM88, you need to re-check your inductor calculations.

I re-checked mine. 0.1 Ohms in series with 0.78uH works out to 0.14 Ohms magnitude at 20 kHz as reported by Stereophile. I double-checked the original back calculation with a SPICE sim of 0.1 Ohms and 0.78uH.
Click to expand...


Hi Andy

Something a little more relevant to the discussion; The DM88's frequency response plot:

806H88FIG01.jpg



Note that the output impledance of the Halcro results in a ~0.2dB ripple when driving a reactive loudspeaker-simulating test load.

How distractingly audible are those amplitude variations going to be?

Cheers,
Glen
 
#128
Glen,

I'm guessing not much since I've yet to see a speaker with a response within 0.2dB.

I'd be more concerned about the non-linearity of the coil itself. Any magnetic and/or electrically conductive material in the vicinity of the coil will induce high frequency non-linearity due to the skin effect. Any high order distortion products would then combine in the coil to produce potentially audible IM products. Of course, if the amplifier has no high order products then this shouldn't be a problem.

This is just a theory although it makes some sense to me. I'd be happy to hear alternate views however.

Tony
 
#129
Cellardoor said:
Glen,

I'm guessing not much since I've yet to see a speaker with a response within 0.2dB.

I'd be more concerned about the non-linearity of the coil itself. Any magnetic and/or electrically conductive material in the vicinity of the coil will induce high frequency non-linearity due to the skin effect. Any high order distortion products would then combine in the coil to produce potentially audible IM products. Of course, if the amplifier has no high order products then this shouldn't be a problem.

This is just a theory although it makes some sense to me. I'd be happy to hear alternate views however.

Tony
Click to expand...


G'day

I essentially agree with everything you have said. The design of the inductor is obviously important. An interesting thing to note about the Halcro DM88, with its ~0.78uH output inductor (as calculated by andy_c) is it’s vanishingly low measured distortion.

This obviously demonstrates that a properly designed output inductor need not be a limiting factor in an amplifiers distortion performance.

There are much, much bigger inductances to worry about in a speaker system (100’s of uH’s in tweeter crossovers and 1000’s of uH’s in woofer crossovers) that are much harder to make very linear and decouple from immediately surrounding electromagnetic influences.

Cheers,
Glen
 
#130
I stick to 0.5uH, even for the DM88. I originally calculated the DM58, but the 88 has essentially the same inductance. They measured the DM88 through 6 feet of wire! .1 ohm was the base resistance and .14 ohms was the 20kHz resistance. Please don't second guess, unless you really know that you have found something important.
 
#131
Where a NFB loop is capable of further reducing high audio frequency distortion at the output node due to a series output choke being fitted, I would concur with your 0.5uH figure.

Though this same figure is often already provided by internal loudspeaker wiring and amplifier to loudspeaker interconnects, even when using an amplifier as a loudspeaker sited monobloc.

So is the internal inductor really necessary ?
Maybe better as a third terminal or user option.
 
#133
John Curl,
---I am now stuck with a .3 ohm resistor in each output of my power amps.---

John Linsley-Hood used a 0.22 Ohm in most of his amp design, all of them have impressive square wave responses with a 2.2µF load.

The technology of the resistor may not less benign than the technology of a foot of wounded copper wire. I remember Peter Baxandall said having been very surprised when designing an audio test intrument, to note that a low value power resistor introduces an unexpected distorsion.
 
#134
I started building amps without coils at the output more than 25 years ago. I have never encountered problems, even with speaker wires of approx 10 meters length. I did not listen to capacitors but speakers however ! 😉

Fun aside: Apart from ringing - which would be high in frequency, much higher then the ringing of some metal domes and probably better damped due to the ubiquitous parallel resistor - the resulting series resonant circuit can sink quite a lot of current at frequencies above the power bandwidth.

Regards

Charles
 
#135
I like 1 ohm resistors. However I am now stuck with a .3 ohm resistor in each output of my power amps.
Click to expand...

I like that. It makes the DF independant of frequency, output voltage, output current and phase angle of the load, which is not that self-evident for circuits with high DF in gerneral. A consistent DF and good load stability is more important than a high DF derived by very high NFB IMO.

Regards

Charles
 
#136
Hi phase_accurate,

When NFB amplifiers have such a low output impedance the loudspeaker system (and series output choke) can become part of a high 'Q' series resonant circuit at some frequencies.

Your series R busts that Q, and so prevents serious (audible) interface reaction.
Bi and tri wiring with separate R (if necessary) per circuit would be even better.

Cheers ........ Graham.
 
#137
andy_c said:
John, I get 0.78uH. I got this from:
L = sqrt(0.14^2 - 0.1^2) / (2 * pi * 20e3)
R = 0.1 and
sqrt(R^2 + X^2) = 0.14 @ 20 kHz and
X = omega * L (neglect the parallel R)
This is for the DM88
Click to expand...

Hi Andy,

As the published impedance figures might be unreliable, you could also estimate the inductance by winding an equally sized coil by yourself, just as I did.
Based on pictures and known outside dimensions of the enclosure, I get: If N=6, length = 6cm, diameter = 2.5cm and wire diam = 2.5mm (actually, I used 1.8mm), Such coil has a measured an inductance of 0.47uH.

Cheers,
 
#138
#139
Nordic said:
http://translate.google.com/transla...&hl=en&ie=UTF-8&oe=UTF-8&prev=/language_tools

I have built this L and C meter... it is very accurate... (+-1%) and very easy to make... (I literaly drew the pcb with a black marker pen). No more inductor guessing.. and handy for matching caps and inductors, as it can show value, as well as diffirence between value and another previously measured component...
Click to expand...

Hi Nordic,

Ask Bruce Candy for sample of his inductors, so you can give us the exact answer 🙂

Good luck,
 
#140
Hi John C

I discovered as far back as 1985 that I did not like the sound of output coils either.

So none of my amplifiers have ever used them.
My latest amplifier which develops 2kw into 4 Ohms has no problem with stability at all and it has no output coil.

I don't know what all the fuss is about, a respected designer tries to give some free information and all the Protagonists come out of the closet and slap him around the head.
Even if you don't agree with him, you guy's could at least pretend to be nice. Instead of flaunting your arrogant attitudes about the place.



So John C, I'm with you mate! :smash:

Flame Suit On:flame:
 
Status
Not open for further replies.
Top Bottom