Thoughts Concerning Cordell, Otala, and Gilbert papers

[PB2]

The easy way "to see " and study the EMF , is using a speaker with two separate bobbin .(like some car speakers)

You inject your signal in one of them ...and you can "see" the EMF in the other coil.

Yes I see your point, there would also be some coupling due to mutual inductance in addition to the motional voltage.

 

[PB2]

I'd like to discuss one of Mikeks' points from the original thread.
from post #246 in this thread:
http://www.diyaudio.com/forums/showthread.php?postid=491090#post491090

The moral of the story really is that there is no 'new' audio amp. impairment out there that does not also manifest itself as an increase in THD.....

viz: In my view, if a design's THD is less than 1000ppm across the audio band, for any conceivable load driven to rated voltage swing, then there can be no doubt that the design is, for practical purposes, free of ALL known error.

I agree with your claim mikeks, to some extent in theory but have you considered real world circuits and test conditions? I didn't notice you offering any more detail concerning your position that low THD is all that's required, did I miss it somewhere on the forum?

I believe that there are cases where IM testing is required to uncover the underlying non-linearity, wouldn't you agree?

 

[jcx]

absolutly positively theoretically possible, but practically how important?

i thought i sorta addressed that...

http://www.diyaudio.com/forums/showthread.php?postid=492499#post492499

but i had to re-explain it a few pages later

anyone interested really must get Cerzwinski's mutlitone testing paper

but furter research needs to be done, multitone amplifier measurements could perhaps put a practical limit on "hidden" intermodulation distortion, i wouldn't be suprised if it was limited to as little as 10 dB over the prediction from single tone distortion identification of the principle nonliearities - the GedLee metric is probably not too far wrong in that aspect

 

[PB2]

Re: absolutly positively theoretically possible, but practically how important?

i thought i sorta addressed that...

http://www.diyaudio.com/forums/showthread.php?postid=492499#post492499

but i had to re-explain it a few pages later

anyone interested really must get Cerzwinski's mutlitone testing paper

but furter research needs to be done, multitone amplifier measurements could perhaps put a practical limit on "hidden" intermodulation distortion, i wouldn't be suprised if it was limited to as little as 10 dB over the prediction from single tone distortion identification of the principle nonliearities - the GedLee metric is probably not too far wrong in that aspect

Yes, sorry, that's such a long thread that I went back to it and notice that most of the points I wanted to make were made later on.

My points are fairly basic, I've not dug too far into those articles you mention, I don't need convincing, I believe that intermod testing is the best way to go. Yet a few keep asking isn't THD across the audio band enough? I just wanted to offer my position for those who keep asking. It is often taught that a test signal of large enough peak amplitude, covering the frequency of interest will uncover the degree of a weak-memoryless nonlinearity, however this assumes wide bandwidth at the output of the device under test. There are many devices - narrow band amplifiers, class D power amps, older 1X audio D to A converters, with output filters near 20 kHz, some with steep "brick wall" filters. If we try to measure the system above 10 kHz with simple THD tests the harmonics are filtered by the output filter and we get a reading that does not reflect the non-linearity of the system. Some say it does not matter since the harmonics are above the audio band but if we had done a multi tone test we'd see that the non-linearity produces intermodulation sum and difference sidebands within the audio band that are indeed audible.

I've heard accounts of equipment that measured very well with THD tests but that produced clearly audible "birdies" (intermodulation distortion) with complex music input and perhaps pilot tone leakage, IIRC.

I believe that PIM is a bit more complex and the spectrum is related to FM analysis and Bessel functions. I did notice yours and andy_c's excellent work and understand that 3rd harmonic is an indicator, but that is not the complete spectrum of a phase modulated distortion as I see it, just an indicator that some form of distortion is present and therefore the PIM does not "hide" or go undetected. Again, this of course assumes no output filter which is not always the case.

I'm also interested in speaker design and multitone testing is routine since there is the problem of the limited bandwidth of the radiator (cone or dome) acting as a filter often after the non-linearity. Here's a practical example by Linkwitz, his non-linear distortion test:
http://www.linkwitzlab.com/mid_dist.htm
Here's another related to test equipment limitations:
http://www.libinst.com/close3rd.htm

Thanks jcx I did notice your answers and simulations after I made the previous post.

 

[mikeks]

Is interesting that the usual miller capacitor in the VAS, is the stabilization method ,that gives the worst null.

A capacitor to ground in the VAS stage produce a much better null.

I suspect there is more to this than meets the eye (ear?)...

Specifics required...

 

[Tube_Dude]

Specifics required...

Yep!...

 

[darkfenriz]

Hi
could anybody send me or link me to the papers mentioned in title?

 

[andy_c]

Hi darkfenriz,

Shoot me an email and I can send them to you. If I just click the email button, the resulting form doesn't allow me to send attachments. At any rate, I'd need to check with you to see how large an attachment your email is allowed. Since there would be three attachments, sending them all at once might overflow your inbox.

Edit: The Gilbert paper is now available on the web here.

 

[darkfenriz]

darkfenriz@interia.pl

thank you

 

[mastertech]

mikeks []

mike ks-

Originally posted by Tube_Dude

Is interesting that the usual miller capacitor in the VAS, is the stabilization method ,that gives the worst null.

A capacitor to ground in the VAS stage produce a much better null.
I suspect there is more to this than meets the eye (ear?)...

"Specifics required..."

specifics of "specifics required" required

 

[PB2]

I was recently looking at some Krell schematics and noticed that most of their designs, KSA-50, the later 100, use an emitter follower between the diff amp and VAS, outside of the Miller loop just as I described. I do not think Self explores this topology, and I've not seen it discussed in the Journals. Here's a schematic for the Krell KSA-50:

http://www.diyaudio.com/forums/attachment.php?postid=490398&stamp=1097610320&

Pete B.

Hi Andy,
I follow Self's logic and agree when Cdom closes the loop, but when I first saw the emitter-follower used there many years ago (I think it was in Wireless World) I don't think there was a Cdom cap in the circuit so it's reasonable to view it as an emitter follower "buffering the input stage". The diff amps stage gain will go up, it will see a more linear load, the emitter follower will reduce distortion by driving the non-linear parasitic Miller capacitance from a lower source impedance. Self's claim of others being wrong is only true when his configuration is assumed. I like the emitter follower idea in both cases.

I've always thought it would be a good way to lower the HF distortion in the Leach. Stability would have to be considered and it would raise the open loop gain when the intention was low open loop gain. Have you considered this?

Do you still have your Leach amps?

 

[Ed LaFontaine]

bump from the archives

 

[Bob Cordell]

I was recently looking at some Krell schematics and noticed that most of their designs, KSA-50, the later 100, use an emitter follower between the diff amp and VAS, outside of the Miller loop just as I described. I do not think Self explores this topology, and I've not seen it discussed in the Journals. Here's a schematic for the Krell KSA-50:

http://www.diyaudio.com/forums/attachment.php?postid=490398&stamp=1097610320&

Pete B.

Hi Pete,

I must admit this topology looks strange to me. It is usually a good idea to put an emitter follower in front of the VAS transistor to increase current gain and avoid the nonlinear influence of the VAS transistor's collector-base capacitance. However, The usual role of Miller compensation cannot be performed if the Miller capacitor is returned to the low-impedance point of the emitter follower. I wonder if there is something else at work here. Has anybody simulated this? The component vales seem a little hard to read.

Cheers,
Bob

 

[PB2]

Hi Pete,

I must admit this topology looks strange to me. It is usually a good idea to put an emitter follower in front of the VAS transistor to increase current gain and avoid the nonlinear influence of the VAS transistor's collector-base capacitance. However, The usual role of Miller compensation cannot be performed if the Miller capacitor is returned to the low-impedance point of the emitter follower. I wonder if there is something else at work here. Has anybody simulated this? The component vales seem a little hard to read.

Cheers,
Bob

Hi Bob,

Sorry for the delay here, I don't get email notices as far as updates go on this thread. Yes, I agree with everything you say there but I was just wondering if I missed something or if there might be any advantage. As you say wondering if there might be something else going on. Looks like emitter degeneration is 100 ohms on the diff pair and VAS, Cdom cap is shown as 390 pF but I seem to remember a comment about the correct value being 39 pF, not positive about it.

No I've not simulated it, I believe that some on here have.

Thanks!