The Jack Bybee NAQ (Never-Asked Questions)

[scott wurcer]

Right, said Fred : New Page 1
(better skip the first 10ch )

Ahh, a hold over from the propagation of sound under water thread.

 

[john curl]

Scott, please don't make a fool of yourself. Read Terman: 'Electronic and Radio Engineering' pp 159-165. Surely, your office library has a copy. Put up, or shut up, please. ;-)

 

[syn08]

The Bybee devices appear to reduce a certain kind of 1/f related modulation noise. This means that you have to have signal, to measure the noise, not just a static noise measurement, and apparently the ear (at least from my experience) is sensitive enough to hear the difference. It is darn difficult to measure, but apparently still possible with the right test conditions, even with my out-of-date HP3563.
Let me explain the problem. Most testing for noise does not require a signal present, so you just measure the Johnson or 1/f noise generated from the resistance or other sources. This does not bring out the noise that the Bybee addresses. Only with a signal, that then, has to be removed almost completely, does the noise difference show up on test instruments. The ear apparently does this naturally. Of course, many think that the ear can't hear such things. Oh well. Back to your zip cord.

I am ready to shoot the messanger, for the crime of Physics slaughter, in which supposingly the messanger has a degree.

Anything you say, can and will be used... I may upgrade the charge to first degree murder.

 

[syn08]

Scott, please don't make a fool of yourself. Read Terman: 'Electronic and Radio Engineering' pp 159-165. Surely, your office library has a copy. Put up, or shut up, please. ;-)

Yes, and call Dick Sequerra

 

[PMA]

This is a good thread. All of that Blowtorch OT's finally at the right place. Thanks, Andy.

 

[john curl]

No, Scott, don't waste your time calling Dick Sequerra, even though he would tell you also that it is a cavity resonator. He likes you for who you are, not what he could teach you. He would have you more comfortable, and in the dark, rather than uncomfortable in the light.

 

[andy_c]

This is a good thread. All of that Blowtorch OT's finally at the right place. Thanks, Andy.

You're very welcome Pavel. And you nailed one of my points right on the money in this post.

 

[MJL21193]

Bybee a Trekkie?

 

[BrianDonegan]

I've got a little pocket book entitled "The Worlds Worst Cars". The AMC Gremlin is #1 in the "Design Disasters" chapter.

Presumably, goin’ by your obvious lack of taste, if you were looking for a larger car at the time you would have bought the equally hideous Pacer?

They say you can't polish a turd, but you sure can try:

Gremlin X- Home Page

 

[PMA]

You're very welcome Pavel.

Thanks - and I have appreciated your sense of humour in the 1st post of this thread. I had not laughed that much for long

 

[pooge]

LAMO

I just looked that one up in the index - Chevrolet Citation (1980-1986), #7 in the "Badly Built" chapter.

Problems:
Engine oil consumption
Overly tight brake hoses prone to rupture (brake failure)
Oversteer (+overly light steering)
Rust (missing underseal)

But NO FEEDBACK!!!

 

[cuibono]

... He would have you more comfortable, and in the dark...

Hopefully, hes not the only one Scott



I thought you guys were supposed to be the most knowledgeable about this technical stuff, and this is what we get?

 

[Edmond Stuart]

This is a good thread. All of that Blowtorch OT's finally at the right place. Thanks, Andy.

 

[PMA]

Doug's 4th edition of Amplifier Handbook

I would like to cite from 4th edition of the Self's book:

/beginning of citation/ "This offers us the interesting possibility that crossover distortion - which has hitherto appeared inescapable - can be reduced to an arbitrary level simply by paralleling enough transistors. To the best of my knowledge this is a new insight" /end of citation/

Page 143, Elsevier 2006, 2008. Audio Power Amplifier Design Handbook.


This is a typical stage 3

 

[Steve Dunlap]

Andy,

I will apologize here also, but the explanation will be posted in the Krill thread.

 

[Edmond Stuart]

noise

[snip]
Most testing for noise does not require a signal present, so you just measure the Johnson or 1/f noise generated from the resistance or other sources. This does not bring out the noise that the Bybee addresses. Only with a signal, that then, has to be removed almost completely, does the noise difference show up on test instruments. The ear apparently does this naturally. Of course, many think that the ear can't hear such things. Oh well. Back to your zip cord.

Hi John,

Normally I trust measurements more than my ears. But during reading your comment, I suddenly remembered the moment when I said **** you to my boss! It happened a long time ago. I was asked to design a pink noise generator. So I built a pseudo random generator, took a (24 bits?) SR plus a couple of XOR gates and wired the stuff as prescribed by the recipes from the mainstream literature in order to get the max. length sequence (i.e. 2^N-1).
What happened? Every 3 seconds I heard a kind of 'hiccup'. So I modified the FB circuit. Now the hiccup has gone and the generator sounded like real noise. However, the max. length sequence was shortened by 25%. So what, I thought. But the PhD's yelled at me: that's not random! Nevertheless, that noise generator went into production.

C.E.

 

[john curl]

I also hoped that you learned something, PMA, you are pretty good, when I can finally make a 'breakthrough' with you, better than most.

 

[scott wurcer]

No, Scott, don't waste your time calling Dick Sequerra, even though he would tell you also that it is a cavity resonator. He likes you for who you are, not what he could teach you. He would have you more comfortable, and in the dark, rather than uncomfortable in the light.

I know them, typically single crystal pure sapphire resonant disks are needed for Q > 10,000. In this case they are a bargain. That thing with all those surface irregularities and a piece of ordinary #14 wire through it has a Q of many thousands at GHz frequencies? Why insult our intelligence this way?

I know what the bear is doing now.

 

[scott wurcer]

"Rust (missing underseal)"

That was the next one, 1988 fire engine red Suburban. The rust out started as you drove off the lot. This one was cursed too, delivered with a door that could not close, it lost its transmission after a month (had to tow a trailer 135mi. in first gear).

 

[syn08]

Hi John,

Normally I trust measurements more than my ears. But during reading your comment, I suddenly remembered the moment when I said f@ck you to my boss! It happened a long time ago. I was asked to design a pink noise generator. So I built a pseudo random generator, took a (24 bits?) SR plus a couple of XOR gates and wired the stuff as prescribed by the recipes from the mainstream literature in order to get the max. length sequence (i.e. 2^N-1).
What happened? Every 3 seconds I heard a kind of 'hiccup'. So I modified the FB circuit. Now the hiccup has gone and the generator sounded like real noise. However, the max. length sequence was shortened by 25%. So what, I thought. But the PhD's yelled at me: that's not random! Nevertheless, that noise generator went into production.

The alternative could be fixing the FB logic; what you got (the hiccup) is a classic problem of logic hazard, pretty common in implementing linear feedback shift registers with long GF(2) primitive polynomials.

Happened to have the same problem in a student project (1979). The teacher showed me how to optimize the FB implementation to avoid hazard. Not always (and/or necessary) easy.