Negative Group Delay Circuit

Unfortunately they found it had already been reported in last month's journal. ;)

I once wasted an entire day trying to do that (back when I was 21 or so). These guys don't seem to have got the memo about causality.

Love to see them build one that works.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

No mistake. They really do appear to advance the signal. They don't actually advance it, but they do appear to.

Jerry

-- Engineering is the art of making what you want from things you can get.

The real input is when the switch is closed.

As the article itself points out, neither signal is actually delayed at all. Both appear as soon as the switch is closed.

The apparent advancement of the output pulse arises because the observer is attaching an inappropriate significance to its shape.

Now, you might think you could put a switch after the signal labelled input, and thereby achieve a reverse causality. But it wouldn't work as you expect. You'd see a discontinuity in the output signal that occurs at the time you open the switch, but not before.

Sylvia.

OMG! Current leads voltage in a capacitor! Who'd a thunk it? Fig 7 shows a square wave input, but fig 8 is not square and therefore not the input as claimed. Fig. 6 has two resistors both labelled R1 and two capacitors both labelled C1, those guys must have had too much saki.

I've read about something very similar, and it's not good:

Why didn't I learn THAT one at age 5 in Sunday School?

...

:-)

John

That's old hat. All it takes is a dollop of thiotimoline. (Google The Endochronic Properties of Thiotimoline et seq.)

...

Unless you're older than I guess, it was old hat then too.

Jerry

-- Engineering is the art of making what you want from things you can get.

Okay, four months ago. Must be a better time machine than I thought. ;) (That's the one we ripped up here a few weeks ago, I remember--or is it just deja vu? Time machines are so confusing.)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

Things are causal This I know 'Cause the Second tells me so...

-- Les Cargill1

I thought a lot about this, and I don't think this is just phase- advance, but common sense tells you it must be. What is not clear is the 2 LEDs in teh circuit, is the second one supposed to light before the first?

Hardy

: IEEE Journal of Selected Topics in Quantum Electronics - IEEE J SEL TOP QUANTUM ELECTR , vol. 9, no. 1, pp. 43-51, 2003

I thought a lot about this, and I don't think this is just phase- advance, but common sense tells you it must be. What is not clear is the 2 LEDs in teh circuit, is the second one supposed to light before the first?

Hardy

=========================================== Sure, no reason it shouldn't. step 1: circuit is steady state. step 2: delta V appears across the non-inverting input and inverting input. step 3: through amplification the output LED lights. step 4: through the feedback capacitor the "input" LED lights. step 5: circuit reverts to steady state as there is no DC in the feedback to the non-inverting input but there is to the inverting input, R3. step 6: the output LED extinguishes. step 7: the "input" LED extinguishes.

The "input" LED actually gets its power from the arse end of the op-amp, not from any input source. Fig 7 doesn't show that.

This is just some students scratching out a term paper, obviously anyone can duplicate their results. I'd deduct marks for incorrect labelling of R1/C1.

As others have said, no effect without cause.

.

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

t.

can

2. >

actually it's an IEEE trans paper. I am willing to bet you have never even written a paper let along had a full one in IEEE.

Hardy

actually it's an IEEE trans paper. I am willing to bet you have never even written a paper let along had a full one in IEEE.

Hardy ========================================= What the f*ck does what I've done have to do with you "don't think this is just phase-advance, but common sense tells you it must be"? I've explained it to you, if you have problem with the explanation then says so, otherwise stick your sour grapes up your arse. I'm willing to bet you don't think at all.

...

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

ck

, e

=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

Because you're talking complete bollocks! It's a phase-advance circuit. Your heath-robinson up a bit down a bit explanation is a red herring and of no use to the central argument. The circuit is in fact a form of predictor and the output of the pulse is before the input as claimed.(but not uncausal). It's not current leading voltage, it's voltage phase-shifted to be more positive than the input. as claimed, the group delay is in fact negative. For example, if you take a pure time-advance exp(jwT) then this has unity magnitude and a positive linear slope for phase. (wT). Over a limited freq range this is what they have (their amplitude is not constant however ).

To overcome this problem one can do the following.

Take a pure integrator and put a pure time-delay exp(-jwt) in the feedback path. For high-enough gain at low frequencies only, the closed-loop transfer function will (by the action of feedback and high gain) be approx exp(+jwt). Of course you must maintain stability which doesn't give you much bandwidth to work with at all. If you then try and stabilise the circuit, you need phase-advance which cancels out the phase-delay of the time-delay. If you put a sine-wave in you get an output which appears to be before the input. This is just a predictor of sorts, but deterministic. It's not uncausal since the output only appears to be advanced in time on the scope just as with the case of a phase-advance (for a single sine wave). As the authors point out, it is "anticipation".

Hardy

1) Publish questionable research. 2) Seek funds for commercial project base upon #1. 3) Advertise 'negative group delay' speaker cables on Amazon. 4) ????? 5) Profit!

--
Paul Hovnanian     mailto:Paul@Hovnanian.com
------------------------------------------------------------------
6 out of 7 dwarfs are not happy.

Because you're talking complete bollocks! ==============================

Oh, ok. The circuit output occurs before the input and causality is denied. Happy now?

*plonk*

Do not reply to this generic message, it was automatically generated; you have been kill-filed, either for being boringly stupid, repetitive, unfunny, ineducable, repeatedly posting politics, religion or off-topic subjects to a sci. newsgroup, attempting cheapskate free advertising for profit, because you are a troll, because you responded to George Hammond the complete fruit cake, simply insane or any combination or permutation of the aforementioned reasons; any reply will go unread.

Boringly stupid is the most common cause of kill-filing, but because this message is generic the other reasons have been included. You are left to decide which is most applicable to you.

There is no appeal, I have despotic power over whom I will electronically admit into my home and you do not qualify as a reasonable person I would wish to converse with or even poke fun at. Some weirdoes are not kill- filed, they amuse me and I retain them for their entertainment value as I would any chicken with two heads, either one of which enables the dumb bird to scratch dirt, step back, look down, step forward to the same spot and repeat the process eternally.

This should not trouble you, many of those plonked find it a blessing that they are not required to think and can persist in their bigotry or crackpot theories without challenge.

You have the right to free speech, I have the right not to listen. The kill-file will be cleared annually with spring cleaning or whenever I purchase a new computer or hard drive. Update: the last clearance was 19/08/10. Some individuals have been restored to the list.

I'm fully aware that you may be so stupid as to reply, but the purpose of this message is to encourage others to kill-file fuckwits like you.

I hope you find this explanation is satisfactory but even if you don't, damnly my frank, I don't give a dear. Have a nice day and f*ck off.

Depends what you mean by "output" and "input". It's not a mistake in the paper. Here is a quote..

"One may think that making use of this twist it is possible to send information to the past despite of the causality. Of course this is wrong"

They go on to explain why it's wrong.