To Phill Hobbs

Smirk >:-} ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

Really Jim ... i can't resist anymore ... i'm laughing, l'm laughing ... and even crying of laugh !

Habib.

heeemmm ! :-)

I spent quite a bit more than that on it. ;)

If you write out the expression for the actual photocurrents, with all the relevant constant factors, you'd get something like

I_signal = A (1 + d cos (omega t)) I_comp = B (1 + d cos (omega t + delta_phi))

We assume B > A so the noise canceller can work. The d factors are the same by hypothesis, because this is a common-mode term, i.e. laser noise and not signal.

The feedback loop will force the DC terms to cancel by adjusting the gain in the compensation path to be A/B. Thus the difference photocurrent is

I_diff = A [ (1+ d cos (omega t)) - (1 + d cos (omega t + delta_phi))]

= A d (cos (omega t) - cos (omega t + delta_phi))

With me so far?

We can expand the right-hand cosine and collect terms:

I_diff = A d(cos(omega t)(1-cos(delta_phi)) + sin(omega t) sin(delta_phi)).

The first term is quadratic in delta_phi, and so can be neglected to leading order. Sin(omega t) and cos(omega t) have the same amplitude, so the minimum normalized residual signal ratio (in amplitude) is

dInorm/ = sin(delta_phi).

In the calculation in the paper, I made the phase shift symmetric, i.e. the phase shifts of I_signal and I_diff were +-delta_phi/2, respectively. That's what leads to the slightly different expression (2 sin (delta_phi/2)), but in practical cases the difference is insignificant.

Cheers

Phil Hobbs

i)).

t.

Hey Phil,

You made many hypothesis in your paper than anyone, even those are experien ced, could easily understand .... it looks like as an IEEE transactions art icle by and to peers,

Habib.

All of those articles are from technical journals or conference proceedings. The OPN one is from the Optical Society's member mag, and is sort of Scientific American level.

I've had hundreds of discussions with people about noise cancellers, and once they got used to the idea, they understood fine. I've been trying to help out, but if it's not worth the effort to you, so be it.

I mean, it's only three transistors and two op amps.

Cheers

Phil Hobbs

ienced, could

and to peers,

gs. The OPN one is from the Optical Society's member mag, and is sort of Sc ientific American level.

once they got used to the idea, they understood fine. I've been trying to h elp out, but if it's not worth the effort to you, so be it.

Sure we'll make it ! The effort to be familiar with.

Many electronics are hard to understang "in fine" although the structure re mains simple. Yes the structure is simple but electronics is far away when we are talking about phase shifting on two signals correlated and noise cancelling .. tha t is not so trivial to prove the "cancelor electronics" limits.

We'll read more articles ! Thank you, Habib.