Noise figure measurements--tire kicking requested

So should I search for Stirling cooler or superconductor filter?

Yes, a friend of mine has a business putting on "science shows" for schools and private parties. With small groups the LN2 instant ice cream is the most popular part of the show, since it is damn good ice cream :-).

He always starts the LN2 demo with a discussion of boiling followed by boiling LN2 in a tea kettle heated by air. One time his lovely assistant poured in a bit too much, and the (no doubt quite brittle) kettle ruptured with a bang. After a brief pause to attend to a bit of bleeding the show went on.

That is primarily to limit heat gain, but the small opening does result in enough flow velocity of evaporated gas to prevent ingress of air.

Perfectly miscible. The air separation plants I am familiar with separate them with a fractional distillation column very similar to the ones in petroleum refineries. (I don't know if any more recent plants separate with membranes prior to liquefaction, couple of decades out of date here.)

I recall a LOX handling safety movie with some interesting demos, such as dropping a LOX trailer tow link on an oily rag - BOOM - flipped the heavy link back vertical (these things vent cold O2 constantly, no spillage required) and some accelerated combustion demos, most notably the bra burning demo. After the narrators teaser, a well stacked model walked on and deftly removed her bra without removing her sweater, lit it and tossed it near a small LOX container. I'm pretty sure it burned, but I only remember the model, and our disappointment at the bra removal method :-).

Got it, thanks. Interesting paper--he used a transformer to provide both the negative FB and impedance transformation, so he could get low noise temperatures even with tubes.

Cheers

Phil Hobbs

Dry ice is easy to come by. I noticed it at a local grocery store just last week.

We got a big block for a environmental test once upon a time, and used the excess in the punch. Alcoholic punch, that is--we threw a party. It made a nice brew.

Bubble, bubble, toil and trouble...

-- Cheers, James Arthur

Backfill with COO or NN.

Our prof used to gargle LN2 to impress the freshers...

Hi Phil,

Does UHF injection reduce this much, or does it just reduce the mode hopping?

Thanks,

UHF injection gets rid of mode hopping but otherwise I think it makes the laser noisier in general--essentially it has to start up from noise on each cycle of modulation, so there's jitter in both amplitude and phase associated with the very brief period when it crosses threshold.

It's much like a superregenerative receiver, and the famous rushing sound of superregens is amplified thermal noise due to their extremely high and signal-dependent gain.

Cheers

Phil Hobbs

Just got a quote from the local Praxair distributor--$5.50 per litre to fill a 25-litre dewar, plus $25 to deliver it. (And $800 to buy the dewar, of course.)

I didn't know it was possible to gold-plate nitrogen.

Cheers

Phil Hobbs

days

I didn't realize you needed that much. Ouch!

One of the amateur science guys had a few sources--Shawn Carlson or maybe Bill Beatty--but I can't find the article.

"Liquid nitrogen is very common and easy to obtain, if you know where to look. It is used by hospitals, physicians, cattle breeders, universities, factories, and welding shops. Liquid nitrogen can be obtained from these sources for a minimal cost."

-- Cheers, James Arthur

It's almost free compared to helium.

I have seen price differentials of better than 4:1 between various industrial gas distributors in my area, with J.F. Martin having the lowest prices in the Philadelphia area. You might be able to find a significantly lower price from another distributor.

Yeah, but spending $1k for a noise figure measurement is a lot. At least you guys get to fly with bush pilots and other manly stuff like that. ;)

Cheers

Phil Hobbs

Phil Hobbs wrote: : On 12/15/2011 1:57 PM, Spehro Pefhany wrote: : > It's almost free compared to helium. : Yeah, but spending $1k for a noise figure measurement is a lot. At : least you guys get to fly with bush pilots and other manly stuff like

Actually, the quote Phil got is quite steep indeed, close to LHe price. The old rule-of-thumb is that LN2 costs about the same as milk, and LHe costs about the same as whisky (if you don't return the gas).

Regards, Mikko

days

Yowser! That's a lot. The delivery fee sounds standard, (it's what we pay, but they will fill up any size dewar (up to 100 liters) at the loading dock for $75. Hmm guess that would be $3/l if I only had a

But I'm not accepting the $800 Dewar price from Dr. Phil 'I got this deal on ebay' Hobbs.

Are you seriously thinking of buying a dewar? I'll see if I can find the name of the 'claimed' several month hold time one I heard of.

George H.

;)

I'm not planning to pay $175 to get it filled, either. I'm phoning around the local welding shops, etc, to see if there's anybody local I can get it from. If so, I can get a nice little 5- or 10-litre dewar and fill it up when I need it.

I'm trying not to buy one, but it might come to that. It would allow me to refrigerate the amp as well as the load, just to see if that helps the noise.

Cheers

Phil Hobbs

in

just

ch,

the

wise

ble

I've had this 'crazy idea' for a while, of measuring the noise from a light bulb, with the current cranked up. And measuring the temperature, by matching colors with some black body standard. (I'm not sure how well a light bulb follows a black body curve.)

Thanks for the reference to actively cooled loads.

I don't suppose anyone has a link/ reference to an article that shows an actual circuit. I've been trolling the web with no luck. It's something I've never quite understood. Is it related to active damping of mechanical systems?

George H.

• You look through a lens at the light and match up a color wheel by eye.

George Herold wrote: > [...]

It's not so mysterious. Say, you want an amplifier with a defined input resistance to terminate a source with impedance Rs. Consider the two arrangements below:

R=Rs -----/\/\/\---- | | | === | Rs | |\ -----/\/\/\--+-------| >------- Uo | |/ A Uin | ===

R=Rs(1-A) -----/\/\/\---- | | Rs | |\ | -----/\/\/\--+-------| >------- Uo | |/ A Uin | ===

Say that the amplifier block has gain A, largish and negative. Both amplifiers have gain Uo/Uin = A/2 and input impedance Rs. Considering only the noise contributions of the resistors, the input-referred noise level of the top circuit is sqrt(4kT Rs/2). For the bottom circuit, it's sqrt(4kT Rs)/2.

That latter value is what the noise would have been with a noiseless termination resistor, i.e., only the noise of the source resistance is seen.

I used this in the head amplifiers for beam position position monitors in an accelerator here at CERN. You can see a snippet of the schematics here: . It ended up having an input referred noise level of 290pV/rt(Hz).

Getting this right in wide-band amplifiers is fiddly.

Jeroen Belleman

=A0 =A0 |

=A0 =A0=3D=3D=3D

(1-A)

=A0 =A0 |

=A0|

Great! Thanks Jeroen. It'll take me a bit to 'process' that.

George H.