Op-Amp noise figures: Generally >=10dB?

We've got arms AND legs if necessary. :-)

I did that yesterday -- he's the guy who's working on getting the lower-frequency specs; he said he had to contact another engineer to get'em.

Nice! Do you use the pool much?

---Joel

Nope. We never watch any shows, or TV in general. What was the question where he blew it? Do you remember his name or ranch?

:-)

Only thing is that many pricey parts vanish once a key account decides not to use them anymore. Or a defense contract runs out. But I do not know the availability history of Sirenza/RFMD since I never used their parts so far. It did happen with other mfgs though. Remember the days when many companies had these fantastic super-muscular buffer amps? One of those plus some ferrite and enameled wire and you had a great RF stage. Gone :-(

That's when my support meter clock begins. Let's see. Infineon is now at

We used to. But since global cooling is happening out here usage has shrunk to four months. Used to be at least six. If I had another choice today I'd buy a house without a pool.

I'm told that RFMD is rather (in)famous for paper launches and generally poor availability if you're not a six-digit account. However, this is coming from someone who dealth with them the better chunk of a decade ago now, so (wishful thinking here) perhaps things have improved in the interim. I actually haven't mentioned to this fellow that RFMD acquired Sirenza last year -- wouldn't want to unduly sully Sirenza's name just yet!

---Joel

Well, Digikey has the SBF-4089 in stock so that's a good sign. Seems you have to buy 1000 though and with that purchase the 1000 they have in stock would be gone.

I've used SGA-3586 as a wideband amp. It's the lowest-noise 50-ohm mmic I know of, and, unlike most other parts, it's a very good 50 ohm match. It's SiGe, low power but a bit fragile... it needs input protection if it's connected to the outside world.

John

On May 7, 2:25 pm, John Larkin wrote: ...

Not as broadband (only a decade) but lower noise: HMC-ALH482. Narrower band and even lower noise: HMC617LP3 (0.5dB NF; octave bandwidth). Better than 20dB input return loss; better than 10dB output return loss.

HMC659: at low freq., 19dB gain, +25dBm P1dB, NF 2.5dB typ., input return loss 19dB typ., output r.l. 18dB typ.; capable of DC-15GHz. This puppy is just a tad expensive, though.

Cheers, Tom

Wow, impressive part!

Right. Anything much below 3dB NF must be tuned [1]. Some discrete phemts hit 0.4 dB in their sweet spots. That's a noise temp around

John

Hi all,

"John Larkin" wrote in = message news: snipped-for-privacy@4ax.com...

ohm

Maybe not -- at least at lower frequencies. See:

--=20 Regards, Howard snipped-for-privacy@ix.netcom.com

Looking at the gain and input impedance curves, it looks like this has a transformer in the front end, which performs the impedance optimizations like a tuned circuit would.

I think you can also parallel a lot of now-noise (like sub 1 nv/rthz) jfets and get arbitrarily low noise performance, up until the gate capacitances getcha. Again, a distributed amp gets interesting; I wonder if anybody makes a super-low-noise distributed jfet amp.

John

You can do what you describe, however, that first stage ought to be a fet or beter a phemt, have seeen < .2 dB NF on avagos latest stuff. But even a 2n 4416 will be lowr than the op amps , which can be used in second or third stage to really porvide gain. Try to get fist stage gain 10-12 db to swamp out alll the stuff that folllows. Good luck.

Marc