MPSA18 : really low noise ??

[CheffDeGaar]

Hi there,
I'm in the process of designing a low noise differential input stage for a MM preamp purpose. As I have difficulties to source low noise FETs (2SK - 389 - 369 -170 - etc...) industrial quantities (for matching), I wanted to give the bipolars a try, since I have quite a lot of them in my junkbox. As I need a high input impedance (47K), high Hfe devices seem to be a good deal.
Then I made a few searches, and came upon the well known MPSA18 (datasheet here), which seems to be quite high gain (typically above 800 for collector currents > 100uA), and with a low noise figure.
Things being made pretty easier by manufacturers , I have not been able to compare with the characteristics of the well known low noise BC550C (lower gain than the MPSA18) ( BC550 datasheet here)... But at first glance, MPSA18 seems to have lower noise than the BC550C... Maybe the simple fact that the manufacturer emphasises on the low noise of the device, providing numerous noise curves, and none for the 550C, indicates that it is really a low noise device .

So the question I wanted to ask you guys is if anybody here have used the MPSA18 in low noise apps, and how does it compares to BC550C or any low noise BJT you have tried ?

Thanks for any hint

[jcx]

both of these are required reading on bipolar tarnsistor noise:

http://www.national.com/an/AN/AN-104.pdf

http://www.national.com/an/AN/AN-222.pdf

(lthough the dc coupling to a mm cart isn't recomended on distortion grounds by most people)

a often recommended "poor man's" low noise transistor is the 2n4403 - cheap, low rbb (rx) - pnp's have a natural advantage with higher conductivity n type base

really low noise actually points to single end input - automatically 3 dB less than a diff front end

op amps can of course be very good here, opa227/8 or ad743/5

[CheffDeGaar]

Jcx,
Sorry for delayed reply...

Thanks for the links. I knew the first one (great explanations) but was unaware of the LM394 Application Note. Thanks.

I've read somewhere else on this forum the 2N4401/4403 comp. pair was highly regarded for its good noise/price ratio. But I've been unable to get noise characteristics for these devices from their datasheet (tried On, Fairchild and Philips...) Any hint ?
The other thing bothering me here is the relatively low Hfe of these BJTs. I have to find a compromise between the gain of the diff pair and its input impedance. With low Hfe devices, I have to increase the emitter resistance to increase the input impedance, but increasing Re lowers the gain... Moreover, I want to parallel 2 of these devices to reduce noise, and the Re has to be twiced to achieve the same Zin... That's why I found the MPSA18 to be interesting...

Thanks again for your reply.