From: "Jeff Verive"
Subject: Re: jellybean audio transistors
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
Date: Tue, 14 Jan 2003 17:07:50 -0600
NNTP-Posting-Date: Tue, 14 Jan 2003 17:07:51 CST
What's your power dissipation going to be? Unless your peak Ic is very
rarely approached, a TO-92 is going to get pretty warm for Vce of 5V or
higher. Of course, push pull helps limit the duty cycle for the current
excursions, but It still seems that the design might be pushing the thermal
limit of a TO-92.
I presume these are driving the load directly? If so, what's the load
impedance and Vpk-pk?
"Tom Bruhns" wrote in message
> "Walter Harley" wrote in message
> > I need a complementary pair of transistors for a low-power push-pull
> > follower stage, operating at audio frequencies. Specs I need are
> > >300mA; Vceo >30v; hfe ~ 100 at Ic = 30mA or so; Pd low enough that a
> > case is acceptable (though not quite ideal). I'm currently using 2n2222
> > 2n2907, mostly because that's what I "imprinted" on in the 1980's.
> > at some spec sheets, I noticed that 2n4401/2n4403 have very similar
> > and identical prices.
> > My question is threefold:
> > 1) With specific respect to these two pairs (2n2222/2n2907 and
> > 2n4401/2n4403), is there some meaningful difference that I'm missing?
> > seem so similar that I'm not sure why they bother manufacturing both
> The 4401/4403 have a bit better holdup of hFE at high current, and a
> bit worse flatness of hFE at lower current.
> > 2) There's a zillion general-purpose complementary NPN/PNP pairs out
> > too many to study all the datasheets. Maybe the technology has improved
> > over the last 20 years? Should I be looking to adopt a new set of audio
> > jellybeans? If so, can someone help me narrow the search by tipping me
> > to what the good lines are these days?
> I think the 4401/4403, 2222/2907 and 5089 (or 5210)/5087 are all still
> good choices for medium-to-low current audio work. There are also
> some good 2S and BC numbers, but the 2N numbers are most readily
> available to me so I use them. The fact that all are readily
> available in SMT tells me that they are still all used in reasonably
> high volume.
> > 3) For this relatively non-critical application, does it even matter?
> > Should I just stick with what I've got?
> I'd stay with them if they work for you. If low noise or low
> distortion are particularly important, you might be able to do better,
> but you likely should also look very carefully at other factors like
> circuit topology and selection of passive components and layout.