LM134/234/334 Current Source

[Kashmire]

This question is for the one-and-only Nelson Pass.

I was wondering why you chose a small BJT in your current source instead of using the LM134/234/334 family of three-terminal current source devices. Seems like these devices are more stable and predictable than the Zetex BJT, and specifically made just for this purpose.

Did you try them and were displeased?

(p.s. if anyone has a LM134/234/334 SPICE model, I'd love to have it)

[Nelson Pass]

Looks like a nice enough part, although I would be concerned
about the voltage in some applications. In truth, I don't really
need that much precision, and the discretes cost less than a
dollar, where the chip is in the $5 range.

From a DIY standpoint, I think it's preferrable to assemble
circuits from simple parts - there's more satisfaction and
understanding in the result.

[Bricolo]

My theory is that ICs are taboo at Pass Labs

PS: nice nick, makes me think about Led Zep

[Nelson Pass]

Originally posted by Bricolo
My theory is that ICs are taboo at Pass Labs

Our policy is that if we have to put an op amp in somewhere,
we deliberately use one that nobody would accuse us of
running signal through.

[Bricolo]

an opamp with infinite input impedance?

[rbroer]

I've used the LM334 several times to supply a couple of mA to zener references (RC filtered after that).

But it's easy as well, to just use a n-channel jfet like J111 or 2N5485 (look for low gos figure in datasheets), or for higher voltage and currents there are these handy self biasing depletion mode mosfets like BSS129 and DN2540, requiring just one more resistor (stopper) compared to using the LM334.