"Tony Moody" <[email protected]> wrote: > But that kind of circuit is a real pain (if not impossible) to get > reliable low current.
>OK, >Tony Tony, The LM317 has a minimum load current of 3.5mA. This is far too high for a typical cs generator. It is also limited to input/output voltage differential of 40V max. This is too low for some cs generators and the LM317 will saturate at the start of the brew. It will mess up the Faraday calculations. However, it is trivial to get a stable, precision temperature-compensated constant current source at microampere current levels. All it takes is two resistors, a TL431, and a high voltage npn transistor such as a 2N5550. The 2N5550 will take up to 160V, which is far more than most cs generators will need. I use it with excellent results. It is perfectly stable without bypass capacitors and needs no compensation. The TL431 has internal compensation that determines the open loop gain margin and stability. The load is isolated from the loop since it is in the collector of the 2N5550, and anything you tack on the output cannot affect the loop gain. The circuit is shown in Figure 29, "Precision Constant-Current Sink", on page 18 of http://www.tkhifi.com/datablade/tl431.pdf The equation is Rs = 2.5 / Ic where Ic = desired cell current Rs = current setting resistor Note, the current source shown as Figure 28, "Precision Current Limiter", on page 17, won't work as a current source. The parts only cost a few pennies. Octopart will help you find vendors. Here is the 2N5550: http://octopart.com/parts/search?q=2n5550 Here is the TL431: http://octopart.com/parts/search?q=tl431 Thanks, Mike -- The Silver List is a moderated forum for discussing Colloidal Silver. Rules and Instructions: http://www.silverlist.org Unsubscribe: <mailto:[email protected]?subject=unsubscribe> Archives: http://www.mail-archive.com/[email protected]/maillist.html Off-Topic discussions: <mailto:[email protected]> List Owner: Mike Devour <mailto:[email protected]>
