On Tue, 5 Nov 2002, Michael Rigby-Jones wrote:

*>Surely the transistor will only just conduct enough to keep 0.6 volts on the
*>base, the transistor will not be able to be entirely switched off until the
*>cap is charged.  This would effectively cause the relay to switch off very
*>slowly.  I certainly don't see how this is going to cause an oscillation as
*>the cap is effectively providing NEGATIVE feedback at anything over DC.

Correct but I have to add again that relays do not switch slowly. As I
said before the magnetic force generated by the coil depends on the
inverse square of the gap and relays have a well known 'catch' and
'release' voltage, far apart from each other. Just try to make an ordinary
relay switch 'slowly'. Use a benchtop regulated psu as accurate as you
wish.

The capacitor on CB is equivalent to a capacitor across BE where Cbe ~=
Ccb * beta I think. This is most useful since it causes negative feedback
in ac there where the transistor has most gain (which in this circuit
amplifies noise). The beta is relatively low when the transistor is off or
saturated and there the capacitor plays a smaller role. So entering and
leaving saturation and off state is less slowed down than with a Cbe cap.
Also a 1uF plastic cap will have significantly lower leakage than a 100uF
electrolytic.

Peter

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.