555 EFI again...

[Wen Yen Chan]

A while back when I was designing a linear PWM for a resonant mode
switching supply (the Prof. would not let us use a PWM chip) my partner
and I used a current source to charge the timing cap. We used a lin. 
voltage reg. (78L05...dirt cheap $0.07) rigged as a current source. We
tied the output of the device to one end of a resistor and the 'ground'
pin to the other end of the resistor. The current was taken off of the
'ground' pin. Since the regulator maintains a constant voltage between its
output and 'ground' pin a constant current can be pulled off the 'ground'
pin. The output current is set by the (reg. voltage)/(reference
resistance) + about 25uA. Assuming the current your planning on sourcing
is much greater than 25uA the regulator's supply current can be ignored. 

We found that this setup was much stiffer than any current mirror we could
design for the same price. Temperature stability was a small nuisance but
it was still better than what we would have achieved with discrete parts. 

l8tr

Wen

On Fri, 10 Jul 1998, Walter Petermann wrote:

> > > > >http://brokersys.com/~corsaro/wiring_diagrams/555fe.htm
> > ...
> > ... The voltage across
> > > > the capacitor rises in an exponential fashion, whereas we require that
> > > > the voltage rises in a linear ramp.  This can be acheived by using a
> > > > constant current source to charge the cap.
> > ..
> > > > Chris Morriss
> > 
> > Clare Snyder wrote:
> > > How do we implement this - modification to the referenced circuit?
> > 
> 
> I wanted to see just how bad the effect of the non linear RC charging
> was...It's BAD. Over 25% error at atmospheric pressure ( reference input
> at max = 2/3*5v). I made a plot of the response and added it to the page. 
> I have a current circuit in mind..someone sugested a current mirror. 
> I'll try that and not use the extra op-amp. 
>  
>  Walter
>