Major problem fixed!

[Chris Morriss]

In message <199703241516.JAA17850 at mail.utexas.edu>, Tom Cloud
<cloud at peaches.ph.utexas.edu> writes

A straight diode across the winding (like you would use on a relay) is
no good.  Don't forget di/dt = V/L  If you use a plain diode, V will be
about 0.6V and therefore the current in the coil will take a long time
to decay, holding the injector open when you want it closed.  A 39V
Zener diode allows the di/dt to be much higher when the driver
transistor turns off, therefore dissipating the energy stored in the
injector coil quicker.

The actual time taken for the Zener itself to start conducting isn't
relevant.  It will be MUCH faster than the injector on-time.


>>I have not really looked at it, but I would not expect too much. The
>>recommened approach is to use the zener instead of a regular diode. I can't
>>remember when I stuck the scope on it and watched the input and outputs, but
>>I don't remember seeing anything fishy.
>>
>>Sandy
>>
>>At 10:57 AM 3/24/97 +1200, you wrote:
>>>I finally got my system running reasonably. Problem turned
>>>out to be that, when using a LM1949 (or in my case a 
>>>discrete copy), Using a diode to absorb the inductive kick
>>>lengthens the pulse by about 2.5ms!. Not good when idle requires
>
>why would a zener work better ???  if the coil is clamped with
>a diode, it could saturate and take a while to come back, but
>the zener will clamp it too (after it is a diode also).  The
>zener is slower than a switching diode (all diodes turn ON fast,
>its t-rr that's slow).  The only advantage I can see from a zener
>is that it limits the voltage in BOTH directions, but it's
>t-rr is VERY slow ???
>
>Tom Cloud <cloud at peaches.ph.utexas.edu>
>

-- 
Chris Morriss