Re: dc/dc converter 12/200V

On Mon, 18 Apr 2005 19:45:20 -0400, Mike Monett <no@xxxxxxxx> wrote:

>John Larkin wrote:
>
>[...]
>
>> >Power dissipation is much lower... device is always saturated.
>>
>> Only if you time the base drive right. Your circuit really needs "on"
>> timing adjustment as a function of battery voltage, or alternately a
>> current sensor.
>
>Jim's circuit really is very elegant, John. Do it in SPICE - then you see
>how well he has done.


I hardly need Spice to understand how a thing like this works.


>You don't need a current sensor or fancy drive electronics. Just make the
>base current sufficient to make a spark at worst-case ambient temp,
>battery voltage, and include a fudge factor for old ladies who never got
>around to switching their summer weight 40W100 oil.


The inductor current builds as something like Vb/Ton, and the spark
energy goes as I^2. So low battery voltage, as in cold cranking, would
seem to need a longer (not higher current) base drive. Spice that!


>
>> >Simple control requirements allow use of star-wheel-magnetic pickup OR
>> >points.
>> >
>> >Inductive storage requires either a shaped pickup or complex control
>> >system for low dissipation and avoidance of coil current all over the
>> >map.
>>
>> That's easy. Your circuit needs base timing, too. 50 cents worth of uP
>> looks like a pretty good deal to eliminate a big inductor and a big
>> cap.
>
>Then how do you get a spark? Where is the energy storage?

Store it in the coil, like people have been doing for a century now.

>In Jim's circuit, all you need is a fixed width base drive to saturate
>the tranny.

Not if you want to get to work every morning.

>His circuit wastes the first trigger pulse. After that, 400V is sitting
>on the collector waiting for the leading edge of a trigger to dump the
>charge into the ignition coil.

Actually, it's tricky to get high stepup ratios in circuits like this.
As the boost ratio increases, energy gets lost in the various
parasitics. You're suggesting a 100:1 boost (4v cold cranking, 400v
peak out) and that's pushing things.

>I love circuits like that! Brilliant. Elegant. Even I can understand how
>they work. Bulletproof. Minimum parts count. Pennies for the components.
>
>Why do it any other way?

To save a lot of size and money; neither the L nor the C will cost
pennies. Does any carmaker actually do it this way?

John


.

Quantcast