From: Joerg on
John Larkin wrote:
>
> I recall somewhere, maybe in this ng, the procedure for making an
> ideal diode in LT Spice. But I can't find it now. Help!
>
> The ideal diode should be a stock pulldown part. I wonder why it's
> not.
>
> And why can't I change the voltage of one of their pull-from-the-list
> zeners? That would sure be convenient.
>

Ok, now that you have the ideal diode, a zener can also be sim'd by a
voltage source and one of your ideal diodes in series with it. Plus a
non-ideal one across if your zener needs to also be used in reverse. The
voltage source lets you dial in the zenering voltage. There's probably
an easier way but I've done it a few times where that was good enough.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
From: oopere on


John Larkin wrote:
> On Sun, 04 Apr 2010 08:02:46 -0700, John Larkin
> <jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote:
>
>>
>> I recall somewhere, maybe in this ng, the procedure for making an
>> ideal diode in LT Spice. But I can't find it now. Help!
>>
>> The ideal diode should be a stock pulldown part. I wonder why it's
>> not.
>>
>> And why can't I change the voltage of one of their pull-from-the-list
>> zeners? That would sure be convenient.
>>
>> John
>
> Never mind, I figured out the ideal diode.
>
> Click the D thing and add a diode to the schematic. Rename it Didl.
>
> Then add a spice directive string nearby...
>
> .model Didl D(Ron=0.1 Roff=1G Vfwd=0)
>
> with whatever values you think are close to ideal. For some reason, my
> sim slows down if I use too low an Ron value. The 0.1 is good enough.
>
> The ideal zener is then just a voltage in series with this diode.
>
>
> I'm doing what the controls guys call a "high select", to create a
> voltage that's the greater of two inputs, to switch my power supply
> from constant-current to constant-voltage mode. I did a clever thing
> with two switch elements, but it did funny things sometimes and was
> too hard to think about.
>
>
> John
>
>

I did not check out your solution yet. I take N=0.001 in the diode
equation to approximate the ideal one.

..model Didl D ( N=0.001)

Pere