Perry Harrington wrote to the megasquirt list:

A "virtual" Zener is a transistor arrangement that uses a low capacity Zener
and some support components to act as a shunt voltage regulator.  It is essentially
an old school Zener regulator.  You can use low voltage components and a high
voltage transistor to shunt large voltages.

When Marcell indicated that they are using software controllable virtual Zeners,
he means they are using a virtual Zener shunt, but with logic enable.  This is
virtually identical to the logic enable dual flyback that Bruce has talked
about.
--

You know, you can eliminate your virtual zener for the power source by using
a 24v MOV.  This will shunt overvoltage transients.  You can get them in
exceedingly high current capacities so they last for a long time.

It would also be interesting to see what an MOV would do in a flyback clamping
situation.  Compare a single component to a software controlled flyback.

For those not following:

An MOV is a Metal Oxide Varistor.  It's a variable resistor that is "inert" at
voltages below it's rating.  Once the power raises to the rated value of the MOV,
it acts as a shunt.  Since it's a resistor, it is non-polarized.  It acts like
a very high current Zener diode, very fast acting.  MOVs are rated in Joules
of energy they can shunt. 24v would be a good choice for the MS.

--Perry
-------------
Perry goes further in his explanation:


Well, the real reason why the logic controlled flyback exists is because PWM
mode turns the injectors into a buck-boost converter.

The normal active flyback is activated when the flyback voltage exceeds 36v,
turning on the transistor.  Because the injectors are generating a constant
flyback current during the injection event, it causes the flyback transistor
to dissipate large amounts of energy.  This is also the cause of the PWM
limit change and VE table changes between the v1.01 flyback and the v2.2
flyback.  The active damping reduces the effective current into the injector
during PWM mode, thus instead of a linear 25% PWM hold, you need 70%.

By using the logic control, you can force the active flyback to only be
on at the end of cycle.  This causes the flyback to recirculate in the injectors
during the injection event.  When the injection event ceases, you turn on the active
flyback to quickly shunt the injector current and close it sooner.  This reduces
the power dissipation of the active flyback and increases the effective current
in the injectors during PWM mode.  You should see that a lower PWM hold value
works with the logic-active flyback.

> [MOV ] we will consider this. (btw price and availability?)

ERZ-V20D220, Digikey lists them @$.75 in quantities of 1, in stock item.
(This is a substitute part from the original recommended to me, this has
the same specs, but different part number).