Several months ago there was a thread about the landing gear on 320/360s
failing to extend.  One of the theories that had been around for a
number of years was that an internal leak was allowing pressure to
escape from the high to low sides in flight thus pressurizing the low
side thus inactivating the down-circuit.  Occasional in-flight blips of
the hydraulic pump may have contributed to this theory.  While both
examination of the pump hydraulic schematic and testing have shown that
such leaks cannot build up any significant pressure in the low side
circuit, a satisfactory explanation was lacking.  Some recent
observations have led me to a theory as to what might be going on and I
offer it up to the list for discussion.
In the way of background information:  Several years ago I also built a
test fixture that duplicates the aircraft hydraulic system in order to
test anything from individual cylinders to the whole aircraft hydraulic
system from outside the plane.  All of this has been most valuable
during troubleshooting.  My plane has pressure gauges for both high and
low sides.  I also have override switches which allow me to bypass the
pressure switches from the panel.
Recently I was able observe the failure of my gear to extend.  Prior to
extension the high side was charged and the low side was zero.  Upon
selecting down, the gear dropped just enough to hear the increase in air
noise.  At this point both pressure gauges were steady and showed
several hundred psi.
To explore what might be going on here we need to jump inside the pump
for a moment.  The pump retains pressure in both high and low circuits
with check valves at the outputs.  Under normal operation, one check
valve is closed while the other is held open by the spool valve.  When
the pump direction is reversed, the spool valve is pushed to the
opposite side, allowing the currently open check valve to close and
forcing the opposite check valve open.  This takes quite a push.   It
must force open the check valve against the fluid pressure of the
hydraulic circuit currently in use.  The unique aspect during extension
is that the fluid in the high side circuit is subjected to the potential
energy stored in the landing gear.  As soon as the spool valve cracks
the high side check valve, a large volume of hydraulic fluid surges back
into the pump pushed along by three gear legs being pulled down by
gravity.  If this surge is greater than the speed at which the pump can
deliver the fluid back to the down side, the spool valve will see a
momentary force reversal.  I am guessing that the spool valve is pushed
back just far enough to allow the high side check valve to close off.
Now the fluid entering the down circuit has something to push against
and the low circuit operating pressure is reached very quickly.  The
pump shuts off even though the gear is not down.  All this happens very
quickly, in a fraction of a second.
Unlocking the system is best accomplished by cracking the dump valves.
Even though the entire system may now be isolated from the pump with
both internal check valves closed, the flow of fluid from the high to
low side through the dump valve lowers the pressure on both sides since
rod volume is being extracted from the system.  The pump fires up again
and the gear continues on its way down.  Over the years, I recall what
sounded like a hesitation in the pump right after initiating the gear
extension.  This may have been the system almost locking up.  
In terms of fixes, I don't have any that I like.  Restricting the
flow rate in the high side circuit or increasing pump speed would
probably do the trick.  I would certainly not recommend adding any
restriction as this would impede the emergency gear extension.  Changing
pumps would be quite expensive.  If anything better comes to mind,
I'll pass it along.  

Chris Zavatson
N91CZ
360 std, >1000 gear extensions