Boost vs. WGFV DC and DC vs. Temperature

[Smeins, Larry]

I know that boost and how to keep or increase it has been beaten to death on
this list.  But, in the 5 years I've been around I don't remember seeing
WGFV duty cycle (DC) and how it is controlled getting much discussion.  I
have an 87 MC / MAC11 5kt that I have driven at 5000 ft. altitude for over
200k miles and have never seen over 1.2 bar on the in dash digital manifold
absolute pressure display.  According to Scott Mockery's site the absolute
manifold pressure should be limited at 1.42 bar by the ECU.  Bentley shows
that boost above ambient pressure increases as ambient decreases for a
nearly constant absolute pressure limit.  So the question I went off to
answer was, where is my missing .22 bar.  I went over Scott's list, fixed a
few things to no avail, and went off on my own.  All indicators pointed to
wastegate so I disabled it and had no trouble hitting fuel pump shut off.
Satisfied that I had good boost capability I reconnected the wastegate,
hooked a duty cycle meter to the WGFV and started taking data.  I learned
some things I wasn't expecting and found my boost followed DC as predicted.

It appears to me that control of maximum boost is open loop.  The ECU puts
out a predetermined WGFV DC dependent on a few parameters other than
measured manifold pressure.  Boost follows    the equation: BOOST = SP/
(1-DC).  Where: BOOST is manifold pressure above ambient atmospheric
pressure, SP is the pressure set by the wastegate spring and DC is the WGFV
duty cycle.  No altitude correction here.  At sea level, with a .2 bar
spring and 50% duty cycle BOOST is 0.4 bar and absolute manifold pressure is
limited to 1.4 bar.  50% duty cycle is the max I see and I believe that is
the predetermined maximum available.  At 5,000 feet, 0.8 bar ambient,
absolute manifold pressure is limited to 1.2 bar.  It looks like if I want
to see 1.4 bar I must move to sea level, change the spring, or modify the
ECU for greater DC.  Might as well go for more.

Something that threw me for a loop was an extreme ambient temperature
dependence on maximum DC.  I find my DC changes with ambient temperature
approximately hitting the following points: <40F=<10%, 50F=25%, 55F=30%,
60F=40% and >65F=50%.  Boost follows that predicted by duty cycle.   This
doesn't seem right but I haven't identified what is causing it.  It
definitely reduces my boost in the winter.  Maybe I need to move to Florida.

OK what do I have wrong here?  I've consulted in Germany and know the design
logic of German engineers is different than that of US engineers.  Did I
miss something that should be obvious?

Larry