UNI Loss Model
Figure 1 shows my estimation of UNI losses where Rtot has been evaluated as (U(0)-U(I))/I.
While I haven't been able to accurately measure the small Rcu at the high current windings
(S1-6), this was easy for S7-S8. An interesting and still somewhat confusing thing is the
fact of Rtot being much higher than Rcu+n2*Rpe (where n=383/230 for S7-S8). I named this
additional loss Ri due to suspected iron-core loss (65%! for S7-S8).
Anyway, using the Rtot concept will always yield the correct total maximum loss and the
power lost in this Rtot is (or should be) in direct proportion to the actual heat increase (<60C).
I have a certain small reservation for this statement though because my original intention
with this double set of primaries was actually BOTH the feature of Mains Isolation (or Step-Up/Down)
AND Enhanced Power Output. Whereas the Mains Isolation feature works quite well, the
Enhanced Power Output was a major disappointment because paralleling the primaries only
increases maximum output power with some pathetic 3% (or 6W)!
I'm however not 100% convinced of this sad result because I wonder where the actual 14C (-6.4W)
lower primary heat went or did it really transfer to the secondaries? I suppose so but in that
case the distance for heat convection has been lowered (due to the secondary windings being
closer to the capsule). This fact along with the actual greater heat generation area (which
lessens heat resistance) should enable more than 3% output power for 75C capsule temperature.
It should however be noted that it's not the actual capsule temperature that is the problem,
it's the temperature of the core (before it reaches it so called "Curie"-point where inductance
simply vanishes and consequently, disaster!)
In any case, as long as lost power is less than some 29W (absolute max, 24W@236/242VA recommended)
the transformer should survive (for a while anyway...). These 29Ws means a maximum output power of
some 283VA@85C (or +20% due to 60C/50C) and at least 290VA@85C for the Standard UNI Configuration
(fig.2) and the High Power UNI Configuration (fig.3) respectively.
Notable is however that these 29 Watts may be lost almost anywhere within the transformer.
It is however important to be resonable and note that lost power never should exceed some 100%
of normal lost power and this is due to the fact of interwinding (plastic) isolation. It is however
fully possible to 100% current overload (P=4P0!) at for instance S1 while drawing normal output power
from the rest of the secondary windings (except S2-S4).
The output voltage at optional 0-200% current is noted in figure 1 and may generally and approximately
be written as U(I)=U(norm)-I/Inorm*Uloss(norm).