Very interesting material on the subject (IMO) ...
Gilles Thse and Jrme Delamare have tested the Rotax alternator and
the Ducati Energia rectifier/regulator and written 3 articles (with
another to follow but that hasn't appeared):
http://foxpapa.com/rubrique.php3?id_rubrique=138
The first article (alternator) shows:
- listing of basic properties
- no-load: output voltage (100Vp-p max.), harmonics composition,
effective value as a function of rpm
- impedance at various stator currents
- short circuit: impedance and effective current as a function of rpm
(21A max.)
- short circuit temperature rise (30 degrees + 17 after stopping -
considered ok)
The second article (battery) shows:
- battery Hawker SBS8, 7.4Ah
- discussion of battery role
- discharged at 5A for 45 minutes
- 40 minutes of recharging at 14.3-14.4V, current starting at 17A
The third article (rectifier/regulator) shows:
- diode bridge rectified, no load
- diode bridge rectified with capacitor, no load
- with shunt regulating resistor at max rpm, no load
- with shunt regulating transistor at max rpm, no load: dissipates 260W!
- with series regulating resistor, worst load, at max rpm
- with series regulating transistor, worst load, at max rpm: dissipates
220W!
- controlled one polarity thyristor rectification, on-off; hard on the
hardware
- controlled one polarity thyristor rectification, each wave; this is
how it works
- (JdJ: and dissipates 80W apparently - see below)
- load budget for various conditions
The fourth article would likely have reported on the marginal thermal
situation of the regulator.
Gilles gave a short version on the aeroelectric forum:
- regulator dissipates max. 80W, at 240W output
- heatsink to sea level ambient air thermal resistance is 2 degrees C/W
(measured)
- junction to heatsink thermal resistance is 0.3 degrees C/W (estimated)
- junction to sea level ambient air thermal resistance is 2.3 degrees
C/W (the sum)
- so 80W gives a 184 degrees C increase in junction temperature above
ambient
- 125 degrees C is a reasonable maximum for a semiconductor junction
- (JdJ: the maximum I have seen on a datasheet was 175 degrees)
- so marginal is hardly the term to use
- but moving air helps a lot - with a PC-type fan the 2.3 number changes
to 0.8
- so then 80W gives a 64 degrees C increase in junction temperature
above ambient
My conclusion:
A solid state device is extremely reliable if its limits are not
exceeded, so that is apparently what we do.
Could we not mount the regulator next to the radiators somehow instead
of on the firewall. That would guarantee moving air. Four meters of
10awg weigh 200g. The mounting should probably allow the capacitor to
join. Can't think of further input.
?
Jan de Jong
461, too slow, starting on the wings..
|