I haven't followed this thread much, but isn't the big capacitor essential
=0Afor the Ducatti system/ The big C prevents voltages from rising and=0Aac
ts as a dump/battery or has my memory tricked me again?=0AGraham=0A=0A=0A
=0A=0A________________________________=0A From: Greg Fuchs <gregoryf.flyboy
@comcast.net>=0ATo: europa-list@matronics.com =0ASent: Wednesday, 22 Januar
y 2014, 8:04=0ASubject: RE: Europa-List: Re: Ducati rectifier/regulators=0A
mcast.net>=0A=0AMaybe not so fast about waste/dump conclusions..=0AWhen I b
uilt my motorcycle regulator about 5 years ago, I made it in the=0Ashunt co
nfiguration, or the simple waste/dump. It still works great to this=0Aday.
It is very similar to the example in Nigel's document (which I plan to=0Ago
over in more detail about the Mosfet versions of regulators), except I=0Au
sed an actual programmable shunt regulator IC to control the high-power=0As
tage (a simple thing..it modeled like a zener with an amazing amount of=0Am
ath to come up with the programming resistors, for what it is).=0A=0AUnfort
unately in this case, I didn't pay attention to the orientation of the=0Asc
r's in the Ducatti regulator.=0AWhen I went to prove to myself that they we
re shorting out the windings of=0Athe generator to itself (similar to conne
cting the two outputs together), I=0Acould only prove what Jan de Jong has
been saying all along..that they=0Aweren't. Thus, seems to not be a shunt c
onfiguration after all.=0A=0AOne downfall in this configuration..with no lo
ad on the generator, its AC=0Apotential may have the ability to climb quite
high. In hindsight, I guess=0Athat was talked about in this thread ..30 vo
lts...(42 volts peak?), but I=0Aassumed probably incorrectly, that the meas
urement was made without a=0Aregulator connected. =0AI wouldn't prefer thos
e high voltages flying around...increasing the danger=0Aof getting zapped a
nd the dielectric needs of the components, but it is what=0Ait is. In this
configuration, the diode AND the scr are conducting at once,=0Ato carry cur
rent to the load...increasing the needs on the heat sink.- =0A=0AStill, i
f the output diodes are a common failure mode, and with the cost of=0Arepla
cements...it would be nice to see some of these things come back to=0Alife.
Stick it to the man..or the overpriced dealer, heehee.=0A=0A=0A=0A-----Ori
ginal Message-----=0AFrom: owner-europa-list-server@matronics.com=0A[mailto
:owner-europa-list-server@matronics.com] On Behalf Of Pete=0ASent: Tuesday,
January 21, 2014 1:42 PM=0ATo: europa-list@matronics.com=0ASubject: Re: Eu
ropa-List: Re: Ducati rectifier/regulators=0A=0A--> Europa-List message pos
ted by: Pete <peterz@zutrasoft.com>=0A=0A>From that schematic, it is indeed
a simple waste/dump regulator,with the=0Atwo failing diodes being there to
prevent battery from reverse feeding=0Aduring the dumps.- It is pretty s
illy that they undersized them for rated=0Aalternator generating current ca
pacity.- "Certified". Wow.=0A=0ACheers,=0APete=0A=0A> On Jan 21, 2014, at
4:12 PM, Bill Henderson <europa10@bellsouth.net>=0Awrote:=0A> =0A> As info
.....=0A> =0A> I opened up one of the Rotax regulators a friend gave me aft
er a failure.=0AThe two diodes were bad and were open so paralleling them w
ith external=0Adiodes does seem to be a viable option. Relatively cheap for
the cost of the=0Adiodes and a heat sink.- No modification to the existi
ng regulator required.=0AAttached is the fix I obtained from another websit
e.- Just need to join the=0Adiodes to the B and two G terminals.=0A> =0A>
Just my two cents worth.....=0A> =0A> Bill=0A> A010 Classic Monowheel=0A>
Still building.=0A> =0A> =0A> =0A> =0A>> On 1/21/2014 3:02 PM, Jan de Jong
e_jong@casema.nl>=0A>> =0A>> =0A>> ad A:=0A>> Don't agree. The only stress
would be somewhat higher reverse voltage.=0AThe Ducati R/R is not a shunt r
egulator.=0A>> =0A>> ad B:=0A>> Agree, but the thyristors are more stressed
than the diodes. They drop a=0Alittle more voltage and dissipate more heat
.=0A>> On the other hand, the custom fitting of the diode pills could be mo
re=0Auncertain than the standard fitting of the thyristors (TO220 I believe
).=0A>> =0A>> Parallelling the diodes with external ones is possible. They
would=0Arequire a heat sink. Are the failure statistics clear about the dio
des?=0A>> I think it would be better to get an R/R without issues.=0A>> =0A
>> Regards,=0A>> Jan=0A>> =0A>> =0A>>> On 1/21/2014 6:30 PM, Greg Fuchs wro
oryf.flyboy@comcast.net>=0A>>> =0A>>> =0A>>> =0A>>> Well, since the generat
or is not a two-legged 'phi' connection (i.e. =0A>>> the midpoint of the ge
nerator coil connected to the ground of the =0A>>> system or vehicle, then
'dumping to ground' may not be the term to =0A>>> use after all. Yes, I thi
nk that is correct.=0A>>> =0A>>> Here are the two main failure modes of the
Ducatti regulator as I see=0Ait..=0A>>> =0A>>> A) If the aircraft has a ve
ry low load usage (i.e. no lights, no =0A>>> instruments, etc.), the scr's
are stressed and 'may' fail =0A>>> eventually. If this happens, the Ducatti
is unrecoverable.=0A>>> =0A>>> B) If the aircraft has a very high load usa
ge (using up most of its =0A>>> 15-18 amps), the output diodes are stressed
and probably will fail =0A>>> eventually, which is backed up by historical
data gathering (they =0A>>> fail a lot at high loads per many forum users)
. The Ducatti would =0A>>> then be recoverable by placing 25 Amp or higher
diodes on the =0A>>> external terminals as referenced earlier.=0A>>> =0A>>>
Scenario (A) is unlikely, since everyone likes power-hungry =0A>>> instrum
ents and the like, where scenario (B) is the most likely for the=0Asame rea
son.=0A>>> =0A>>> Thus, if someone wants to 'beef up' their ducatti, then t
hey have =0A>>> the option to place the external diodes. I guess it doesn't
matter =0A>>> what kind, as long as they can withstand the current from th
e =0A>>> generator. They will either help the internal diodes survive almos
t =0A>>> forever, or take over for them when they fail (It depends on the
=0A>>> turn-on points between the internal and external diodes).=0A>>> =0A>
>> Jan, do you agree with this?=0A>>> =0A>>> Best Regards,=0A>>> Greg=0A>>>
=0A>>> -----Original Message-----=0A>>> From: owner-europa-list-server@mat
ronics.com=0A>>> [mailto:owner-europa-list-server@matronics.com] On Behalf
Of Jan de =0A>>> Jong=0A>>> Sent: Tuesday, January 21, 2014 4:59 AM=0A>>> T
o: europa-list@matronics.com=0A>>> Subject: Re: Europa-List: Re: Ducati rec
tifier/regulators=0A>>> =0A>>> Equally respectfully - "dumping to ground" i
s not dumping, but =0A>>> connection to the negative DC output terminal.=0A
>>> I attach a little OpenOffice drawing that hopefully clarifies (I =0A>>>
hope that works).=0A>>> =0A>>> Regards,=0A>>> Jan=0A>>> =0A>>>> On 1/20/20
reg Fuchs "=0A>>>> --> <gregoryf.flyboy@comcast.net>=0A>>>> =0A>>>> Respect
fully, I believe that the circuit configuration that you are =0A>>>> thinki
ng of is incorrect, Jan de Jong.=0A>>>> =0A>>>> The SCR's do EXACTLY 'dump'
the AC output DIRECTLY to ground.=0A>>>> When one end of the thyristor con
nects to the alternating current =0A>>>> output of the coil, and the other
end to ground...then that is all =0A>>>> they can do but short the AC outpu
t to ground. It is not acting as =0A>>>> a PWM-like device at all, simply a
grounding device. Call it a=0A>>> direct-connect.=0A>>>>- You mention h
eat. The direct-connect method is the best way to =0A>>>> deal with the hea
t. The forward voltage drop of the 'dumping' =0A>>>> device is low, and the
current does not need to go through the =0A>>>> diodes at all, which are b
ypassed...hence your lowest heat output.=0A>>>> =0A>>>> Since the 'dumping
device' connects the alternator output directly =0A>>>> to ground, without
going through the diode, what is the power=0Adissipated?=0A>>>> While in th
e dumping mode, lets say the current to dump was 10 =0A>>>> amps(I'll use y
our generously supplied information, so I don't have =0A>>>> to look up spe
cs) it would be 1.1V * 10 amps. That's it.- 11 watts. =0A>>>> That is muc
h better than the 18 Watts in your configuration.=0A>>>> =0A>>>>> (=0A>>>>>
a failure mode for the Ducati device is reportedly the loss of =0A>>>>> co
ntinuity=0A>>>> of a diode pill attachment through thermal >cycling; fittin
g =0A>>>> external parallel diodes has been proposed as a solution; replaci
ng =0A>>>> the whole device seems more=0A>>>>> sensible=0A>>>>> )=0A>>>> We
ll, maybe so!- However, if the failure mode is a loss in =0A>>>> continui
ty, that would mean the diodes are opening. That would also =0A>>>> mean th
at connecting the external diodes to the regulator would =0A>>>> bring it b
ack to=0A>>> life!=0A>>>> Regards,=0A>>>> Greg=0A>>>> =0A>>>> =0A>>>> =0A>>
>> -----Original Message-----=0A>>>> From: owner-europa-list-server@matroni
cs.com=0A>>>> [mailto:owner-europa-list-server@matronics.com] On Behalf Of
Jan de =0A>>>> Jong=0A>>>> Sent: Monday, January 20, 2014 7:40 AM=0A>>>> To
: europa-list@matronics.com=0A>>>> Subject: Re: Europa-List: Re: Ducati rec
Jong =0A>>>> --> <jan_de_jong@casema.nl>=0A>>>> =0A>>>> Re two-phase bridge
rectifier/regulator.=0A>>>> The 2 thyristors in the rectifier bridge do no
t short the AC input =0A>>>> or=0A>>> "dump"=0A>>>> anything.=0A>>>> On the
contrary, they disconnect the DC output from the AC input =0A>>>> when the
output voltage threatens to exceed the set voltage.=0A>>>> All output curr
ent passes through one branch of one diode and one =0A>>>> thyristor in ser
ies half the time and through the other identical =0A>>>> branch also half
the time.=0A>>>> Heat development is proportional to the output current and
the =0A>>>> voltage drop across a branch.=0A>>>> Voltage drops increase so
mewhat with current, so heat development =0A>>>> increases more than linear
ly with output current.=0A>>>> In the Ducati device the diode is reportedly
a MR2510 pill =0A>>>> (typically 0.75V at 10A, 0.8V at 20A), the thyristor
can be a =0A>>>> 2N6504 (typically 1.1V at 10A, 1.3V at 20A).=0A>>>> So at
10A the two branches of the bridge are each expected to =0A>>>> generate
=0A>>>> 0.5 x- (7.5 + 11) = 9.25W (total 18.5W) of heat.=0A>>>> And at
20A the two branches of the bridge are each expected to =0A>>>> generate=0A
>>>> 0.5 x (16 + 26) = 21W (total 42W) of heat.=0A>>>> =0A>>>> (=0A>>>> a
failure mode for the Ducati device is reportedly the loss of =0A>>>> conti
nuity of a diode pill attachment through thermal cycling; =0A>>>> fitting e
xternal parallel diodes has been proposed as a solution; =0A>>>> replacing
the whole device seems more sensible=0A>>>> )=0A>>>> =0A>>>> Cheers,=0A>>>>
=============
|