Caution - InterAv "Spike Guard" Capacitor
MikeremlaP
wish to inspect it immediately and perhaps remove it.
I was noodling around the engine the other day, shaking baffles, when I heard a
"rattle." Investigating further, I discovered the noise was coming from our
"Big Fat Capacitor" (a Mallory brand 6000 mFd Electrolytic), mounted on the
front baffle next to the alternator. (Lyc O-320.) Ours was offered by InterAv
(Alternator conversion kit maker) and purchased through Chief. It's basically
a big filter capacitor that hangs on the output of the alternator to catch
noise and voltage transients.
My guess is the electrolyte has dried up, leaving the guts of the capacitor to
rattle inside the can. It measures zero capacitance. Fortunately, it failed
in an open, but could have just as easily failed in a short - and a short
directly across the alternator output could be catastrophic, especially while
flying. (Hopefully a diode would open in the alternator, but who knows?)
I've yet to pull the logs, but I think this has been in service for about 5
years now (roughly 1000 hours). The capacitor was mounted vertically, with the
lugs down. It was mounted on the "cool" side of the front baffle, but gets
baked during shutdown by cylinder #1.
I suspect an electrolytic capacitor is not a good choice for hot engine
compartments, but it's hard to get 6000 mFd any other way. If they continue to
be sold, perhaps there should be a life limit (time-wise) on them? As a
minimum, an in-line fuse should be added to the circuit, so that the fuse will
blow if the capacitor shorts.
In any event, the potential for disaster seems high if you have one of these,
since most people don't mount fuses in line with filter caps, and as most
people don't shake capacitors as part of a routine inspection, I thought I'd
better alert you all.
Hope this helps,
Mike Palmer <><
Excellence in Ergonomics
papenfus...@juneaudotmedotvt.edu
: rattle inside the can. It measures zero capacitance. Fortunately, it failed
: in an open, but could have just as easily failed in a short - and a short
: directly across the alternator output could be catastrophic, especially while
: flying. (Hopefully a diode would open in the alternator, but who knows?)
This is the usual mode of failure for an electrolytic. If it does
fail as a "short," it's not that big of a deal either. I've seen enough
students either over-voltage or reverse-bias electrolytic caps to see them
fail, and when they do it's a bit messy, but not that major. For the ones
the size on a plane, they'll pop like a firecracker if they suddenly
decide to take a bunch of current. Remeber, that's a firecracker, not a
hand-grenade, so things should still chug along happily. Might need to
clean it up when you notice it though, as the electolyte is fairly
corrosive.
-Cory
--
*************************************************************************
* The prime directive of Linux: *
* - learn what you don't know, *
* - teach what you do. *
* (Just my 20 USm$) *
*************************************************************************
MikeremlaP
They say 1) the capacitor should NOT be connected to the output of the
alternator, but to their voltage regulator. They referred me to their
schematic at www.inter-av.com. [My fault - there was no schematic shipped with
the capacitor when I bought it from Chief. There was no documentation at all.]
2) they recommend mounting the capacitor on the firewall.
So false alarm on the first part... although I know one other Glasair builder
who has his "Spike Guard" mounted across the alternator output also. It seems
like the natural location for a "spike" guard.
Inter-Av claims very few failures of this capacitor... but I'm wondering.
Unless you physically shake the can, you won't know if the electrolyte
evaporated. (That means you have to take it out of the Adel clamp... which is
probably never done during inspection.) While a firewall isn't as severe a
heat location as a front baffle, our firewall still gets fairly hot after a few
hours of flying. Don't know if the electrolyte drying out is a function of
boiling point or long term evaporation. Sounds like further investigation is
warranted.
But if wired according to Inter-Av, the problem isn't as dire as it could be.
Aaron Coolidge
: Inter-Av claims very few failures of this capacitor... but I'm wondering.
: evaporated. (That means you have to take it out of the Adel clamp... which is
: probably never done during inspection.) While a firewall isn't as severe a
: heat location as a front baffle, our firewall still gets fairly hot after a few
: hours of flying. Don't know if the electrolyte drying out is a function of
: boiling point or long term evaporation. Sounds like further investigation is
: warranted.
The cap should not rattle even if the electrolyte dries out. The cap is not
constructed like a flooded cell battery. It's made of two pieces of aluminum
foil with a very thin, porous, paperlike material sandwiched between them.
This porous material is soaked in electrolyte like a damp cloth. This whole
mess is then rolled up like a set of blueprints so as to increase the surface
area (equation of capacitance is Capacitance = Area * Dielectric Constant).
If the electrolyte evaporates, the capacitance decreases. If the electrolyte
evaporates quickly, the capacitor can will bulge out, explode, or "vent"
the overpressure. There is probably a small vent opening on the top of
your capacitor next to the screw terminals. If this is not popped open the
electrolyte is probably still contained in the cap.
It sounds to me like your capacitor has failed due to mechanical vibration.
This is the only way I can think of that would leave the guts of an
electrolytic cap rattling.
--
Aaron Coolidge (N9376J)
"100% of the world is covered by air. Use it all." - Citation ad
Doug Carter
"MikeremlaP" <miker...@aol.comorgnet> wrote in message
news:20030630141301...@mb-m27.aol.com...
>While a firewall isn't as severe a > heat location as a front baffle, our
firewall still gets fairly hot after a few > hours of flying. Don't know if
the electrolyte drying out is a function of > boiling point or long term
>evaporation. Sounds like further investigation is
> warranted.
While I've no idea what the specs are for this particular, large
electrolytic capacitors are used in automotive designs. The example in this
reference: http://images.powerelectronics.com/files/285/ElectrolyticCap.pdf
discusses estimating the life of an electrolytic capacitor operating at 85
degrees C. Caps can operate at these temperature levels, just depends on
the specific one.
MikeremlaP
See foto at
http://www.fotolog.net/palmer_mp/
No foil that I can see. In fact, I'm not sure how this thing works. I thought
the idea was to create maximum surface area.
I probably should have buzzed the negative terminal to the can to see if
connected. (Can do that tomorrow).
Black stuff has a strong order to it...
In any event, it seems that, with vertical mounting, lugs down, either
component could rattle its way to short the inputs. I'm surprised it didn't
happen.
Mike Palmer <><
Excellence in Ergonomics, but apparently obsolete in Electrical Engineering.
MikeM
MikeremlaP wrote:
>
> If you have a "Spike Guard" filter capacitor on your alternator output,
Mike, according to the Interav installation documents, the capacitor is
supposed to be wired across the "Diode-Trio" output, and serves to filter
the ripple out of the voltage which the voltage regulator senses. The field
current is effectively supplied from this capacitor, so without it, the
VR sees a lot of ripple on what it is "sensing". Why the fuXX they dont
just sense the main bus like 100s of Millions of other automotive and
aircraft alternator systems is beyond me??? Maybe to get around Motorola's
patents???
> you may
> wish to inspect it immediately and perhaps remove it.
A year of two ago, I got into it with InterAv about why they think that this
capacitor is even needed. They could only say that they don't know why its
there. They bought the STC from a third party, who developed the STC with
the capacitor and got it approved with the FAA. In Interav's view, removing
it would invalidate the STC, and they would have to redo the STC application,
which they are not willing to do.
I think that the Interav alternator STC and the way they are sensing the
system voltage is brain dead, and as you just pointed out, it may actually
be dangerous. Putting a 60,000uF Mallory electrolytic capacitor in an
aircraft engine room is stupid, and almost guarantees a heat/vibration related
failure of the capacitor. The FAA had its head in a warm, moist place when
they approved it in the first place. Gives you a warm, fuzzy feeling about
other FAA approved STCs, doesn't it?
Wiring any big electrolytic across the main output of the alternator
compounds the problem! Such capacitors have a "Ripple Rating", which will
likely be exceeded when it is connected across the main bus. The excessive
ripple will cause a meltdown in the guts of the capacitor. That is what
your picture shows...
MikeM, PhD EE
Skylane '1mm
Pacer '00z
passthetunaporfavor
jwill...@gmail.com
jwills8606_at_gmail.com
Firstly, I can make NO sense out of the big cap. I haven't checked mine, but it's at least twenty years old, and I bet it's dead. InterAV shows the cap between the "regulator" terminal and ground - I assume that terminal is some kind of voltage output sensing wire - it shows that terminal can be hooked to a "warning light" also. I had the alternator rebuilt about 7 years ago, so I don't remember exactly where that wire is hooked inside the alternator.
What I DO remember is that the 3-phase diode array output then all feeds into ANOTHER diode, which is then connected to the alt. output lug. WTF? Nice single-point of failure - a single diode for all the alt output.
On my 182, I have developed an alternator whine that is worse with increasing electrical load. I put an oscilloscope on it, and interestingly, it is not the alt. output - it doesn't increase with rpm or change frequency with engine speed - it's the square waves being fed into the "field" from the voltage regulator. I wonder if somehow the big cap has something to do with that. I am told that reg is a very poor design and should be replaced whenever possible.
And another thing: I've never seen a spike snubber made with such a big cap. It's usually a smaller tantalum in parallel with a very small ceramic. I think somebody doesn't know what they are doing, even though it (somewhat) works.
Thoughts?