Perhaps an evil spell??

On Mar 2, 2:02*pm, Andy wrote:
On Mar 2, 12:05*pm, Andy wrote:



On Mar 2, 11:25*am, seventripleone wrote:

On Mar 1, 2:35*am, Charlie Papa wrote:

I have a Discus 2cT with about 650 hours on it. *Schempp-Hirth did the
instrument installation, and a more professional looking job you will
not find. *But there was one problem that confounded even the avionics
guy: when the engine was running, pressing the push-to-talk button
would kill the engine until the button was released.

The engine's ignition is a very simple magneto (Solo 2350), and the
radio a Dittel FSG2T radio, - yes, the one with a recall. *How to
explain the ability of the radio to ground the magneto?? *No one even
had a theory. *But when I reinstalled the radio after the recall
service (which was very efficiently and courteously handled by Dittel,
BTW), the problem was gone. *I tried to induce the problem several
different days, but it exists no more.

So what flaw in this radio can cause such an effect? *A black curse?

No, the engine electronics will shut the ignition down when max. revs
for the prop are exceeded. The revs are present in the instrument as a
proportional voltage signal (1V = 1000 RPM) and i'd bet that somehow
the electronics got tricked into thinking that it was the case. A
voltage drop in the power supply from the battery during radio
transmission or an induced spike from the radio's RF power could have
been the reason.

Anyone that designed a system to transmit a critical parameter as a
fraction of a reference voltage but did not also provide the reference
voltage *to the receiving system would be better employed flipping
burgers. *In other words, a properly designed engine speed monitoring
system would be independent of the battery voltage.

RFI could have strange results though. *Maybe the incorrectly
installed capacitor allowed RF on the radio power line and that
coupled into the engine monitor.

Anyone know the function of the problem radio capacitor?

Andy

I know only that the capacitor was installed incorrectly and that it
affected transmission output. Here's the link:

http://www.eaa.se/dokument/tekniska/...tel_SB_FSG2T-1...

9B

Our company makes a very good living tracking down electromagnetic
compatibility issues like this. Radio frequencies impinging on
unshielded control circuits can end up being rectified and appearing
as a spurious dc signal. We've seen this a lot in electrical control
and industrial process control equipment (a few years ago I watched a
robot throwing gearboxes across a car assembly plant instead of
handing them gently to the next robot in line!). I would strongly
suspect something similar is happening here - 120 MHz is a great
frequency for the purpose.

Mike

On Mar 2, 9:05*pm, Andy wrote:

Anyone that designed a system to transmit a critical parameter as a
fraction of a reference voltage but did not also provide the reference
voltage *to the receiving system would be better employed flipping
burgers. *In other words, a properly designed engine speed monitoring
system would be independent of the battery voltage.

RFI could have strange results though. *Maybe the incorrectly
installed capacitor allowed RF on the radio power line and that
coupled into the engine monitor.

I'd generally fully agree with you on this one, but there is no signal
outside of the electronics box (case mostly metal and made by Ilec
BTW). Except that you can connect an external voltmeter to act as RPM
indicator on older units and on new ones there is perhaps an internal
DVM module to do the same.

The box senses the RPM as ignition pulses via the earthing lines from
the two ignition modules (coils) and does the conversion internally.
Depending on said signal three LEDs are lit to indicate in which PRM
range (green, yellow, red) the engine is turning and the ignition is
switched off (earthed) when a threshold of ~7000 RPM is exceeded.

The above does contain some informal reverse engineering concerning
the internal workings of the box on my side to be honest but i'd
happily buy you all a round if i'm wrong.

(IMHO it's much better to try to contain the RF in the radio and
antenna cable than try to harden everything in the glider against it.
Otherwise i'd end up with tinfoil around my b**** :-)

On Mar 4, 3:27*am, seventripleone wrote:
On Mar 2, 9:05*pm, Andy wrote:

Anyone that designed a system to transmit a critical parameter as a
fraction of a reference voltage but did not also provide the reference
voltage *to the receiving system would be better employed flipping
burgers. *In other words, a properly designed engine speed monitoring
system would be independent of the battery voltage.

RFI could have strange results though. *Maybe the incorrectly
installed capacitor allowed RF on the radio power line and that
coupled into the engine monitor.

I'd generally fully agree with you on this one, but there is no signal
outside of the electronics box (case mostly metal and made by Ilec
BTW). Except that you can connect an external voltmeter to act as RPM
indicator on older units and on new ones there is perhaps an internal
DVM module to do the same.

The box senses the RPM as ignition pulses via the earthing lines from
the two ignition modules (coils) and does the conversion internally.
Depending on said signal three LEDs are lit to indicate in which PRM
range (green, yellow, red) the engine is turning and the ignition is
switched off (earthed) when a threshold of ~7000 RPM is exceeded.

The above does contain some informal reverse engineering concerning
the internal workings of the box on my side to be honest but i'd
happily buy you all a round if i'm wrong.

(IMHO it's much better to try to contain the RF in the radio and
antenna cable than try to harden everything in the glider against it.
Otherwise i'd end up with tinfoil around my b**** :-)

Thanks for the replies, - at least the reasoned and helpful ones. I
am at Seminole Lake and spent a half hour yesterday with the very well
informed TA discussing the mystery. Additional information, that may
be relevant:
• I don't know if the problem has existed from the beginning, as with
the engine running I can't hear myself think. It was only recently,
while flying with a friend, I got low and advised I was going to light
the engine. He asked me to advise him when it was running, and so
even though I knew I would not hear a reply, when it lit as it very
reliably has, I pressed the PTT and had an Oh Sh*t moment. I released
the button and the engine caught again. I again tried to transmit,
and the engine died. The following day I verified while just off tow
that each push killed the engine.
• I did not look for the effect on the LED's which might have been
showing an 'overspeed' signal
• I tried the PTT when the radio was removed for the recall, and there
was no effect
• I don't have the knowledge to trace ground leads and the like, and
so did not

TA theorized, if I can explain it accurately, that RF interference was
picked up by the wires from the rev sensor and the sensing unit
interpreted these as excess RPM, and grounded the magneto. This is
similar to the speculation from seventripleone. 711 has offered me
his unrepaired radio to try to again induce the problem, and eliminate
the possibility that it arose from frayed wires in the harness.

Me, I like the explanation that I let the smoke out. Thanks all.

Charlie Papa