Today's Message Index:
----------------------
1. 12:39 AM - Re: Info on MGL Enigma (Jay Hyde)
2. 03:36 AM - Off Delay Cabin Light (user9253)
3. 05:24 AM - Re: Off Delay Cabin Light (Charlie England)
4. 05:55 AM - Economical solid state pressure sensors (Ken)
5. 06:41 AM - Re: Info on MGL Enigma (Eric M. Jones)
6. 08:07 AM - Re: Bonding Straps (Robert L. Nuckolls, III)
7. 08:54 AM - Re: Bonding Straps (stearman456)
8. 11:23 AM - Re: Economical solid state pressure sensors (Robert L. Nuckolls, III)
9. 12:00 PM - Re: Info on MGL Enigma (Robert L. Nuckolls, III)
10. 12:05 PM - Re: Re: Bonding Straps (Robert L. Nuckolls, III)
11. 01:24 PM - Re: Re: Bonding Straps (Daniel Jones)
Message 1
INDEX | Back to Main INDEX |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Info on MGL Enigma |
Hello again listers,
Here is a further response from Rainier:
I did not say that spikes or similar ARE the cause of what it going on.
It might be something completely unrelated. It's a pretty complex system.
There are no old style linear regulators in the system.
There are three high frequency switch mode supplies and all input power is
protected by a heavy duty transorb apart from other filtering (mostly
inductive/capacitive).
Nevertheless, we are talking about a system that has multiple connectivity
to other electrical systems on the aircraft - anything going on anywhere and
the EFIS gets it. We are taking about transient events at RF frequencies.
Do-160 simply does not cater for this on a long term bases. But that is
another story.
So, honestly, without actually having the system on a bench and
investigating - and conclusively finding the cause, going on like that is
not called for here.
Getting back to the Flash memory - yes it does fail if it gets transient
power events that make it through the system by whatever path. It may take
years. However that is by no means the only possible cause. Flash is known
to wear over time. Manufacturers of these devices guarantee 10 years of data
retention typically, even if no writes take place - yet these devices are
used throughout the industry for missions where it is assumed that they will
never fail. Usually they don't.
In this case we don't even know if this memory has failed. It might be
something else - perhaps the communications link between main CPU and the
processor that contains this memory has gone faulty. It could be anything.
I'm guessing. That is all I can do without having the Enigma on my bench.
Electronics do fail - does not matter how much money you spend on the
components. Service history of the Enigma has been excellent. The number of
Enigmas coming back to the factory for repairs is almost nil. That is not
too surprising - it has been designed for open cockpit ultralights and it is
quite tough. But they are getting older now and failures of electronic
components become more of a factor. There is something called a "bathtub
failure curve" - because it looks like it. Failures tend to occur when the
device is new (partly the reason we run them for 48 hours at elevated
temperatures during manufacture) - then things settle down and after some
time failure rates start creeping up slowly as things age.
Components that do age are crystals (They are mechanical devices - high
frequency types in particular have a higher failure rate in our experience).
Then we have ceramic capacitors - a typical EFIS has hundreds of them. Some
start leaking after a while. Most of the time that is not an issue but
sometimes it is. We don't use tantalums because of reliability concerns
(good caps but they have an annoying tendency to blow up). Our electrolytics
are top quality devices - can't even remember any failures. Then we get to
the semiconductors. Well, here we have a pretty random failure pattern. They
do die for no apparent reason. The rate is low but it does happen. Yes, we
use aerospace, automotive and industrial spec devices - we do not use
consumer graded devices at all.
Most failures are perhaps related to PCB failures (usually thermal stress
cracking vias). Some of our PCBs are high tech multi-layer. They are more
sensitive - part of the process. Some assemblies are lead free and the
entire industry has been battling for years with "tin-wiskers".
Microscopic growths of tin between tracks and pads. Only visible under a
microscope. Takes years to cause a problem. That we have had. Lucky this is
not so much of an issue anymore as lead free processes and fluxes have
improved and I have yet to see such an issue on anything produced in the
last 4 or 5 years. "Cold solder joints" - that happens either via repeated
thermal stress cracking a solder joint or a manufacturing defect that may
take years to manifest itself. Yes, it happens. It passes all tests and
optical quality controls (we use an automated AOI machine plus visual checks
at various stages during the assembly
process) - yet it still manages to avoid early detection. Sure, the rate is
low but we make thousands of devices every month. Including devices not
intended for aircraft (and not under our brand name). I would love a zero
failure rate. Failures are bad. Everybody gets annoyed - from the customer
to us. We can't have a zero failure rate - that is a pipe dream. But we can
always work to reduce it to the barest minimum. If we did not do that we'd
be out of business long ago.
Most of our repairs are directly related to what we call "blow ups".
Reverse polarity issues, connecting power or grounds to the wrong places
etc. Interface driver chips (RS232, CAN, switches etc) blown for various
reasons - the items directly connected to the outside World.
All of that happens typically during installation. Mostly with the so called
"professional installers, AMOs etc". I'm not kidding. We find the home
builders are the ones with the least issues - perhaps because most take the
time to study the docs and plan their wiring. There is no rush, no
deadlines. That helps.
Our statistics (which we do keep for obvious reasons) shows this trend very
clearly.
Instruments get dropped, might have a hard time during transport - sometimes
they get soaked in water or oil.
Anyway, I can't really comment on this Enigma until I see it myself (if that
ever happens). If there is a problem, whatever the cause may be - we can fix
it. That is the bottom line.
Rainier
-----Original Message-----
From: owner-aeroelectric-list-server@matronics.com
[mailto:owner-aeroelectric-list-server@matronics.com] On Behalf Of Robert L.
Nuckolls, III
Sent: 02 June 2014 03:45 PM
Subject: RE: AeroElectric-List: Info on MGL Enigma
--> <nuckolls.bob@aeroelectric.com>
At 04:02 AM 6/2/2014, you wrote:
Here is the reply from Rainier, the owner and founder of MGL:
The CR2032 battery is only used to maintain items that change frequently
like calculated fuel levels, local pressure settings etc.
Setup is stored in a flash memory section of a secondary processor chip in
the EFIS. This does not need a backup battery.
If setup data stored here is not maintained there is usually only one
cause: The supply voltage to that processor is raised briefly (micro-seconds
to mill-seconds) above about 8V. This can happen if the supply to the EFIS
contains sharp voltage spikes (typically caused by bad regulators or high
voltage ignition is coupling into ground or supply). This can also happen
if there are static discharges typically caused by airflow over metal parts
that are not grounded.
Eventually, this will damage the flash memory which is the part that dies
first in our experience.
When this happens we replace the processor chip.
If you have a techie handy - it is a type ATMEL ATMega128. It is a SMD
package but easy to solder/desolder if you have the experience. Once
replaced I would also recommend replacing the 16Mhz crystal right next to
this chip - it tends to wear out over a long time (we replace these as a
matter of course when we get an Enigma in for repairs). No specific
programming of the new processor needs to be done as the firmware will
detect a new processor and do the programming and setup of this chip
automatically (the first start-up will take a bit longer and you will see
some messages on the display to this effect).
I am exceedingly skeptical of these assertions.
I don't know of a single 5v regulator offered to
the task of conditioning power for electronics
that doesn't handily mitigate any of the commonly
known variables on the ship's bus. DO-160 offers
a clear and concise path to living confidently
in the world of DC powered vehicles.
Crystals that 'wear out'?
Static discharges ? ! ? ! . . . again, for any
piece of hardware to be qualified for installation
on a TC aircraft, it has to be capable of withstanding
discharges of a 150pF capacitor through 330 ohm
resistor fed directly to every input/output pin
on the device under test. TEN times for positive
spikes, TEN more times for negative spikes.
Sound brutal . . . but the components and architecture
for designing to this level of stress is rudimentary
and has been as common to the designer's toolbox as
a hammer is to a carpenter's toolbox. I quit doing
the test 25 years ago after learning what it took
to pass . . . easily . . . every time.
If the stresses cited are genuine risks to MGL's
products, then they've failed to understand
and embrace the real world stresses found in
mobile DC powered system of all stripe . . . not
the least of which are airplanes.
Bob . . .
Message 2
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Off Delay Cabin Light |
The July 2014 issue of Kitplanes Magazine has an article by Jim Weir describing
a circuit to delay turning off a light after the master switch is shut off.
http://www.kitplanes.com/issues/31_7/builder_spotlight/aero_lectrics_july_2014_21035-1.html
There is no power wire connected to Vcc pin 8 of the 555 IC. Was a wire inadvertently
left off or will the circuit work as drawn?
Joe
--------
Joe Gores
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=424276#424276
Message 3
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Re: Off Delay Cabin Light |
Perhaps it's been fixed; the online version has the connection.
On Tue, Jun 3, 2014 at 5:35 AM, user9253 <fransew@gmail.com> wrote:
>
> The July 2014 issue of Kitplanes Magazine has an article by Jim Weir
> describing a circuit to delay turning off a light after the master switch
> is shut off.
>
> http://www.kitplanes.com/issues/31_7/builder_spotlight/aero_lectrics_july_2014_21035-1.html
> There is no power wire connected to Vcc pin 8 of the 555 IC. Was a wire
> inadvertently left off or will the circuit work as drawn?
> Joe
>
> --------
> Joe Gores
>
>
> Read this topic online here:
>
> http://forums.matronics.com/viewtopic.php?p=424276#424276
>
>
Message 4
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Economical solid state pressure sensors |
I only seem to get a few hundred hours out of name brand
electromechanical engine pressure sensors so I tried a 30 psi $25. solid
state Chinese sensor for coolant pressure monitoring. Quite a bit
cheaper than what I was using and seemingly good specifications which
I've copied below. While it calibrates perfectly at room temperature, it
reads at least 5 psi low when it gets hot (ie. 4 psi when it should be
9). The zero point changes similarly and I confirmed all this with raw
voltage readings and scope traces to make certain. This sensor was
mounted on the firewall and tee'd into a 5/16" coolant hose so the
coolant likely warmed it up to about 80*C. Seems it would need to be
installed on a stub line on the cool side of the firewall to be useful.
The temperature specification of Maximum:0.05%FS/C would indicate up to
1.5 psi zeroing error over a 100*C range which would be acceptable to me
if that was the case and other errors were minor. This particular sensor
may be outside of stated tolerances. In any case, the compensation
temperature range of -20 to +80*C seems to be meaningless. I'd also
hoped to replace the oil pressure sensor but it seems these economical
sensors are not the answer for under the cowl.
Ken
Brand new pressure transducer. 316 Stainless steel body.
Input: 0-30 psi(Gauge Pressure);
Output: 0.5-4.5V linear voltage output. 0 psi outputs 0.5V, 15 psi
outputs 2.5V, 30 psi outputs 4.5V.
Works for oil,fuel or air pressure.Can be used in oil tank,gas
tank,diesel tank etc.
Accuracy: +/-1%FS;
Thread: 1/8 NPT;
Wiring Connector: Packard plug-in unit is included;
Wiring: Red: +5V; Black: ground; Green: signal output;
It's an advance pressure sender than traditional mechanical pressure sender;
Overload Capacity: 2-4 times;
Working Temperature: -40+120C;
Compensation Temperature: -20+80C;
Protection Class: IP67;
Pressure Medium: The gas and liquid which is compatible with 316L
stainless steel;
Load Resistance: ((supply power-6.5V/0.02A);
Long-term stability: Less than 0.1%FS/year;
Temperature Effect on Zero: Typical:0.02%FS/C; Maximum:0.05%FS/C;
Temperature Effect on Sensitivity: Typical:0.02%FS/C; Maximum:0.05%FS/C;
Shock Resistance: 1000g;
Anti-Shock: +/-0.01%FS(X,Y,Z axes, 200Hz/g);
Response Time: 1ms;
Insulation Resistance: >100m 500VDC;
Weight: 55g;
Explosion-proof Class: ExiaTTCT6;
Electromagnetism Compatibility: EN50051-1.
Message 5
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Re: Info on MGL Enigma |
It is hard to say that a 3V battery measuring 2.9V is dead or not depending on
load. The nominal discharge current of the battery is 0.2 mA. Putting a 150k in
parallel will show 2.9V or above for a still good battery. But since getting
there is half the work, change it.
http://www.adafruit.com/datasheets/maxell_cr2032_datasheet.pdf
See DX.com (Deal Extreme) for a 10 pack of Sony 3V CR2032 Lithium Ion
$5.98. Free shipping
--------
Eric M. Jones
www.PerihelionDesign.com
113 Brentwood Drive
Southbridge, MA 01550
(508) 764-2072
emjones(at)charter.net
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=424294#424294
Message 6
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Re: Bonding Straps |
At 12:26 PM 6/2/2014, you wrote:
>
>In a mixed construction airplane (steel tube fuselage, wooden wings,
>aluminum structure flight controls, everything fabric covered) what
>would the down side be in not having everything electrically
>bonded? My electrical system is a two wire, 14v system with
>everything running to a "forest of tabs" ground bus on the
>firewall. The avionics (SL40 com, GTX 327 txdr) run to their own,
>dedicated ground bus which in turn is wired to the firewall ground
>bus. I'm doing the bonding per the original design, but I'm just
>curious. Would all the various assemblies of the airplane create
>static charges that would cripple the radios?
>
>Dan
Bonding for static control, grounding for power distribution
and grounding for antenna performance are largely separate
issues.
P-static is not an issue for low speed aircraft cruising
in the lower atmospheres. Sounds like you've got the
lights lit up so the only thing left is to provide
adequate ground planes under antennas. But in no case
is there value in tying every 'grounded' device to
all other grounded devices.
Bob . . .
Message 7
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Re: Bonding Straps |
Thanks Bob. Would there be any value in tying the avionics ground bus directly
to the battery instead of to the main ground bus? (Putting the two ground buses
in parallel instead of in series).
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=424303#424303
Message 8
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Re: Economical solid state pressure sensors |
At 07:55 AM 6/3/2014, you wrote:
I only seem to get a few hundred hours out of
name brand electromechanical engine pressure
sensors so I tried a 30 psi $25. solid state
Chinese sensor for coolant pressure monitoring.
Quite a bit cheaper than what I was using and
seemingly good specifications which I've copied
below. While it calibrates perfectly at room
temperature, it reads at least 5 psi low when it
gets hot (ie. 4 psi when it should be 9). The
zero point changes similarly and I confirmed all
this with raw voltage readings and scope traces
to make certain. This sensor was mounted on the
firewall and tee'd into a 5/16" coolant hose so
the coolant likely warmed it up to about 80*C.
Seems it would need to be installed on a stub
line on the cool side of the firewall to be
useful. The temperature specification of
Maximum:0.05%FS/C would indicate up to 1.5 psi
zeroing error over a 100*C range which would be
acceptable to me if that was the case and other
errors were minor. This particular sensor may be
outside of stated tolerances. In any case, the
compensation temperature range of -20 to +80*C
seems to be meaningless. I'd also hoped to
replace the oil pressure sensor but it seems
these economical sensors are not the answer for under the cowl.
I have one of those transducers in a 300 psi version . . .
in fact . . . it's in my pocket right now. We ordered
it in to consider for a future development program.
I'll see if I can get a quick look-see on accuracy
and temperature effects.
The numbers you're citing are pretty poor . . . I can't
imagine anyone putting so much into a stainless case
and cutting short the electronics . . . electronics
that are pretty much jelly-bean parts. If you're going
to round-file the aberrant transducer, drop it to
me in the mail . . . I'd like to peek at it on the
bench too.
Bob . . .
Message 9
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Info on MGL Enigma |
At 02:37 AM 6/3/2014, you wrote:
Hello again listers,
Here is a further response from Rainier:
I did not say that spikes or similar ARE the cause of what it going on.
It might be something completely unrelated. It's a pretty complex system.
. . . then why even mention it? The customer
does not benefit from a narration of possible
causes . . . especially when those causes include
things that 'might' be wrong with his airplane.
But to assert that stresses external to the
product's enclosure can zap a micro-controller
does nothing to engender confidence in the rest of
hardware under the best of circumstances.
There are no old style linear regulators in the system.
There are three high frequency switch mode supplies and all input power is
protected by a heavy duty transorb apart from other filtering (mostly
inductive/capacitive).
<SNIP>
Electronics do fail - does not matter how much money you spend on the
components. Service history of the Enigma has been excellent. The number of
Enigmas coming back to the factory for repairs is almost nil.
When we conduct a reliability study on an assembly,
powers-that-be don't really start smiling until
your aggregate MTBF drops below 1 x 10 minus 7
failures per flight hour. For critical systems
1 x 10 minus 9 is the target.
So if your device has 100 components in it, INDIVIDUAL
failure rates have to be 100x better than the aggregate.
If this guy has crystals that warrant routine replacement
base on some fielded experience . . . he is well
advised to change brands.
<snip>
intended for aircraft (and not under our brand name). I would love a zero
failure rate. Failures are bad. Everybody gets annoyed - from the customer
to us. We can't have a zero failure rate - that is a pipe dream. But we can
always work to reduce it to the barest minimum. If we did not do that we'd
be out of business long ago.
Most of our repairs are directly related to what we call "blow ups".
Reverse polarity issues, connecting power or grounds to the wrong places
etc. Interface driver chips (RS232, CAN, switches etc) blown for various
reasons - the items directly connected to the outside World.
One cannot necessarily design every i/o port
to withstand the vagaries of neophyte installations
but failures of components not connected directly
to the outside world is cause for concern.
All of that happens typically during installation. Mostly with the so called
"professional installers, AMOs etc". I'm not kidding. We find the home
builders are the ones with the least issues - perhaps because most take the
time to study the docs and plan their wiring. There is no rush, no
deadlines. That helps.
Our statistics (which we do keep for obvious reasons) shows this trend very
clearly.
Instruments get dropped, might have a hard time during transport - sometimes
they get soaked in water or oil.
Anyway, I can't really comment on this Enigma until I see it myself (if that
ever happens). If there is a problem, whatever the cause may be - we can fix
it. That is the bottom line.
To be sure, nobody can offer "zero failure rates" but
they can be exceedingly small. Builders anticipating operations
over long stretches of unfriendly terrain or in IMC would be well
advised to craft 'plan-b', 'plan-c' or whatever in the
event the whole system goes belly-up. "Barest minimum"
failure rates are un-quantified.
If we had any hardware coming back from the field
suffering some of the failures cited above . . .
things around here would, shall we say, get
very 'tense'?
A display with this kind of data . ..
Emacs!
implies certain capabilities. Capabilities that
require TC aircraft designers to go Level A
software/hardware, careful partitioning of
functions and probably dual screens.
Do your own Failure Modes Effects Analysis
based on how you need to use this hardware.
Have a solid alternative for every situation
when it isn't working.
Bob . . .
Message 10
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Re: Bonding Straps |
At 10:52 AM 6/3/2014, you wrote:
>
>Thanks Bob. Would there be any value in tying the avionics ground
>bus directly to the battery instead of to the main ground
>bus? (Putting the two ground buses in parallel instead of in series).
No . . . and especially not the battery(-) post.
The battery is but one of many components that
benefit by sharing a centralized ground but its
(-) terminal point should not be that
centralize ground location.
If you have a panel ground for potential vicitims
that is in turn tied ot the central ground on the
fire wall, you're good to go.
Bob . . .
Message 11
INDEX | Back to Main INDEX |
PREVIOUS | Skip to PREVIOUS Message |
NEXT | Skip to NEXT Message |
LIST | Reply to LIST Regarding this Message |
SENDER | Reply to SENDER Regarding this Message |
|
Subject: | Re: Bonding Straps |
Ok - thanks. That was the way I understood it.
----- Original Message -----
From: Robert L. Nuckolls, III <nuckolls.bob@aeroelectric.com>
Sent: Tue, 03 Jun 2014 13:03:48 -0600 (MDT)
Subject: Re: AeroElectric-List: Re: Bonding Straps
At 10:52 AM 6/3/2014, you wrote:
>
>Thanks Bob. Would there be any value in tying the avionics ground
>bus directly to the battery instead of to the main ground
>bus? (Putting the two ground buses in parallel instead of in series).
No . . . and especially not the battery(-) post.
The battery is but one of many components that
benefit by sharing a centralized ground but its
(-) terminal point should not be that
centralize ground location.
If you have a panel ground for potential vicitims
that is in turn tied ot the central ground on the
fire wall, you're good to go.
Bob . . .
Other Matronics Email List Services
These Email List Services are sponsored solely by Matronics and through the generous Contributions of its members.
-- Please support this service by making your Contribution today! --
|