Today's Message Index:
----------------------
1. 12:35 PM - Current Sensor basic question (donjohnston)
2. 12:52 PM - Re: Current Sensor basic question (C&K)
3. 01:32 PM - Re: Current Sensor basic question (Bill Putney)
4. 01:43 PM - Re: Current Sensor basic question (Daniel Hooper)
5. 01:46 PM - Re: Current Sensor basic question (Alec Myers)
6. 02:00 PM - Re: Current Sensor basic question (C&K)
7. 02:12 PM - Re: Current Sensor basic question (donjohnston)
8. 02:13 PM - Re: Current Sensor basic question (donjohnston)
9. 04:06 PM - Re: Current Sensor basic question (Robert L. Nuckolls, III)
10. 04:27 PM - Re: Z-13/8 Questions (MMiller)
11. 04:27 PM - Re: Re: Current Sensor basic question (C&K)
12. 05:13 PM - Re: Re: Z-13/8 Questions (Charlie England)
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: | Current Sensor basic question |
I'm using the 50a hall effect sensor which came with my Grand Rapids EIS 6000.
But the current numbers don't jibe with the VPX. It's not a huge difference,
but enough that I don't like it.
I've talked with GRT and verified the offset and scale factor configuration but
it's still not right.
While searching to see if anyone else was having this issue I ran across an aftermarket
sensor.
High accuracy current sensor
http://www.crazedpilot.com/high-accuracy-current-sensor-for-5v-input-7-30v-supply-engine-monitor-systems-dynon-mgl-grand-rapids-garmin-and-more-ems-efis-systems/
Interestingly, it looks just like the GRT sensor.
I contacted them to make sure it would work with the EIS6000 and if it came with
the offset and scale factor values that needed to be configured on the EIS.
They said yes and that the correct values were based on the wire size being used.
There is nothing in the GRT documentation about the wire size having an effect
on the offset and scale factor settings.
Is it true that the size of the wire effects the reading of a hall effect sensor?
Don
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=457996#457996
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: | Re: Current Sensor basic question |
I can believe that wire size affects that sensor but not the hollow
round 60 amp sensor that GRT sold me several years ago.
The wonderful thing about the GRT sensor is that you can adjust the
offset and scale for almost any sensor. IMO the values from GRT are
simply a recommended starting value and all of my sensors benefited from
some tweaking when I calibrated them.
Ken
On 09/07/2016 3:34 PM, donjohnston wrote:
>
> I'm using the 50a hall effect sensor which came with my Grand Rapids EIS 6000.
But the current numbers don't jibe with the VPX. It's not a huge difference,
but enough that I don't like it.
>
> I've talked with GRT and verified the offset and scale factor configuration but
it's still not right.
>
> While searching to see if anyone else was having this issue I ran across an aftermarket
sensor.
>
> High accuracy current sensor
> http://www.crazedpilot.com/high-accuracy-current-sensor-for-5v-input-7-30v-supply-engine-monitor-systems-dynon-mgl-grand-rapids-garmin-and-more-ems-efis-systems/
>
> Interestingly, it looks just like the GRT sensor.
>
> I contacted them to make sure it would work with the EIS6000 and if it came with
the offset and scale factor values that needed to be configured on the EIS.
They said yes and that the correct values were based on the wire size being
used. There is nothing in the GRT documentation about the wire size having an
effect on the offset and scale factor settings.
>
> Is it true that the size of the wire effects the reading of a hall effect sensor?
>
> Don
>
>
> Read this topic online here:
>
> http://forums.matronics.com/viewtopic.php?p=457996#457996
>
>
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: Current Sensor basic question |
If it's a Hall Effect sensor the wire size doesn't matter. Magnetic
field is what the Hall Effect device is measuring. What dictates the
magnetic field strength is Ampere-Turns. I assume you have the wire only
passing through the sensor once (one turn).
A smaller wire will have more voltage drop across it for any given
current flow than a larger one, but won't change the current measurement.
Look at your installation. If the wire and the sensor is laying tightly
with other wires, the field may be effected. This should be a very
localized effect so if you have about 1/2" from other wires carrying
current you should be OK.
Bill Putney - WB6RFW
Chief Engineer
KPTZ - Port Townsend, WA
PP-SEL/A&P-IA
"...you know me to be a very smart man. Don't you think if I were wrong, I'd know
it?" -Sheldon Cooper
On 7/9/16 12:34 PM, donjohnston wrote:
>
> I'm using the 50a hall effect sensor which came with my Grand Rapids EIS 6000.
But the current numbers don't jibe with the VPX. It's not a huge difference,
but enough that I don't like it.
>
> I've talked with GRT and verified the offset and scale factor configuration but
it's still not right.
>
> While searching to see if anyone else was having this issue I ran across an aftermarket
sensor.
>
> High accuracy current sensor
> http://www.crazedpilot.com/high-accuracy-current-sensor-for-5v-input-7-30v-supply-engine-monitor-systems-dynon-mgl-grand-rapids-garmin-and-more-ems-efis-systems/
>
> Interestingly, it looks just like the GRT sensor.
>
> I contacted them to make sure it would work with the EIS6000 and if it came with
the offset and scale factor values that needed to be configured on the EIS.
They said yes and that the correct values were based on the wire size being
used. There is nothing in the GRT documentation about the wire size having an
effect on the offset and scale factor settings.
>
> Is it true that the size of the wire effects the reading of a hall effect sensor?
>
> Don
>
>
> Read this topic online here:
>
> http://forums.matronics.com/viewtopic.php?p=457996#457996
>
>
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: | Re: Current Sensor basic question |
In general that is 100% correct, but this sensor doesnt appear to have a split
ring to concentrate the magnetic field into a Hall sensor.
When youre not collecting the whole field, geometry has a large effect on field
strength.
Im curious how this sensor works or if they expect some sort of calibration step
after installation.
> On Jul 9, 2016, at 3:29 PM, Bill Putney <billp@wwpc.com> wrote:
>
>
> If it's a Hall Effect sensor the wire size doesn't matter. Magnetic field is
what the Hall Effect device is measuring. What dictates the magnetic field strength
is Ampere-Turns. I assume you have the wire only passing through the sensor
once (one turn).
>
> A smaller wire will have more voltage drop across it for any given current flow
than a larger one, but won't change the current measurement.
>
> Look at your installation. If the wire and the sensor is laying tightly with
other wires, the field may be effected. This should be a very localized effect
so if you have about 1/2" from other wires carrying current you should be OK.
>
> Bill Putney - WB6RFW
> Chief Engineer
> KPTZ - Port Townsend, WA
>
> PP-SEL/A&P-IA
>
> "...you know me to be a very smart man. Don't you think if I were wrong, I'd
know it?" -Sheldon Cooper
>
> On 7/9/16 12:34 PM, donjohnston wrote:
>>
>> I'm using the 50a hall effect sensor which came with my Grand Rapids EIS 6000.
But the current numbers don't jibe with the VPX. It's not a huge difference,
but enough that I don't like it.
>> I've talked with GRT and verified the offset and scale factor configuration
but it's still not right.
>> While searching to see if anyone else was having this issue I ran across an
aftermarket sensor.
>> High accuracy current sensor
>> http://www.crazedpilot.com/high-accuracy-current-sensor-for-5v-input-7-30v-supply-engine-monitor-systems-dynon-mgl-grand-rapids-garmin-and-more-ems-efis-systems/
>>
>> Interestingly, it looks just like the GRT sensor.
>> I contacted them to make sure it would work with the EIS6000 and if it came
with the offset and scale factor values that needed to be configured on the EIS.
They said yes and that the correct values were based on the wire size being
used. There is nothing in the GRT documentation about the wire size having
an effect on the offset and scale factor settings.
>> Is it true that the size of the wire effects the reading of a hall effect sensor?
>> Don
>>
>>
>>
>>
>> Read this topic online here:
>>
>> http://forums.matronics.com/viewtopic.php?p=457996#457996
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
>
>
>
>
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: Current Sensor basic question |
It appears from the photograph that it's sensing the magnetic field at only one
or two points; The field decays in strength with distance away from the conductor,
and measured at the surface surface a larger conductor has a smaller field
than a small conductor carrying the same current (because of the 1/r rule)
so I would expect the size of the conductor to make a difference.
If the current sensor were a transformer (which effectively integrates the field
strength around a closed loop surrounding the conductor) then no, the size
of the conductor wouldn't make a difference; but this doesn't appear to be such
a thing. Of course a transformer wouldn't be able to detect a DC current so
you'd have to be doing something very smart like integrating the transformer output
which would be expensive, and probably inaccurate.
If it's a Hall Effect sensor the wire size doesn't matter. Magnetic field is what
the Hall Effect device is measuring. What dictates the magnetic field strength
is Ampere-Turns. I assume you have the wire only passing through the sensor
once (one turn).
A smaller wire will have more voltage drop across it for any given current flow
than a larger one, but won't change the current measurement.
Look at your installation. If the wire and the sensor is laying tightly with other
wires, the field may be effected. This should be a very localized effect so
if you have about 1/2" from other wires carrying current you should be OK.
Bill Putney - WB6RFW
Chief Engineer
KPTZ - Port Townsend, WA
PP-SEL/A&P-IA
"...you know me to be a very smart man. Don't you think if I were wrong, I'd know
it?" -Sheldon Cooper
On 7/9/16 12:34 PM, donjohnston wrote:
>
> I'm using the 50a hall effect sensor which came with my Grand Rapids EIS 6000.
But the current numbers don't jibe with the VPX. It's not a huge difference,
but enough that I don't like it.
> I've talked with GRT and verified the offset and scale factor configuration
but it's still not right.
> While searching to see if anyone else was having this issue I ran across an
aftermarket sensor.
> High accuracy current sensor
> http://www.crazedpilot.com/high-accuracy-current-sensor-for-5v-input-7-30v-supply-engine-monitor-systems-dynon-mgl-grand-rapids-garmin-and-more-ems-efis-systems/
>
> Interestingly, it looks just like the GRT sensor.
> I contacted them to make sure it would work with the EIS6000 and if it came
with the offset and scale factor values that needed to be configured on the EIS.
They said yes and that the correct values were based on the wire size being
used. There is nothing in the GRT documentation about the wire size having an
effect on the offset and scale factor settings.
> Is it true that the size of the wire effects the reading of a hall effect sensor?
> Don
>
>
>
>
> Read this topic online here:
>
> http://forums.matronics.com/viewtopic.php?p=457996#457996
>
>
>
>
>
>
>
>
>
>
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: Current Sensor basic question |
Bill
I certainly agree for what I'd call a common hall effect transformer
with the sensor in a gap of a magnetic torus and the wire through the
center hole but the pic that came up when I clicked on Don's link just
appeared to be hall sensors strapped directly on both the sides of the
wire. I would expect such an arrangement to be affected by wire diameter?
Ken
On 09/07/2016 4:29 PM, Bill Putney wrote:
>
> If it's a Hall Effect sensor the wire size doesn't matter. Magnetic
> field is what the Hall Effect device is measuring. What dictates the
> magnetic field strength is Ampere-Turns. I assume you have the wire
> only passing through the sensor once (one turn).
>
> A smaller wire will have more voltage drop across it for any given
> current flow than a larger one, but won't change the current measurement.
>
> Look at your installation. If the wire and the sensor is laying
> tightly with other wires, the field may be effected. This should be a
> very localized effect so if you have about 1/2" from other wires
> carrying current you should be OK.
>
> Bill Putney - WB6RFW
> Chief Engineer
> KPTZ - Port Townsend, WA
>
> PP-SEL/A&P-IA
>
> "...you know me to be a very smart man. Don't you think if I were
> wrong, I'd know it?" -Sheldon Cooper
>
> On 7/9/16 12:34 PM, donjohnston wrote:
>> <don@velocity-xl.com>
>>
>> I'm using the 50a hall effect sensor which came with my Grand Rapids
>> EIS 6000. But the current numbers don't jibe with the VPX. It's not
>> a huge difference, but enough that I don't like it.
>> I've talked with GRT and verified the offset and scale factor
>> configuration but it's still not right.
>> While searching to see if anyone else was having this issue I ran
>> across an aftermarket sensor.
>> High accuracy current sensor
>> http://www.crazedpilot.com/high-accuracy-current-sensor-for-5v-input-7-30v-supply-engine-monitor-systems-dynon-mgl-grand-rapids-garmin-and-more-ems-efis-systems/
>>
>>
>> Interestingly, it looks just like the GRT sensor.
>> I contacted them to make sure it would work with the EIS6000 and if
>> it came with the offset and scale factor values that needed to be
>> configured on the EIS. They said yes and that the correct values were
>> based on the wire size being used. There is nothing in the GRT
>> documentation about the wire size having an effect on the offset and
>> scale factor settings.
>> Is it true that the size of the wire effects the reading of a hall
>> effect sensor?
>> Don
>>
>>
>>
>>
>> Read this topic online here:
>>
>> http://forums.matronics.com/viewtopic.php?p=457996#457996
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>
>
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: Current Sensor basic question |
Can you elaborate on the tweaking process? For example did you adjust the scale
factor or offset? Which direction did you change for what effect?
yellowduckduo(at)gmail.co wrote:
> I can believe that wire size affects that sensor but not the hollow
> round 60 amp sensor that GRT sold me several years ago.
> The wonderful thing about the GRT sensor is that you can adjust the
> offset and scale for almost any sensor. IMO the values from GRT are
> simply a recommended starting value and all of my sensors benefited from
> some tweaking when I calibrated them.
> Ken
>
> On 09/07/2016 3:34 PM, donjohnston wrote:
>
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=458012#458012
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: Current Sensor basic question |
Only one pass through the sensor. No other wires within about 4-6".
billp(at)wwpc.com wrote:
> If it's a Hall Effect sensor the wire size doesn't matter. Magnetic
> field is what the Hall Effect device is measuring. What dictates the
> magnetic field strength is Ampere-Turns. I assume you have the wire only
> passing through the sensor once (one turn).
>
> A smaller wire will have more voltage drop across it for any given
> current flow than a larger one, but won't change the current measurement.
>
> Look at your installation. If the wire and the sensor is laying tightly
> with other wires, the field may be effected. This should be a very
> localized effect so if you have about 1/2" from other wires carrying
> current you should be OK.
>
> On 7/9/16 12:34 PM, donjohnston wrote:
>
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=458013#458013
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: | Re: Current Sensor basic question |
>
>High accuracy current sensor
>http://www.crazedpilot.com/high-accuracy-current-sensor-for-5v-input-7-30v-supply-engine-monitor-systems-dynon-mgl-grand-rapids-garmin-and-more-ems-efis-systems/
>
This is obviously not a 'cored' sensor.
Because there is no control over the
concentration of flux in the measured
wire, there is similarly no physical filtering
of external magnetic fields normally afforded
by the use of a gapped core.
Hence, the need for two sensors on opposite
sides of the wire. Magnetic lines of force
will be equal/opposite in the two sensors
where the sum of the two will represent
a field proportional to flux surrounding
the wire.
Flux from outside the sensor pair will
affect each sensor approximately the same
and since they're wired for adding opposing
fields, non-opposing fields tend to cancel
each other.
As long as the instrumentation system has adequate
accommodation of scale factor and offset, then
this 'coreless' approach seems reasonable.
I use a closed loop, gapped core sensors in
my instrumentation packages. One of my favorites
is this critter
http://tinyurl.com/zxsdyj4
[]
[]
Emacs!
It's core function is a plus-minus 6A sensor
but easy to 'jeep' into fractional multiples and
sub multiples of 6A. Of course, it's designed to
mount on an ECB but I've converted dozens of them
to remotely installed sensors at the end of pendant
cables.
These are gapped-core, closed-loop sensors with
great linearity and calibration numbers. They
too could probably be adapted to the popular
displays. The price sure is right.
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: Z-13/8 Questions |
Bob
Here are the details of about a dozen Rotax regulator failures. Draw your own
conclusion(s).
First some background. There are several versions of the Rotax 912/Ducati regulator. Older versions include Rotax PN 965-345/Ducati 383475, I believe this is the version examined on the http://contrails.free.fr/elec_ducati_en.php website. This was replaced by Rotax PN 965-347/Ducati 343620. This version was manufactured until late 2009 and was replaced by the current version, Rotax PN 965-349/Ducati 362001. I believe this change was driven by component obsolescence, the button case diode was no longer produced. My findings show there are two failure modes, depending on the version of the regulator.
The older pre 2010 regulators fail from the inability to adequately transfer heat
from the diode assembly to the case. This causes the diodes to unsolder themselves
from the assembly. Both diodes show signs of unsoldering, but, If look
closely, the inboard diode always shows more heat damage then the outboard
diode. This asymmetrical failure is likely caused by the asymmetrical geometry
of the enclosure/heat sink. This is the only failure I have seen in this version
regulator. This includes regulators with melted supply terminals (G pin,)
the solder connections on these diodes show the same failure mode. I have
not seen any component fail out of spec in any regulator examined.
Here a schematic/assy for this version
https://drive.google.com/open?id=0ByPxADyU-DwPdTQ3NWp5LW9iOHc
Typical failure mode
https://drive.google.com/open?id=0ByPxADyU-DwPTnlXcUVEQTdfcVU
Current production since 2010 uses a different diode package. This resolved the
internal heatsinking issue of the older regulator versions. Unfortunately, at
this time they changed to gel type potting compound. All the examined failures
of this version are likely caused by vibration. This failure is unique to
the regulators with the new gel potting compound, this failure mode is nonexistent
on the older version regulators.
Here a schematic/assy for this version
https://drive.google.com/open?id=0ByPxADyU-DwPWTBBcGlUeENTcHc
Here are some failure photos
https://drive.google.com/open?id=0ByPxADyU-DwPYm1rYUVsbUJyQ3M
Heat is often the cause of failure in electronics, but these regulators fail from
poor mechanical design and poor material choices, long before heat would kill
them.
nuckolls.bob(at)aeroelect wrote:
> At 06:37 AM 7/5/2016, you wrote:
>
> >
> > Mike Miller has done autopsies on several failed Ducati regulators and he
attributes the failures to poor manufacturing and vibration,
>
> vibration in a potted assembly?
>
>
> > not from heat.
>
> Has he published a teardown report?
>
>
> > Some RV-12 owners have reported a gradual failure where they get low voltage
alarms at RPMs where they used to have normal voltage. Each regulator could
fail differently.
>
> Color me skeptical . . . these regulators have
> been in production for decades. There must have
> been thousands sold. Over that period of time
> and numbers of examples, I find it curious that
> a constellation of failure modes would emerge.
>
> .....
>
> Bob . . .
Read this topic online here:
http://forums.matronics.com/viewtopic.php?p=458023#458023
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: Current Sensor basic question |
Don
The formula is in Section 7.1.5 "Transfer Function" in my GRT manual.
From a quick refresher it seems that increasing the Aux Scale Factor
increases your reading by MULTIPLYING all sensor outputs by a larger
scale factor. Increases or decreases each reading by a similar
percentage so to speak.
Changing the offset simply adds (or subtracts) the same value to all
readings. For example you could change this by some value that adds say
5 amps to all indications. It would then add 5 amps to the indication
even when no current is flowing, and 5 amps to all readings. Tweak this
by an even number such as -2 or +2. Changing it by an odd number such as
1 or 3 reverses the direction for example from adding to subtracting.
If you describe what you are seeing on the display and what it actually
should be at one point preferably somewhere near a normal cruising
current and also with zero current flowing (such as with no wire or a
disconnected wire in the sensor) we can make a suggestion as to what
tweaking might help. There is an interaction in that we get the
multiplier correct and then use the offset to make it read zero when no
current is flowing. Sometimes it takes a couple of tries.
The suggested values from GRT for my sensor were Aux Scale Factor = 93
and Aux Offset = 177 so you probably have those set as a starting point
if it is the same sensor and you are setting it for +/- 50 amps.
Increasing the Aux Scale Factor to 94 would slightly raise all displayed
readings especially the highest readings. If it then reads above zero
with no current flowing you'd try tweaking the Aux Offset to maybe 175
to lower the zero current display a bit.
Ken
On 09/07/2016 5:11 PM, donjohnston wrote:
>
> Can you elaborate on the tweaking process? For example did you adjust the scale
factor or offset? Which direction did you change for what effect?
>
>
> yellowduckduo(at)gmail.co wrote:
>> I can believe that wire size affects that sensor but not the hollow
>> round 60 amp sensor that GRT sold me several years ago.
>> The wonderful thing about the GRT sensor is that you can adjust the
>> offset and scale for almost any sensor. IMO the values from GRT are
>> simply a recommended starting value and all of my sensors benefited from
>> some tweaking when I calibrated them.
>> Ken
>>
>> On 09/07/2016 3:34 PM, donjohnston wrote:
>>
>
>
> Read this topic online here:
>
> http://forums.matronics.com/viewtopic.php?p=458012#458012
>
>
> .
>
Message 12
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: Z-13/8 Questions |
FWIW, I spent a lot of years doing electronics repair, first on consumer
products and then industrial products. The post-2010 failure pics all
look like what electronics techs call 'cold solder joints'. Either bad
prep (dirty contacts or leads), or inadequate heat when soldering. I've
seen joint failures on circuit boards from TV sets that never move after
installation to industrial equipment valued at 6-7 figures. Every time I
dealt with it, cleaning the lead either mechanically (scraping) or by
using a good electronics grade flux during the re-soldering process,
cured the problem.
I have no reason to doubt your word that the components can move if
embedded in 'soft' potting compound, but.... in my opinion, a properly
soldered joint on a pc board pad like that would not fail due to
vibration. The pad/trace would peal up from the substrate, or the lead
would break, before the solder joint to the lead would fail. Properly
executed solder joints are much stronger than most people think.
Charlie
On 7/9/2016 6:24 PM, MMiller wrote:
>
> Bob
> Here are the details of about a dozen Rotax regulator failures. Draw your own
conclusion(s).
>
> First some background. There are several versions of the Rotax 912/Ducati regulator. Older versions include Rotax PN 965-345/Ducati 383475, I believe this is the version examined on the http://contrails.free.fr/elec_ducati_en.php website. This was replaced by Rotax PN 965-347/Ducati 343620. This version was manufactured until late 2009 and was replaced by the current version, Rotax PN 965-349/Ducati 362001. I believe this change was driven by component obsolescence, the button case diode was no longer produced. My findings show there are two failure modes, depending on the version of the regulator.
>
> The older pre 2010 regulators fail from the inability to adequately transfer
heat from the diode assembly to the case. This causes the diodes to unsolder
themselves from the assembly. Both diodes show signs of unsoldering, but, If
look closely, the inboard diode always shows more heat damage then the outboard
diode. This asymmetrical failure is likely caused by the asymmetrical geometry
of the enclosure/heat sink. This is the only failure I have seen in this
version regulator. This includes regulators with melted supply terminals (G pin,)
the solder connections on these diodes show the same failure mode. I have
not seen any component fail out of spec in any regulator examined.
> Here a schematic/assy for this version
> https://drive.google.com/open?id=0ByPxADyU-DwPdTQ3NWp5LW9iOHc
> Typical failure mode
> https://drive.google.com/open?id=0ByPxADyU-DwPTnlXcUVEQTdfcVU
>
> Current production since 2010 uses a different diode package. This resolved
the internal heatsinking issue of the older regulator versions. Unfortunately,
at this time they changed to gel type potting compound. All the examined failures
of this version are likely caused by vibration. This failure is unique
to the regulators with the new gel potting compound, this failure mode is nonexistent
on the older version regulators.
> Here a schematic/assy for this version
> https://drive.google.com/open?id=0ByPxADyU-DwPWTBBcGlUeENTcHc
> Here are some failure photos
> https://drive.google.com/open?id=0ByPxADyU-DwPYm1rYUVsbUJyQ3M
>
> Heat is often the cause of failure in electronics, but these regulators fail
from poor mechanical design and poor material choices, long before heat would
kill them.
>
>
> nuckolls.bob(at)aeroelect wrote:
>> At 06:37 AM 7/5/2016, you wrote:
>>
>>>
>>> Mike Miller has done autopsies on several failed Ducati regulators and he
attributes the failures to poor manufacturing and vibration,
>> vibration in a potted assembly?
>>
>>
>>> not from heat.
>> Has he published a teardown report?
>>
>>
>>> Some RV-12 owners have reported a gradual failure where they get low voltage
alarms at RPMs where they used to have normal voltage. Each regulator could
fail differently.
>> Color me skeptical . . . these regulators have
>> been in production for decades. There must have
>> been thousands sold. Over that period of time
>> and numbers of examples, I find it curious that
>> a constellation of failure modes would emerge.
>>
>> .....
>>
>> 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! --
|