Today's Message Index:
----------------------
1. 07:51 AM - Toggle Switches with Fast-On Tabs (Robert L. Nuckolls, III)
2. 08:04 AM - Re: Switch problem ***MAJOR UPDATE*** (Robert L. Nuckolls, III)
3. 08:51 AM - Re: AeroElectric-List Digest: 11 Msgs - 09/12/08 (Scott Klemptner)
4. 10:00 AM - Re: Re: AeroElectric-List Digest: 11 Msgs - 09/12/08 (Richard E. Tasker)
5. 10:09 AM - Re: Re: AeroElectric-List Digest: 11 Msgs - 09/12/08 (Robert L. Nuckolls, III)
6. 10:28 AM - Re: Re: AeroElectric-List Digest: 11 Msgs - 09/12/08 (Robert L. Nuckolls, III)
7. 10:30 AM - Re: Re: AeroElectric-List Digest: 11 Msgs - 09/12/08 (ROGER & JEAN CURTIS)
8. 10:39 AM - Re: Re: AeroElectric-List Digest: 11 Msgs - 09/12/08 (Vernon Little)
9. 11:03 AM - Re: AC vs DC ratings???? (Robert L. Nuckolls, III)
10. 11:03 AM - Re: Toggle Switches with Fast-On Tabs (Vernon Little)
11. 11:16 AM - Switch Ratings versus Root Cause of Failure (Robert L. Nuckolls, III)
12. 11:33 AM - Re: Toggle Switches with Fast-On Tabs (Robert L. Nuckolls, III)
13. 12:20 PM - Re: Toggle Switches with Fast-On Tabs (Dan Reeves)
14. 02:14 PM - Re: Toggle Switches with Fast-On Tabs (Vernon Little)
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: | Toggle Switches with Fast-On Tabs |
Let's lay out all the simple ideas behind the design
fabrication and operation of a toggle switch fitted with
fast-on tabs for the purpose of discovering a failure mode
and deducing a remedy. I'll refer the serious students to
the quick-n-dirty sketch at:
http://aeroelectric.com/Pictures/Switches/Toggle_Switch_with_Fast-On_Tabs.jpg
This cross-section will allow us to trace the path of
current flow through the switch as follows:
Electrons come in via Wire(B) and pass to the fast-on
terminal through wire grip(1) and then on to the fast-on
tab(C) through a high-pressure metal-to-metal
terminal/tab interface(2). Current must then pass
through a high-pressure, metal-to-metal joint(3) to
the contact(D). A low-pressure, metal-to-metal
contact/contact interface(4) carries current to
the teeter-totter(F) via another high-pressure,
metal-to-metal joint(5).
Current flows through the teeter-totter to a
low-pressure, metal-to-metal sliding joint(6)
at the top of the saddle(G) and then down to
a high-pressure, metal-to-metal joint at the
saddle to rivet interface(7), through the plastic
housing(A) to another high-pressure, metal-to-metal
staked joint(8) and thence on to the fast-on-tab.
From the fast-on-tab, we find another high-pressure,
metal-to-metal joint at the terminal/tab interface(9)
and finally, another high-pressure, metal-to-metal
joint at the terminal's wire grip(10).
So if you count them up, there are TEN, conductor-
to-conductor joints that carry current through this
switch installation when the switch is closed.
By inspection we can deduce that the weakest links
in this conductor chain are at the low-pressure,
metal-to-metal, NON GAS TIGHT joints at (4) and (6).
Indeed, the first switch failure we considered gave
us physical evidence of a failure at (6) that produced
a slowly progressing failure of the switch. This
study was described in detail at:
http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/Anatomy_of_a_Switch_Failure.html
Now we are pondering a demonstrated case of repeated
failures over a period of years in one aircraft. Here
are some of the hard data points:
(a) All of the switches involved are Carlingswitch
toggles with fast-on tabs. These switches are
an exceedingly mature design that dates back
at least 50 years. IF the root cause of failure
lies with the switches, then it's most likely
a failure of process and not of design.
(b) All switches showed signs of heating on terminals
that are also on "loose rivets" at (3) or (8).
The fast-on terminals also showed signs of
over-heating in the form of discolored insulators
over the wire-grips.
(c) While the majority of switch failures were used
in circuits that carry substantial amounts of
current (strobe and landing lights) the first
failure reported was in a master switch that
carries 1A of contactor current and field
current of perhaps 4A max with an in-flight
average current on the order of 1A.
(d) A photo offered at:
http://aeroelectric.com/Pictures/Switches/VL_Switch_Failure_2.jpg
shows distinct signs of over-heating at the
terminal's insulator but no overt signs of
overheating in the metal parts under the terminal.
I posited the hypothesis a few days ago that IF
the source of heating came from within the switch
and IF temperatures rose high enough to discolor
the terminal's insulation, then temperatures on
the metal parts under the terminal would be high
enough to discolor their surfaces as well. For
example, in the photo at:
http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/b.jpg
We see an overheated and loose rivet. We also see
signs of discoloration due to accelerated corrosion
of the fast-on tab adjacent to the rivet head.
These were the overt signs visible from OUTSIDE a
switch that was in serious trouble from heating
effects INSIDE . . . the fast-on-terminal was not
hot enough to distort the shape or color of the
terminal's wire grip.
Now, the loose rivet phenomenon is easily explained
by a degradation of structural integrity of the
switch's plastic housing(A). Further, since the
failed switches have obviously been running hot,
it follows that the plastic has lost structural
integrity due to heating . . . what is NOT obvious
is whether the initial heat-source came from INSIDE
or OUTSIDE the switch.
(e) We know that hundreds of thousands of switches using
this design and process are flying on aircraft. IF
there is a problem with the switches, then the BIG
puzzle to be solved is why we find a suite of failures
spanning years of switch production batches and many
flight hours of the subject aircraft. The astute
investigator is obligated to consider all features
of the current path study cited above and either
confirm or discard each of the TEN metal-to-metal
connections as candidates for root cause of
the failures.
Okay, this dissertation illustrates our of understanding
at the time of this writing.
It's not only useful but necessary to discount or
confirm the integrity of wire grip joints on the terminals
ESPECIALLY in light of localized heating observed
on the wire-grips of the terminals. This line of
investigation is further encouraged by analysis
of the probability of such concentration of switch
failures having root cause in design or construction of
the switches. This consideration alone is strong suggestion of
an ALTERNATIVE EFFECT COMMON TO ALL THE FAILURES.
On a related topic it has been suggested in the
pages of this forum that Fast-On terminals have
no "Gas Tight" qualities. For clarity let us agree
on the meaning of gas tight. In the dissertation above
I've used the terms high-pressure and metal-to-metal
to describe the interface between two conductors.
By high pressure, I'm speaking of conditions severe
enough to deform metal, i.e. upset its surface or
shape. Keep in mind that this kind of activity
implies pressures in the tens of thousands of pounds
per square inch. In the context of gripping the
strands of wire in the crimp of a terminal, the
term "gas tight" is very descriptive of the
design goal.
Consider the sketched cross-section of a fast-on
joint which I've posted at:
http://aeroelectric.com/Pictures/Terminals/Fast-On_Physics.jpg
We speak to this drawing during the weekend seminars
and point out that most individuals look at a Fast-On
terminal and incorrectly deduce that the spring
forces at the ends of the grips(A) provide an
enduring connection to the tab(B) at the flat interface
between terminal and tab at (2).
Consider that when you push the Fast-On terminal onto
a tab and pull it off, an examination of the area under
the tips of the grips at (1) will show bright lines
or scratches in the tab metal surface. Tiny? yes.
Pressure? Pushing the terminal onto the tab plows
furrows in the surface of the tab i.e. exerts pressures
in the tens of thousands of PSI. The pressures on the
back side of the interface at (2) are a tiny fraction
of those found on the front side.
Therefore, I suggest that not only are the interfaces
at (1) gas-tight (due to the intimate contact of terminal
and tab) the interface at (2) is not gas tight. While
(2) may contribute significantly to joint conductivity
when shiny and new, it's contribution ten years hence is
a small fraction of the total.
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: | Switch problem ***MAJOR UPDATE*** |
At 05:23 PM 9/12/2008 -0700, you wrote:
><rv-9a-online@telus.net>
>
>Bob: A couple of switches, the burnt fast-ons, and some sample terminations
>on your way by airmail today. Hope you have time to add a learned second
>opinion after you have a look.
Great. I've published a working document that lists all
the simple-ideas to be considered in this study of
your switch problems. Taking a good look at joint
integrity of the wire-grips is a good place to start.
Bob . . .
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: AeroElectric-List Digest: 11 Msgs - 09/12/08 |
Vern,
you quoted Carling switch ratings, but if you read further down on the page, you
find the following:
"Types of Loads:
An electric load is the
amount of electric power delivered or required at any specific point or
points on a system. The requirement originates at the energy consuming
equipment of the consumers. More simply put, a load is the piece of
equipment you turn on and off.
Resistive loads primarily offer resistance to the flow
of current. Examples of resistive loads include electric heaters,
ranges, ovens, toasters, and irons. If the device is supposed to get
hot and doesn't move, it's most likely a resistive load.
Inductive loads are usually devices that move and
normally include electric magnets, like an electric motor. Examples of
inductive loads include such things as power drills, electric mixers,
fans, sewing machines, and vacuum cleaners. Transformers also produce
inductive loads.
High Inrush loads draw a higher amount of current or
amperage when first turned on, compared to the amount of current
required to continue running. An example of a high inrush load is a
light bulb, which may draw 20 or more times its normal operating
current when first turned on. This is often referred to as lamp load.
Other examples of loads that have high inrush are switching power
supplies (capacitive load) and motors (inductive load).
.
.
.
.
.
L & T Ratings
An "L" rating denotes the ability of a switch to handle the initial
high inrush characteristics of a Tungsten Filament Lamp on AC voltage
only. A "T" rating is the equivalent lamp load for DC.
H Rating
An "H" rating denotes a non-inductive resistive rating. Ratings listed
in Carling Technologies' product information may appear with the symbol
"H" or with the words "non-inductive" or "resistive". "H" ratings are
typically required for switches used in commercial oven applications."
Are you sure you are using the switches within their rated specs???????????
My experience using AC rated switches on DC circuits with high inrush currents
(landing/taxi lights and strobes in particular) is failure after failure.
Eric....this includes ROCKER SWITCHES TOO!.... Cessna rocker switches used in the
70's for example (AC rated "Mr Coffee Maker switches).
My $.02 worth......
Scott A Klemptner
bmwr606 on Yahoo IM
What if the Hokey-Pokey IS what it's all about?
----- Original Message ----
From: AeroElectric-List Digest Server <aeroelectric-list@matronics.com>
Sent: Saturday, September 13, 2008 1:55:46 AM
________________________________ Message 7 _____________________________________
Time: 09:42:41 AM PST US
From: "Vernon Little" <rv-9a-online@telus.net>
Subject: RE: AeroElectric-List: Re: Switch problem ***MAJOR UPDATE***
...
> Despite what Bob says, 115 VAC switches should never be used
> for LV DC high-current applications. Do what the manufacturer
> suggests. That's what's printed on the side of the switch
> Bubela. Otherwise you're on your own.
>
...
> --------
> Eric M. Jones
> www.PerihelionDesign.com
...
Hi again Eric. Just to comment on what you said:
"DC Rule of Thumb
For those switches that list an AC voltage rating only, the "DC Rule of
Thumb" can be applied for determining the switch's maximum DC current
rating. This "rule" states the highest amperage on the switch should perform
satisfactorily up to 30 volts DC. For example, a switch which is rated at
10A 250VAC; 15A 125VAC; 3/4HP 125-250VAC, will be likely to perform
satisfactorily at 15 amps up to 30 volts DC (VDC)"
http://www.carlingtech.com/products/switches/learn_more.asp?page=switches_amp-rating
Therefore, all of my Carling switches are being used according to the
manufacturers recommendations. The only gotcha is that the switches are not
rated for operating below 0 degrees C. This is probably because the
switches are not sealed from condensing moisture. This is not a problem
where I live.
Vern
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: AeroElectric-List Digest: 11 Msgs - 09/12/08 |
A comment on the below comments:
If you look at the schematic Vern posted, you would see that he is using
NTC surge limiters on each light. These minimize startup surges caused
by the very low cold filament resistance of the lamps.
On the other hand, it depends on the time constants of the NTC surge
limiters vs the time constants of the lamp filaments as to whether these
are effective when in wig-wag mode.
Dick Tasker
Scott Klemptner wrote:
> Vern,
>
> you quoted Carling switch ratings, but if you read further down on the
> page, you find the following:
>
> "Types of Loads:
> An *electric load* is the amount of electric power delivered or
> required at any specific point or points on a system. The requirement
> originates at the energy consuming equipment of the consumers. More
> simply put, a load is the piece of equipment you turn on and off.
>
> *Resistive loads* primarily offer resistance to the flow of current.
> Examples of resistive loads include electric heaters, ranges, ovens,
> toasters, and irons. If the device is supposed to get hot and doesn't
> move, it's most likely a resistive load.
>
> *Inductive loads* are usually devices that move and normally include
> electric magnets, like an electric motor. Examples of inductive loads
> include such things as power drills, electric mixers, fans, sewing
> machines, and vacuum cleaners. Transformers also produce inductive loads.
>
> *High Inrush loads* draw a higher amount of current or amperage when
> first turned on, compared to the amount of current required to
> continue running. An example of a high inrush load is a light bulb,
> which may draw 20 or more times its normal operating current when
> first turned on. This is often referred to as lamp load. Other
> examples of loads that have high inrush are switching power supplies
> (capacitive load) and motors (inductive load).
>
> .
>
> .
>
> .
>
> .
>
> .
>
> L & T Ratings
>
> An "L" rating denotes the ability of a switch to handle the initial
> high inrush characteristics of a Tungsten Filament Lamp on AC voltage
> only. A "T" rating is the equivalent lamp load for DC.
>
>
> H Rating
>
> An "H" rating denotes a non-inductive resistive rating. Ratings listed
> in Carling Technologies' product information may appear with the
> symbol "H" or with the words "non-inductive" or "resistive". "H"
> ratings are typically required for switches used in commercial oven
> applications."
>
>
> Are you sure you are using the switches within their rated
> specs???????????
>
> My experience using AC rated switches on DC circuits with high inrush
> currents (landing/taxi lights and strobes in particular) is failure
> after failure.
>
> Eric....this includes ROCKER SWITCHES TOO!.... Cessna rocker switches
> used in the 70's for example (AC rated "Mr Coffee Maker switches).
>
> My $.02 worth......
>
> Scott A Klemptner
> bmwr606 on Yahoo IM
>
> What if the Hokey-Pokey IS what it's all about?
>
> ----- Original Message ----
> From: AeroElectric-List Digest Server <aeroelectric-list@matronics.com>
> To: AeroElectric-List Digest List <aeroelectric-list-digest@matronics.com>
> Sent: Saturday, September 13, 2008 1:55:46 AM
>
>
> ________________________________ Message 7
> _____________________________________
>
>
> Time: 09:42:41 AM PST US
> From: "Vernon Little" <rv-9a-online@telus.net
> <mailto:rv-9a-online@telus.net>>
> Subject: RE: AeroElectric-List: Re: Switch problem ***MAJOR UPDATE***
>
>
> ...
> > Despite what Bob says, 115 VAC switches should never be used
> > for LV DC high-current applications. Do what the manufacturer
> > suggests. That's what's printed on the side of the switch
> > Bubela. Otherwise you're on your own.
> >
> ...
> > --------
> > Eric M. Jones
> > www.PerihelionDesign.com
> ...
>
> Hi again Eric. Just to comment on what you said:
>
> "DC Rule of Thumb
> For those switches that list an AC voltage rating only, the "DC Rule of
> Thumb" can be applied for determining the switch's maximum DC current
> rating. This "rule" states the highest amperage on the switch should
> perform
> satisfactorily up to 30 volts DC. For example, a switch which is rated at
> 10A 250VAC; 15A 125VAC; 3/4HP 125-250VAC, will be likely to perform
> satisfactorily at 15 amps up to 30 volts DC (VDC)"
> http://www.carlingtech.com/products/switches/learn_more.asp?page=switches_am
> <http://www.carlingtech.com/products/switches/learn_more.asp?page=switches_amp-rating%20>p-rating
> <http://www.carlingtech.com/products/switches/learn_more.asp?page=switches_amp-rating%20>
> Therefore, all of my Carling switches are being used according to the
> manufacturers recommendations. The only gotcha is that the switches
> are not
> rated for operating below 0 degrees C. This is probably because the
> switches are not sealed from condensing moisture. This is not a problem
> where I live.
>
> Vern
>
>
> *
>
>
> *
--
Please Note:
No trees were destroyed in the sending of this message. We do concede, however,
that a significant number of electrons may have been temporarily inconvenienced.
--
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: AeroElectric-List Digest: 11 Msgs - 09/12/08 |
At 08:48 AM 9/13/2008 -0700, you wrote:
>Vern,
>
>you quoted Carling switch ratings, but if you read further down on the
>page, you find the following:
>
>"Types of Loads:
> An electric load is the amount of electric power delivered or required
> at any specific point or points on a system. The requirement originates
> at the energy consuming equipment of the consumers. More simply put, a
> load is the piece of equipment you turn on and off.
<snip>
>Are you sure you are using the switches within their rated specs???????????
>
>My experience using AC rated switches on DC circuits with high inrush
>currents (landing/taxi lights and strobes in particular) is failure after
>failure.
>
>Eric....this includes ROCKER SWITCHES TOO!.... Cessna rocker switches used
>in the 70's for example (AC rated "Mr Coffee Maker switches).
>
>My $.02 worth......
>
>Scott A Klemptner
>bmwr606 on Yahoo IM
Scott,
Keep in mind that RATINGS have to do with SWITCH
SERVICE LIFE generally given in the tens of thousands of
cycles with various conditions. Our use of switches in
light aircraft is exceedingly light-duty in terms
of service life.
For example, a master switch (one of those cited in
the constellation of failures) operates ONCE PER FLIGHT-CYCLE
and carries very small currents compared to the switch's
RATINGS.
Scott, please review the arguments offered in the
piece at:
http://www.aeroelectric.com/articles/Switch_Ratings.pdf
and explain where you find discrepancies of fact or
logic. I'll suggest that had Vern sat in his airplane
one afternoon and put 1000 cycles on any of the failed
switches, he would NOT have experienced a
failure. Yet, over a period of many flight hours
and a mere handful of operations he experienced
failures that speak to long term effects of ENVIRONMENT
that include DEFICIENCIES of process either in fabrication
of the switch as described here . . .
http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/Anatomy_of_a_Switch_Failure.html
or installation of the switch as hypothesized in
my earlier treatise published this morning.
The old AC vs. DC as a critical driving force in
switch life is a myth. Further, it is clear that the
study before us has nothing to do with switch
life in terms of ratings, operating cycles, or
application. Root cause of these failures will be
loss of conduction integrity in one or more of the
ten metal-to-metal joints that make up the switch's
current path.
Bob . . .
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: AeroElectric-List Digest: 11 Msgs - 09/12/08 |
At 12:52 PM 9/13/2008 -0400, you wrote:
><retasker@optonline.net>
>
>A comment on the below comments:
>
>If you look at the schematic Vern posted, you would see that he is using
>NTC surge limiters on each light. These minimize startup surges caused by
>the very low cold filament resistance of the lamps.
>
>On the other hand, it depends on the time constants of the NTC surge
>limiters vs the time constants of the lamp filaments as to whether these
>are effective when in wig-wag mode.
>
>Dick Tasker
A couple of years ago, the effectiveness/value of surge
limiters in Wig-Wag systems was discussed here on the List.
I went to the workbench and took this trace from the
operating current of a 55W halogen lamp flashed at
about 2 flashes per second.
http://www.aeroelectric.com/Pictures/Curves/Wig_Wag_Currents.jpg
Note that the first time the lamp gets power we
see the characteristic cold-filament inrush current.
However, on subsequent turn-ons, the lamp does not
have enough time to cool such that the 7x inrush
experience is repeated.
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: AeroElectric-List Digest: 11 Msgs - 09/12/08 |
Vern,
you quoted Carling switch ratings, but if you read further down on the
page,
you find the following:
"Types of Loads:
An electric load is the amount of electric power delivered or required
at
any specific point or points on a system. The requirement originates at
the
energy consuming equipment of the consumers. More simply put, a load is
the
piece of equipment you turn on and off.
Resistive loads primarily offer resistance to the flow of current.
Examples
of resistive loads include electric heaters, ranges, ovens, toasters,
and
irons. If the device is supposed to get hot and doesn't move, it's most
likely a resistive load.
Inductive loads are usually devices that move and normally include
electric
magnets, like an electric motor. Examples of inductive loads include
such
things as power drills, electric mixers, fans, sewing machines, and
vacuum
cleaners. Transformers also produce inductive loads.
High Inrush loads draw a higher amount of current or amperage when first
turned on, compared to the amount of current required to continue
running.
An example of a high inrush load is a light bulb, which may draw 20 or
more
times its normal operating current when first turned on. This is often
referred to as lamp load. Other examples of loads that have high inrush
are
switching power supplies (capacitive load) and motors (inductive load).
.
.
.
.
.
L & T Ratings
An "L" rating denotes the ability of a switch to handle the initial
high
inrush characteristics of a Tungsten Filament Lamp on AC voltage only. A
"T"
rating is the equivalent lamp load for DC.
H Rating
An "H" rating denotes a non-inductive resistive rating. Ratings listed
in
Carling Technologies' product information may appear with the symbol "H"
or
with the words "non-inductive" or "resistive". "H" ratings are typically
required for switches used in commercial oven applications."
Are you sure you are using the switches within their rated
specs???????????
My experience using AC rated switches on DC circuits with high inrush
currents (landing/taxi lights and strobes in particular) is failure
after
failure.
Eric....this includes ROCKER SWITCHES TOO!.... Cessna rocker switches
used
in the 70's for example (AC rated "Mr Coffee Maker switches).
My $.02 worth......
Scott A Klemptner
bmwr606 on Yahoo IM
Since there is no reference to DC switch ratings above, does this mean
that
the DC switches are immune to different load types and L, T, & H
ratings?
What, may I ask, is the difference between a DC switch and an AC
switch???
Roger
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: AeroElectric-List Digest: 11 Msgs - 09/12/08 |
Hi Scott. The landing and taxi light circuits have inrush current
limiters
installed, for this very reason. The objective was to save wear and
tear on
the switches. Steady state draw is about 5 amps.
As for the strobe supply, there was some debate on my theory of why this
is
the worst load in the a/c. My hypothesis is that strobe supplies have a
negative voltage-current relationship (negative resistance). If you
reduce
the voltage to a strobe supply (or, in fact many switchmode power
supplies),
its current actually increases. Once a switch or terminal fails, this
leads
to thermal runaway and the results are what I have seen (twice).
Not everyone agrees, but I've had this failure twice and Bob has
documented
another one-- all in the strobe circuits.
Thanks for your feedback.
-----Original Message-----
From: owner-aeroelectric-list-server@matronics.com
[mailto:owner-aeroelectric-list-server@matronics.com] On Behalf Of Scott
Klemptner
Sent: September 13, 2008 8:48 AM
Subject: AeroElectric-List: Re: AeroElectric-List Digest: 11 Msgs -
09/12/08
Vern,
you quoted Carling switch ratings, but if you read further down on the
page,
you find the following:
"Types of Loads:
An electric load is the amount of electric power delivered or required
at
any specific point or points on a system. The requirement originates at
the
energy consuming equipment of the consumers. More simply put, a load is
the
piece of equipment you turn on and off.
Resistive loads primarily offer resistance to the flow of current.
Examples
of resistive loads include electric heaters, ranges, ovens, toasters,
and
irons. If the device is supposed to get hot and doesn't move, it's most
likely a resistive load.
Inductive loads are usually devices that move and normally include
electric
magnets, like an electric motor. Examples of inductive loads include
such
things as power drills, electric mixers, fans, sewing machines, and
vacuum
cleaners. Transformers also produce inductive loads.
High Inrush loads draw a higher amount of current or amperage when first
turned on, compared to the amount of current required to continue
running.
An example of a high inrush load is a light bulb, which may draw 20 or
more
times its normal operating current when first turned on. This is often
referred to as lamp load. Other examples of loads that have high inrush
are
switching power supplies (capacitive load) and motors (inductive load).
.
.
.
.
.
L & T Ratings
An "L" rating denotes the ability of a switch to handle the initial
high
inrush characteristics of a Tungsten Filament Lamp on AC voltage only. A
"T"
rating is the equivalent lamp load for DC.
H Rating
An "H" rating denotes a non-inductive resistive rating. Ratings listed
in
Carling Technologies' product information may appear with the symbol "H"
or
with the words "non-inductive" or "resistive". "H" ratings are typically
required for switches used in commercial oven applications."
Are you sure you are using the switches within their rated
specs???????????
My experience using AC rated switches on DC circuits with high inrush
currents (landing/taxi lights and strobes in particular) is failure
after
failure.
Eric....this includes ROCKER SWITCHES TOO!.... Cessna rocker switches
used
in the 70's for example (AC rated "Mr Coffee Maker switches).
My $.02 worth......
Scott A Klemptner
bmwr606 on Yahoo IM
What if the Hokey-Pokey IS what it's all about?
----- Original Message ----
From: AeroElectric-List Digest Server <aeroelectric-list@matronics.com>
<aeroelectric-list-digest@matronics.com>
Sent: Saturday, September 13, 2008 1:55:46 AM
________________________________ Message 7
_____________________________________
Time: 09:42:41 AM PST US
From: "Vernon Little" <rv-9a-online@telus.net>
Subject: RE: AeroElectric-List: Re: Switch problem ***MAJOR UPDATE***
...
> Despite what Bob says, 115 VAC switches should never be used
> for LV DC high-current applications. Do what the manufacturer
> suggests. That's what's printed on the side of the switch
> Bubela. Otherwise you're on your own.
>
...
> --------
> Eric M. Jones
> www.PerihelionDesign.com
...
Hi again Eric. Just to comment on what you said:
"DC Rule of Thumb
For those switches that list an AC voltage rating only, the "DC Rule of
Thumb" can be applied for determining the switch's maximum DC current
rating. This "rule" states the highest amperage on the switch should
perform
satisfactorily up to 30 volts DC. For example, a switch which is rated
at
10A 250VAC; 15A 125VAC; 3/4HP 125-250VAC, will be likely to perform
satisfactorily at 15 amps up to 30 volts DC (VDC)"
http://www.carlingtech.com/products/switches/learn_more.asp?page=switch
es_am
<http://www.carlingtech.com/products/switches/learn_more.asp?page=switc
hes_a
mp-rating%20> p-rating
<http://www.carlingtech.com/products/switches/learn_more.asp?page=switc
hes_a
mp-rating%20>
Therefore, all of my Carling switches are being used according to the
manufacturers recommendations. The only gotcha is that the switches are
not
rated for operating below 0 degrees C. This is probably because the
switches are not sealed from condensing moisture. This is not a problem
where I live.
Vern
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: AC vs DC ratings???? |
At 01:28 PM 9/13/2008 -0400, you wrote:
>Since there is no reference to DC switch ratings above, does this mean
>that the DC switches are immune to different load types and L, T, & H ratings?
>What, may I ask, is the difference between a DC switch and an AC switch???
Roger, check the article at:
http://www.aeroelectric.com/articles/Switch_Ratings.pdf
ALL switches have CAPABILITY in either DC or
AC systems. Not all switches are RATED for all
applications . . . especially if one limits their
enlightenment to the tiny print on the side of the
switch. There's not enough room to state the full
constellation of a switch's capabilities in that little
space. So what's a mother to do? Everyone prints
ratings on the device that address the majority of
applications where that switch would be used. Service
in household and industrial AC line operated appliance.
Does the lack of DC numbers mean that the device
has no capability in DC operations? Not at all.
But in spite of publication of engineering data there
are individuals who cannot or choose not to
interpret that data in useful ways. Such is the
case for the author of the article in Van's Air Force
newsletter that prompted the article cited above.
Yes, there ARE switch designs that will exhibit
optimized service life in DC or AC applications.
However, the fact that any given switch is not
optimized for one service or the other does not
mean that it does not offer useful service in our
airplanes.
Because of a very low duty-cycle in light aircraft,
ratings are only loosely tied to suitability to task.
Failure of a switch in our airplanes is more likely
to be related to environmental conditions as opposed
to service stresses.
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: | Toggle Switches with Fast-On Tabs |
Hi Bob: One more variable: The serious switch failures (the one you
analyzed, plus my two) have all been in strobe circuits. We've had some
debate about this in the past, but perhaps the analysis of the I-V
characteristics of Whelen strobe supply would be in order. I'm curious if
there is a combination of elements problem here with switches, terminals and
load.
When my master switch failed (loose rivets), it manifested itself with
alternator overvoltage alarms. The voltage regulator was boosting
alternator output to compensate for a voltage drop in the master switch.
Normally, you'd suspect a regulator, but it was the switch. The ultimate
boilerplate fix was to use the new switch to control a relay that connected
the alternator output to the regulator directly (through appropriate
fuselink).
I'm wondering if a similar band-aid fix for the strobe supply is in order.
Perhaps use the switch to control a good automotive relay for this circuit.
I don't have a setup for testing a strobe supply myself. Perhaps Whelen has
some insight.
Vern
> -----Original Message-----
> From: owner-aeroelectric-list-server@matronics.com
> [mailto:owner-aeroelectric-list-server@matronics.com] On
> Behalf Of Robert L. Nuckolls, III
> Sent: September 13, 2008 7:49 AM
> To: aeroelectric-list@matronics.com
> Subject: AeroElectric-List: Toggle Switches with Fast-On Tabs
>
>
>
> --> <nuckolls.bob@cox.net>
>
> Let's lay out all the simple ideas behind the design
> fabrication and operation of a toggle switch fitted with
> fast-on tabs for the purpose of discovering a failure mode
> and deducing a remedy. I'll refer the serious students to the
> quick-n-dirty sketch at:
>
> http://aeroelectric.com/Pictures/Switches/Toggle_Switch_with_F
> ast-On_Tabs.jpg
>
> This cross-section will allow us to trace the path of
> current flow through the switch as follows:
>
> Electrons come in via Wire(B) and pass to the fast-on
> terminal through wire grip(1) and then on to the fast-on
> tab(C) through a high-pressure metal-to-metal
> terminal/tab interface(2). Current must then pass
> through a high-pressure, metal-to-metal joint(3) to
> the contact(D). A low-pressure, metal-to-metal
> contact/contact interface(4) carries current to
> the teeter-totter(F) via another high-pressure,
> metal-to-metal joint(5).
>
> Current flows through the teeter-totter to a
> low-pressure, metal-to-metal sliding joint(6)
> at the top of the saddle(G) and then down to
> a high-pressure, metal-to-metal joint at the
> saddle to rivet interface(7), through the plastic
> housing(A) to another high-pressure, metal-to-metal
> staked joint(8) and thence on to the fast-on-tab.
>
> From the fast-on-tab, we find another high-pressure,
> metal-to-metal joint at the terminal/tab interface(9) and
> finally, another high-pressure, metal-to-metal joint at the
> terminal's wire grip(10).
>
> So if you count them up, there are TEN, conductor-
> to-conductor joints that carry current through this
> switch installation when the switch is closed.
>
> By inspection we can deduce that the weakest links
> in this conductor chain are at the low-pressure,
> metal-to-metal, NON GAS TIGHT joints at (4) and (6). Indeed,
> the first switch failure we considered gave us physical
> evidence of a failure at (6) that produced a slowly
> progressing failure of the switch. This study was described
> in detail at:
>
> http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failu
> re/Anatomy_of_a_Switch_Failure.html
>
> Now we are pondering a demonstrated case of repeated
> failures over a period of years in one aircraft. Here
> are some of the hard data points:
>
> (a) All of the switches involved are Carlingswitch
> toggles with fast-on tabs. These switches are
> an exceedingly mature design that dates back
> at least 50 years. IF the root cause of failure
> lies with the switches, then it's most likely
> a failure of process and not of design.
>
> (b) All switches showed signs of heating on terminals
> that are also on "loose rivets" at (3) or (8).
> The fast-on terminals also showed signs of
> over-heating in the form of discolored insulators
> over the wire-grips.
>
> (c) While the majority of switch failures were used
> in circuits that carry substantial amounts of
> current (strobe and landing lights) the first
> failure reported was in a master switch that
> carries 1A of contactor current and field
> current of perhaps 4A max with an in-flight
> average current on the order of 1A.
>
> (d) A photo offered at:
>
> http://aeroelectric.com/Pictures/Switches/VL_Switch_Failure_2.jpg
>
> shows distinct signs of over-heating at the
> terminal's insulator but no overt signs of
> overheating in the metal parts under the terminal.
>
> I posited the hypothesis a few days ago that IF
> the source of heating came from within the switch
> and IF temperatures rose high enough to discolor
> the terminal's insulation, then temperatures on
> the metal parts under the terminal would be high
> enough to discolor their surfaces as well. For
> example, in the photo at:
>
> http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/b.jpg
>
> We see an overheated and loose rivet. We also see
> signs of discoloration due to accelerated corrosion
> of the fast-on tab adjacent to the rivet head.
> These were the overt signs visible from OUTSIDE a
> switch that was in serious trouble from heating
> effects INSIDE . . . the fast-on-terminal was not
> hot enough to distort the shape or color of the
> terminal's wire grip.
>
> Now, the loose rivet phenomenon is easily explained
> by a degradation of structural integrity of the
> switch's plastic housing(A). Further, since the
> failed switches have obviously been running hot,
> it follows that the plastic has lost structural
> integrity due to heating . . . what is NOT obvious
> is whether the initial heat-source came from INSIDE
> or OUTSIDE the switch.
>
> (e) We know that hundreds of thousands of switches using
> this design and process are flying on aircraft. IF
> there is a problem with the switches, then the BIG
> puzzle to be solved is why we find a suite of failures
> spanning years of switch production batches and many
> flight hours of the subject aircraft. The astute
> investigator is obligated to consider all features
> of the current path study cited above and either
> confirm or discard each of the TEN metal-to-metal
> connections as candidates for root cause of
> the failures.
>
> Okay, this dissertation illustrates our of understanding
> at the time of this writing.
>
> It's not only useful but necessary to discount or
> confirm the integrity of wire grip joints on the terminals
> ESPECIALLY in light of localized heating observed on the
> wire-grips of the terminals. This line of investigation is
> further encouraged by analysis of the probability of such
> concentration of switch failures having root cause in design
> or construction of the switches. This consideration alone is
> strong suggestion of an ALTERNATIVE EFFECT COMMON TO ALL THE FAILURES.
>
> On a related topic it has been suggested in the
> pages of this forum that Fast-On terminals have
> no "Gas Tight" qualities. For clarity let us agree
> on the meaning of gas tight. In the dissertation above
> I've used the terms high-pressure and metal-to-metal
> to describe the interface between two conductors.
> By high pressure, I'm speaking of conditions severe
> enough to deform metal, i.e. upset its surface or
> shape. Keep in mind that this kind of activity
> implies pressures in the tens of thousands of pounds
> per square inch. In the context of gripping the
> strands of wire in the crimp of a terminal, the
> term "gas tight" is very descriptive of the
> design goal.
>
> Consider the sketched cross-section of a fast-on
> joint which I've posted at:
>
> http://aeroelectric.com/Pictures/Terminals/Fast-On_Physics.jpg
>
> We speak to this drawing during the weekend seminars
> and point out that most individuals look at a Fast-On
> terminal and incorrectly deduce that the spring
> forces at the ends of the grips(A) provide an
> enduring connection to the tab(B) at the flat interface
> between terminal and tab at (2).
>
> Consider that when you push the Fast-On terminal onto
> a tab and pull it off, an examination of the area under
> the tips of the grips at (1) will show bright lines
> or scratches in the tab metal surface. Tiny? yes.
> Pressure? Pushing the terminal onto the tab plows
> furrows in the surface of the tab i.e. exerts pressures
> in the tens of thousands of PSI. The pressures on the
> back side of the interface at (2) are a tiny fraction
> of those found on the front side.
>
> Therefore, I suggest that not only are the interfaces
> at (1) gas-tight (due to the intimate contact of terminal
> and tab) the interface at (2) is not gas tight. While
> (2) may contribute significantly to joint conductivity
> when shiny and new, it's contribution ten years hence is
> a small fraction of the total.
>
> 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: | Switch Ratings versus Root Cause of Failure |
At 10:38 AM 9/13/2008 -0700, you wrote:
>Hi Scott. The landing and taxi light circuits have inrush current
>limiters installed, for this very reason. The objective was to save wear
>and tear on the switches. Steady state draw is about 5 amps.
>As for the strobe supply, there was some debate on my theory of why this
>is the worst load in the a/c. My hypothesis is that strobe supplies have
>a negative voltage-current relationship (negative resistance). If you
>reduce the voltage to a strobe supply (or, in fact many switchmode power
>supplies), its current actually increases. Once a switch or terminal
>fails, this leads to thermal runaway and the results are what I have seen
>(twice).
>
>Not everyone agrees, but I've had this failure twice and Bob has
>documented another one-- all in the strobe circuits.
It is true that the "negative resistance" or more appropriately
"constant power" characteristics of switchmode power supplies may
have contributed to the pace of failure. However, in terms
of having abused the switch's ratings I'll refer the readers to
the photo at:
http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/g.jpg
This is the working contact in a failed strobe supply switch
we examined a few years ago. Note that the contact - the major
life-limiting component is almost pristine in comparison with
the carnage going on elsewhere . . .
http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/e.jpg
There's more to this failure analysis stuff than beating
on the ratings drum. I'll have to sift through the gray matter
but I don't think I have NEVER encountered a failure of
a component aboard an aircraft wherein root cause was a failure
to observe ratings. It's the FIRST question that should be
asked but it's also the easiest concern to put to rest.
Bob . . .
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: | Toggle Switches with Fast-On Tabs |
At 10:58 AM 9/13/2008 -0700, you wrote:
><rv-9a-online@telus.net>
>
>Hi Bob: One more variable: The serious switch failures (the one you
>analyzed, plus my two) have all been in strobe circuits. We've had some
>debate about this in the past, but perhaps the analysis of the I-V
>characteristics of Whelen strobe supply would be in order. I'm curious if
>there is a combination of elements problem here with switches, terminals and
>load.
>
>When my master switch failed (loose rivets), it manifested itself with
>alternator overvoltage alarms. The voltage regulator was boosting
>alternator output to compensate for a voltage drop in the master switch.
>Normally, you'd suspect a regulator, but it was the switch. The ultimate
>boilerplate fix was to use the new switch to control a relay that connected
>the alternator output to the regulator directly (through appropriate
>fuselink).
Yes, that drops the current through the switch to some
value on the order of 100 milliamps on the alternator side
while leaving the contactor load of about 1A on the battery
side. This raise some big red flags suggesting that ratings
abuse is not root cause of the failures you've suffered.
>I'm wondering if a similar band-aid fix for the strobe supply is in order.
>Perhaps use the switch to control a good automotive relay for this circuit.
>
>I don't have a setup for testing a strobe supply myself. Perhaps Whelen has
>some insight.
Assuming that root cause turns out to be some form of
ratings abuse, then yes . . . the fix is get back within
the bounds of ratings.
This could include things like buffer relays . . .
http://www.aeroelectric.com/PPS/Adobe_Architecture_Pdfs/Z32K.pdf
http://www.aeroelectric.com/PPS/Adobe_Architecture_Pdfs/Z22-23K.pdf
. . . or other adjustment to architecture.
But it's not clear to me that we're dealing with a ratings
issue. You've had failures in the master switch . . . this is
probably the most lightly loaded device on your switch panel!
If we are looking at a ratings issue, I think it will be a
first in my career of chasing the sources of smoke in airplanes.
Bob . . .
Message 13
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: | Toggle Switches with Fast-On Tabs |
Perhaps this was already discussed but is it possible that you have a short
through the panel itself somewhere?- Kind of like if you lost your engin
e ground strap and hit the starter, the current finds a ground through thro
ttle cables, etc and fries whatever is in it's path.--Maybe a cigarette
lighter style recepticle shorting through the panel and through the switch
housings mounted to the panel.- I-wonder if something is vibrating and
maybe loosly or intermittantly laying across the fast-on tabs.- Any chan
ce of a parking brake cable or something like that making contact with some
thing-electrical when it's set or released.- Maybe all irrelevant,,,but
just thoughts that came to mind while reading through the recent posts and
not hearing of others experiencing similar switch failures.
-
--- On Sat, 9/13/08, Vernon Little <rv-9a-online@telus.net> wrote:
From: Vernon Little <rv-9a-online@telus.net>
Subject: RE: AeroElectric-List: Toggle Switches with Fast-On Tabs
<rv-9a-online@telus.net>
Hi Bob: One more variable: The serious switch failures (the one you
analyzed, plus my two) have all been in strobe circuits. We've had some
debate about this in the past, but perhaps the analysis of the I-V
characteristics of Whelen strobe supply would be in order. I'm curious if
there is a combination of elements problem here with switches, terminals an
d
load.
When my master switch failed (loose rivets), it manifested itself with
alternator overvoltage alarms. The voltage regulator was boosting
alternator output to compensate for a voltage drop in the master switch.
Normally, you'd suspect a regulator, but it was the switch. The ultimate
boilerplate fix was to use the new switch to control a relay that connected
the alternator output to the regulator directly (through appropriate
fuselink).
I'm wondering if a similar band-aid fix for the strobe supply is in order.
Perhaps use the switch to control a good automotive relay for this circuit.
I don't have a setup for testing a strobe supply myself. Perhaps Whelen
has
some insight.
Vern
> -----Original Message-----
> From: owner-aeroelectric-list-server@matronics.com
> [mailto:owner-aeroelectric-list-server@matronics.com] On
> Behalf Of Robert L. Nuckolls, III
> Sent: September 13, 2008 7:49 AM
> To: aeroelectric-list@matronics.com
> Subject: AeroElectric-List: Toggle Switches with Fast-On Tabs
>
>
>
III"
> --> <nuckolls.bob@cox.net>
>
> Let's lay out all the simple ideas behind the design
> fabrication and operation of a toggle switch fitted with
> fast-on tabs for the purpose of discovering a failure mode
> and deducing a remedy. I'll refer the serious students to the
> quick-n-dirty sketch at:
>
> http://aeroelectric.com/Pictures/Switches/Toggle_Switch_with_F
> ast-On_Tabs.jpg
>
> This cross-section will allow us to trace the path of
> current flow through the switch as follows:
>
> Electrons come in via Wire(B) and pass to the fast-on
> terminal through wire grip(1) and then on to the fast-on
> tab(C) through a high-pressure metal-to-metal
> terminal/tab interface(2). Current must then pass
> through a high-pressure, metal-to-metal joint(3) to
> the contact(D). A low-pressure, metal-to-metal
> contact/contact interface(4) carries current to
> the teeter-totter(F) via another high-pressure,
> metal-to-metal joint(5).
>
> Current flows through the teeter-totter to a
> low-pressure, metal-to-metal sliding joint(6)
> at the top of the saddle(G) and then down to
> a high-pressure, metal-to-metal joint at the
> saddle to rivet interface(7), through the plastic
> housing(A) to another high-pressure, metal-to-metal
> staked joint(8) and thence on to the fast-on-tab.
>
> From the fast-on-tab, we find another high-pressure,
> metal-to-metal joint at the terminal/tab interface(9) and
> finally, another high-pressure, metal-to-metal joint at the
> terminal's wire grip(10).
>
> So if you count them up, there are TEN, conductor-
> to-conductor joints that carry current through this
> switch installation when the switch is closed.
>
> By inspection we can deduce that the weakest links
> in this conductor chain are at the low-pressure,
> metal-to-metal, NON GAS TIGHT joints at (4) and (6). Indeed,
> the first switch failure we considered gave us physical
> evidence of a failure at (6) that produced a slowly
> progressing failure of the switch. This study was described
> in detail at:
>
> http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failu
> re/Anatomy_of_a_Switch_Failure.html
>
> Now we are pondering a demonstrated case of repeated
> failures over a period of years in one aircraft. Here
> are some of the hard data points:
>
> (a) All of the switches involved are Carlingswitch
> toggles with fast-on tabs. These switches are
> an exceedingly mature design that dates back
> at least 50 years. IF the root cause of failure
> lies with the switches, then it's most likely
> a failure of process and not of design.
>
> (b) All switches showed signs of heating on terminals
> that are also on "loose rivets" at (3) or (8).
> The fast-on terminals also showed signs of
> over-heating in the form of discolored insulators
> over the wire-grips.
>
> (c) While the majority of switch failures were used
> in circuits that carry substantial amounts of
> current (strobe and landing lights) the first
> failure reported was in a master switch that
> carries 1A of contactor current and field
> current of perhaps 4A max with an in-flight
> average current on the order of 1A.
>
> (d) A photo offered at:
>
> http://aeroelectric.com/Pictures/Switches/VL_Switch_Failure_2.jpg
>
> shows distinct signs of over-heating at the
> terminal's insulator but no overt signs of
> overheating in the metal parts under the terminal.
>
> I posited the hypothesis a few days ago that IF
> the source of heating came from within the switch
> and IF temperatures rose high enough to discolor
> the terminal's insulation, then temperatures on
> the metal parts under the terminal would be high
> enough to discolor their surfaces as well. For
> example, in the photo at:
>
> http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/b.jpg
>
> We see an overheated and loose rivet. We also see
> signs of discoloration due to accelerated corrosion
> of the fast-on tab adjacent to the rivet head.
> These were the overt signs visible from OUTSIDE a
> switch that was in serious trouble from heating
> effects INSIDE . . . the fast-on-terminal was not
> hot enough to distort the shape or color of the
> terminal's wire grip.
>
> Now, the loose rivet phenomenon is easily explained
> by a degradation of structural integrity of the
> switch's plastic housing(A). Further, since the
> failed switches have obviously been running hot,
> it follows that the plastic has lost structural
> integrity due to heating . . . what is NOT obvious
> is whether the initial heat-source came from INSIDE
> or OUTSIDE the switch.
>
> (e) We know that hundreds of thousands of switches using
> this design and process are flying on aircraft. IF
> there is a problem with the switches, then the BIG
> puzzle to be solved is why we find a suite of failures
> spanning years of switch production batches and many
> flight hours of the subject aircraft. The astute
> investigator is obligated to consider all features
> of the current path study cited above and either
> confirm or discard each of the TEN metal-to-metal
> connections as candidates for root cause of
> the failures.
>
> Okay, this dissertation illustrates our of understanding
> at the time of this writing.
>
> It's not only useful but necessary to discount or
> confirm the integrity of wire grip joints on the terminals
> ESPECIALLY in light of localized heating observed on the
> wire-grips of the terminals. This line of investigation is
> further encouraged by analysis of the probability of such
> concentration of switch failures having root cause in design
> or construction of the switches. This consideration alone is
> strong suggestion of an ALTERNATIVE EFFECT COMMON TO ALL THE FAILURES.
>
> On a related topic it has been suggested in the
> pages of this forum that Fast-On terminals have
> no "Gas Tight" qualities. For clarity let us agree
> on the meaning of gas tight. In the dissertation above
> I've used the terms high-pressure and metal-to-metal
> to describe the interface between two conductors.
> By high pressure, I'm speaking of conditions severe
> enough to deform metal, i.e. upset its surface or
> shape. Keep in mind that this kind of activity
> implies pressures in the tens of thousands of pounds
> per square inch. In the context of gripping the
> strands of wire in the crimp of a terminal, the
> term "gas tight" is very descriptive of the
> design goal.
>
> Consider the sketched cross-section of a fast-on
> joint which I've posted at:
>
> http://aeroelectric.com/Pictures/Terminals/Fast-On_Physics.jpg
>
> We speak to this drawing during the weekend seminars
> and point out that most individuals look at a Fast-On
> terminal and incorrectly deduce that the spring
> forces at the ends of the grips(A) provide an
> enduring connection to the tab(B) at the flat interface
> between terminal and tab at (2).
>
> Consider that when you push the Fast-On terminal onto
> a tab and pull it off, an examination of the area under
> the tips of the grips at (1) will show bright lines
> or scratches in the tab metal surface. Tiny? yes.
> Pressure? Pushing the terminal onto the tab plows
> furrows in the surface of the tab i.e. exerts pressures
> in the tens of thousands of PSI. The pressures on the
> back side of the interface at (2) are a tiny fraction
> of those found on the front side.
>
> Therefore, I suggest that not only are the interfaces
> at (1) gas-tight (due to the intimate contact of terminal
> and tab) the interface at (2) is not gas tight. While
> (2) may contribute significantly to joint conductivity
> when shiny and new, it's contribution ten years hence is
> a small fraction of the total.
>
> Bob . . .
>
>
>
>
>
>
=0A=0A=0A
Message 14
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: | Toggle Switches with Fast-On Tabs |
Hi Dan. Not likely. I checked for for shorts. What I measured was a 2
volt drop across the switch when operating (terminal to terminal).
About
12.5V in, 10.5V out. This does not mean I don't have a wiring problem
downstream somewhere, but there is no evidence of it, the breaker is
intact
and the strobes operate. It also means that the switch is dissipation
about
14 Watts. No wonder it's hot.
Vern
-----Original Message-----
From: owner-aeroelectric-list-server@matronics.com
[mailto:owner-aeroelectric-list-server@matronics.com] On Behalf Of Dan
Reeves
Sent: September 13, 2008 12:17 PM
Subject: RE: AeroElectric-List: Toggle Switches with Fast-On Tabs
Perhaps this was already discussed but is it possible that you have a
short
through the panel itself somewhere? Kind of like if you lost your
engine
ground strap and hit the starter, the current finds a ground through
throttle cables, etc and fries whatever is in it's path. Maybe a
cigarette
lighter style recepticle shorting through the panel and through the
switch
housings mounted to the panel. I wonder if something is vibrating and
maybe
loosly or intermittantly laying across the fast-on tabs. Any chance of
a
parking brake cable or something like that making contact with something
electrical when it's set or released. Maybe all irrelevant,,,but just
thoughts that came to mind while reading through the recent posts and
not
hearing of others experiencing similar switch failures.
--- On Sat, 9/13/08, Vernon Little <rv-9a-online@telus.net> wrote:
From: Vernon Little <rv-9a-online@telus.net>
Subject: RE: AeroElectric-List: Toggle Switches with Fast-On Tabs
<rv-9a-online@telus.net>
Hi Bob: One more variable: The serious switch failures (the one you
analyzed, plus my two) have all been in strobe circuits. We've had some
debate about this in the past, but perhaps the analysis of the I-V
characteristics of Whelen strobe supply would be in order. I'm curious
if
there is a combination of elements problem here with switches, terminals
and
load.
When my master switch failed (loose rivets), it manifested itself with
alternator overvoltage alarms. The voltage regulator was boosting
alternator output to compensate for a voltage drop in the master switch.
Normally, you'd suspect a regulator, but it was the switch. The
ultimate
boilerplate fix was to use the new switch to control a relay that
connected
the alternator output to the regulator directly (through appropriate
fuselink).
I'm wondering if a similar band-aid fix for the strobe supply is in
order.
Perhaps use the switch to control a good automotive relay for this
circuit.
I don't have a setup for testing a strobe supply myself. Perhaps Whelen
has
some insight.
Vern
> -----Original Message-----
> From: owner-aeroelectric-list-server@matronics.com
> [mailto:owner-aeroelectric-list-server@matronics.com] On
> Behalf Of Robert L. Nuckolls, III
> Sent: September 13, 2008 7:49 AM
> To: aeroelectric-list@matronics.com
> Subject: AeroElectric-List: Toggle Switches with Fast-On Tabs
>
>
>
III"
> --> <nuckolls.bob@cox.net>
>
> Let's lay out all the simple ideas behind the design
> fabrication and operation of a toggle switch fitted with
> fast-on tabs for the purpose of discovering a failure mode
> and deducing a remedy. I'll refer the serious students to the
> quick-n-dirty sketch at:
>
> http://aeroelectric.com/Pictures/Switches/Toggle_Switch_with_F
> ast-On_Tabs.jpg
>
> This cross-section will allow us to trace the path of
> current flow through the switch as follows:
>
> Electrons come in via Wire(B) and pass to the fast-on
> terminal through wire grip(1) and then on to the fast-on
> tab(C) through a high-pressure metal-to-metal
> terminal/tab interface(2). Current must then pass
> through a high-pressure, metal-to-metal joint(3) to
> the contact(D). A low-pressure, metal-to-metal
> contact/contact interface(4) carries current to
> the teeter-totter(F) via another high-pressure,
> metal-to-metal joint(5).
>
> Current flows through the teeter-totter to a
> low-pressure, metal-to-metal sliding joint(6)
> at the top of the saddle(G) and then down to
> a high-pressure, metal-to-metal joint at the
> saddle to rivet interface(7), through the plastic
> housing(A) to another high-pressure, metal-to-metal
> staked joint(8) and thence on to the fast-on-tab.
>
> From the fast-on-tab, we find another high-pressure,
> metal-to-metal joint at the terminal/tab interface(9) and
> finally, another high-pressure, metal-to-metal joint at the
> terminal's wire grip(10).
>
> So if you count them up, there are TEN, conductor-
> to-conductor joints that carry current through this
> switch installation when the switch is closed.
>
> By inspection we can deduce that the weakest links
> in this conductor chain are at the low-pressure,
> metal-to-metal, NON GAS TIGHT joints at (4) and (6). Indeed,
> the first switch failure we considered gave us physical
> evidence of a failure at (6) that produced a slowly
> progressing failure of the switch. This study was described
> in detail at:
>
> http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failu
> re/Anatomy_of_a_Switch_Failure.html
>
> Now we are pondering a demonstrated case of repeated
> failures over a period of years in one aircraft. Here
> are some of the hard data points:
>
> (a) All of the switches involved are Carlingswitch
> toggles with fast-on tabs. These switches are
> an exceedingly mature design that dates back
> at least 50 years. IF the root cause of failure
> lies with the switches, then it's most likely
> a failure of process and not of design.
>
> (b) All switches showed signs of heating on terminals
> that are also on "loose rivets" at (3) or (8).
> The fast-on terminals also showed signs of
> over-heating in the form of discolored insulators
> over the wire-grips.
>
> (c) While the majority of switch failures were used
> in circuits that carry substantial amounts of
> current (strobe and landing lights) the first
> failure reported was in a master switch that
> carries 1A of contactor current and field
> current of perhaps 4A max with an in-flight
> average current on the order of 1A.
>
> (d) A photo offered at:
>
> http://aeroelectric.com/Pictures/Switches/VL_Switch_Failure_2.jpg
>
> shows distinct signs of over-heating at the
> terminal's insulator but no overt signs of
> overheating in the metal parts under the terminal.
>
> I posited the hypothesis a few days ago that IF
> the source of heating came from within the switch
> and IF temperatures rose high enough to discolor
> the terminal's insulation, then temperatures on
> the metal parts under the terminal would be high
> enough to discolor their surfaces as well. For
> example, in the photo at:
>
> http://www.aeroelectric.com/articles/Anatomy_of_a_Switch_Failure/b.jpg
>
> We see an overheated and loose rivet. We also see
> signs of discoloration due to accelerated corrosion
> of the fast-on tab adjacent to the rivet head.
> These were the overt signs visible from OUTSIDE a
> switch that was in serious trouble from heating
> effects INSIDE . . . the fast-on-terminal was not
> hot enough to distort the shape or color of the
> terminal's wire grip.
>
> Now, the loose rivet phenomenon is easily explained
> by a degradation of structural integrity of the
> switch's plastic housing(A). Further, since the
> failed switches have obviously been running hot,
> it follows that the plastic has lost structural
> integrity due to heating . . . what is NOT obvious
> is whether the initial heat-source came from INSIDE
> or OUTSIDE the switch.
>
> (e) We know that hundreds of thousands of switches using
> this design and process are flying on aircraft. IF
> there is a problem with the switches, then the BIG
> puzzle to be solved is why we find a suite of failures
> spanning years of switch production batches and many
> flight hours of the subject aircraft. The astute
> investigator is obligated to consider all features
> of the current path study cited above and either
> confirm or discard each of the TEN metal-to-metal
> connections as candidates for root cause of
> the failures.
>
> Okay, this dissertation illustrates our of understanding
> at the time of this writing.
>
> It's not only useful but necessary to discount or
> confirm the integrity of wire grip joints on the terminals
> ESPECIALLY in light of localized heating observed on the
> wire-grips of the terminals. This line of investigation is
> further encouraged by analysis of the probability of such
> concentration of switch failures having root cause in design
> or construction of the switches. This consideration alone is
> strong suggestion of an ALTERNATIVE EFFECT COMMON TO ALL THE FAILURES.
>
> On a related topic it has been suggested in the
> pages of this forum that Fast-On terminals have
> no "Gas Tight" qualities. For clarity let us agree
> on the meaning of gas tight. In the dissertation above
> I've used the terms high-pressure and metal-to-metal
> to describe the interface between two conductors.
> By high pressure, I'm speaking of conditions severe
> enough to deform metal, i.e. upset its surface or
> shape. Keep in mind that this kind of activity
> implies pressures in the tens of thousands of pounds
> per square inch. In the context of gripping the
> strands of wire in the crimp of a terminal, the
> term "gas tight" is very descriptive of the
> design goal.
>
> Consider the sketched cross-section of a fast-on
> joint which I've posted at:
>
> http://aeroelectric.com/Pictures/Terminals/Fast-On_Physics.jpg
>
> We speak to this drawing during the weekend seminars
> and point out that most individuals look at a Fast-On
> terminal and incorrectly deduce that the spring
> forces at the ends of the grips(A) provide an
> enduring connection to the tab(B) at the flat interface
> between terminal and tab at (2).
>
> Consider that when you push the Fast-On terminal onto
> a tab and pull it off, an examination of the area under
> the tips of the grips at (1) will show bright lines
> or scratches in the tab metal surface. Tiny? yes.
> Pressure? Pushing the terminal onto the tab plows
> furrows in the surface of the tab i.e. exerts pressures
> in the tens of thousands of PSI. The pressures on the
> back side of the interface at (2) are a tiny fraction
> of those found on the front side.
>
> Therefore, I suggest that not only are the interfaces
> at (1) gas-tight (due to the intimate contact of terminal
> and tab) the interface at (2) is not gas tight. While
> (2) may contribute significantly to joint conductivity
> when shiny and new, it's contribution ten years hence is
> a small fraction of the total.
>
> Bob . . .
>
>
>
>
>
>
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3
D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3
D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3
D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3
D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
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! --
|