Today's Message Index:
----------------------
1. 09:58 AM - Bridge diodes VS Schottky (Paul)
2. 10:25 AM - Re: Electrical Mystery (Chuck Jensen)
3. 12:25 PM - Re: Re: AeroElectric-Battery/SD8 problem (Frank Stringham)
4. 01:24 PM - Re: Strobe shielded wire splice (Robert L. Nuckolls, III)
5. 02:51 PM - power to handheld gps (bob noffs)
6. 03:10 PM - Re: Bridge diodes VS Schottky (Robert L. Nuckolls, III)
7. 03:17 PM - Re: Re: AeroElectric-Battery/SD8 problem (Robert L. Nuckolls, III)
8. 03:33 PM - Re: power to handheld gps (Ernest Christley)
9. 03:54 PM - Re: Re: AeroElectric-List Digest: 29 Msgs - 07/03/08 (Robert L. Nuckolls, III)
10. 08:08 PM - Re: Bridge diodes VS Schottky (Paul)
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: | Bridge diodes VS Schottky |
I have seen the debate over the low cost Bridge diode VS the Schottky diode
come up year after year mostly based on the opinion the bridge is good
enough and easy to get at RS, mount it anywhere to a bracket (implying no
significant heat sinking is needed etc). Perhaps I have missed it but I do
not recall a single technical analysis based on engineering evaluation of
the specified values based on data sheet analysis. Further many suggested
uses show two diodes of the bridge in parallel. This does not increase the
current capacity much as diodes do not share current well as one will get
hotter and hog the current making it hotter etc. Its not clear the
reliability is increased as internal bridge package connections are not
known and vary with the manufacturer. Diodes typically fail short and only
open if the current is high enough.
Over this nearly 15 year period the bridge diode has been used for several
applications but most often as I recall is to isolate the emergency (now
endurance) bus. This bus originally was apparently designed for 2-3 amp max.
Today many users have a 8 amp load and still use the bridge diode. Notably,
Eric has promoted the Schottky diode and been shouted down over and over
with no real data to support the opposing positions.
15 years ago I selected the Schottky diode as I had dual batteries and
wanted to use the diode to provide battery isolation and and also single
source battery charging. The bridge diode simply would not charge the
battery to full charge while the Schottky did.
Data sheet worst case design analysis can be difficult as the typical data
sheet is intended to make the part look good and typically shows information
that is at least misleading with the real worst case information hard to
find.
Today the bridge diode part is very old and many manufacturers have
shortened the data sheet to the extent its not possible to determine the
worst case parameters. Some data sheets are simply scans from another
Manufacturer's old data sheet.
I evaluated both types of bridge diodes available today. The GBPC series and
the older MB series. The MB series has a metal base with metal sides. The
GBPC series has a smaller metal base and no metal sides. The electrical
parameters are essentially the same and both are available with faston type
connections. Remember some data is shown as nominal, not max ratings.
Remember that bridge diodes are rated as a full wave rectifier which means
some data must be converted to DC.
Here are the worst case values for Vf at 1 and 10 amps and at 25C and 150C.
This is for a single diode.
Note the data is basically a straight line from 1A to 10A which allows for
simple intermediate current calculations
Vf 1A 25C Vf 10A 25C Vf 1A 150C Vf 10A 150C
0.83V 1.05V 0.63V 0.85V
Power dissipation as follows at 25C
1A 2.5A 5.0A 7.5A 10A
0.83W 2.23W 4.7W 7.5W 10.5W
To figure case temp consider 1.9deg/c per watt from die temp. For example
case temp max at 10 amps is 10.5 times 1.9 approx 20C or 130C max. If
ambient temp is 40C near the mount then the heat sink must dissipate 10.5W
with a temp drop of 90C across the heatsink.
The above does not allow for the case to heat sink temp resistance depending
on heat sinking surface material IE grease. The above does not allow for
service derating. The aerospace industry recommends no more that 100C case
temp which increases the heat sink size for a temp gradient of from 90 to 60
c max
Its clear that above a current of 1A, some (or a lot of) heat sink is
required. I do not recall any specified requirement other than to simply
mount it on a small bracket. No mention of the surface smoothness of the
bracket nor what it's thickness or size is. No mention of the required heat
sink grease etc. At 10 amps a couple of manufacturers require a heat sink
size of 6" x 2.2" x 2.2" finned plate plus additional requirements to keep
the junction temperature below 150C (the max rated temperature). To
calculate case temp use 1.9deg/c per watt from die temp. For example case
temp max at 10 amps is 10.5 times 1.9 approx 20C or 130C max.
Thus its not wrong to state a 10W heat sink is needed but there is a string
of calculations required to determine the physical heat sink needed.
Now lets look at the current Schottky diode available. In this case the
diode selected is the DSS2x121-0045B made by IXYS. Two independent 120 amp
diodes. In the case of Schottky diodes the higher the current rating the
lower the Vf at low currents as seen in this application. IR used to be the
low Vf leader but this IXYS is slightly better and lower cost and easy to
get. ( www.findchips.com )
Vf 1A 25C Vf 10A 25C Vf 1A 150C Vf 10A 150C
0.29V 0.34A 0.15V 0.24V
Power dissipation as follows at 25C
1A 2.5A 5.0A 7.5A 10A
0.29W 0.78W 1.58W 2.48W 3.40W
Clearly a far better part to use. Just as easy to mount and wiring support
is easier.
If battery charging is intended thru this diode, not only is a full charge
possible but there is no current overload as is likely to occur with the
bridge if the battery being charged is significantly discharged and where
the charging current can be well above 10 amps.
Why worst case analysis? Because its not possible to know the actual
parameters of the parts used by the various builders. Manufacturers do
produce worst case parts and are often sold to resellers like RS with the
higher quality parts are distributed to the industry market.
Why not just do a test and use that as proof of the design? A test only
proves the worst case analysis is valid and is only good for the exact parts
used and does not indicate what results will be with similar parts from the
identical manufacturers production batch. Proper engineering design always
starts with a worst case analysis of the design followed with a test to show
the results are inside the worst case analysis. Only then is the design
released for production or in this case released for the general builder to
build and be sure it will perform to expectations.
Availability? Its strange to me that the availability of the bridge at RS is
promoted but many of the recommended switches are only available thru B&C.
Modern electronics use switching supplies which use more or less power not
amps as input baded on input voltage. Thus the higher the Vf the more amps
required and thus shorter battery life as the battery life in amp hours goes
down as the amp drain goes up.
The cost of the superior Schottky part is very small compared to the cost of
the electrical wiring/terminal etc system and insignificant to the total
cost of the aircraft.
Paul
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: | Electrical Mystery |
EPILOG: After conducting an investigation of the previously submitted 'electrical
mystery', it was determined that the battery had two bad cells. Oddly, the
battery started the plane fine, helped by it being a good-starting Lyc 360,
but with electrical load, the voltage (apparently) dropped to the point that the
430 would start to drop out. At elevated RPMs, the generator was able to provide
sufficient flow of electrons to keep everybody happy, but as soon as RPMs
dropped, the generator output followed suit and the battery simply would not
pickup the load to keep voltage up.
Thanks to everyone for their ideas and suggestions.
Chuck Jensen
-----Original Message-----
From: owner-aeroelectric-list-server@matronics.com
[mailto:owner-aeroelectric-list-server@matronics.com]On Behalf Of Robert
L. Nuckolls, III
Sent: Thursday, July 03, 2008 10:14 AM
Subject: Re: AeroElectric-List: Electrical Mystery
At 07:44 AM 7/2/2008 -0400, you wrote:
>
>This was related to me by a fellow hangar mate:
>
>While taxing to the runway I got a message on my 430 Garmin and the
>nav/com2 started blinking as if it where loosing power.
>The Mooney is powered by an 0360 A2F lycoming with a generator charging
>system.. I had a normal start-up and all was normal. Engine RPM directly
>after start-up was 700. After a couple of minutes of warm up I take it to
>1000 to allow vacuum to increase. I set my radios and instruments then
>began my taxi.
>
>All indications were normal until half way to the runway. I retrieved a
>message from the 430 that said "no altitude input" and the Nav/Com2
>started blinking. RPM was 1200 during taxi. When I stopped for run-up the
>garmin had shut off completely. I took the RPM to 2000 and it came back on
>and I was able to reset my destination. All other indications were normal.
>Amp gauge was showing a positive charge as it normally does. I turned off
>all my lights and when I dropped the RPM to approximately 1500 the garmin
>and nav/com2 once again lost power.
>
>Is this symptomatic of an electron illness of a particular sort?
The FIRST thing to do in these situations is to KNOW
what the bus voltage is with the engine at rpms conducive
to full output from the alternator/generator. When you
say "showing positive charge as it normally does", was
there any indication that the MAGNITUDE of that reading
was lower than normal but still positive?
Since you have a generator and in light of the poor
service-life/reliability of generators compared to
alternators, the highest probability for cause of your
observed symptoms is that the generator is not functioning.
Indeed, it may have been out of service for several flights.
I had alternators go bad on some of our training fleet
at 1K1 and the instructors were able to complete a day's
worth of training flights battery only.
Put a battery maintainer on your ship's battery and
leave it attached until the green light comes on (the
charger says the battery has accepted all the energy
it's capable of taking on). Then put a voltmeter on
the bus and start the engine. Minimize loads on the
bus . . . turn off everything and/or pull breakers.
With engine at 2000 your ammeter should show a pretty
healthy charge (generator is replacing energy used
to start the engine) and bus voltage should be above
13.8 and preferably on the order of 14.2 volts.
After the ammeter drops to less than 10% of
plus-side full scale, turn on landing lights.
Ammeter should not appreciably change reading
and bus voltage should stay at or above 13.8
volts.
Failure to meet this requirement suggests that
the regulator is not supporting the the loads
due to either poor voltage or current regulation
-OR- the generator's brushes are too worn. In the
later case, you want to get the situation fixed
ASAP.
Failed brushes take commutator surfaces down
with them and can make repairs more expensive.
If generator brushes were changed out
BEFORE failure based on some time-line, your
overall cost of ownership for generators can be
MUCH less than allowing brushes to go to failure.
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-Battery/SD8 problem |
While investigating my SD8/Battery problem it allowed me the opportunity to
double check my wiring scheme per Z25 and the location of the components a
nd wire runs. D- to near F....especially once the fire wall to sub panel to
skin is riveted on the RV7A.
In short it would have been real difficult to work on these electrical conn
ections and components even with the access panel I put in this skin.
So with that said=2C back to the drawing board to rewire and relocate some
of the components to make the install better.........There are some silver
linings in this problem storm!!!!!
Bob=2C was also correct in my wiring of the AUX ALT switch to the SD-8. #3
switch terminal to CB instead of correctly going to #2. I also left off the
PNL ground to the diode rectifier (20-50KuFd @ 15 to 50V). Still not sure
how I missed this one???????
By the way=2C since I have had the SD-8 disconnected from the battery conta
ctor the battery has maintained it's charge.
Finally my plan is to relocate the S704-1/OVM-14/diode bridge/ with the var
ious components/wires compactly nested to a standoff panel attached to the
port aft upper firewall. Leave the voltage regulator and the diode rectifie
r on the original base that spans the horizontal angle on the fire wall to
the forward side of the sub panel just above the rudder peddle bars. mid fu
selage. I will reverse the direction of the diode rectifier so the wire att
ach points face forward. All in all this will make for a much cleaner insta
ll and easier to work on wiring scheme for Z25 in my aircraft. .......Stay
tuned. If I rebuild one more thing it will be offical .......In building on
e plane I really built two of them!!!!!!!!!!!!!!!!!!!!
Thanks for the help and advice....especially Bob.
Frank @ 1L8 ....RV7A....it may fly some day.........last 901 details!!!!!!!
> Date: Thu=2C 3 Jul 2008 10:46:59 -0500> To: aeroelectric-list@matronics.c
om> From: nuckolls.bob@cox.net> Subject: Re: AeroElectric-List: RE: AeroEle
rt L. Nuckolls=2C III" <nuckolls.bob@cox.net>> > At 01:14 PM 6/27/2008 -060
0=2C you wrote:> > >Hi to All> >> >Last week I asked a question about my Od
yessey 680 battery not being able > >to keep a charge. I am using the Z13/8
with SD8 self excitation. I thought > >I had checked all the suspects and
eliminated them one by one as the > >problem. So as a final step I figured
the battery itself was bad. So I > >recharged and disconnected the ground f
rom the firewall. Let it set and > >wouldn't ya know it. It maintained a ch
arge. So back to the drawing board. > >In checking all the possible problem
s again I noted that that SD8 self > >excitation may be the problem. I have
connected the ground back to the > >battery and the firewall. Will wait to
see if it draws down again.> >> >With that said who is using the SD8 with
self excitation as pre Z25. And > >if so any problems.......> >> >Being tec
h. challenged in this stuff I am not quite sure how to check the > >SD8 wir
ing to find the problem if in fact it is. If the battery discharges > >with
the SD8 as configured then maybe it is the problem. I will recharge > >the
battery and disconnect the SD8 from the battery contactor and see what > >
happens. If the battery maintains charge with the SD8 removed I will know >
>that the SD8 is the problem if not....................????????> > If you'
ve followed the wiring suggested in> > http://www.aeroelectric.com/PPS/Adob
e_Architecture_Pdfs/Z13-8N.pdf> > then the self-excitation circuitry . . .>
> http://aeroelectric.com/Pictures/Schematics/PM_Regulator/Self_Excitation
_Experiment.jpg> > http://aeroelectric.com/Pictures/Alternators/SD8_SelfExc
ite_A.jpg> > http://aeroelectric.com/Pictures/Alternators/SD8_SelfExcite_B.
jpg> > . . . is disconnected when the airplane is parked.> > If your SD-8 c
ontrol relay is connected UPSTREAM> of the filter capacitor=2C then the res
istor across the> capacitor WOULD produce a constant load on the battery.>
> Did you make a determination of magnitude on discharge> current? The resi
stor cited would load the battery to> the tune of 12 milliamperes. Assuming
an 17 a.h.> battery=2C this load would take about 750 hours to> use about
========================> >
>
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: Strobe shielded wire splice |
At 11:21 PM 7/4/2008 -0500, you wrote:
><say.ahh1@verizon.net>
>
>My RV-9A wing strobe wires will be a little short in reaching the fuselage
>mounted power supply. I could buy some longer wires, but I have a spare
>set that can be used to splice on some extra length. Any problem splicing
>as per this:
>http://www.aeroelectric.com/articles/Shielded_Wire_Splicing/S_Wire_Splice.html
That would be just fine. The shielded wires that come with most strobe
kits is the Beldfoil with drain wire . . .
http://www.aeroelectric.com/Mfgr_Data/Wire/Belden/8770_Beldfoil_Shielded_Trio.pdf
. . . as opposed to the overall braided cover but
principals are the same. The very short exposure
of adjacent wires in bundle to the "uglies" that live
in the strobe cables is insignificant. It's a common misconception
that a break in the shielding of an antagonist wire is an
opening that allows ALL that is bad within to escape. Not
so. The splicing technique described will function as advertised.
>In case I want to have a wing to fuselage strobe wire connector (I know
>the pros and cons), is there an issue about having the Molex connector
>unshielded? In this case, do folks just run solder the shield ends to once
>of the Molex pins?
Normally, shields are taken through a connector on its
own pin as if it were one of the signal carrying conductors.
Pigtail the shields on either side as described in . . .
http://aeroelectric.com/articles/pigtail/pigtail.html
. . . or in the case of the Beldfoil shields, trim the
shielding foil back to the outer jacket and take the
drain wire through the connector as a separate
conductor.
Bob . . .
----------------------------------------)
( . . . a long habit of not thinking )
( a thing wrong, gives it a superficial )
( appearance of being right . . . )
( )
( -Thomas Paine 1776- )
----------------------------------------
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: | power to handheld gps |
hi all,
i want to mount a plug in into my panel for my handheld gps for 12 volt
power. the gps unit would have a 6 inch pigtail left on it to connect to
the plug in. the gps unit will be mounted but still portable. i am
looking something along the lines of a d sub connector but i only need 2
pins, not 9. the smaller the better. i dont want to put the power cord
of the gps thru the panel because the plug in on the back of the gps
looks like bent pins waiting to happen[i want to leave this plug in
attached to the gps unit].
any suggestions would be appreciated.
bob noffs
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: Bridge diodes VS Schottky |
At 09:53 AM 7/5/2008 -0700, you wrote:
>
>I have seen the debate over the low cost Bridge diode VS the Schottky
>diode come up year after year mostly based on the opinion the bridge is
>good enough and easy to get at RS, mount it anywhere to a bracket
>(implying no significant heat sinking is needed etc). Perhaps I have
>missed it but I do not recall a single technical analysis based on
>engineering evaluation of the specified values based on data sheet
>analysis. Further many suggested uses show two diodes of the bridge in
>parallel. This does not increase the current capacity much as diodes do
>not share current well as one will get hotter and hog the current making
>it hotter etc. Its not clear the reliability is increased as internal
>bridge package connections are not known and vary with the manufacturer.
>Diodes typically fail short and only open if the current is high enough.
>
>Over this nearly 15 year period the bridge diode has been used for several
>applications but most often as I recall is to isolate the emergency (now
>endurance) bus. This bus originally was apparently designed for 2-3 amp
>max. Today many users have a 8 amp load and still use the bridge diode.
>Notably, Eric has promoted the Schottky diode and been shouted down
. . . you've mis-interpreted the discussions. Nobody from
this side of the list-server has ever shouted at anyone . . .
> over and over with no real data to support the opposing positions.
Exactly! The superior performance of the Schottky diode is
not and never has been in question. What WAS in question was
the offering of an alternative to the lowly silicon rectifier
with equal dearth of data to support the promotion of a part
that was more expensive, had to be acquired from a source perhaps
less convenient than Radio Shack and was no easier to install.
Further, this had nothing to do with me, B&C or even Radio Shack
making a buck on the occasional sale of a bridge rectifier. When
the "new and improved" product came to the market, the lack of
data left many folks who already had bridge rectifiers installed
wondering if they'd done a bad thing . . . "Gee, should I rip it
out and put in the really good part?"
While I was consistently suggesting that meeting design goals
with a minimum cost of ownership opens a lot of alternative
doors, the proponents of
>15 years ago I selected the Schottky diode as I had dual batteries and
>wanted to use the diode to provide battery isolation and and also single
>source battery charging. The bridge diode simply would not charge the
>battery to full charge while the Schottky did.
One COULD quibble over what the design goals for achieving "full charge".
Clearly, if you have say 24 hours to wait, 13.8 volts will top off a room
temperature lead-acid battery of any pedigree . . . but as a practical
matter, we'd like to get our ship's battery topped off in the first few
minutes of flight and under variety of temperatures. So we force the
battery to suffer acceptable indignities by boosting the bus voltage
under a perfectly rational idea that the goal is to store energy to
meet design goals . . . not squeeze a few extra flights of service
life out of a battery that is difficult to qualify. But yes, if one
chooses to go the diode isolator route for batteries, the Schottky
device offers a demonstrable advantage over silicon junction
rectifiers. I'll point out further that your design goals for
incorporation of a diode were decidedly different than for normal
feed path steering of the e-bus power.
>Data sheet worst case design analysis can be difficult as the typical data
>sheet is intended to make the part look good and typically shows
>information that is at least misleading with the real worst case
>information hard to find.
Shucks . . . you're not suppose to say that in front of
everybody!!! :-) Yes, unless one makes a living at the
interpretation, validation and application of manufacture's
data sheets, it CAN be a daunting and sometimes frustrating
task. We had some pretty good discussions here on the List
a few years back when curves for guaranteed performance
on breakers was misinterpreted as LIMITS to performance.
As a designer with decades of experience of working with
those breakers (knowing about their demonstrated capabilities
outside the guaranteed performance), I was able to
exploit their capabilities in ways that got some folks
really excited . . . for no good reason based on the physics.
>Today the bridge diode part is very old and many manufacturers have
>shortened the data sheet to the extent its not possible to determine the
>worst case parameters. Some data sheets are simply scans from another
>Manufacturer's old data sheet.
>
>I evaluated both types of bridge diodes available today. The GBPC series
>and the older MB series. The MB series has a metal base with metal sides.
>The GBPC series has a smaller metal base and no metal sides. The
>electrical parameters are essentially the same and both are available with
>faston type connections. Remember some data is shown as nominal, not max
>ratings. Remember that bridge diodes are rated as a full wave rectifier
>which means some data must be converted to DC.
>
>Here are the worst case values for Vf at 1 and 10 amps and at 25C and
>150C. This is for a single diode.
>Note the data is basically a straight line from 1A to 10A which allows for
>simple intermediate current calculations
>
>Vf 1A 25C Vf 10A 25C Vf 1A 150C Vf 10A 150C
>0.83V 1.05V 0.63V 0.85V
>
>Power dissipation as follows at 25C
>1A 2.5A 5.0A 7.5A 10A
>0.83W 2.23W 4.7W 7.5W 10.5W
>
>To figure case temp consider 1.9deg/c per watt from die temp. For example
>case temp max at 10 amps is 10.5 times 1.9 approx 20C or 130C max. If
>ambient temp is 40C near the mount then the heat sink must dissipate 10.5W
>with a temp drop of 90C across the heatsink.
Exactly. And that works out to a heatsink with about 8.5C/Watt
thermal resistance which is NOT a very big heatsink.
>The above does not allow for the case to heat sink temp resistance
>depending on heat sinking surface material IE grease. The above does not
>allow for service derating. The aerospace industry recommends no more that
>100C case temp which increases the heat sink size for a temp gradient of
>from 90 to 60 c max
>
>Its clear that above a current of 1A, some (or a lot of) heat sink is
>required. I do not recall any specified requirement other than to simply
>mount it on a small bracket.
Hmmmm . . . don't know about "small bracket" . . . . I've
suggested any flat surface of aircraft sheet metal which is
generally a firewall, shelf or perhaps skin.
> No mention of the surface smoothness of the bracket nor what it's
> thickness or size is. No mention of the required heat sink grease etc. At
> 10 amps a couple of manufacturers require a heat sink size of 6" x 2.2" x
> 2.2" finned plate plus additional requirements to keep the junction
> temperature below 150C (the max rated temperature). To calculate case
> temp use 1.9deg/c per watt from die temp. For example case temp max at 10
> amps is 10.5 times 1.9 approx 20C or 130C max.
>
>Thus its not wrong to state a 10W heat sink is needed but there is a
>string of calculations required to determine the physical heat sink needed.
You stole my thunder sir! I'm preparing a Shop Notes document
that speaks to power and temperature ratings as is applies to
combinations of devices that dissipate energy and products that
are intended to keep those devices from self-destruction from
self-heating. The REAL specification of heat-sink performance
is stated in temperature rise per watt of heat dissipated.
A heat sink of ANY physical size can dissipate 10 watts. Depending
on how big it is, the temperature may go up 2 degrees or 200
degrees. Further, the slightest breeze will have a profound
effect on heat sink performance. Needless to say, there are
few places in our airplanes that are NOT breezy! But the
short story is that suggesting a builder use a heatsink rated
for 10W is not the whole story.
>Now lets look at the current Schottky diode available. In this case the
>diode selected is the DSS2x121-0045B made by IXYS. Two independent 120 amp
>diodes. In the case of Schottky diodes the higher the current rating the
>lower the Vf at low currents as seen in this application. IR used to be
>the low Vf leader but this IXYS is slightly better and lower cost and easy
>to get. ( www.findchips.com )
>
>Vf 1A 25C Vf 10A 25C Vf 1A 150C Vf 10A 150C
>0.29V 0.34A 0.15V 0.24V
>
>Power dissipation as follows at 25C
>1A 2.5A 5.0A 7.5A 10A
>0.29W 0.78W 1.58W 2.48W 3.40W
>
>Clearly a far better part to use. Just as easy to mount and wiring support
>is easier.
>
>If battery charging is intended thru this diode, not only is a full charge
>possible but there is no current overload as is likely to occur with the
>bridge if the battery being charged is significantly discharged and where
>the charging current can be well above 10 amps.
Absolutely. I've seen several battery powered systems
fall short of design goals because the folks forgot that
a battery intended to supply perhaps 5-10 amps in use
can DRAW several times that current while being recharged.
>Why worst case analysis? Because its not possible to know the actual
>parameters of the parts used by the various builders. Manufacturers do
>produce worst case parts and are often sold to resellers like RS with the
>higher quality parts are distributed to the industry market.
Do you have some reference to support this? In years gone
by there have been countless suppliers of parts to the
hobby electronics market that indeed offered "floor sweepings"
in bags. I discovered the photo-sensitivity of junction
transistors when devices I'd purchased as "2N3904" transistors
had a terrible amount of hum that disappeared when the florescent
light over the bench was shut off. The plastic was not quite
as opaque as one might wish for a "real" 2N3904.
However, the real cost of good parts has become so small
compared to all the other costs of distributing and marketing
that there is little value in spending labor to create a
separate product stream for sub-standard parts. Further, the
idea that sub-standard parts is even exist is suspect. Modern
semiconductor houses have so many process checks in place along
the assembly line that sub-standard parts are going to be
pitched long before they're put into final packages and marked
with the manufacturer's brand and part-number. My sense
is that the days of Poly-Pak's and Olson Radio Warehouse's
"bargain bags" are long behind us.
>Why not just do a test and use that as proof of the design? A test only
>proves the worst case analysis is valid and is only good for the exact
>parts used and does not indicate what results will be with similar parts
>from the identical manufacturers production batch. Proper engineering
>design always starts with a worst case analysis of the design followed
>with a test to show the results are inside the worst case analysis. Only
>then is the design released for production or in this case released for
>the general builder to build and be sure it will perform to expectations.
>
>Availability? Its strange to me that the availability of the bridge at RS
>is promoted but many of the recommended switches are only available thru B&C.
I "promote" selection of parts adequate to the task
with some notion of $time$ expended to acquire those
parts. A $1.50 switch from the store is no bargain
if you hop in the car and drive 7 miles to buy it . . .
while saving your $5.00 switch in the drawer for a
"more demanding" task.
The part numbers cited in my drawings are artifacts
of the catalog numbers I assigned to those parts
while I was supplying them to the OBAM aviation
industry. When I decided to move that activity to
B&C . . . many of those numbers were retained and
are most often cited in my work much of which dates
back 20 years.
However, I've never attempted to hide the fact that
there are MANY alternate sources for equal or even
better parts if the builder has the willingness, $time$
and skill to acquire them. I had a builder drop by the
booth at OSH several times during the week with small
scraps of wire in hand. He was scouring the Fly-Market
for "wire bargains". I don't know how his project
was ultimately wired. Given that I saw a lot of
him during the week suggests he would have been
WAAAYyyy better off buying new, M22759 wire at Aircraft
Spruce prices than to spend the $time$ necessary to
insure that his "bargains" were appropriate to the
task.
I submit that the parts I've oft suggested are an
aid for those who believe their time is better spent
bucking in a few more perfect rivets and bending
one more ideal stiffener. This is opposed to combing
the 'Net' in some attempt to save a few dollars on parts
selection when their personal expertise is orthodontics
or managing a grocery store.
>Modern electronics use switching supplies which use more or less power not
>amps as input baded on input voltage. Thus the higher the Vf the more amps
>required and thus shorter battery life as the battery life in amp hours
>goes down as the amp drain goes up.
>
>The cost of the superior Schottky part is very small compared to the cost
>of the electrical wiring/terminal etc system and insignificant to the
>total cost of the aircraft.
This discussion has never been about cost of parts.
It has always been about selection of parts adequate
to the design goals with the minimum total expenditure
of $time$ to select, acquire, install, and maintain
those parts.
In situations where the design goals were changed
(like battery isolation, and boss-hog e-bus loads)
then some re-evaluation of suggested parts is in
order. I've take data in the chamber on a variety
of diodes, done the stress tests to ascertain thermal
resistance of the parts and done all the photographs.
Measured the practical thermal resistance of small
sheets of aluminum.
Just need some time to put it all together into a
set of Shop Notes that speaks to the BIG PICTURE
when it comes to making any heat generating part
perform to design goals. Just understand that the
simple-ideas that support these discussions have
almost nothing to do with the additional cost
of a Schottky versus silicon-junction rectifiers.
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-Battery/SD8 problem |
At 01:21 PM 7/5/2008 -0600, you wrote:
<snip>
>Finally my plan is to relocate the S704-1/OVM-14/diode bridge/ with the
>various components/wires compactly nested to a standoff panel attached to
>the port aft upper firewall. Leave the voltage regulator and the diode
>rectifier on the original base that spans the horizontal angle on the fire
>wall to the forward side of the sub panel just above the rudder peddle
>bars. mid fuselage. I will reverse the direction of the diode rectifier so
>the wire attach points face forward. All in all this will make for a much
>cleaner install and easier to work on wiring scheme for Z25 in my
>aircraft. .......Stay tuned. If I rebuild one more thing it will be
>offical .......In building one plane I really built two of
>them!!!!!!!!!!!!!!!!!!!!
Frank, that happens with EVERY project I've ever participated in.
In fact, by the time one device makes it to a customer airplane on the
production line, I've built as many as a dozen variants. The
brass-board and simulators is where we start. A chop-n-hack package
will go into the temperature chamber for thermal evaluations.
A flight-worthy package will go through engineering shake-n-bake
before another flight-worthy package gets bolted to the first
flight airplane. During flight tests, more modifications to improve
performance may be identified. The test articles used to qualify
the product (usually three more) are trash when we're finished
with them.
So if you can get your bird into the air with only a second
or third iteration on the design, you're waaaayyyy ahead of
me!
>
>Thanks for the help and advice....especially Bob.
No problem my friend. It's what we do.
Bob . . .
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: power to handheld gps |
bob noffs wrote:
> hi all,
> i want to mount a plug in into my panel for my handheld gps for 12 volt power.
the gps unit would have a 6 inch pigtail left on it to connect to the plug
in. the gps unit will be mounted but still portable. i am looking something along
the lines of a d sub connector but i only need 2 pins, not 9. the smaller
the better. i dont want to put the power cord of the gps thru the panel because
the plug in on the back of the gps looks like bent pins waiting to happen[i
want to leave this plug in attached to the gps unit].
> any suggestions would be appreciated.
>
bob noffs
>
Put a layer of tape on the back and bottom of the GPS.
Plug in the power cord.
Mix up some epoxy and stir in half micro and half flox, kneeded to the
consistency of dough.
Pack the epoxy dough around the connector and form a cradle for the GPS.
Cast in some mounting hardware, or an aluminum tube for a strap. Even a
straw from McDonald's would work (not that I ever eat there, mind you).
Shape with a file if the results isn't pleasing to the eye.
--
http://www.ronpaultimeline.com
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: AeroElectric-List Digest: 29 Msgs - 07/03/08 |
At 08:23 PM 7/4/2008 -0400, you wrote:
>Bob, you made a point yesterday, that I had not thought through. I do not
>have an electrical/charging/output performance card in my flight test
>deck. I do of course, have operational performance checks planned on all
>my avionics, but I don't have a specifice plan for checking out the
>performance of my "as designed" electrical system (not flying yet).
>
>Wonder if you or anyone else can recommend specific flight test procedures
>to accomplish a reasonable system test in a sequenced and logical
>form? My system is all electric with a Plane Power 60 amp, internally
>regulated main alternator, an SD-8 backup, all wired a la Z-13/8 with
>"auto excitement" of the SD-8. I have an EDM-900C engine monitor with
>voltage and amperage readings and the B&C BC207-1 over/under voltage
>sensor (not yet wired up). Any recommendations as to wiring changes that
>might facilitate the above would also be appreciated.
You wouldn't want to know what we go through to prove
the concept for a fresh system design on a TC aircraft.
Those-who-know-more-about-airplanes-than-we-do can
ask a LOT of questions for which only an approved test
plan with lots of data gathering followed by a test
report weighing at least 5 pounds is indicated.
The key questions for your as-installed equipment
are pretty simple.
What RPM on the engine will produce rated output
of the main alternator? I would test this on the
ground with a snap-on DC ammeter around or (shunt in
series with) the alternator b-lead. Perhaps you have
altenrator load-meters already installed that can
serve this purpose? Hook a battery tester . . .
http://aeroelectric.com/Pictures/Battery_Testers/HF_Carbon_Pile_Load.jpg
across the ship's battery. Use the battery tester
to add load until the alternator produces full
output at say 2200 rpm and then reduce rpm until
alternator output begins to fall off. The reading
on the battery tester is irrelevant. You're only
using it as a load bank to ADD to ship's loads
until the alternator output is maxed out. The
bus voltage should not fall below 14.0 volts
until the rpms are too low to support full output
from the alternator.
The SD-8 has to be tested in cruising flight.
Turn the SD-8 on, turn main alternator OFF,
adjust ships loads until to show that you can
maintain 13.0 volts or more at 10.0 amps
of SD-8 output.
Now, if I were putting this system into a TC
aircraft, I would have to demonstrated cooling
under hot-day, Vx climb conditions. Of course,
heating is a slowly changing condition that
is almost never demonstrable in a hot-day
environment. So we do a lot of testing. Average
a lot of data and then correct that data to
expected hot-day performance 'cause today just
wasn't that hot.
When you're testing 600A generators, it can get
pretty warm in the cabin when those load banks
are tossing out the heat produced by those big
honk'n generators!
>
>BTW , I don't see over/under voltage detection or indications on
>schematics for the SD-8 typically. Is the expectation that the BC207-1
>will provide adequate early warning of main alternator failure and that
>then battery and SD-8 power will be adequate for any reasonable endurance
>bus/fuel remaining scenario---such that further warning that the SD-8 has
>failed too is considered an unnecessary complication?
Some form of low voltage indication on the e-bus
is a good thing. Not low voltage warning, only a
readout. You won't be concerned about e-bus voltage
EXCEPT after a main alternator failure. Further, you
MIGHT just choose to load the SD-8 down to the level
where a low voltage warning might come on (13.0 volts)
so just a meter is fine.
>
>
>If the SD-8 failed after the main alternator had already failed on my
>aircraft, my battery (Odyssey 925) is the final arrow in the quiver.
Yes . . . but dual failures of any kind in the electrical
system on any one tank of fuel is exceedingly rare. So rare
in fact that we don't even consider probability rates for
loss of both alternators in Part 23 airplanes and only
in Part 25 aircraft used in flight-for-hire in IMC. Even
then, failure rates for dual loss are always comfortably
small.
>The 925 is a monster though and personally, I would already be headed for
>good weather/divert/home base/whatever if the SD-8 let go. I don't think
>I need another warning system, but thought I'd ask the group. What's the
>consensus thinking on this one?
Exactly. Z-13/8 with a properly maintained SVLA battery
has better overall system reliability than that which flies
in any Part 23 aircraft and most Part 25 aircraft. LV warning
for the main bus, voltmeter for the e-bus and good preventative
maintenance for the battery will let you launch with zero
concerns for electrical energy sufficient to the task.
Keep in mind that there are thousands of OBAM aircraft
flying where the SD-8 is the only engine driven power
source. I would not attempt to dissuade you from unplanned
termination of flight should you have only the SD-8 working.
But for my airplane, I'd KNOW that I can do it battery only.
The three-layer energy development system embodied in Z-13/8
offers an opportunity to enjoy 100.0% reliability of the
electrical system for the purpose of completing the intended
mission . . . IN SPITE OF INEVITABLE FAILURES. That's what
failure tolerance is all about.
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: Bridge diodes VS Schottky |
WOW Bob, Its my turn to point out you have misunderstood my points in a few
cases
For example
sniped to shorten the post
> . . . you've mis-interpreted the discussions. Nobody from
> this side of the list-server has ever shouted at anyone . . .
Not you, but others have suggested there is no point to Eric's
recommendation, Shout is my opinion perhaps not yours. Shouting down is in
my opinion a series of comments indicating that Eric's suggestion was not
needed along with suggested Bridge diode Vf well below worst case values
which may not be repeatable by other builders. Its important to go much
farther in parts selection, even suggestions years old. Bridge diodes were
useful back in the days of single battery and e bus loads of 2-3 amps. Its
important to update and or put real limits on the use of Bridge diodes. Most
thought the Bridge diode was a 25 amp diode until Eric pointed out it was
really a 12.5 amp diode. Further showing two in parallel is understood to
non electrical engineers as splitting the current which is better? right :-)
> Exactly! The superior performance of the Schottky diode is
> not and never has been in question. What WAS in question was
> the offering of an alternative to the lowly silicon rectifier
> with equal dearth of data to support the promotion of a part
> that was more expensive, had to be acquired from a source perhaps
> less convenient than Radio Shack and was no easier to install.
> Further, this had nothing to do with me, B&C or even Radio Shack
> making a buck on the occasional sale of a bridge rectifier. When
> the "new and improved" product came to the market, the lack of
> data left many folks who already had bridge rectifiers installed
> wondering if they'd done a bad thing . . . "Gee, should I rip it
> out and put in the really good part?"
The Schottky diode has decades of industry wide documentation in support of
its superior performance and the diode bridge is simply not used in any
recent design application for DC circuit controlling as we use it in your
excellent application suggestions. Further I never suggested or implied
anyone was making a buck on any part ever! Where did that comment come
from??? As for should I rip it out? Without more information such as heat
loss, battery charging etc the answer is maybe it depends. PS the new and
improved product existed in the 1980's Also my local RS store does not carry
these parts.
Today's circuits include battery charging as several posters have mentioned,
another had stated his bus load was 8 amps and the bridge recommended
attached to a simple bracket is not a good solution at higher currents. As
many have clearly interpreted the recommended bridge it will not meet these
design goals.
>>15 years ago I selected the Schottky diode as I had dual batteries and
>>wanted to use the diode to provide battery isolation and and also single
>>source battery charging. The bridge diode simply would not charge the
>>battery to full charge while the Schottky did.
> One COULD quibble over what the design goals for achieving "full
> charge".
> Clearly, if you have say 24 hours to wait, 13.8 volts will top off a
> room
> temperature lead-acid battery of any pedigree . . . but as a practical
> matter, we'd like to get our ship's battery topped off in the first few
> minutes of flight and under variety of temperatures. So we force the
> battery to suffer acceptable indignities by boosting the bus voltage
> under a perfectly rational idea that the goal is to store energy to
> meet design goals . . . not squeeze a few extra flights of service
> life out of a battery that is difficult to qualify. But yes, if one
> chooses to go the diode isolator route for batteries, the Schottky
> device offers a demonstrable advantage over silicon junction
> rectifiers. I'll point out further that your design goals for
> incorporation of a diode were decidedly different than for normal
> feed path steering of the e-bus power.
True but as above many current list builders use the diode to charge the
battery. Setting the alternator to 14.3V will not, in a reasonable time,
charge a backup battery as the bridge Vf can reach nearly 1 volt leaving the
battery terminal voltage around 13.3 V which is below a reasonable voltage
for charging. Further if the battery is really discharged the charging
current is likely to exceed the allowable bridge maximum current.
I recall the delta V to charge a battery is 0.2V per cell at room temp. This
amounts to 1.2V for a 12V battery. Thus 13.3V charging V will eventually
charge the battery to 12.1V hardly adequate!
>>Data sheet worst case design analysis can be difficult as the typical data
>>sheet is intended to make the part look good and typically shows
>>information that is at least misleading with the real worst case
>>information hard to find.
>
> Shucks . . . you're not suppose to say that in front of
> everybody!!! :-) Yes, unless one makes a living at the
> interpretation, validation and application of manufacture's
> data sheets, it CAN be a daunting and sometimes frustrating
> task. We had some pretty good discussions here on the List
> a few years back when curves for guaranteed performance
> on breakers was misinterpreted as LIMITS to performance.
Only by exceeding the specified data sheet parameters. The data sheet
specifications are safe guaranteed LIMITS used in proper and safe design.
Yes in some cases the actual part as being made at the time can exceed these
limits safely but as that discussion went, there are many different mfgrs
making equivalent CB and now many are made overseas to questionable
standards. The extrapolation of data sheet parameters or using data sheet
parameters from one mfgr and assume other mfgrs parts will meet these
parameters is not permitted in aerospace industry. Further there is no
reason to exceed data sheet parameters in a good design. As I recall your
example was assumed by many to simply justify an omission of a simple
current limiting resistor that would have safely limited the current in the
CB and not detracted in anyway the overall CB function but that is a
distraction to today's topic.
Yes data sheets tend to mislead to get a slaes advantage. One part is stated
to have a max current of 200 amps while there is a not to a n app sheey that
limits that same part to 75 amos due to the die to case bond wires. Most
specify parameters at 25C where in any real world the temp is much higher
with any reasonable heat sink.
> As a designer with decades of experience of working with
> those breakers (knowing about their demonstrated capabilities
> outside the guaranteed performance), I was able to
> exploit their capabilities in ways that got some folks
> really excited . . . for no good reason based on the physics.
I understand the widespread of use of parts outside their design because one
can usually get away with it. This is a poor excuse for proper design and
again simply not allowed in aerospace where it must work with any part from
any approved mfgr. In the above example there is no need to extrapolate as
the addition of a simple series resistor will limit the current to
specification values and not detract with the designs purpose or function.
>>Today the bridge diode part is very old and many manufacturers have
>>shortened the data sheet to the extent its not possible to determine the
>>worst case parameters. Some data sheets are simply scans from another
>>Manufacturer's old data sheet.
>>
>>I evaluated both types of bridge diodes available today. The GBPC series
>>and the older MB series. The MB series has a metal base with metal sides.
>>The GBPC series has a smaller metal base and no metal sides. The
>>electrical parameters are essentially the same and both are available with
>>faston type connections. Remember some data is shown as nominal, not max
>>ratings. Remember that bridge diodes are rated as a full wave rectifier
>>which means some data must be converted to DC.
>>
>>Here are the worst case values for Vf at 1 and 10 amps and at 25C and
>>150C. This is for a single diode.
>>Note the data is basically a straight line from 1A to 10A which allows for
>>simple intermediate current calculations
>>
>>Vf 1A 25C Vf 10A 25C Vf 1A 150C Vf 10A 150C
>>0.83V 1.05V 0.63V 0.85V
>>
>>Power dissipation as follows at 25C
>>1A 2.5A 5.0A 7.5A 10A
>>0.83W 2.23W 4.7W 7.5W 10.5W
>>
>>To figure case temp consider 1.9deg/c per watt from die temp. For example
>>case temp max at 10 amps is 10.5 times 1.9 approx 20C or 130C max. If
>>ambient temp is 40C near the mount then the heat sink must dissipate 10.5W
>>with a temp drop of 90C across the heatsink.
> Exactly. And that works out to a heatsink with about 8.5C/Watt
> thermal resistance which is NOT a very big heatsink.
Not sure I agree and clearly at least a couple of bridge mfgrs disagree as
they specify (to me) a large heatsink as noted elsewhere. However each case
of mounting location etc is different and one needs to make a worst case
heat sink or do the design specific analysis. Thermal resistance is
different mounted behind the engine vs. mounted behind the panel vs. mounted
behind the baggage. Also today some aircr4qaft fly well above 20,000 where
convection transfer is much lower.
>>The above does not allow for the case to heat sink temp resistance
>>depending on heat sinking surface material IE grease. The above does not
>>allow for service derating. The aerospace industry recommends no more that
>>100C case temp which increases the heat sink size for a temp gradient of
>>from 90 to 60 c max
>>
>>Its clear that above a current of 1A, some (or a lot of) heat sink is
>>required. I do not recall any specified requirement other than to simply
>>mount it on a small bracket.
> Hmmmm . . . don't know about "small bracket" . . . . I've
> suggested any flat surface of aircraft sheet metal which is
> generally a firewall, shelf or perhaps skin.
Most firewalls are or should be SS and that is a very poor heatsink The skin
is better but no mention of the need for heat sink grease used on the
bracket to bridge and bracket to skin etc. Many skins are also very thin
relative to normal heat sinks leading to high thermal resistance.
>> No mention of the surface smoothness of the bracket nor what it's
>> thickness or size is. No mention of the required heat sink grease etc. At
>> 10 amps a couple of manufacturers require a heat sink size of 6" x 2.2" x
>> 2.2" finned plate plus additional requirements to keep the junction
>> temperature below 150C (the max rated temperature). To calculate case
>> temp use 1.9deg/c per watt from die temp. For example case temp max at 10
>> amps is 10.5 times 1.9 approx 20C or 130C max.
>>
>>Thus its not wrong to state a 10W heat sink is needed but there is a
>>string of calculations required to determine the physical heat sink
>>needed.
>
> But the
> short story is that suggesting a builder use a heatsink rated
> for 10W is not the whole story.
I agree but my point is that is the start of the analysis along with the
delta T across the heat sink
>>Now lets look at the current Schottky diode available. In this case the
>>diode selected is the DSS2x121-0045B made by IXYS. Two independent 120 amp
>>diodes. In the case of Schottky diodes the higher the current rating the
>>lower the Vf at low currents as seen in this application. IR used to be
>>the low Vf leader but this IXYS is slightly better and lower cost and easy
>>to get. ( www.findchips.com )
>>
>>Vf 1A 25C Vf 10A 25C Vf 1A 150C Vf 10A 150C
>>0.29V 0.34A 0.15V 0.24V
>>
>>Power dissipation as follows at 25C
>>1A 2.5A 5.0A 7.5A 10A
>>0.29W 0.78W 1.58W 2.48W 3.40W
>>
>>Clearly a far better part to use. Just as easy to mount and wiring support
>>is easier.
>>
>>If battery charging is intended thru this diode, not only is a full charge
>>possible but there is no current overload as is likely to occur with the
>>bridge if the battery being charged is significantly discharged and where
>>the charging current can be well above 10 amps.
>
> Absolutely. I've seen several battery powered systems
> fall short of design goals because the folks forgot that
> a battery intended to supply perhaps 5-10 amps in use
> can DRAW several times that current while being recharged.
Yes but posts stating the use of the bridge as a trickle charger were not
challenged as not safe with the bridge diode.
>>Why worst case analysis? Because its not possible to know the actual
>>parameters of the parts used by the various builders. Manufacturers do
>>produce worst case parts and are often sold to resellers like RS with the
>>higher quality parts are distributed to the industry market.
>
> Do you have some reference to support this? In years gone
> by there have been countless suppliers of parts to the
> hobby electronics market that indeed offered "floor sweepings"
> in bags. I discovered the photo-sensitivity of junction
> transistors when devices I'd purchased as "2N3904" transistors
> had a terrible amount of hum that disappeared when the florescent
> light over the bench was shut off. The plastic was not quite
> as opaque as one might wish for a "real" 2N3904.
I am not talking about out of spec parts. Worst case but meeting
specification. Today parts are generally automatically tested and binned
based on performance and selected and sold at a higher price to
manufacturers.
> However, the real cost of good parts has become so small
> compared to all the other costs of distributing and marketing
> that there is little value in spending labor to create a
> separate product stream for sub-standard parts. Further, the
> idea that sub-standard parts is even exist is suspect. Modern
> semiconductor houses have so many process checks in place along
> the assembly line that sub-standard parts are going to be
> pitched long before they're put into final packages and marked
> with the manufacturer's brand and part-number.
I never stated sub standard parts I did say parts that were worst case which
in this discussion had the max Vf for example. See comment just above your
reply There is no special labor today, its 99% automated today. A given
production run generally has a very tight distribution. Thus one run may
produce 99% parts that are within 1% of max and another run may produce most
parts that are 25% better than max. The better parts are often selected for
higher performance requirements.
There is a HUGE problem with remarked and parts made with different Die
current plaguing the industry. Its bit even major manufacturers. This is a
wide spread discussion in industry pubs this year and its apparently getting
to be a larger issue as more and more parts are sub standard or not even the
correct die inside.
>>Why not just do a test and use that as proof of the design? A test only
>>proves the worst case analysis is valid and is only good for the exact
>>parts used and does not indicate what results will be with similar parts
>>from the identical manufacturers production batch. Proper engineering
>>design always starts with a worst case analysis of the design followed
>>with a test to show the results are inside the worst case analysis. Only
>>then is the design released for production or in this case released for
>>the general builder to build and be sure it will perform to expectations.
>>
>>Availability? Its strange to me that the availability of the bridge at RS
>>is promoted but many of the recommended switches are only available thru
>>B&C.
>
> I "promote" selection of parts adequate to the task
> with some notion of $time$ expended to acquire those
> parts. A $1.50 switch from the store is no bargain
> if you hop in the car and drive 7 miles to buy it . . .
> while saving your $5.00 switch in the drawer for a
> "more demanding" task.
My point is there are special contact/switching configurations not normally
available anywhere to the normal builder other than you then, and now B&C. I
would suggest you consider electrical redesign and restrict the switches to
the commonly available contact configurations. Not all builders want to use
large bat handle toggles but other style switches are not available in the
special contact configurations..
> The part numbers cited in my drawings are artifacts
> of the catalog numbers I assigned to those parts
> while I was supplying them to the OBAM aviation
> industry. When I decided to move that activity to
> B&C . . . many of those numbers were retained and
> are most often cited in my work much of which dates
> back 20 years.
>
> However, I've never attempted to hide the fact that
> there are MANY alternate sources for equal or even
> better parts if the builder has the willingness,
> I submit that the parts I've oft suggested are an
> aid for those who believe their time is better spent
> bucking in a few more perfect rivets and bending
> one more ideal stiffener. This is opposed to combing
> the 'Net' in some attempt to save a few dollars on parts
> selection when their personal expertise is orthodontics
> or managing a grocery store.
My point was specific towards some of you specified switches that simply are
not stocked or made in other mechanical configurations. I know as I have
looked for myself and others who wanted to use that specification switching
function but not the large bat handle actuator.
>>Modern electronics use switching supplies which use more or less power not
>>amps as input baded on input voltage. Thus the higher the Vf the more amps
>>required and thus shorter battery life as the battery life in amp hours
>>goes down as the amp drain goes up.
>>
>>The cost of the superior Schottky part is very small compared to the cost
>>of the electrical wiring/terminal etc system and insignificant to the
>>total cost of the aircraft.
>
> This discussion has never been about cost of parts.
> It has always been about selection of parts adequate
> to the design goals with the minimum total expenditure
> of $time$ to select, acquire, install, and maintain
> those parts.
I only pointed out there is no significant cost difference in the big
picture and there is significant performance improvements to many builders.
Today the cost of part acquation using the internet is lower cost that a
auto trip to RS. Mouser is a good source for example and they will special
order if the mfgr makes it. Shipping now is lower cost than auto gas and in
many cases the internet delivers a no sales tax order.
> In situations where the design goals were changed
> (like battery isolation, and boss-hog e-bus loads)
> then some re-evaluation of suggested parts is in
> order. I've take data in the chamber on a variety
> of diodes, done the stress tests to ascertain thermal
> resistance of the parts and done all the photographs.
> Measured the practical thermal resistance of small
> sheets of aluminum.
The problem with testing parts is the testing does not prove equivalent
parts will perform the same way. Testing in the aerospace industry is only
done as a proof of concept and only after a detailed design. Then the design
analysis NOT the test results are used for production. Repeatable testing is
only a valid statement using the exact same parts over and over.
The design requirements have changed and its been several years since I
noticed battery charging and higher e bus currents on your group or in the
"book". It would have been helpful had there been some original limits
defined in the "Book" regarding the use of bridge diodes as one example
> Just need some time to put it all together into a
> set of Shop Notes that speaks to the BIG PICTURE
> when it comes to making any heat generating part
> perform to design goals. Just understand that the
> simple-ideas that support these discussions have
> almost nothing to do with the additional cost
> of a Schottky versus silicon-junction rectifiers.
I agree and that is why I never have considered bridge diodes since years
ago far better Schottky diodes were available. In my example I illustrated
the best currently available schottky, much lower cost schottky diodes are
also available if one only wants double the performance. Today and 15 years
power Schottky diodes were available that were insulated bolt down.
You are doing an incredible job helping the masses, but you seem to over
react to comments that intend to improve and or update the design. There are
many of us with knowledge of more modern parts that in our opinion a large
improvement to your evolving designs
Paul
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! --
|