2-Batteries

Tony

You would start by insulating the battery well. It should only take
about an inch of foam insulation (R7) to get the heat leak down to a
watt or so but it needs to be carefully made.Startting with a warm
battery, good insulation and a little self heating from the battery
during discharge should get you a long way. Of course the battery
likely won't fit into its mount with all that insulation. After doing
all that adding a heater may make sense but you need to know what you
are doing with a termperature regulation circuit, and good luck finding
somebody who will sign off the work. And yes the expectation for an
insulated battery should be that the cost of running the heater is
well worth it in terms of getting more capacity out of the battery. It
would take two heater plates or nichrome wire on the battery end (or
sides) that parallel the internal plates. It is those two outward most
facing cells that probably conduct most heat into/out of the battery
(in a well insulated battery the terminals and cables may be a
significant leak as as well).

Before looking a the complexity of a heater I'd look at a solar panel
of the glider. They work great cold.

If you need to go to the effort of adding a heater it may also be time
to investigate alteratives to lead acid batteries (which may also need
heaters).

Lead acid batteries have a pretty large thermal mass and the AGM
bateries cores tend to relatively well insulated because there is
little electrolyte or metal contact with the case so in practice they
probably don't get as cold as ambient on typical flights, especially if
like my batteries they sit up on a parcel shelf packed with stuff
around them. I've been curious about the temperature issues and have
thought about drilling thermocouples into the inside of AGM batteries
for some tests but I've never got around to it (just for ground
measuremnts I'd not want to fly with it).

Darryl Ramm

Tony Verhulst wrote:
wrote:
Nice presentation snipped
... One of the curves shows the
rapid voltage drop of a cell freezing as it discharges (some thing to
worry about on that next really long and cold wave flight :-)

If we had properly designed heater elements surrounding the battery,
powered by the battery itself, could we expect more "useful capacity"
from the battery on those cold flights? Useful capacity being defined as
the amount of current delivered to the avionics.

Tony V.

OxAero wrote:

The problem was that the large regulator couldn't handle the start up
load from the transponder. So, my quick fix was to bypass that
regulator with the second switch. I expect to revisit the project some
day and finish it properly.

If you are using a transponder like the Becker ATC 4401, the bypass is
the best setup. The Becker, like many other modern transponders and
radios, has a wide range input voltage regulator (9-32 volts for the
Becker), so regulating it's input voltage has no advantage. As you
discovered, it may even prevent the equipment from working properly, as
it is designed to operate from a battery, not a regulated supply.

I suggest you determine exactly which instrument(s) really benefit from
input voltage regulation and put only those on the regulator. Most of
the newer equipment we use was specifically designed for battery use, so
adding a regulator just increases the things that can fail.

--
Eric Greenwell - Washington State, USA
* Change "netto" to "net" to email me directly
* "Transponders in Sailplanes" http://tinyurl.com/y739x4
* "A Guide to Self-launching Sailplane Operation" at www.motorglider.org

COLIN LAMB wrote:
Most diodes have about a .6 volt drop, which may be significant. Therefore,
you will need diodes with a low threshold. Schottky diodes should be about
.3 volts drop. You can measure the actual drop using a digital
volt-ohmmeter. There is a diode test range and it will show the voltage
drop.

I'm flying with the diode setup. The best diodes I was able to find in
the UK were Fairchild MBR1035 Schottky diodes. These handle 35 amps and
come in TO-220 packages. Their spec quotes a 0.57 v drop, which seems
about right: with 2 x GPS II+, EW-D logger, SDI C4 and B.40 varios all
on the C4's internal voltmeter shows 11.6 v on two fully charged 7 AH cells.

I need to fit a radio and am thinking of fitting a Filser ATR-500. Can
anybody tell me if that will be OK on the end of the diodes or would I
be better to discard the diodes and use one battery to drive the ATR-500
and the other to run the GPS, logger, and varios?

Another reason I'm wondering about rewiring to separate the batteries is
that if we get landed with transponders I assume I'd be better off using
one for radio and transponder and reserving the other for the
GPS/logger/vario setup. Comments?

The other choice of radio would be a Microair 760. I assume that, as its
rated for 12-14v it would be quite unhappy on the end of the diodes and
just barely usable on a separate battery. I see that Maxim sell a range
of solid state voltage boosters (there is a model that can output 14v at
2 amps). Has anybody tried using one of these to drive a Microair radio?
If so, how well did it work?




Colin Lamb




--
martin@ | Martin Gregorie
gregorie. | Essex, UK
org |

wrote:
This is great discussion. I was doing a bit of research myself (it has
been over 10 years since I did any HW design, so am pretty rusty).

Another option that sounds even better is to use the new chips designed
for exactly this to control DirectFETs. That way the voltage drop and
power waste will be significantly reduced with very little additional
cost. The chips I were looking at were IR5001S:
http://www.irf.com/product-info/data...ta/ir5001s.pdf or LT4351:
http://www.linear.com/pc/productDeta...,C1079,P21 73


I have used DirectFET stuff in robotics and found them to be very
reliable and very simple to design with.

My biggest problem is all these components are only available in
surface mount these days and I have never tried to build a circuit
using them.

That should be easy enough to do with a pair of suitable power MOSFETs
and an LM358 dual op-amp to control the MOSFETs. If I don't split my
supply (see earlier post) I'm thinking about making this type of circuit
this because the voltage drop in the MOSFET will be very small.

Size isn't an issue - suitable MOSFETs are in TO-220 type packages and
the LM358 is an 8 pin DIP package.


--
martin@ | Martin Gregorie
gregorie. | Essex, UK
org |

Steve Paavola wrote:

If both batteries are on-line all the time, how do you know when one is
getting weak and needs to be replaced? Or do you replace both batteries
when voltage is marginal at the end of a flight?

You can get fairly inexpensive automatic peak detecting battery
cycler/charger units ($60-$100 at a guess - around GBP 69.00 in the UK)
that will measure the capacity during a discharge cycle. These units
will charge lead-acid, NiCd, NiMH and Li-poly with settable charge and
discharge currents. You can run them off 12v DC (mains or batteries)
they'll charge 12v lead acid and 14 cell NiCd batteries.

Check your local friendly RC model supply shop if you're interested in
this type of charger.


--
martin@ | Martin Gregorie
gregorie. | Essex, UK
org |

The Filsner is unlikely to have a problem through the diode setup you
have now. (not that I like the diode setup but it should work - BTW you
should be able to find lower voltage drop diodes).

For the Microair I would not read too much into a nominal voltage spec.
It is hard to tell wether the manufacturer is quoting a real absolute
voltage range or the nominal voltage. On the other hand I'd not want to
inflict anybody with a Microair radio (oops I can feel the flames
already). They have a very bad reputation where I fly. The several I've
seen installed have had problems, especially apparent heat related
problems with the displays getting garbled and then finally the whole
radio going out to lunch (and I'm not talking really hot days). Other
brands seems to be much more reliable. Why not go with Becker?

Personally I'd stay away from trying to get too fancy with regulators
etc. I want the supply to be as dumb and simple as possible. Just
batteries and circuit breaker or fuse and master switches. As few
connectors or solder joints as possible. And I personally I like
running one battery at a time - I want to see the health of each
battery and know about how much capacity I have in both batteries and
know I can switch in a reserve if I run a battery down -- which I may
not notice until it is too late. Like many of us in the Western USA I
fly over pretty desolate areas and I really want to know roughly what
battery capacity in reserve.

If you are goigg to install a transponder, have lots of avionics toys,
expect long cold flights it may well be you need to look at moving up
to larger capacity batteries or installing solar panels.

Darryl Ramm


Martin Gregorie wrote:
COLIN LAMB wrote:
Most diodes have about a .6 volt drop, which may be significant. Therefore,
you will need diodes with a low threshold. Schottky diodes should be about
.3 volts drop. You can measure the actual drop using a digital
volt-ohmmeter. There is a diode test range and it will show the voltage
drop.

I'm flying with the diode setup. The best diodes I was able to find in
the UK were Fairchild MBR1035 Schottky diodes. These handle 35 amps and
come in TO-220 packages. Their spec quotes a 0.57 v drop, which seems
about right: with 2 x GPS II+, EW-D logger, SDI C4 and B.40 varios all
on the C4's internal voltmeter shows 11.6 v on two fully charged 7 AH cells.

I need to fit a radio and am thinking of fitting a Filser ATR-500. Can
anybody tell me if that will be OK on the end of the diodes or would I
be better to discard the diodes and use one battery to drive the ATR-500
and the other to run the GPS, logger, and varios?

Another reason I'm wondering about rewiring to separate the batteries is
that if we get landed with transponders I assume I'd be better off using
one for radio and transponder and reserving the other for the
GPS/logger/vario setup. Comments?

The other choice of radio would be a Microair 760. I assume that, as its
rated for 12-14v it would be quite unhappy on the end of the diodes and
just barely usable on a separate battery. I see that Maxim sell a range
of solid state voltage boosters (there is a model that can output 14v at
2 amps). Has anybody tried using one of these to drive a Microair radio?
If so, how well did it work?




Colin Lamb




--
martin@ | Martin Gregorie
gregorie. | Essex, UK
org |

OxAero wrote:
Gary,

I developed a circuit card with isolation diodes for two batteries and
a pair of efficient voltage regulators, one for high current devices
(radio and transponder) and the other for the rest. As Dick indicated
the panel has two SPST switches for two batteries. The diodes drop .3
Volts, but the regulators operate down to low voltages. There is a
price to pay for operating the regulators. But, I felt that
considering that I use 12 Ah batteries, I had plenty of capacity and I
was more interested in supplying regulated power to my instruments.

The problem was that the large regulator couldn't handle the start up
load from the transponder. So, my quick fix was to bypass that
regulator with the second switch. I expect to revisit the project some
day and finish it properly.

Jim Hendrix


Gary Emerson wrote:

wrote:

By all means a second battery should be installed in our electrically
driven modern sailplanes. After many years of quickly flipping my
3-position battery switch, and trying not to have my logger to
momentarily dropout, I have concluded that is best to use 2 single-pole
battery switches. That way one can have either or both batteries
connected at the same time.

I saw the light when Jim Hendrix brought his sailplane to Caddo Mills
for Wing Deturbulator flight testing, and it was wired like that. You
will hear much more about that amazing new invention at the coming SSA
Convention.

Thermally,

Dick Johnson


Better yet is to use diodes so that both batteries will always be "on"
in parallel and you're always pulling from the best battery with no
fiddling required from the pilot. Relatively low voltage drop diodes
are available with 18 Amp forward capacity. For redundancy, I used two
in parallel on each battery. The diodes are available in the TO-220
package and it's easy to incorporate a small heat sink, but I have no
reason to believe they ever attained any temperature at all.

With two batteries connected with switches, if one battery does really
"die" then any time spent with both switches in the "on" position causes
the good battery to attempt to charge the "dead" battery to no avail, so
ultimately you're wasting what power you do have during this time. The
diodes eliminate any chance for cross charging...or discharging. Used
this system for several years, never experienced a single power issue.



Just curious what made you want to add a regulator? Pulling off a
battery should be pretty regulated supply in the first place. Shouldn't it?

Martin Gregorie wrote:
wrote:

This is great discussion. I was doing a bit of research myself (it has
been over 10 years since I did any HW design, so am pretty rusty).

Another option that sounds even better is to use the new chips designed
for exactly this to control DirectFETs. That way the voltage drop and
power waste will be significantly reduced with very little additional
cost. The chips I were looking at were IR5001S:
http://www.irf.com/product-info/data...ta/ir5001s.pdf or LT4351:
http://www.linear.com/pc/productDeta...,C1079,P21 73



I have used DirectFET stuff in robotics and found them to be very
reliable and very simple to design with.

My biggest problem is all these components are only available in
surface mount these days and I have never tried to build a circuit
using them.

That should be easy enough to do with a pair of suitable power MOSFETs
and an LM358 dual op-amp to control the MOSFETs. If I don't split my
supply (see earlier post) I'm thinking about making this type of circuit
this because the voltage drop in the MOSFET will be very small.

Size isn't an issue - suitable MOSFETs are in TO-220 type packages and
the LM358 is an 8 pin DIP package.


How much lower thatn 0.3V drop will the MOSFETs get you?

Tinwings wrote:
Being the incurably curious type, I decided to test this theory. I
took two known good 7Ah 12V SLA batteries and discharged one to 8 volts
(resting) with a 12V light bulb. The other battery I topped off with a
charger to 13.6 volts. I connected the two using less than two feet of
18 gauge wire and a ordinary toggle switch. Using a 60 Mhz bandwith
oscilloscope and a hall-effect type current probe I looked at the
resulting waveform when I closed the switch; a nice square edged rise
to about 3 amps, tapering down to 2.5 amps in a few seconds. Because I
didn't know what the frequency response of this current probe was, I
inserted a precision .001 ohm current shunt in line (very high
frequency response) and used the scope to watch the voltage drop across
it. The results were identical; no current spike, no inrush of current
- just a nice square edged waveform rising to about 3 amps. This simply
isn't going to weld contacts, burn out switches or blow (properly
sized) fuses. As for "wasting energy" by dumping from the good battery
into the dead battery when switching over - just do the math. Even if
the two batteries were connected for as much as 5 seconds while
switching from one to the other (two switch or "make before break"
switch arrangement), you will be using less than one thousanth of the
good battery's capacity to charge the "dead" battery.

If the batteries are both "good", but not charged equally, then the last
part is true. If a battery happens to go "bad" and won't take a charge
then the situation is different. Then the "bad" battery will just suck
down the extra power from the good battery and that power will not be
recovered.

Gary Emerson wrote:
Tinwings wrote:

Being the incurably curious type, I decided to test this theory. I
took two known good 7Ah 12V SLA batteries and discharged one to 8 volts
(resting) with a 12V light bulb. The other battery I topped off with a
charger to 13.6 volts. I connected the two using less than two feet of
18 gauge wire and a ordinary toggle switch. Using a 60 Mhz bandwith
oscilloscope and a hall-effect type current probe I looked at the
resulting waveform when I closed the switch; a nice square edged rise
to about 3 amps, tapering down to 2.5 amps in a few seconds. Because I
didn't know what the frequency response of this current probe was, I
inserted a precision .001 ohm current shunt in line (very high
frequency response) and used the scope to watch the voltage drop across
it. The results were identical; no current spike, no inrush of current
- just a nice square edged waveform rising to about 3 amps. This simply
isn't going to weld contacts, burn out switches or blow (properly
sized) fuses. As for "wasting energy" by dumping from the good battery
into the dead battery when switching over - just do the math. Even if
the two batteries were connected for as much as 5 seconds while
switching from one to the other (two switch or "make before break"
switch arrangement), you will be using less than one thousanth of the
good battery's capacity to charge the "dead" battery.


If the batteries are both "good", but not charged equally, then the last
part is true. If a battery happens to go "bad" and won't take a charge
then the situation is different. Then the "bad" battery will just suck
down the extra power from the good battery and that power will not be
recovered.
Sorry, didn't read your post fully. You were only talking about the
very short period when the batteries are still connected while switching
from A to B. The nice thing about the diode setup is you'll never have
to fiddle with a battery switch again.