Avoiding diode losses in battery-backed system

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
         America: Land of the Free, Because of the Brave

...Jim Thompson

--
|  James E.Thompson, P.E.                           |    mens     |
|  Analog Innovations, Inc.                         |     et      |
|  Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
|  Phoenix, Arizona            Voice:(480)460-2350  |             |
|  E-mail Address at Website     Fax:(480)460-2142  |  Brass Rat  |
|       http://www.analog-innovations.com           |    1962     |
             
         America: Land of the Free, Because of the Brave

Hmm, why not simply use a basic relay to insert the commercial power in which would use the commercial power it self? When commercial power goes away, so wont the relay contacts that made the connection to the commercial power.

Sounds like basic concepts to me.

--
"I\'d rather have a bottle in front of me than a frontal lobotomy"

http://webpages.charter.net/jamie_5"

Linear has stuff like the LTC4357 a lossless diode

martin

Its not hard to make a discrete version of the LTC circuit, but the IC has several attractive features. First, there's the charge pump, which provides a voltage far enough above the input/output to power the n-channel mosfet. Second, there's a fast 350ns shutoff if the current is reversed due to a source fault. Third, there's an active 25mV load-drop function, useful for equal diode-OR load sharing.

Yep, a CW multiplier sneaking more volts from an smps is pretty good, but the LT stuff is easier if you dont have a smps

martin

ace

s while

at

er is

This would be even nicer if the FET were integrated into the IC, or, more likely, the circuit were integrated into some FET.

while

The advantage of having an external fet is that you can size it to your requirements, otherwise Linear would have to keep zillions of different lumps of precisely melted beach sand in stock, in neat little plastic packages, with really bad thermal thingummies Not good sense, the size of fets can be so small these days

martin

Somehow I initially missed about half of the responses here -- thanks to everyone who did so; I'll be trying out the various approaches.

---Joel

Because at 3A, ~15V you need some pretty hefty capacitors to hold up the power rails during the perhaps... 100ms?... time the relay takes to switch over.

Diodes are much faster.

Thanks Martin, great find!

You seem to be ignoring the usual role of the source in charging the battery. This places the battery directly across the load, with only the source being effectively diode-orred. The battery is only switched out by a charge controller, wich needs be a little more complicated than a diode.

For diode losses in battery mode only for the non-charging situation you're talking about, try the basic configuration below.

RL

Version 4 SHEET 1 880 680 WIRE -176 -64 -288 -64 WIRE -288 32 -288 -64 WIRE -144 32 -288 32 WIRE -96 32 -96 -64 WIRE -160 80 -208 80 WIRE -32 80 -80 80 WIRE 208 80 -32 80 WIRE 400 80 208 80 WIRE 496 80 400 80 WIRE 208 128 208 80 WIRE -32 208 -32 80 WIRE 400 224 400 80 WIRE 208 240 208 208 WIRE 496 240 496 80 WIRE -208 256 -208 80 WIRE -32 320 -32 288 WIRE 160 320 -32 320 WIRE -288 400 -288 32 WIRE -208 400 -208 336 WIRE -208 400 -288 400 WIRE -32 400 -32 320 WIRE -32 400 -208 400 WIRE -16 400 -32 400 WIRE 208 400 208 336 WIRE 208 400 48 400 WIRE 400 400 400 304 WIRE 400 400 208 400 WIRE 496 400 496 304 WIRE 496 400 400 400 WIRE 208 416 208 400 FLAG 208 416 0 SYMBOL nmos 160 240 R0 SYMATTR InstName M1 SYMATTR Value FDR4420A SYMBOL diode 48 384 R90 WINDOW 0 0 32 VBottom 0 WINDOW 3 32 32 VTop 0 SYMATTR InstName D1 SYMBOL res -48 192 R0 SYMATTR InstName R1 SYMATTR Value 1E3 SYMBOL voltage -208 240 R0 WINDOW 123 0 0 Left 0 WINDOW 39 24 132 Left 0 SYMATTR SpiceLine Rser=.1 SYMATTR InstName V1 SYMATTR Value PULSE(0 12 0 1E-3 1E-3 1E-4 2.2E-3 1) SYMBOL voltage 208 112 R0 WINDOW 123 0 0 Left 0 WINDOW 39 24 132 Left 0 SYMATTR SpiceLine Rser=.02 SYMATTR InstName V2 SYMATTR Value 10 SYMBOL sw -176 80 M90 SYMATTR InstName S1 SYMATTR Value INT SYMBOL voltage -80 -64 R90 WINDOW 0 -32 56 VBottom 0 WINDOW 3 32 -232 VTop 0 WINDOW 123 0 0 Left 0 WINDOW 39 0 0 Left 0 SYMATTR InstName V3 SYMATTR Value PULSE(0 5 0 1E-6 1E-6 9E-5 1E-4 1000) SYMBOL res 384 208 R0 SYMATTR InstName R2 SYMATTR Value 3 SYMBOL cap 480 240 R0 SYMATTR InstName C1 SYMATTR Value 220µ SYMATTR SpiceLine Lser=.05 TEXT -64 24 Left 0 !.model INT SW(Ron=.001 Roff=1Meg Vt=2.5 Vh=-.4) TEXT -294 450 Left 0 !.tran 2.2E-3

The longer the better, makes you immune to power glitches: View in a fixed-width font such as Courier.

. . . . ~ . .-----------o o----------. . | | . | | . | | . | RELAY | . | - - - - - - - - - - | . | | ------- | | . '----------|///////|-------' . | ------- | . NO . PRIME PWR>-------------------o---- | . | ----o----------------. . .------o---- COM | . | | NC | | . -| - - - - - - - - - | . | | . | | . +----------|>|------------------+ . | DIODE | . | | . | + ------ . --- | | . - BATT | LOAD | . | - | | . | ------ . | | . | | . PWR GND >-----------+-------------------------------+ . | . --- . /// . .

-or-

View in a fixed-width font such as Courier.

. . . . . . . . . RELAY . - - - - - - - - - - . | ------- | . .---------------|///////|-------. . | | ------- | | . | NO | . PRIME PWR>--+--|>|-----------o---- | | . DIODE | ----o----------------. . .------o---- COM | | . | | NC | | | . -| - - - - - - - - - | | . | | | . | | | . +----------|>|------------------+ . | DIODE | | . | | | . | + | ------ . --- | | | . - BATT | | LOAD | . | - | | | . | | ------ . | | | . | | | . PWR GND >-----------+-------------------------------+ . | . --- . /// . .

Nice Fred, that is both pretty elegant and simple -- thanks.

---Joel

I was ignoring it primarily the sake of discussion -- it's a LiO battery so it has the usual CC-CV charger controller associated with it.

Mmm... is V2 the battery and V1 the primary power there?

---Joel

V2 is the battery. M1 just acts as a diode (parasitic body) with enhancment..

You can do the same thing with the diodes in the positive rail, if you want, but it means using a PMosfet.

RL

:Nice Fred, that is both pretty elegant and simple -- thanks. : :---Joel :

Except that the diode is the wrong way round..... the load is powered from the battery ALL the time.

:On Sun, 10 Feb 2008 14:17:48 -0800, "Joel Koltner"

:wrote: : ::Nice Fred, that is both pretty elegant and simple -- thanks. :: ::---Joel :: : : :Except that the diode is the wrong way round..... the load is powered from the :battery ALL the time.

My bad... didn't see bottom diag at first glance.

However, in the top diag, if the battery voltage exceeds the prime voltage by

0.7V, while prime is selected, the battery will effectively drain into the load in parallel with the prime voltage. The bottom diag is better.

That would be a brownout scenario where line is too low to maintain regulation of the DC output derived from it, but the relay does not drop out, so the battery consequently powers the load with the diode loss. This may or may not be significant to the required backup time.