Maker Pro
Maker Pro

UPS external battery hook up


rick stanford

Jan 1, 1970
I have a Tripplite SU series UPS that allows external batteries to be
connected to get longer run time in case of power outage. It gives
specs. of 30 amps, 36 volts on the external battery connection area
(so I assume Tripplite has 36 volts total from internal batteries).

I want to use some godzilla batteries, like three 12 volt car
batteries in series (which often give 13 volts or more each fully

I'm nervous about connecting though, because I already toasted my
DELTEC UPS connecting a 12 volt car battery to the internal battery
leads (I removed 12 volt internal battery first and then connected
using proper polarity, of course).

It seems such a powerful battery(s) makes like a large inrush upon
connection ?
It toasted two 30 amp fuses hard soldered on the circuit board.

Should I use a diode, or a resistor, or something to control/buffer
the "inrush" ? What specs would you suggest ?

Please cc to my private email at [email protected]




Jan 1, 1970

I gave this some thought - and think you may have induced the problem
in the Deltec UPS. The charge circuitry should protect the UPS
against a low-battery, high-inrush condition, as SLA batteries exhibit
the same phenomenon as a car battery.

The design of the inverter is quite marginal in low cost units. When
you removed the internal battery - you added lots of distributed
inductance to the switching circuit - which expected the battery to
function like a really large, low impedance capacitor. Lack of a
local, large capacitor probably resulted in severe over/undershoot of
the switching FET supply voltage as it was hammered down by the FET
switch, and rang for relative days afterwards.

Long leads also induce voltage drop - which will INCREASE the current
drawn from the batteries - aiding in the above, and in causing other

You should have no problems with a UPS that features an external
battery connector - with the internal battery present. Play it safe,
and confirm this with the manufacturer. They will have harness and
battery guidelines to follow.

---- HINTS TO ALL ----

DO NOT remove internal batteries from a UPS, and fail to provide a low
ESR path (local battery or bank of low ESR capacitors, that can handle
the high ripple current) for the switch current.

Verify any modifications with a scope and a minimal load, such as a
25W light bulb. Verify there is not excessive 'ringing' on the DC
supply lines at the battery, or on the drain of the switch FET(s).
Verify this again at a slightly higher load (100W) - then at your
desired load. Use an isolation transformer to power the scope while
doing this test! (The battery negative FLOATS -and can smoke the
probe in an instant if the scope is grounded!)

Remember that longer run times were not envisioned in most UPS designs
- and if you modify a low-cost unit ADD AT LEAST ONE FAN to blow air
directly over the FET(s) and switch diodes. The fans should be fused,
connected directly to the battery, and operate at the battery voltage.
Occasionally, a larger heat sink may also be required for the FETs
and diodes.

Current sharing is difficult in some conditions. When paralleling
batteries, the current path for all batteries must be equal in length,
and join together in one point (a heavy terminal block) for each
polarity. Clean the block carefully, and use a thin grease to prevent
oxidation. Check voltage drop with a DVM to ensure the drops are
equal. Series connected batteries do not have this restriction.

USE LARGE WIRE AND SHORT RUNS. Wattage in equals wattage out - plus
of a 12 volt battery-supplied UPS - given this formula. If you have a
200W load, this is 25A or so from the battery. If the battery voltage
drops, the battery current increases proportionally - and can create a
runaway condition - destroying the FETs in seconds. 4GA wire is a
minumum for a 300W inverter - given a 12V battery, a full 2.5A-120V
load, and three feet of battery leads.

Never use 'modified sine wave' UPS's with line filters on the OUTPUT.
The output waveform on these units resembles a square wave - which
the line filters will try to correct (short) - resulting in the
destruction of the inverter.

You can see that inverters are not simple devices - given the physics
of what they do - and what people want to pay for them. If you are
unsure about any of this diatribe, - BUY AN APPROPRIATE INVERTER for
your needs.

Good Luck!