24VDC Backup

[topadge]

I have some large machines that have 24Vdc control boards that reset
during storms and other power anomalies. I would like to keep the
controller running which draws @1.5A. I figure a huge cap like the
ones used in car audio (.5 or 1F) would keep me up for .25-.5 seconds
which should be plenty of time. I could also make a cheap battery
back up circuit. Does anyone have any opinion on which is the right
route to take. And if the battery backup is the way to go, any
circuit suggestions? Thanks in advance.
-topadge

 

[CFoley1064]

I have some large machines that have 24Vdc control boards that reset

during storms and other power anomalies. I would like to keep the
controller running which draws @1.5A. I figure a huge cap like the
ones used in car audio (.5 or 1F) would keep me up for .25-.5 seconds
which should be plenty of time. I could also make a cheap battery
back up circuit. Does anyone have any opinion on which is the right
route to take. And if the battery backup is the way to go, any
circuit suggestions? Thanks in advance.
-topadge

Actually, you don't need that much cap. A 10,000uF 35V cap will sag about 1.2V
at 1.5A load for each second the power is interrupted. That should be enough
to keep most industrial electrics (which usually require +/- 10% at their
pickiest) going through a brief power outage.

If your 24VDC supply is linear/overcurrent protected, you're in luck -- just
put it across the power supply (I use the standard open frame linears on
control panels whenever I can get away with it). If it's a switcher, you might
have problems at turn-on with that big of a cap -- many are designed to
interpret that kind of load as a short, and will not turn on. You might want
to try something like this to limit turn-on surge:

.------------------.
| |
| | D
| + o-----o---------------->|---o--o
| | | |
| | | R |
| | | ___ D |
| 24VDC Switcher | '--|___|--o------>|---'
| | |+
| | ---
| | --- 10000uF
| | |
| - o---------------|--------------o
| |
| |
'------------------'

created by Andy´s ASCII-Circuit v1.22.310103 Beta www.tech-chat.de

D is dependent on your peak current (I'd start with 1N5402), and R depends on
how rapidly you want your cap to charge up. Since we're industrial here, try
220 ohms, 1 Watt.

A UPS driving a second power supply seems to be overkill if all you need is a
second or two of protection.

Good luck.
Chris

 

[John Fields]

I've used a large cap to ride through browndowns, as below.

24Vdc-------+-----+------> Load (1.5A)
| |
\ _|_
R / /_\ D
\ |
| |
+--+--+
|
C === (0.5F)
|
0v-----------+---------

D is probably a Schottky diode, and R is there to limit the
initial charging current. C= 0.5F and R= 120 ohms would
be a 60sec time constant, giving 180sec to charge the cap
(at switch-on) up to 22V and 300 sec up to 23.75V.

 

[CFoley1064]

Sorry -- one more thing. 220 ohms, 3W, not 1 watt. Thanks.

 

[Howard Henry Schlunder]

in message

Actually, you don't need that much cap. A 10,000uF 35V cap will sag about 1.2V
at 1.5A load for each second the power is interrupted. That should be enough
to keep most industrial electrics (which usually require +/- 10% at their
pickiest) going through a brief power outage.

Umm, you might wish to rerun your calculation. According to i=c(dv/dt),
there should be a voltage drop of 150 Volts/second when a 1.5A constant
current draw is attached to a 0.01F capacitor. A 22.8V, 1.5A load would
draw 34.2 Joules of energy in one second, while a 0.01F capacitor at 24V
seems to have only 2.88 Joules in it.

 

[John Popelish]

I have some large machines that have 24Vdc control boards that reset
during storms and other power anomalies. I would like to keep the
controller running which draws @1.5A. I figure a huge cap like the
ones used in car audio (.5 or 1F) would keep me up for .25-.5 seconds
which should be plenty of time. I could also make a cheap battery
back up circuit. Does anyone have any opinion on which is the right
route to take. And if the battery backup is the way to go, any
circuit suggestions? Thanks in advance.
-topadge

The capacitor is certainly lower maintenance, though the battery can
hold a lot longer. Make sure the cap charges through a resistor (to
limit the inrush current) and dumps through a diode.

We do this in a big way for boiler control circuits. We have a farad
or so at 125 volts and a 120 volt battery bank of lead acid batteries,
with both feeding the 135 volt bus (normally supplied by 3 phase
transformers with 6 phase rectifiers) through big diodes. The
batteries have their own chargers that are independent of the 135
volt bus rectifiers.

 

[Spehro Pefhany]

in message


Umm, you might wish to rerun your calculation. According to i=c(dv/dt),
there should be a voltage drop of 150 Volts/second when a 1.5A constant
current draw is attached to a 0.01F capacitor. A 22.8V, 1.5A load would
draw 34.2 Joules of energy in one second, while a 0.01F capacitor at 24V
seems to have only 2.88 Joules in it.

His calculation is about right for 1/2 cycle of 60Hz AC (full-wave
rectified) rather than 1 second.

Best regards,
Spehro Pefhany

 

[topadge]

The capacitor is certainly lower maintenance, though the battery can
hold a lot longer. Make sure the cap charges through a resistor (to
limit the inrush current) and dumps through a diode.

We do this in a big way for boiler control circuits. We have a farad
or so at 125 volts and a 120 volt battery bank of lead acid batteries,
with both feeding the 135 volt bus (normally supplied by 3 phase
transformers with 6 phase rectifiers) through big diodes. The
batteries have their own chargers that are independent of the 135
volt bus rectifiers.

Thank you all for the responses. Does anyone know where I can find a
1F cap rated for 30Vdc?

 

[Bob Minchin]

Thank you all for the responses. Does anyone know where I can find a
1F cap rated for 30Vdc?

You should be able to find 100,000uF capacitors Ok and use 10 of them in parallel

 

[Chris Quayle]

You should be able to find 100,000uF capacitors Ok and use 10 of them in parallel

That rather begs the question as to how big and expansive the caps would
be in relation to a small lead acid battery ?. 10 100kUf caps is
expensive and a lot of space .

A pair of 1 Amp/hour 12v in series, on float, would be smaller, cheaper
and keep the system up for much longer.

Chris