I'm trying to come up with a simple battery backup for an aquarium pump. If someone would be kind enough to take a look at my drawing and tell me if it is likely to work or not, I'd appreciate it.

 

It should work.
But with the relay you don't need the diode connected in series with relay output from the battery.

 

Thank you very much! Now that I looked at it for another minute, I see why only one diode is needed-good catch.
The pump draws about 1.5A, I have 3A diode 1N5408. Is that a good choice?

From the 1N5408 data sheet.

Peak Forward Voltage
1.2 V

Peak Reverse Current
5 uA

Peak Non-Repetitive Surge Current
200 A

Peak Reverse Repetitive Voltage
1000 V

Operating Temperature
-55 to 150 °C

Configuration
Single

Peak Average Forward Current
3@Ta=75C A

 

I think you want to use the common and NC terminals on the relay. Loss of power will cause the relay to drop out and the common and NC contacts will close.

 

The pump draws about 1.5A, I have 3A diode 1N5408.

That should be fine.

I think you want to use the common and NC terminals on the relay. Loss of power will cause the relay to drop out and the common and NC contacts will close.

Good catch.

 

Oops, I was thinking normally open meant open when power is applied. Thanks for pointing that out Ylli. And thanks crutschow for letting me know I was on the right track,
Now, I'm not so worried about frying things to death.
THANKS GUYS!

 

I decided to use DC relays and add an alarm. The problem is, 12V is enough to hold the (fuse) relay partially open when the 20V power supply is unplugged. I can feel the relay vibrate which explains the chattering buzzer. If both 12V and 20V are disconnected and I connect 12V, the fuse relay stays closed and the alarm works perfectly.

I like the idea of an alarm if the fuse blows. Is there a way to delay 12V for a second or two when power is lost to let the fuse relay close before it gets 12V? Or maybe another fix without getting crazy complicated?

This is the relay I'm using (both main relay and fuse relay). https://www.ebay.com/itm/NEW-10pcs-...e=STRK:MEBIDX:IT&_trksid=p2057872.m2748.l2649

I left the extra diode on the main relay in case it ever stuck closed, 20V would not get into the 12V battery. Do I need another fuse to prevent 12V and 20V mixing if the main relay stuck in the closed position?

 

I don't understand.
If the fuse blows, where is the coil getting 12V from?
Or do you want the alarm to also sound when it goes on battery?

It also confusing when you talk about the relay closing. Normally that means energized.
Say energized or deenergized to make it clear.

 

Energized or deenergized it is.

If the fuse blows, I feel sure the alarm will work properly.
After the fuse relay has been energized with 20V and I unplug the 20V power supply, 12V is introduced to the fuse relay. At that point, 12V causes that relay to remain partially energized. If the 12V is disconnected and reconnected with the 20V power supply unplugged, that relay remains deenergized and the alarm works as planned.

My plan is to get an alarm if I the 20V power supply fails, I lose power, or the fuse blows.

 

Do you have any thoughts as to how the 12 volt source is getting to the relay coil? From your drawing, I don't see any path. What is the voltage across the relay coil when the 20 volts is disconnected but the 12 volt battery remains?

 

How about adding a diode from he main relay 12V output to the buzzer?
That way the buzzer will sound when the 12V is applied, independent of the state of the relay.
It won't hurt the relay to stay partially energized from the 12V.

The buzzer will still also sound if the fuse blows, since that definitely causes the fuse relay to deenergize.

 

Like this? The diode was already there, I just connected the main relay 12V output directly to the buzzer.

I assumed a relay that is vibrating because it can't fully deenergize would tear the contact points up, IDK though.

When the 20V power supply is unplugged, 12V goes to the pump and the fuse relay as well. I was bench testing with a 12V wall wart in place of the battery. IIRC, the wall wart put out 12.28V and by the time I measured it across the relay coil, it was a little lower, still above 12 though.

 

Like this? The diode was already there, I just connected the main relay 12V output directly to the buzzer.

That should work.

I assumed a relay that is vibrating because it can't fully deenergize would tear the contact points up, IDK though.

Not if it's not carrying any current.

 

A beautiful solution-thank you very much!

 

What if a relay failed and got stuck in the closed position? Would the mixture of 12V and 20V be a fire hazard?

 

What if a relay failed and got stuck in the closed position? Would the mixture of 12V and 20V be a fire hazard?

The diodes prevent any problem for that failure.

 

But both diodes have electricity flowing to the pump. If the relay failed in the closed position and the 20V power supply was on, wouldn't 12V and 20V mix?

 

But both diodes have electricity flowing to the pump. If the relay failed in the closed position and the 20V power supply was on, wouldn't 12V and 20V mix?

No.
The 20V will reverse bias the 12V diode to the 12V so no mixing occurs.
Actually you don't need the main relay to control the 12V, since the diode will apply the 12V when the 20V stops.
You still need it (or some other method) to control the buzzer when the 12V is supplying current, however.

 

Well, that's pure luck-I'll take it . Thanks.

 

Under normal working conditions, the fuse relay can't completely de energize when 12V is applied. It opens and closes so fast it vibrates. That bothers me. If the fuse relay is de energized and 12V is applied, it remains de energized.

I am considering adding a diode and a capacitor to the main relay. When power is lost, I want the capacitor to keep the main relay energized long enough to allow the fuse relay to completely de energize before it gets 12V.

The relay consumes 0.45W.

Would this 10,000 uf 50V cap do the trick?
https://www.ebay.com/itm/NOVER-Audi...m=272858941492&_trksid=p2045573.c100506.m3226

Data sheet for the 24V relay.

http://datasheetcafe.databank.netdna-cdn.com/wp-content/uploads/2015/10/SRD-05VDC-SL-C-Datasheet.pdf