Short term battery backup circuit

[Swinkels]

Hello everyone,

In my car I have an Android navigation system, which takes some time to start up. Now when I get into my car I always first put in my car keys and make contact. Then I put my phone away, put my safety belt on and a couple more things, in the mean time the Android system begins turning on. But now when I start the car, the car cuts the power to the system until the car has started, so the whole system has to start all over again, which I find annoying.

I have found a circuit like this one which would bridge the 2-3 second power cut of, but it would also drain the 12v backup battery every time I turn off the car. So I'm looking for a circuit that provides 12v backup power for something like 10 seconds, and after that turns off. How would I do that?

Maybe I can wire a capacitor and a second transistor somewhere in the circuit so when the capacitor runs empty the backup battery turns off again? I can perfectly solder together something, but I don't have the knowledge to design something like this...

Thanks in advance!

TLDR: In my case the 14v is briefly cut off when starting the car, and during these seconds I want the 12v battery to take over. But only for a maximum duration of around 10 seconds, so that when the 14v is cut off completely the backup battery doesn't drain.

 

[danadak]

Do you know how much power/current the Android system requires ?

Regards, Dana.

 

[Swinkels]

Good question, I tried to find it, but I dont know.

This is the radio though

 

[Martaine2005]

If it’s anything like my son’s (Chinese) facia stereo, it has a yellow accessory wire. It was necessary to wire that directly to the battery side rather than through the ignition side.

Martin

 

[Swinkels]

Hello Martin,
Thanks for the reply, but I already swapped the yellow and red wires. This makes the radio retain its memory and therefore start a lot faster, but the power loss when starting is still too long for not making it restart.

 

[Martaine2005]

Try the yellow wire directly to the battery. It used to be illumination of the stereos lighting plus memory. The illumination should turn off when the ignition is turned off, and the cars lights.Therefore saving the battery. But a direct connection of the yellow may well help in keeping your unit alive only when the ignition is switched on.

Martin

 

[Swinkels]

Yellow directly to the battery? You mean always keeping the radio turned on? That would drain the battery, wouldn't it?

How it is now is how it is supposed to be, memory connected directly to the battery and the on/off connected to the key contact.

The problem is that when I start the car both red and yellow are disconnected, cutting all electricity from the radio.

 

[AnalogKid]

The first schematic in the page you linked to is both more and less than what you need. The second schematic seems overly complex. And, a relay is overkill. The current or power draw of the device you are powering should be in a datasheet, product page, molded into the case near the power inlet, etc. Wih that information, the transistor type can be optimized for the power being controlled. Without it, we can make a good guess.

More: You do not need a 12 V regulator. Your device has been running on vehicle power just fine without it with no problems, so why add one?

Keep: D1, R2, D2, and the battery.

Less: Add a 1-transistor shutoff switch with a 10-second timer. The timer is a simple R-C delay powered by the 12 V input *before* D1. It sees the 12 V go away, waits 10 seconds, then turns off the output transistor switch. This part is between the schematic + output and the device. This is called a high-side switch, a common circuit and technique.

The battery is charged continuously whenever ignition is on, while the device runs on vehicle power. The backup battery powers the device only when the ignition voltage is less than the backup battery voltage + 0.8 V (the D1 forward voltage drop), and then for only 10 seconds. As a test that the circuit is working, the device should stay on for 10 seconds -ish after the ignition key is removed.
Which are you more comfortable with, or have access to - a small PNP power transistor or a p-channel lower MOSFET?

Yard work now, schematic later.

ak

 

[davenn]

The problem is that when I start the car both red and yellow are disconnected, cutting all electricity from the radio.

which means that the yellow wire ISNT connected direct to the battery as it is supposed to be.

 

[Bluejets]

That would drain the battery, wouldn't it?

Simple solution...add a switch.......direct to battery supply.
If you've tapped into the ignition switch wiring, the permanent live cannot be too far away.

 

[AnalogKid]

Here is a first pass at a *concept* schematic. Most of the components have to be adjusted to fit the actual device parameters.

R2 trickle-charges the backup battery whenever the engine is running. D1 disconnects the vehicle electrical system when ignition is off, so those devices do not drain the backup battery. D3 disconnects the timer from the vehicle electrical system so those devices do not discharge C1, the timing capacitor.

When the ignition is on, C1 is charged up, Q1 is on, Q2 is on, and the device is powered through D1.

When the ignition is turned off, C1 provides 10 seconds of base current for Q1, which keeps Q2 turned on. The battery powers the device through D2. When C1 is discharged down below 0.6 V, Q1 and Q2 turn off, removing battery power from the device.

ak

 

[AnalogKid]

The schematic in #11 is missing one or two resistors. Is there any way to update that post, or do I have to create another post. None of the other fora I'm on have a fixed time period for updates and corrections.

Also - if the device current is low enough, B1 could be replaced by a large electrolytic capacitor.

ak

 

[Swinkels]

which means that the yellow wire ISNT connected direct to the battery as it is supposed to be.

Ah, it was getting late and got confused. I measured the connections and yellow is connected directly and stays high no matter starting the car or not. The memory is kept when disconnecting the key contact at the moment.


@AnalogKid: Thanks you very much!

I simulated the circuit and it works! I can even just use the yellow wire instead of the backup battery, which makes things easier.

Now I just need to figure out the power consumption of the radio to figure out the parts I need right? So D1 and Q2 need to be able to take the load that the radio requires I guess. But what about R2 and D2? Together the full load, but how much of the current does each element receive?

D3 should only be able to take the load to charge the capacitor. Which is 0.0158 joules in 10 seconds, which is almost nothing, so a small diode should be OK.

But what wattage will be running through R3? Something small right?

Edit: didn't see the second post of AnalogKid stating there are resistors missing yet when replying.

 

[AnalogKid]

I simulated the circuit

The link brings up a page that requires registration.

Also, depending on the current required by the radio, there might be a much more simple way. Basically, it is a version of what is going on with the energy supply to the Q1 base.

As mentioned in #12 - if the device current is low, such as 50 mA or less, then the battery can be replaced by a large capacitor at a not-totally-crazy cost. Now the whole circuit is just 4 parts. The cap is selected to have about 5 seconds of runtime holdup energy. After that, it is discharged and the radio is off. When the vehicle voltage is applied, it charges up in less than one second.

This raises an unasked question - what are you using, or what do you want to use, for the holdup battery?

ak

 

[AnalogKid]

Here is a minimalist concept of a short-term backup with automatic switching. An operating assumption for these numbers is that the 12 V device will operate down to 10 V.

R1 limits the peak charging current to something that won't disturb other devices in the vehicle. D1 is beefed up for the higher charging current. With a 50 mA device load, a 0.1 F cap discharges from 12 V to 10 V in approx. 4 seconds, and charges up from 0 V to 10 V in 2 seconds.

For R1, a 1 W part is what's in my design library. The peak power surge in R1 is short, but the part should be rated for 3 to 5 W. The value should not be less than 4.7 ohm.

ak

For R1,

 

[Kiwi]

I assume that your system has two power wires. A constant battery supply, and an accessory/ignition wake-up wire.
The wake-up wire is usually a low current wire that is just used to turn the system on. Most current comes from the battery supply wire.

The accessory circuit on most vehicles usually drops out during cranking, but the ignition circuit stays live.

A cheap possible solution is to feed the wake-up wire through a pair 1N4007 or similar diodes.
One diode is connected to accessory and the other to an ignition supply that remains alive during cranking.

 

[AnalogKid]

Here is the updated schematic from post #11. R3 and R5 are added to assure that the transistors turn off completely.

Note again that some of the component values will change when more is known about the device being powered, the size of the battery, etc.

Note also that no lithium battery of any type can be used in this circuit.

ak

 

[Swinkels]

Thanks again for the scheme!

I did indeed see during the simulation that the voltage to the device turned off slowly without the two extra resistors. With the extra resistors the drop is more sudden.

I measured the amperage of the wake-up wire, which is about 400 mA. I now realize I don't have to feed the whole radio with this circuit, only the wake-up current. As the backup battery I want to use the car battery itself, so the two 12v supplies are both going to be the same car battery.

 

[AnalogKid]

Here is #17 reworked for using the car battery as the holdup power source. I adjusted the values of some components, but without real information about the device being powered they all are guesswork.

Did you know back in post #1 that the car battery would be the backup energy source?

ak

 

[Swinkels]

Thank you very much, no I didn't realize I could use the battery to power the wake-up connection when I posted the first question. A lot became clear when I was drawing the circuit based one your first sketch, this really is the first time I'm working on something like this.