Battery based tank circuit (tested) - Page 12

Arathranar · 10-18-2004, 12:50 AM

Quote: Originally Posted by Ricky327

However, I'm considering adding a number of diodes to drop the tank battery voltage to nearer 12V as all the DC-DC psu's do not regulate this (bar the Opus).

Are you kidding?

No. This is a crank survival circuit, not intended to run the PC for any significant length of time. So the output of a 5.1Ah tank should be pretty stable for that short time as it won't discharge much in the 5 seconds it supplies power (hence my different emphasis on charging requirements). In addition, I want to avoid over voltage into the PC. ~14V into the +12V line is way out of spec. So I need someway to drop the car alternator voltage down a few notches. I'll be driving an AMD micro Atx m/b so getting that +12V Near spec is a requirement. The 13.8-14.5V fluctuation I see when driving could be too much even if I drop it to 11.8-12.5V. I'm hoping the diode and tank battery combo will also help reduce that fluctuation a little but I suspect not.

justICE · 10-18-2004, 08:05 PM

Quote: Originally Posted by Arathranar

no mather what you do you aint gonna get away with that voltage drop.
True. But I'm still far from convinced the resistor is better than just a diode.

Here's the circuit I'm going to try:

Dangerous circuit

Arathranar · 10-18-2004, 08:06 PM

A few more details would be nice. Why is it dangerous?

justICE · 10-18-2004, 09:11 PM

Maybe dangerous is an innapropriate word. It will be risky for your components assuming you are not going to regulate that power, which I believe is your intention. Either side of the 0.7V swing of your "regulated" alternator is going to cause 1 problem or another. On the low side you will starve your tank. On the high side you will destroy your components.

Spikes could be a problem

Arathranar · 10-18-2004, 09:18 PM

I'll be either using the switched 12V (off during crank) or using a relay to achieve the same effect using the switched & unswitched 12V. So spikes during crank will not be an issue.

I'll actually have a bunch of diodes between the tank and the PC psu to drop the voltage closed to the +12V it needs. I found a schottky diode last night with a typical voltage drop of only 0.25V at 20A. I'll be using one of those before the tank and 6 to 7 after the tank. That should take the voltage down from the 13.8-14.5V I see when driving down to 11.8-12.5V which should be close enough to spec for my AMD mAtx m/b to cope with (we'll see - it's a cheap m/b!). There'll also be an in-line fuse between the tank and the PC for obvious reasons.

Are there spikes during driving? If so, then there is indeed risk to my PC but I'm still game to try it. Hmm. Maybe I could hook up a suitable zener diode to limit the voltage my psu will see.

justICE · 10-18-2004, 09:41 PM

can you update your pic? I'm confused.

Arathranar · 10-18-2004, 09:46 PM

Here it is. This is assuming the 0.25V forward voltage drop for the 20A diodes I've found is reasonably accurate.

justICE · 10-18-2004, 10:00 PM

During crank the voltage at the node with the tank (+) will be at ~12.5V no?

Arathranar · 10-18-2004, 10:03 PM

I won't know until I try it. With no current being used, the tank battery sits at 13.8V (with an unknown state of charge). I'm hoping that it will not drop much when I pull 10-15A out of it for a few seconds, but there's certainly a possibility it could drop too much.

justICE · 10-18-2004, 10:18 PM

You can test it easily enough. Buy me .

Note: the picture depicted in that link is NOT the right resistor. You should put 2 of the 10W 1-ohm resistors in series between the terminals.

tna0601 · 10-18-2004, 10:51 PM

Why cant you use the diode in-between the two batterys, The current will flow into the tank battery and will be blocked when starting the vehicle.

Then use an old UPS to convert the 12v to 120 and use a standard psu for the computer and the perhips. You will also have 120v AC in the car for other various uses. The inverter in the UPS will filter out any noise and would also send windows the shutdown signal to windows.

Ricky327 · 10-18-2004, 10:57 PM

The diode is between the main battery and the tank, so will not cause the above problem. However, I'm considering adding a number of diodes to drop the tank battery voltage to nearer 12V as all the DC-DC psu's do not regulate this (bar the Opus). The diodes should result in a less variable voltage drop than a resistor.

From the discussion we had earlier on the thread you agree that you can live with the battery not being charge at 100% and much happy to go the diode route. You also said that the diode after the tank is not a good idea because it reduces the running time?

If the battery never get charged at 100% and you are also reducing the running time by droping 1.75V Are you aware that the tank battery may not even work?

You are assuming that the alternator is always pumping at 13.4V as well as both main battery and tank pumping 13.4V constantly...that aint gonna happen.

If lets say the tank is undercharged and outputing 12V (with load) and then you are dropping it by another 1.75V you be left with 10.25V into your PC when you crank...crash/reboot.

I suggest you read the voltage tolerance of the +12V component in the PC before you do this.

Ricky327 · 10-18-2004, 11:06 PM

FYI

Intel PSU design recomendation should be within +-5%

So depending how your main/tank battery and alternator are feeling on the day you probably be fine.

Arathranar · 10-19-2004, 12:24 AM

You also said that the diode after the tank is not a good idea because it reduces the running time?

I'm not sure where I said that. I thought that was your worry.

If the battery never get charged at 100% and you are also reducing the running time by droping 1.75V Are you aware that the tank battery may not even work?

Yes. I'm going to experiment with the number of diodes to compromise on the voltage when the power is coming from the car battery vs the tank. Even if I only get an 80% charge (and I expect to get much nearer 100% than that), that's still plenty to give practically the full voltage because of the nature of lead acid batteries.

You are assuming that the alternator is always pumping at 13.4V as well as both main battery and tank pumping 13.4V constantly...that aint gonna happen.

Why not? The alternator outputs 13.8-14.5V whilst idlng or driving. The tank has to output that for up to 5s during the crank whilst nearly 100% charged.

If lets say the tank is undercharged and outputing 12V (with load) and then you are dropping it by another 1.75V you be left with 10.25V into your PC when you crank...crash/reboot.

True - *if* the tank is very low on charge. With a 7.2Ah battery only being used to survive crank, that should never be remotely true.

I suggest you read the voltage tolerance of the +12V component in the PC before you do this.

That applies just as much to your setup as mine.

Arathranar · 10-19-2004, 12:27 AM

Quote: Originally Posted by Ricky327

FYI
Intel PSU design recomendation should be within +-5%
So depending how your main/tank battery and alternator are feeling on the day you probably be fine.

So 11.4-12.6V. I suspect AMD's may be more finnickity than an Intel - especially the cheap Soyo m/b I hope to use. We'll see.