[Moved] SLA Battery charger for 7 AH 12 V battery powering 2 routers

[Macrophage]

Greetings.

Sorry for the weird thread title, didn't know how else to explain it

I've recently been pretty annoyed by power outages in my area, and that's what prompted me to install a backup power source for my router and modem.

1) First off, the modem and router are each rated for 600 mA draw at 9 volts.

2) The battery I'm working with is a 7 AH 12 volt Leoch SLA battery.

3) I have a wall-wart that's rated for 1 ampere at 12 V (actually outputs around 16-18 volts without a load attached).

4) The backup only has to last for a couple of minutes at most, because my generator kicks in by then. So the battery will spend most of it's life fully charged.

Now, I was looking at this circuit: https://www.circuitstoday.com/lead-acid-battery-charger , and it seems to be what I'm looking for. Simple, and easy enough to assemble.

Also, both my router and modem have inbuilt switching voltage regulators. One has an mp1482 (which can take a maximum input of 18 volts) and the other has an fr9888 (max input 23 V).

On to the questions:

1) This means I should be fine powering the router and modem directly from the battery, right?

2) Would powering the router and modem directly from the battery be right way to do it, since the wall-wart power source I'm using is rated for just 1 ampere?

3) Are there any improvements you would suggest?

All responses will be greatly appreciated.

 

[KlausST]

Hi,

one problem can be "isolation"

when using individual wall-warts the routers are isolated from each other because the wall-warts have built in isolation to mains supply.

when using a battery as common supply for both routers then they are not isolated from each other.

Klaus

 

[Macrophage]

Oh! Thanks KlausST. Would that be solved by installing two lm7809's regulator circuits to the battery and powering the devices separately?

And would isolation be that big of a problem anyway?

 

[KlausST]

Hi,

Would that be solved by installing two lm7809's regulator circuits to the battery and powering the devices separately?

No, linear voltage regulators are non isolating.

And would isolation be that big of a problem anyway?

I can´t give you an answer to that. It depends on the devices and their connections.

Usually ethernet data connections are isolated with transformers.

But what about shielding of cables?
And maybe inside the routers the "-" is connected to the case, and the case is mounted into racks, then this may cause ground loops.
It is not predictable. I´d say it is hardly probable, but it may cause problems.

Klaus

 

[Macrophage]

Oh, OK. Well, once I've set it all up, I'll report back whether I'm experiencing isolation issues. It's unlikely though, since the routers don't have metal bodies (they're regular plastic-body routers).

If anyone has other improvements in mind, I'd love to hear them.

 

[red_alert]

For the isolation sake, you could build the simplest flyback/forward converter with two (identical?) secondary windings. All you need is a cheap MOSFET switch, a small ferrite core and a LM555 as a PWM generator.

Of course, you also need two rectification diodes and capacitors but it's an economical solution anyway.

 

[harvie]

I don't think that you can run in any isolation issues. We do exact same thing in ISP company. We are powering multiple consumer and profesional grade equipment using single supply. Never had problems in years. I don't understand where exactly the ground loop should happen in such case.

And most of SMPS wall bricks does not have any isolation. I've learned that hard way when sparks started to appear between tip of my soldering iron and DC powered hi-fi amplifier which had power switch only on + side.

 

[KlausST]

Hi,

And most of SMPS wall bricks does not have any isolation.

For sure they have a safety isolation. But i know what you mean.
I know about the sparks. And measuring the wallwart outputs against earth you will see about half of mains voltage.

This effect is often caused by wall warts without earth connection. Inside they usually have a capacitive connection from both mains inputs to gnd output. Often there is a high ohmic resistor in parallel, too.
If you touch the output there will flow current through your body to earth (in an earthed mains system), you can even measure it with a current meter. But it is limited to a safe value.
Caused by the capacitors ( they must be safety Y-types) you have a short but high inrush current, therefore the spark you speak of. You even may hear a quiet sound of the spark when using low ohmic (metallic) connetion to earth.

You might sense the current with your fingers, especially when feet and fingers are wet. AVOID TO TEST IT WITH YOUR TONGUE, although it won' t kill you, it might be painful.

Klaus

 

[D.A.(Tony)Stewart]

Power both from the battery should be fine. Use heavy wiring. 2mm

As there is no need to fast charge the battery with CC, then CV then float , you can just use an adjustable regulator at the float voltage of 13.5V.

Most old LDO's are bipolar with 2.5V dropout (Vin-Vout min) whereas the best FET LDO's (modern ones) are < 0.5V dropout, so choose an adjustable 13.5V design with 1A rating minimum or 13.5W out.

If routers take 600mA @9V or 4.5W at 12V they should draw under 400mA each.

The 12V 1A is very marginal as the 1A rating has insufficent voltage and 800mA may only be 12.5, which is enough to drive 9V LDO's and charge the battery with a very low LDO get >12V. That would not be full capacity but would work for short outages.

But 12V >2A may be sufficient. or a proper matched supply for the load + LDO drop.

Unregulated DC supplies are always 1.4x rated voltage with no load.

 

[harvie]

Also put polyfuses (600mA - 2A) in series with each router (as close to battery as you can), so short circuit will not result in router exploding with rain of sparks (battery can provide way more current than wallwart adapter, so it can make your wires evaporate on short circuit). also it will prevent failure of one router to put down the other one...

 

[D.A.(Tony)Stewart]

The LDO's will be current limited but fuses are called for by Murphy's Law.

 

[Macrophage]

Hello.

Just thought I'd update this post with what I did.

Didn't make any charging circuit or anything.

Simply hooked up both routers directly to the battery, and connected a 12 V unregulated wall wart to the battery terminals with a diode in between. The battery idles at 14.4-14.8 volts and the routers have been running without a problem for 2 months.

 

[D.A.(Tony)Stewart]

IF you can modify the Battery voltage for 14.2 or less, the electrolyte will last longer.
14.4-14.8 is too high. Also H2 outgassing is considered flammable if concentration level exceeds 5% in air.

 

[harvie]

Even 14.2V is too high for floating mode operation. Floating voltage for 12V SLAs is 13.6V - temperature compensation per cell*degree over 25C. See this:
https://en.wikipedia.org/wiki/Float_voltage

 

[Macrophage]

IF you can modify the Battery voltage for 14.2 or less, the electrolyte will last longer.
14.4-14.8 is too high. Also H2 outgassing is considered flammable if concentration level exceeds 5% in air.

Good advice, I'll try that. I have several 2 watt 2 ohm resistors lying around, I'll try splicing them in, and see if that makes a difference.

And does H2 outgassing occur even with SLA batteries?

Here's the battery I'm using:

**broken link removed**

 

[D.A.(Tony)Stewart]

Yes SLA's must have a safety vent for H2 outgassing from overcharging and if put into a sealed box, then a teflon plug is used to release H2.

If you have a zener or LED controlled power Q, then you can clip the peaks.

 

[Macrophage]

Yes SLA's must have a safety vent for H2 outgassing from overcharging and if put into a sealed box, then a teflon plug is used to release H2.

If you have a zener or LED controlled power Q, then you can clip the peaks.

Apologies for my ignorance, did you mean a device like an LM317?

 

[D.A.(Tony)Stewart]

No these LDO's are not really Low dropout

I meant using an active load. But if you have an automotive 12V lampule, that will work. ( looks like a fuse)

 

[Macrophage]

No these LDO's are not really Low dropout

I meant using an active load. But if you have an automotive 12V lampule, that will work. ( looks like a fuse)

Ah! I get it now. I actually removed the fan that's cooling the wall wart (in the picture I posted earlier), and that may be contributing to the higher voltage to the battery. I'll reattach the fan or another load, and then check the voltage after a few hours.

 

[Macrophage]

Hello.

An update. Everyday, I get around 3 hours of electricity drop-outs regularly. The peak that the battery reaches is about 13.8 volts. Is that OK, or should I try to lower this voltage further?