12V UPS (float) - voltage drop concern

[jst3712]

Hi, I’m still learning, so this may seem to be a dumb question to some of you, but I think it’s valid.

I am building a small “general purpose” 12-volt Uninterruptable Power Supply (UPS) that allows for any low current 12 volt device to be hooked up to it (up to 1A, depending on the battery and input power source/regulator).

I have an LM317 setup that keeps an SLA battery on float charge, and also powers the 12-volt device at the “Load Output” at the same time. Should the mains power fail, the battery keeps the load powered as a backup.

When the external load is switched off/not connected, voltage from the LM317 to the battery sits comfortably around 13.5V, which is normal for float charging. But connecting a load causes the voltage to the battery to drop down a fair bit (for example, from 13.5 to 13.0V). I know there’s got to be a voltage drop somewhere whenever a load is connected (and how much current it’s consuming), but, doesn’t that mean the battery will suffer? I was told a while back from somebody that the voltage drop is normal, and that the battery will be fine, as long as it is above 12.6V with the load connected (or somewhere along those lines), but I really would like someone to confirm that.

• Input: 18V DC (Regulated Switchmode type).
• Regulator: LM317 (1.5A); voltage output set to 14.2V (before series diode; 14.2 – 0.7 = 13.5).
• Current limiting resistor to the battery: 4.7R, 5W (up to 0.36A @ 11.8V)^^.
• Battery: 12V SLA, 2.2Ah (0.66A max charging current).
• Load (external device): Up to 1A (absolute maximum).

^^ When mains power fails, I have an N.C. relay contact that shunts/bypasses the 5W resistor so that there is minimal current loss from the battery to the load. Load will automatically be disconnected if voltage falls to 11.8V.

I would appreciate any assistance / feedback on my current design (see attachment).

JASON.

 

[crutschow]

The normal open circuit voltage of a fully charged 12V battery is 12.6V and as it discharges it can drop below that. There are many articles on the web about lead acid batteries such as Battery Tutorial | Batterystuff.com to help you understand their operation.

 

[jst3712]

The normal open circuit voltage of a fully charged 12V battery is 12.6V and as it discharges it can drop below that. There are many articles on the web about lead acid batteries such as Battery Tutorial | Batterystuff.com to help you understand their operation.

Hi Crutshow, thank you. I'm aware of the battery's "open circuit" voltage when it's full I've read a few articles about charging SLA's, but most of them don't talk about what happens when a load is connected in parallel to the battery when the charger is operating.

The battery is always connected to the charger (float charge)... without any load it's at 13.5V when the battery's full - equal to my charger's output voltage.

Then, connecting a load (12V device) causes the voltage to the battery to drop below the minimum float voltage of 13.5V... So are you saying even if it showed 12.6V on the multimeter it'd still be ok?

I just don't want my battery to become under-charged whenever a load is connected.

Thanks
JASON

 

[Leftyretro]

Hi Crutshow, thank you. I'm aware of the battery's "open circuit" voltage when it's full I've read a few articles about charging SLA's, but most of them don't talk about what happens when a load is connected in parallel to float chargers.

The battery is always connected to the charger (float charge)... without any load it's at 13.5V when the battery's full - equal to my charger's output voltage.

Then, connecting a load (12V device) causes the voltage to the battery to drop below the minimum float voltage of 13.5V... So are you saying even if it showed 12.6V on the multimeter it'd still be ok?

I just don't want my battery to become under-charged whenever a load is connected.

Thanks
JASON

"I just don't want my battery to become under-charged whenever a load is connected."

Hi Jason;

That's quite an elaborate design you have there, must be a very critical or expensive load to justify it? As far as becoming under-charged at floating at13V Vs 13.5V, I think it not that big a deal. It may mean that your battery is only holding say 96% of capacity Vs 100%. There is probably more charge capacity variation with normal ambient temperature variations.

You could always design a test to measure the actual battery discharge capacity with a resistive dummy load. Set/Adjust the regulator to 13.5 battery float voltage, let if charge to full and then disconnect the AC and measure how long the load is powered before your voltage monitor switches the load off. Then adjust the regulator back to it's normal value of 13V at the battery, let fully charge, and time it again. You will then know how much time the .5 volt drop take off.

Lefty

 

[jst3712]

"I just don't want my battery to become under-charged whenever a load is connected."

Hi Jason;

That's quite an elaborate design you have there, must be a very critical or expensive load to justify it? As far as becoming under-charged at floating at13V Vs 13.5V, I think it not that big a deal. It may mean that your battery is only holding say 96% of capacity Vs 100%. There is probably more charge capacity variation with normal ambient temperature variations.

You could always design a test to measure the actual battery discharge capacity with a resistive dummy load. Set/Adjust the regulator to 13.5 battery float voltage, let if charge to full and then disconnect the AC and measure how long the load is powered before your voltage monitor switches the load off. Then adjust the regulator back to it's normal value of 13V at the battery, let fully charge, and time it again. You will then know how much time the .5 volt drop take off.

Lefty

Hi Lefty. No, the load is not expensive, haha; it's just that I want the UPS to be reliable - no matter what I have connected - otherwise there is little point in using a poorly charged battery when the AC mains fails, right?

Oh, I have allowed my battery to become as full as possible before I connected a load in parallel, but the voltage still drops instantly as soon as a load is connected (with the charger still connected). It must be normal???

 

[Nigel Goodwin]

Oh, I have allowed my battery to become as full as possible before I connected a load in parallel, but the voltage still drops instantly as soon as a load is connected (with the charger still connected). It must be normal???

Have you not heard of ohms law?, as soon as you draw current the voltage drops - how much depends on the resistance (internal and external) and the current drawn.

How much does the voltage drop, what current are you drawing (and how big is the battery).

 

[jst3712]

Have you not heard of ohms law?, as soon as you draw current the voltage drops - how much depends on the resistance (internal and external) and the current drawn.

How much does the voltage drop, what current are you drawing (and how big is the battery).

Of course I have, but unfortunately I don't have a lot of knowledge when it comes to two power sources connected in parallel with a load that could be anywhere up to 1A, which is why I don't know if it's ok for the battery or not.
See, I told you it might be a dumb question!

Well my "Test" load is a 3W 12V bulb that draws 200-250mA. Before it was connected, meter showed 13.55V. Connected the bulb, it dropped down to about 13.38V. But the load could draw anything up to 1A (absolute maximum) depending on what I decide to hook up to it (hence the name 'General Purpose UPS'), so I need to make sure the battery won't suffer, especially with slightly bigger loads, consequently bigger voltage drops. Float = 13.5-13.8V.
My battery is a 2.2Ah.

 

[Nigel Goodwin]

It's a pretty small battery, that tiny a drop sounds quite likely.

 

[jst3712]

It's a pretty small battery, that tiny a drop sounds quite likely.

So will it be ok or not? (I am more concerned about a bigger load connected, which will probably cause a larger voltage drop than that).

When should I start to worry?

Thanks.

 

[Nigel Goodwin]

So will it be ok or not? (I am more concerned about a bigger load connected, which will probably cause a larger voltage drop than that).

Again, refer to ohms law - it's dead simple - more current will cause a bigger voltage drop.

When should I start to worry?

When it becomes a problem.

 

[jst3712]

Again, refer to ohms law - it's dead simple - more current will cause a bigger voltage drop.


When it becomes a problem.

That's it?

You can't make any recommendations at all?

You see, I'd rather tackle a potential problem now, not later when I find out my battery is dying far too quick.

You are saying the more current drawn, the bigger voltage drop... fine, I get it. BUT, how is that going to affect the float charging process? Or is it?

 

[Leftyretro]

You are focusing too much on the battery float voltage value. The more important value is what is the battery terminal voltage when it is supplying the load current after AC power is removed. Most lead acid batteries will quickly decrease to 12.6 volts as it starts to supply the load and gradually decrease in terminal voltage as it's discharging into the load. You already have a voltage monitor function that disconnects the battery at some voltage value (at what value?) and that is good. So your circuit is OK in concept and should not cause any damage to the battery or the load.

So your question should be is my device charging the battery to 100% of capacity or not? You can only really determine that by load testing after you have completed construction. At 2.2 AH rating you should expect to be able to supply 1 amp load for around 2.2 hours give or take depending on your low voltage monitor cutoff. Testing is part of design and constructing projects. A good design should show fewer problems when first testing, but designs are seldom 100% perfect so actual circuit testing and measurement should be the final validation of a design.

Lefty

 

[Nigel Goodwin]

That's it?

You can't make any recommendations at all?

It depends on exactly what it's been used for, and how long it's expected to last - but a bigger battery would help all round.

If you're drawing 1A for a 2.2AH battery, I doubt it would last an hour - the battery ratings are normally at a far lower discharge rate.

You see, I'd rather tackle a potential problem now, not later when I find out my battery is dying far too quick.

You are saying the more current drawn, the bigger voltage drop... fine, I get it. BUT, how is that going to affect the float charging process? Or is it?

If the battery if charging, then there's no current been drawn from it, so iy shouldn't have any effect.

 

[jst3712]

You are focusing too much on the battery float voltage value.

Yeah probably, haha.

The more important value is what is the battery terminal voltage when it is supplying the load current after AC power is removed. Most lead acid batteries will quickly decrease to 12.6 volts as it starts to supply the load and gradually decrease in terminal voltage as it's discharging into the load. You already have a voltage monitor function that disconnects the battery at some voltage value (at what value?) and that is good. So your circuit is OK in concept and should not cause any damage to the battery or the load.

Probably 11.8V.

So your question should be is my device charging the battery to 100% of capacity or not? You can only really determine that by load testing after you have completed construction. At 2.2 AH rating you should expect to be able to supply 1 amp load for around 2.2 hours give or take depending on your low voltage monitor cutoff. Testing is part of design and constructing projects. A good design should show fewer problems when first testing, but designs are seldom 100% perfect so actual circuit testing and measurement should be the final validation of a design.

Lefty

Thanks Lefty. See red text.
It sounds like I won't really need to ammend my schematic very much then. I'll definately perform some tests with various loads with AC disconnected.

Jason

 

[jst3712]

It depends on exactly what it's been used for, and how long it's expected to last - but a bigger battery would help all round.

If you're drawing 1A for a 2.2AH battery, I doubt it would last an hour - the battery ratings are normally at a far lower discharge rate.

Well I have 2 projects that will utilise this UPS design... one for emergency LED lighting (using a 2.2Ah; maybe 0.5A would be max... not sure yet), and another for an alarm in the garage (1.3Ah; somewhere between 0.2 and 0.5A).
I'm using smallish batteries because of their perfect physical shape and size... I know they won't last for that long.

The reason why I mentioned "1A" earlier because maybe further down the track I might want to connect something that consumes a little more current than the above-mentioned projects. However, I'd most likely be looking at a higher capacity battery and better regulator.

 

[fsmyth]

For a proven design, see:
Power Supply Schematics - page 5
Works for 12v apps, also. Scale values accordingly.
<als>