LM317 based SLA Charge & Float Circuit

[info]

Hello there fellow ee-enthusiasts!

I found this circuit here while looking for a replacement of my chinese brand SLA-Charger decided to go for a smoke.



Source:https://electronicsforu.com/electronics-projects/hardware-diy/12v-battery-charger-absorb-float

I modified it a bit since I was aiming for about a charge current of 2A for a 24AH 12V battery (charging is to be @0.1C) like this:



Now I have a couple of questions:
- could I have made this easier than the suggested circuit? I haven't seen one that you can actually set to charge and when charge terminates, a.k.a. voltage hits ~14.1V, it goes over to float
- Is a diode really necessary across the input & output from output towards input (dotted lines)
- I copied the 0.7V virtual ground for the LM358 off some circuit I saw here on the forum. Is that a cheap and dirty way of allowing the OP-Amp to go down to zero?
- Perhaps a bit of critique on my schematic. I do want to improve overall, so please don't hesitate.


Will that work as per the datasheet from TI http://www.ti.com/lit/ds/symlink/lm317.pdf on page 17?

[MrAl]

Hi,

It does not have to be that complicated.  A LM317 with a few parts werks well too without that extra LM chip.
I can find the design if you need it but it's just a LM317 set up as voltage regulator with series resistor to get a taper charge.

[Seekonk]

Just use a TL431 to light the LED

[0xfede]

Hello info,

there are a few things that should be considered:

Now I have a couple of questions:
- Is a diode really necessary across the input & output from output towards input (dotted lines)

The diode is there to protect LM317 that may die if Vo is much higher than Vi.

- I copied the 0.7V virtual ground for the LM358 off some circuit I saw here on the forum. Is that a cheap and dirty way of allowing the OP-Amp to go down to zero?

With this schematic the LM358 will never go to 0V. Instead it will go to Vforward(D3) that is about .6V. Simply short D3 and it will go to 0volt.

- Perhaps a bit of critique on my schematic. I do want to improve overall, so please don't hesitate.

You modified the original schematic to obtain more current but D4 is just a 1A diode and probably it will die. Furthermore there is no current regulation and you will probably end with a much higher current than 2A when the battery is empty. The original design rely on the ability of the LM317 to limits its current and, since you added an external series pass transistor (Q1), you have lost the current limiting capability.

Note that this circuit is not protected if the battery is connected in reverse polarity since the supply of the LM358 is taken directly from the battery itself.

As others said there are many different and easier ways to accomplish a SLA charger but if you want to continue to this route you need a bit more refinement of your project.

Best,
0xfede

[MrAl]

Here is a simpler schematic...

This was an Li-ion charger so you need to modify a few things.  Not much though.
Of course you adjust the output to over 12v, like maybe 13.9v or a little higher.

[info]

Just use a TL431 to light the LED

Right, but that's not really my priority tbh.

Hi,

It does not have to be that complicated.  A LM317 with a few parts werks well too without that extra LM chip.
I can find the design if you need it but it's just a LM317 set up as voltage regulator with series resistor to get a taper charge.

Do you mean the current limited circuit in the LM3XX datasheets? with one resistor on the output and adjust connected after the resistor?

Hello info,

there are a few things that should be considered:

Now I have a couple of questions:
- Is a diode really necessary across the input & output from output towards input (dotted lines)

The diode is there to protect LM317 that may die if Vo is much higher than Vi.

- I copied the 0.7V virtual ground for the LM358 off some circuit I saw here on the forum. Is that a cheap and dirty way of allowing the OP-Amp to go down to zero?

With this schematic the LM358 will never go to 0V. Instead it will go to Vforward(D3) that is about .6V. Simply short D3 and it will go to 0volt.

- Perhaps a bit of critique on my schematic. I do want to improve overall, so please don't hesitate.

You modified the original schematic to obtain more current but D4 is just a 1A diode and probably it will die. Furthermore there is no current regulation and you will probably end with a much higher current than 2A when the battery is empty. The original design rely on the ability of the LM317 to limits its current and, since you added an external series pass transistor (Q1), you have lost the current limiting capability.

Note that this circuit is not protected if the battery is connected in reverse polarity since the supply of the LM358 is taken directly from the battery itself.

As others said there are many different and easier ways to accomplish a SLA charger but if you want to continue to this route you need a bit more refinement of your project.

Best,
0xfede

Gotcha. Will remove the diode D3 and D4 is an MBR1550CT, I didn't have a footprint of it so  I just used the 1N4007 as placeholder. Otherwise, I could also use STPS2045C which I salvaged from a dead SMPS. Either should handle the 2A I was aiming for but then I am reconsidering adding a current limit to the circuit. Will update the schematic and post.

Here is a simpler schematic...

This was an Li-ion charger so you need to modify a few things.  Not much though.
Of course you adjust the output to over 12v, like maybe 13.9v or a little higher.

Yes, I thought of going down that road but I wanted something that could fully charge and/or float in one circuit. I know I could set the LM317 to a voltage of 13.6V and leave the battery attached for a permanent float but that doesn't really charge up the battery to full capacity which is what I was aiming for.


I chose this route because:
a. I can charge the battery fully and
b. If it's fully charged, it can be *theoretically* floated indefinetly

I will change the circuit and post the newer version for feedback

[Seekonk]

I fail to see how this circuit works at all as intended.

The 317 can put out two voltages.

The opamp senses the voltage on the regulator, not the battery.

When it senses the higher voltage it switches to the lower voltage.

when it senses the lower voltage it switches back to the higher voltage.

It never stops. It ought to latch in some way.

Only thing I can see is it operates by using the internal limiting of the 317 which is bad design.

And yes, the 431 could replace the op amp.

[woodchips]

Yes, a 0.5 ohm or something resistor on Vout with the adjust pin and battery connected after the resistor. It is an amazingly subtle circuit.

More current then more 317's.

[Seekonk]

If anyone gets that circuit to work, tell me. 

[MrAl]

Just use a TL431 to light the LED

Right, but that's not really my priority tbh.

Hi,

It does not have to be that complicated.  A LM317 with a few parts werks well too without that extra LM chip.
I can find the design if you need it but it's just a LM317 set up as voltage regulator with series resistor to get a taper charge.

Do you mean the current limited circuit in the LM3XX datasheets? with one resistor on the output and adjust connected after the resistor?

Hello info,

there are a few things that should be considered:

Now I have a couple of questions:
- Is a diode really necessary across the input & output from output towards input (dotted lines)

The diode is there to protect LM317 that may die if Vo is much higher than Vi.

- I copied the 0.7V virtual ground for the LM358 off some circuit I saw here on the forum. Is that a cheap and dirty way of allowing the OP-Amp to go down to zero?

With this schematic the LM358 will never go to 0V. Instead it will go to Vforward(D3) that is about .6V. Simply short D3 and it will go to 0volt.

- Perhaps a bit of critique on my schematic. I do want to improve overall, so please don't hesitate.

You modified the original schematic to obtain more current but D4 is just a 1A diode and probably it will die. Furthermore there is no current regulation and you will probably end with a much higher current than 2A when the battery is empty. The original design rely on the ability of the LM317 to limits its current and, since you added an external series pass transistor (Q1), you have lost the current limiting capability.

Note that this circuit is not protected if the battery is connected in reverse polarity since the supply of the LM358 is taken directly from the battery itself.

As others said there are many different and easier ways to accomplish a SLA charger but if you want to continue to this route you need a bit more refinement of your project.

Best,
0xfede

Gotcha. Will remove the diode D3 and D4 is an MBR1550CT, I didn't have a footprint of it so  I just used the 1N4007 as placeholder. Otherwise, I could also use STPS2045C which I salvaged from a dead SMPS. Either should handle the 2A I was aiming for but then I am reconsidering adding a current limit to the circuit. Will update the schematic and post.

Here is a simpler schematic...

This was an Li-ion charger so you need to modify a few things.  Not much though.
Of course you adjust the output to over 12v, like maybe 13.9v or a little higher.

Yes, I thought of going down that road but I wanted something that could fully charge and/or float in one circuit. I know I could set the LM317 to a voltage of 13.6V and leave the battery attached for a permanent float but that doesn't really charge up the battery to full capacity which is what I was aiming for.


I chose this route because:
a. I can charge the battery fully and
b. If it's fully charged, it can be *theoretically* floated indefinetly

I will change the circuit and post the newer version for feedback

Hello again,

Well that circuit is for reference, you can adjust the output to whatever is appropriate for your battery.
If you want to charge at 14.4v that is ok too, and then later turn down to 13.8v.
If you want it to be automatic, then we'd have to add a circuit to switch it.