# SLA charger

#### k7elp60

Joined Nov 4, 2008
562
In the past I have built a number of 12V lead-acid battery chargers. Most of them were used on SLA type at or near room temperature. I decided to try building one with built in temperature compensation. I remember copying the attached schematic from this website.
I built it today and I am impressed with the temperature compensation provided by the 2N3904(Q2) I have a fluke 87-V with a bead temperature probe and used a fluke 8062 DVM to measure the voltage. I fastened the bead probe and the 2N3904 to a 20W 20 resistor, with black cable ties.
Room temperature was 32.2° C and the output voltage was adjusted to 13.65V. I applied 6.1VDC @ .29A to the resistor and monitored the temperature rise. The results were 42.0 C output voltage 13.46. 47.1° and output voltage 13.35V. My calculations show that the average change in charge voltage for a float charger/C° is 20mV.

My question is: Is this the recommended change in charge voltage for the SLA battery?

#### Attachments

• 127.2 KB Views: 80

#### SgtWookie

Joined Jul 17, 2007
22,221
It's approximately -3mV per cell per °C offset from 25°C.
So, with a 6 cell battery, you would need about -3mV * 6 = -18mV per °C offset from 25°C, with the float being ~13.65v @ 25°C. See the datasheet for your battery to confirm, but as a general rule this is good for lead-acid batteries.

You could "tweak" the thermal response by adjusting the ratios of R5 and R3, but already the circuit is much closer to what you should have vs what a non-compensated charger would put out.

You need to thermally couple Q2's package to the battery positive terminal for best results.

Ambient temperature is not important; the battery core temperature is.

A very similar version to that circuit you posted is attached; I later added an isolation diode so that if the charger were powered off, it would not drain the battery. It makes the charge taper off more slowly when nearly fully charged, but that's not all bad - eventually it reaches the float voltage.

#### Attachments

• 68 KB Views: 153
Last edited:

#### k7elp60

Joined Nov 4, 2008
562
It's approximately -3mV per cell per °C offset from 25°C.
So, with a 6 cell battery, you would need about -3mV * 6 = -18mV per °C offset from 25°C, with the float being ~13.65 @ 25°C. See the datasheet for your battery to confirm, but as a general rule this is good for lead-acid batteries.

You could "tweak" the thermal response by adjusting the ratios of R5 and R3, but already the circuit is much closer to what you should have vs what a non-compensated charger would put out.

You need to thermally couple Q2's package to the battery positive terminal for best results.

Ambient temperature is not important; the battery core temperature is.

A very similar version to that circuit you posted is attached; I later added an isolation diode so that if the charger were powered off, it would not drain the battery. It makes the charge taper off more slowly when nearly fully charged, but that's not all bad - eventually it reaches the float voltage.
Thank you very much Sgt. Wookie, I really appreciate your comments and your design. Your post has confirmed what I thought after doing a lot of research. I generally put a schottky diode in series with the output for the reason you posted.
Thankyou again.

#### THE_RB

Joined Feb 11, 2008
5,438
At the risk of sounding like a butcher, I charge all mine with a regulated DC supply at 13.8v and long leads which have about 0.5 ohms resistance.

They draw less than an amp first up, then the current slowly drops to maybe 200mA over a couple of hours and I leave them on overnight, as much as 8 to 10 hours.

#### k7elp60

Joined Nov 4, 2008
562
At the risk of sounding like a butcher, I charge all mine with a regulated DC supply at 13.8v and long leads which have about 0.5 ohms resistance.

They draw less than an amp first up, then the current slowly drops to maybe 200mA over a couple of hours and I leave them on overnight, as much as 8 to 10 hours.
I don't you are a butcher. Your way of charging your batteries seems okey if the ambient temperature doesn't vary in large excursions or the batteries are less than 2A capacity. I am interested in the temperature compensation as I am working on a project to charge a battery where the outside temperatures vary from -5 to +52C°.

On another subject I went to your website and found the alternative energy grid tie information very interesting. I added it as an attachment so other forum members would have an idea what we are talking about.
Anyway I have several questions for you in this regard:
Is the transformer a standard 120V to 24V CT type transformer?
Would a higher current like 2A work?
Where would the best place to connect the output to the mains?

#### Attachments

• 68 KB Views: 55
Last edited:

#### k7elp60

Joined Nov 4, 2008
562
It's approximately -3mV per cell per °C offset from 25°C.
So, with a 6 cell battery, you would need about -3mV * 6 = -18mV per °C offset from 25°C, with the float being ~13.65v @ 25°C. See the datasheet for your battery to confirm, but as a general rule this is good for lead-acid batteries.

You could "tweak" the thermal response by adjusting the ratios of R5 and R3, but already the circuit is much closer to what you should have vs what a non-compensated charger would put out.

You need to thermally couple Q2's package to the battery positive terminal for best results.

Ambient temperature is not important; the battery core temperature is.

A very similar version to that circuit you posted is attached; I later added an isolation diode so that if the charger were powered off, it would not drain the battery. It makes the charge taper off more slowly when nearly fully charged, but that's not all bad - eventually it reaches the float voltage.
SgtWookie, I am trying to really understand the charger ckt. I have a few questions for you:
If I wanted to charge a 6V battery would I change the value of R5 to 560Ω?
I understand the base emitter junction of a NPN transistor has a temperature coefficient forward voltage of -1.8mV/C°. It is the ratio of R3 and R5 that gives the 18mV/C° for the 12V battery?
How do you recommend fastening Q2 to the battery for both temporary and permanent installations?

#### SgtWookie

Joined Jul 17, 2007
22,221
SgtWookie, I am trying to really understand the charger ckt. I have a few questions for you:
If I wanted to charge a 6V battery would I change the value of R5 to 560Ω?
Yes, and the pot R2 would need to be reduced to ~100 Ohms, and be initialized to around 20%.
I understand the base emitter junction of a NPN transistor has a temperature coefficient forward voltage of -1.8mV/C°. It is the ratio of R3 and R5 that gives the 18mV/C° for the 12V battery?
Well, the BE junction does not have a tempco of -1.8mV/°C; just the circuit consisting of Q2, R3 and R5 is a "Rubber Zener" aka "VBE Multiplier". The resistors R3, R5 starting ratio was selected as VBB = VBE {(R5+R3)/R3}; knowing that we needed a certain voltage between the bottom of R2 and B-; and then iteratively trying standard E24 resistor values to see what the best "fit" was for the tempco of the transistor vs what the desired tempco was. Like I mentioned above, the values I came up with did not result in an exact tempco match, but it's a 90% improvement over a non-tempco'ed charger. If someone wishes to go through more iterations and get a better tempco match, they are more than welcome to do so.
http://www.st-andrews.ac.uk/~www_pa/Scots_Guide/audio/part2/page3.html
http://www2.engr.arizona.edu/~brew/ece304spr07/Pdf/VBE Design.pdf
As you can see, VBE multipliers are quite popular for use in audio circuits.

How do you recommend fastening Q2 to the battery for both temporary and permanent installations?
You might use some JB-Weld to attach the face of the transistor to a metal tab to connect to the battery terminal, or epoxy it directly to the battery terminal. Those 2N3904 transistors are cheap; a few pennies each. Even if you go to Radio Shack and buy the 15 NPN transistor assortment for $2, you generally get five 2N3904's, so they'd be around$0.14/ea.

Make certain to use heat shrink tubing on the transistors' leads so they don't get shorted out.

Last edited:

#### THE_RB

Joined Feb 11, 2008
5,438
...
On another subject I went to your website and found the alternative energy grid tie information very interesting. I added it as an attachment so other forum members would have an idea what we are talking about.
Anyway I have several questions for you in this regard:
Is the transformer a standard 120V to 24V CT type transformer?
Would a higher current like 2A work?
Where would the best place to connect the output to the mains?
Well that was a simple sketch of an idea for a topology, ie a "method" for simple minimalist grid tie inverter that might be worth exploring.

No, the transformer is a high frequency type as used in SMPS mains offline designs, and it is designed to supply a regulated current back into the mains over half a mains cycle. The whole system regulates at a high freq (SMPS) type control.

It is an unbuilt and yet unproven design, and like any mains connected electronics it would be dangerous to experiment with for anyone that is not skilled in that field! Also the policy on this forum is to ban discussion of direct mains connected electronics and I will comply with that policy by refusing to discuss it further on this forum. Sorry!

#### k7elp60

Joined Nov 4, 2008
562
Well that was a simple sketch of an idea for a topology, ie a "method" for simple minimalist grid tie inverter that might be worth exploring.

No, the transformer is a high frequency type as used in SMPS mains offline designs, and it is designed to supply a regulated current back into the mains over half a mains cycle. The whole system regulates at a high freq (SMPS) type control.

It is an unbuilt and yet unproven design, and like any mains connected electronics it would be dangerous to experiment with for anyone that is not skilled in that field! Also the policy on this forum is to ban discussion of direct mains connected electronics and I will comply with that policy by refusing to discuss it further on this forum. Sorry!
I undstand no need to applogize, thanks.

#### k7elp60

Joined Nov 4, 2008
562
Yes, and the pot R2 would need to be reduced to ~100 Ohms, and be initialized to around 20%.

Well, the BE junction does not have a tempco of -1.8mV/°C; just the circuit consisting of Q2, R3 and R5 is a "Rubber Zener" aka "VBE Multiplier". The resistors R3, R5 starting ratio was selected as VBB = VBE {(R5+R3)/R3}; knowing that we needed a certain voltage between the bottom of R2 and B-; and then iteratively trying standard E24 resistor values to see what the best "fit" was for the tempco of the transistor vs what the desired tempco was. Like I mentioned above, the values I came up with did not result in an exact tempco match, but it's a 90% improvement over a non-tempco'ed charger. If someone wishes to go through more iterations and get a better tempco match, they are more than welcome to do so.
You might use some JB-Weld to attach the face of the transistor to a metal tab to connect to the battery terminal, or epoxy it directly to the battery terminal. Those 2N3904 transistors are cheap; a few pennies each. Even if you go to Radio Shack and buy the 15 NPN transistor assortment for $2, you generally get five 2N3904's, so they'd be around$0.14/ea.