Wire size to charging battery?

[oddbudman]

Interesting reading. I may as well explain a little about what i do.

I design electronic products, many of which use battery chargers.

My most recent product involves a Microprocessor controlled 3 stage Buck converter for charging SLA batteries. Using the processor is great as you can easily set the battery charging parameters using software.

I'm happy to see you did the tests oldboot, as you got to learn something new .

Be interested to read your circuit suggestion TheRealAndy. Here's a simple one i thought was quite nifty without going into dc:dc conversion.

http://www.siliconchip.com.au/cms/A_109625/article.html


oddbudman

[oldboot]

I am a technician...I learn therfore I am.....The day I cant learn something new is the day I need to curl up & die.

Designing a simple workable charger isnt too dificult a thing where you have plenty of voltage to spare, give me a good solid 18V dc & its a straight application notes job.

The problem here is that we are already at ideal charge voltage when the motor is running.......when the motor stops so does all effective charge.

Even putting a diode into the situation brings you back to 13.5V........ the only reasonable option would be a fet based, switching, series pass sort of thing & that is starting to get complicated.

I had a look at that note book application & it looks over complicated to me for what it achieves.
I can take or leave the series diode..... low & behold there is a 3.3 ohm series resistor & another switched to double the charge current.
I dont see any point in reducung charge current as the battery nears charge.... it will do that it's self.

Why not just select a resistor for maximum permisable initail charge current or a bit less and let nature take its course.
I cant see why have series resistors and a PTC thermistor....unless it was glued to the battery as an over temp device

It occurs to me that the circuit would not be able to deliver the permisable 2.5 amps for a fast charge.
In theory the two 3.3 ohm resistors should limit the charge current to about 2 amps assuming battery at 10v and charge terminal at 13.8..... but in reality the treminla voltage of the battery will rise rapidly in the first few minutes & the charge current will end up being more like 600mA after not very long.
so you are going to take at least 12 hours to charge the battery..... that is a long drive in any terms

The use of a PTC thermistor as a series limiting eliment (this one IS a limiter) would seem worthwhile.... however i've had bad experiences with these things.
They arent fast....in fact they are slow as a wet week
their spread of tolerance particularly with temperature is very wide
and they come in a limited range of values.
And they are expensive for what they are.

The more I look at a properly selected resistor, the better it looks.... a 5 or 10 watt resistor will cost watt $1- $1.50 if you pay retail.....Crikey..I've paid more than that for 2 crimp lugs in a time of despiration.

In reality how often do we need to charge a SLA in a vehicle......most of the time we can charge on 240......and there are plenty of good cheap chargers available.
It would be often simpler & cheaper to take either a larger cell or a couple of charged replacements.

Now if a dc to dc converter allowed charging from a stationary larger battery....AHH now that is a lot more use.

cheers

[driveon]

Hi old boot, with this type of battery you will find the light bulb does work and works better than a resistor.

I started working with domestic security installations a little over 30 ago and at that time it depended on what brand of security panel you bought as to whether you got a resistor or a globe as the charge regulator.

Within the next couple of years every panel you bought had a globe fitted as the charge regulator.

There are a number of advantaged to using a globe but the main one over a resistor, relating to this thread and some findings you have posted, is that you don’t have to match the globe to the size of battery. One size globe fits all size batteries up to say about 15 A/H.

The one thing that you do have to take into consideration is that a globe will take longer to fully charge the battery but this is intended as the battery can be left indefinitely connected to the power supply.

This is the main reason resistors were eventually phased out. If the resistor was not matched exactly, based on both the battery size and the charging voltage, the battery was either never fully charged or was over charged and in both cases, the battery had a short life.

Using a globe simply gets around all these problems, particularly when relating to the varying voltages that are going to be a common place in the type of situation being covered in this thread.

Cheers

[oddbudman]

driveon, you raise some good points about using a globe,

The globe would have a positive temperature coefficient. This would be useful in both the high current stages of charging and the low current stages towards the end of charge.

Also each globe is likely to have a reasonable level of tolerance.

I would be interested to know the charge voltage used in the security systems as well as the globes that are used.

oldboot,

I'm interested to know what type of technician you are? You are right about that circuit not lasting very long in the high current stage. With a charging voltage of 13.8V, its likely to switch over at around 12.7V on the sla battery.

cheers,
oddbudman

[driveon]

Hi oddbudman, I can give you exact info because from 1988 to 2000, I manufactured my own Security panels.

My panels were specifically designed for domestic installations and as such were designed to use very little power as compered to the commercial panels that were and still are available.

Commercial panels are designed to last between 8 and 24 hour on their backup battery, mine were designed to last between 1 and 2 week on the same size 7 A/H 12v backup battery.

The panel was supplied by a two wire plug pack that supplied 15 VAC to a rectifying bridge and then to a 12 DC volt regulator, tricked up to 13.7 DC volts.

This 13.7 DC supplied power to the detectors and siren driver and to a 5 volt reg that powered the PCB which included a wireless remote control receiver.

13.7v was also fed via a 5w 24v festoon globe, to the 7 A/H backup battery.

A low battery would take up to a week to fully recharge but as posted earlier, was connected permanently to the power supply and as many of the systems I not only made but also fitted, are still powered by the original batteries, the globe set up obviously works.

Cheers.

[oldboot]

my background is orriginaly Telecom working in an obscure section that dealt with lots of weird stuff including modems when they were as big as 2 phone books.
But thats a long time ago. They called it electronics but I seemd to do a lot of electro mechanics as an apprectice.

I've done all sorts of stuff, security , intercoms, mining phones, nurse call, fire evac, parking equipment TV antenna / matv, closed circuit TV, wired specilaised truck bodies, and lots of stuff no one else would touch.
These days I work mostly in pro audio and commercial AV, & do a bit of related timber & metal work.......the weirdest thing Ive done recently is some electro mechanical support for an Illusion builder.


Yeh I remember servicing some of those early alarm panels, ( OH there was some horrible stuff) they were notorious for being hard on batteries and not comming up to charge quick enough after black outs...( remember the SEQEB dispute).
Fortunatly all the recent stuff arround my time was 3 terminal reg based & charged at 1 to 1.5 amps & relied on the fold back current limiting on the chip to keep the plug pack in one piece. That is were all the panels sit today...straight simple application notes 3 term reg. That configuration will take a full 10 hours to come up back to full charge from flat.

drive on, you describe precisely what I don't like about the light bulb thing.....slow charging..... unless you specificlay select the bulb for the battery AND the the lamp will be very large.
If you want a slow charge that will work with anything, you might as well spend under a dollar an a 3.3 ohm resistor.

Light bulbs were very popular in the sixties, seventies and early eighties for all sorts of uses...It was a cheap dynamic eliment when we didn't have VCA's and a whole pile of other components we have now.
But all of these designs required some fiddling with different lamps to get good results.

My favorite use is speaker protection, 50 watt 8 ohm speaker, 12 V 21 W lamp in series.....near impossible to blow it up even with a 500 watt amp....Why is there a light shining inside me speaker???

The problem in this application is that unless you use a low voltage high wattage lamp it is not going to function dynamicaly....the best part its postiive temperiture coeficient is when the lamp glows a bit more than half bright under load.


Also a low voltage high wattage globe is needed to achieve a low enough cold resistance to get a practical charge....Id be thinking a 6 volt 18 watt globe would be probly of use... but it would need testing.

The other issue with using lamps as a dynamic ( quasi active) component is that we have far less variety of lamps easily available to us now than we did 20 years ago, and frankly the quality and reliability is very poor these days.

In this situation we don't have a full week to charge a battery......we'd be back home by then...... we need something that will give us some usefull charge in the hour or so run back to camp or along the highway to the next days fishing.

The fastest practical charge with SLA is 4 hours from flat at the 1/3 rate.

If the item is recharged long before it is flat then we are in the race for the short drive or boat run.

cheers

[driveon]

Hi oldboot, I don’t know what your going on about, having to match the globe. The whole point of using a globe is that you DON’T have to match it, one globe fits all, no matter what size globe it is.

The lower the battery is, the higher the voltage difference across the globe, the greater the current flowing through the globe.

As the battery charges, the voltage difference reduces and so does the amount of current flowing through the globe.

Varying the wattage of the globe will vary the amount of current going to a low charged battery but once the battery is near fully charged, no matter what size globe you have the trickle current will be nearly the same whether it’s a small wattage globe or a large wattage globe.

As originally posted, a light globe used as like this is self regulating.

Next, if you fast charge these types of batteries you are going to shorten their operating life, mine were specifically designed for slow recharging but anything quicker than a 10 hour recharge is not recommended for this type of battery.

Last but not least, you can’t use less than a 12 volt globe in a 12 volt charging circuit or the first time the battery goes dead flat ( and is usually stuffed ) or if you reverse the leads, your 6 volt globe is going to have 12 volts through it and it’s going to be stuffed at the same time.

The joys of having a 24 volt globe is that if a customer did reverse the leads all that happens is the globe lights up. No damage to the charging circuit and no damage to the battery but it will still charge the battery with out any problems.

If you want fast charging then you should not be using this type of battery in the first place, they are just not designed to take a fast charge with out doing damage to the battery.

Cheers

[TheRealAndy]

Oddbudman, I was thinking about using SEPIC or flyback topology, as this means input voltage is irrelevant. If the input battery is at 12V, it is still possible to generate 13.8 volts or more for charging. I had in my mind using a LT1513 from linear, but this may be a tad difficult to come across, will need to look at this further. The great thing about Linear Technology parts is that they come with free simulation models and LT supplies a free simulation program. That means i can nut out the best part before buying (and smoking) prototype parts!

The microcontroller design is right up my alley these days, as I do software for a quid now. The only problem with this means that you can no longer just take a part of the shelf and solder it to a board. Most ppl dont have the tools to program a micro. In saying that, manufacturer like Microchip now have (and have for some time) devices that have low voltage programming, making it possible to download to a device from a serial port. Problem there is that serial ports on PC's are getting hard to come by these days. Catch 22.


Oldboot, the problem with SLA batteries is they dont take to fast charging. Fast charging is typically reserved for chemistries such as NiCad and NiMh where you can set the charger current to C1 or greater. I have even seen C4. So charging an SLA to capacity is alwasy going to take around 12hours, that is one of the shortfalls of this chemistry.


As for me, since everyone else has spilled there guts on their profession i better do the same. I started out as an electronics tech. As a tech i did a lot of remote environment monitoring gear, most of it used SLA and solar panels as chargers. I started to design some of the gear and got the bug, so i decided to study. I moved into electronics engineering and did that for quite a few years. I did mainly digital design and power supplies, both linear and switchmode. Also did some RF, but that stuff made you use to much of your brain. Toward the end, was doing a lot of assembler and c programming, and I decided that there was more money in programming, so thats wht i do now.

Interesting reading. I may as well explain a little about what i do.

I design electronic products, many of which use battery chargers.

My most recent product involves a Microprocessor controlled 3 stage Buck converter for charging SLA batteries. Using the processor is great as you can easily set the battery charging parameters using software.

I'm happy to see you did the tests oldboot, as you got to learn something new .

Be interested to read your circuit suggestion TheRealAndy. Here's a simple one i thought was quite nifty without going into dc:dc conversion.

http://www.siliconchip.com.au/cms/A_109625/article.html


oddbudman

[driveon]

the problem with SLA batteries is they dont take to fast charging.

Hi TheRealAndy, my point exactly, the general rule for these types of batteries is charging and just as importantly, discharging of these batteries should be no more than 1/10th the battery’s total capacity, so for say a typical 7 A/H battery, 700ma is the maximum recommended charge and load current.

BTW, I use a USB to Serial converter to program some of the micros I use.

Cheers.

[oldboot]

Lets look at the lamps.

If we have a 24 volt lamp regardless of wattage in series with a 12V battery of any capacity.....unless the battery is faulty...the maximum voltage you will get across it will be 5 volts....assuming the battery is discharged to 8.8 volts which is way below recomended..........the lamp filament MIGHT barely glow... consider the battery will rise in terminal voltage rapidly in the first minute or so to arround 12 volts probly more like 13.4/5 volts once charging has stabalised.
there will be less than 2 volts across the lamp.... it will be nowhere near glowing & will be acting as a static resistance.

for a lamp to be an effective regulator it needs to be operating in a area of its characteristic that causes it to glow at some time.

Ok we have 20AH battery.... if we want to charge at the 10 hour rate.. we want 2 amps......insert a 6 watt 12 volt lamp.......the maximum current it can possibly pass is less than .5 amps........ too little, the lamp is clearly inapropriate... the lamp needs to be selected for battery capacity..... not any lamp will work properly.

While some may consider charging SLA batteries at the 10 hour rate is required...specks sheets I have quickly to hand do not support this... for cyclic use.

Typiclay a 7 ah battery will have a specified maximum initial charge rate of somewhere arround 2 to 3 amps depending on the manufacturer and the detail of the technology..... which is arround the 3 or 4 hour rate.

Talking about slow charge rates may be all fine and beaut for stationary or float applications but is completely unworkable for an on demand cyclic application were it is necessary to charge a battery in a vehicle whilst it is being driven.

Afterall we can't run our cars engine for 14 hours evrey time we want to charge a battery.

so there is a nesesity to charge at the 4 hour rate at least for any practicality..... therfore we need to get optimum charge rates ..... therefore we need to select the charge control eliment for the battery capacity used.

BTW I have never seen a one size fits all component of any type that actulay worked optimumly, I dont care if its a shirt, a hammer or a charge controller.

It is necessary to select the lamp for the size battery, or the lamp will not allow sufficient charge current or the lamp will be acting as a static resstance and you might as well use a resistor.

cheers

[driveon]

Hi oldboot, the problem here is that we don’t know what brand of battery bungie has, so to say that some of these types of batteries can be charged with a higher current is not the way to go about giving safe charging info.

Your info is based on a specific brand while the majority of these types of batteries would be stuffed in quick time if you charge them regularly at too high a current rate.

I will be more precise about one thing I posted, in my reference to one globe fits all, as posted earlier, I was referring to batteries up to 7 A/H and if you want to charge a 20 A/H battery you would need to increase the number of globes to do the same job but it would still be on the basis of a SLOW CHARGE.

Most of these batteries just can not tolerate fast charging.

Also as a an additional point of interest to how the globe regulator can be used, if you have a smallish solar panel ( 10w to 20w ) and want to keep automotive batteries topped up with out cooking them, you can use globe(s) to regulate the current and voltage going to the batteries and again leave them set-up indefinitely.

With larger solar panels you need to go to proper solar regulators but with smaller panels, this can be a cheap and easy way to trickle charge conventional automotive batteries.

Cheers.

[oldboot]

Lets look at current practice and what is currently available. ( I'm excluding the expensive & exotix batteries like the so called AGM and Gates Cyclon and the like.) we are talking about what you would commonly buy at the local allarm supplier or at an electronics supplier.

Almost without exception the SLA batteries we are finding commonly used now are starved electrolite batteries, most of them are pretty similar in performance and capacity, because the industries that use them demand some sort of comodity status.... The better brands will typiclay perform marginaly better and spec marginaly better.... but we re looking at a product that performs THE SAME with in reasonable engineering limits.

Gone are the true gell cells, gone are the calims of some sort of exclusive process or claims of overwhelming superiority the SLA bayttery is a comodity.....unlike in the eighties when they became a real alternative to small wett cell batteries.

I use the 7 AH 12 V battery because it is the buggest volume battery sold.
In the past it would have been a 6 or 6.5 Ah battery the case is the same and consistent between all brands withing a couple of mm
Almost without exception this battery form is now claimed as arround 7 AH and has been for the last 10 years...this reflects an improvement ( or perhaps a change in the way they are specified) and also reflects a comonality across manufactuers
just on spec sheets I don't have to look for
Panasonic...
capacity rated at the 20 hour rate 7 AH
maximum initial charge current for cyclic use 2.8 amps
Quoted initial charge current float use 1.05 amps

Dia mec
capacity also rated at the 20 hour rate 7.2 AH
Maximum initial charge current 2.1 amps
no float charge quoted.

Look to standard industry practice......there are all sorts of off the shelf specific SLA chargers available, there are also lots of current pieces of equipment with SLA chargers incoprirated in them.
It is very common no almost a defacto standard to charge 7Ah gell cell at 1 to 1.5 amps at 13.8 volts...this is predominlatly because stock application notes 3 terminal reg designs are being used and they rely on the foldback current limiting in the chip to provide a limit to prevent the 16V 1.5 amp plug pack from being overloaded.
Most of the dedicated small multistage chargers follow a similar form.
I have been servicing chayo and mipro portable PA systems now for over 15 years... they have either two 7AH SLA (in the earlier models) or 4AH in the later models.... these typiclay charge at arround 1 amps and typiclay charge in between 6 and 10 hours forom flat.... when chayo reduced the battery size from 7AH to 4 AH to save weight they did not change their charging circuit.
All of these units now ship with 4AH batteries.
In schools these things get a hammering, 5 days a week, they get run down till they wont go any more and the charge cycles are not supervised by anybodyu that knows much at all.
While the clients can easily kill the batteries by leaving the unit turned on over the weekend, if they avoid this and charge the units reasonably regularly they get 5 years out of the batteries....i've seen 7.

Now the maximum recomended charge rate on panasonic 4 AH is1.68 amps and DiaMec 1.2 amps. These units usulay come with some obscure brand of chineese SLA. I have unsed both the above brands as replacments as well as others... performance is similar.


So I see no problem repeatedly charging a 7AH battery at 2amps.

You simple dismission that tw lamps would be needed for a 20 AH confirms that the lamp needs to be selected for the battery capacity.

cheers

cheers

[oldboot]

Oh and
As far as discharge both panasonic and DiaMec
quote capacities
at 20 , 10, 5 and 1 hour rates.

Current SLA technology ( that has been with us for at least 15 years) in a substantial improvement on the early gelled electrolite batteries of the early eighties and before and a small improvement on the early starved electrolite.

On the matter of a solar panel
If you were to spend $150 to 300 on a solar panel why woulndt you send the extra $20 on an off the shelf solar regulator, that will give you maximum output for you money.........the other option being a lamp that will give you an indeterminate low charge rate..........sorry not if its my money.

Remember also that the terninal voltage of a 12 volt solar panel can exceed 18 volts....an entirely different story.

cheers

[driveon]

Remember also that the terninal voltage of a 12 volt solar panel can exceed 18 volts....an entirely different story.

And a combination of the globes current limitation and the internal resistance of the battery, even a fully charged battery will guaranty that the battery is never over charged, even when the solar panel output goes up to 18+ volts.

As posted above, this is not a means of fast charging but a cheap trickle charger, that works.

Cheers.

[oldboot]

The point I was making about a solar panel is that with a maximum of 18 volts you have up to 4.2 clear volts to play with... in the original senario we only have........ none.... desired charge voltage 13.8 volts....voltage available 13.8 volts.

Again you will have to select a globe for the purpose, even with more than 4 volts latitude available, a 24 volt lamp is still going to be a static resistance.
You might as well sepnd less that $1 on a resistor.


I keep comming back to the issue that regardless of what eliment it is it will have to be selected for battery capacity, for acceptable results.

The combination of a solar penel a lamp and a battery, will most certaily not prevent overcharging....once the battery has reacged full charge, charge current will reduce considerably and the lamp will definitely be a static resistance at its minimum resistance..........this resistance will be insignificant and the battery will overcharge as it might as well be connected direct to the pannel.

Unless you are relying on the lamp resistance to make the charge current so low that the battery doesn't have a chance to charge in a reasonable time scale.

.
cheers