Electronic circuit design for battery charger - PWM/PLL? - help please

  • Thread starter Thread starter Anonymous
  • Start date Start date
A

Anonymous

Guest
Background

I am about to buy a small 75Ah AGM 12V battery mainly for use in my tender to drive my hookah diving compressor but for any portable 12Vdc applications including emergency engine start, etc. I want to store the battery in a deck locker normally disconnected, no load, no charge, as they hold charge well enough for my purposes. I do not want to charge this battery directly across the main batteries as there would be no independent charging control and these batteries need great care in charging to avoid gassing. I have a variable voltage 'laptop' inverter of about 75W and with that I can provide a 3.75A supply of 19.5V from a cigar lighter socket. I want to build an interface from the 19.5V to charge the 12V AGM battery. My question here does not relate to the charging envelope, only the driver stage.

Design

My design philosophy is a common emitter bipolar npn output transistor (collector to the battery -ve) driven by a variable pulse width source. The pulse width will be adjustable from near zero to 100% controlled by a signal from control circuitry to be designed at another time. The frequency will be essentially fixed at about 1kHz which is a frequency popular with battery anti-sulphation 'pulsers'. The current will be limited by a wire-wound resistor between the batt +ve and the +19.5V rail.

I have limited development facilities (limited component stocks and just a scope and a DMM, no surface mount - Veroboard construction only). I don't desperately need help with the control electronics design (though any ideas or proven designs welcome) but I am struggling a bit to choose a suitable PWM. I've looked at using a 4046 PLL but that gets a bit silly with component count and is an overkill. I've also looked at using couple of comparators or a 555 but nothing is inspiring me at the moment. I'm sure that I designed something very similar many years ago but I just can't remember the details!

I don't want to buy an expensive marine charger - that would be silly and OTT but I would be delighted to find a design, kit or manufactured module that would do what I want it to at low cost - say £15 including the box and all bits which is what my Maplins or RS order will probably be to buy the bits I need for my own design.

Once this is up and running I will be happy to post the circuit and details here for others to use.

Many thanks.
 
LM338T, variable voltage regulator, farnell part number 948 9428, down load a data sheet, shows you the circuit and components.

Give you a fixed voltage at variable amp upto 5 amp, over temp protected, short circuit protected.

Brian
 
p.s.

Maplin or CPC may do a kit with that regulator in, including PCB, thought you could make up the circuit on VERO board. total components should be around £3

Brian
 
Brian, thanks. That is a dc solution and the IC would need to dissipate a lot of power, rather more than it can. I know that it has a thermal shutdown but I don't really want it to shut down, I want it to charge. In particular I am hoping for a pulsed solution as I believe that will improve the resistance to sulphation - I will be using it deep-cycle.
 
He's done this before, ask a question and then argue against the answer. Suspect he wants to get in a back and to to show how technically knowledgeable he thinks he is.
Stu
 
Why dont you save all the hassle of the analogue electroncis and use a simple microcontroler. Use and A-D converter to monitor the battery voltage and a timer to create the PWM signal you require. Simple and cheap, can also use any charging slope you want.
 
Thanks, yes, that would be an idea - a PIC or something? I'd have to buy a standard board, tho' as I couldn't be making multilayer PTH boards which presumably one needs for the busses? I don't have a development system of any sort. Any idea which controller I might base the design on?
 
Have a look at some of the Atmel devices, maybe this one. Circuit can be very simple and built on vero. Atmel have a free complier and instruction to make simple programmers. What do you think? Will have a look for the programmer.
 
Looking at the application note for the constant current generator (which would be a sound basis for my project) is a trip down memory lane.... mapping an S domain transfer function into the z domain. Haven't done that since my finals very many years ago!

Do you think there would be enough on-board programming space - 4k? I could have managed with that with the first and second generation microprocessors but I don't have any experience of programming thing like this. Most programmable devices seem to expect huge amounts of memory now. Extra EPROM would be a pain to have to hard-wire in and anyway I haven't got any development tools or EPROM programmer.

Whatever else, it would be fun to do though I don't want to be collecting stuff as I live on the boat and I am hoping to keep the cost to £15 or so all in. I shall be interested to see the programmer if you find one. Thanks for your help.
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

I have had a lot of success with a SG3525 switching power converter chip.
It will drive a pair of FETs to a transformer rectifier ie essentially your computer power converter you want to use but with easy access to current and voltage controls.
So you could have simple resistive feedback of voltage with current limiting.
You could however incorporate into the control circuit one of the purpose designed 3 stage charging chips for SLA batteries.

Is there going to be a lot of benefit of charging with 1Khz pulses versus DC? It will certainly be tricky to monitor charge state by voltage.
good luck olewill
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

I'll look into the SG3525, thanks. Meanwhile, can you give me the part numbers of any SLA charging ICs that are worth looking at? I think that pulsed mode will reduce dissipation but mostly I believe that it will be good for sulphation (but not my SSB).

Many thanks.
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

David,
Do you want fully automatic or manual ?
You could yse a 555 as a pwm gen to a fet, the old BUZ 11 or similar.. much better than bipolar.
But manual only.
Very simple and low component count.
If you need a pic (sounds overkill) let me know what you need it to do and I will sort one for you as I have the full development kit on board.
Joe
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

Joe, that's looking good and when I did a search for PWM 555 I came across a design (the switch mode circuit in this pdf) http://www.edn.com/contents/images/111402di.pdf

This circuit is set for 60kHz but it will adapt happily to 1kHz and looks as though it has been properly developed by an engineer (we hope). Manual or Auto? I want auto but not the usual sort. I plan to monitor the charging current and keep the bulk charging going until the charging current has fallen to a set limit (say < 0.75A with a small battery like this) then a time delay (say 20 minutes) followed by a fall-back to the float level. I am wondering whether it would be better to have thermal compensation but doing that with discreets would be a pain - maybe I could look for a three stage charger IC but then again, I do want my own (current-sensed) charge characteristics (above) not the usual (voltage-sensed) type. I am totally convinced that a falling current at the end of the bulk charge is the best way to determine the end of charge. I've never heard of a professional charger having this characteristic so I don't suppose that any three stage charger ICs would do it.

Good thinking on the 555s - the article uses two so a 556 will do nicely. I agree that the PIC is overkill but it would be fun. OTOH I have plenty of other 'fun' things to be getting on with!
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

We used to charge Nicads that way David, the chargers sensed the current drop and decided on the end of peak charge.

It gets a tad more complex if you want auto, but.... you COULD sense the current though a load resistor, hence v drop, then use the vdrop to signal a switch over to a constant voltage. You can use the 7812 vreg ith a pot in the common lead to deck to raise the voltage, regulation is still good.. but linear as opposed to pwm
A schmitt would work, or a monostable 555..
Gets messy though.
A Pic is a good option if you KNOW the flow diagram of what you need.
Joe
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

Joe, that circuit pdf I posted shows a voltage-controlled PWM using a 556 so I only need to put an op amp into the feedback path with the appropriate transfer function. I can probably compensate the temperature quite easily as an approximation. I would smooth the measured voltage over ten cycles and measure the current in the middle of the charge simply by gating the current sensing circuit. Probably a couple of op amps, or use one quad. Could be quite easy, I will have a play around on paper.

You'd want a lot more circuitry with a big charger of 20A or more or you could run into trouble with loads coming on but a little 75Ah battery being charged usually with no load from a 3.75A supply should not present a problem. I don't really want to go down the PIC route - I have plenty of stuff that I need to do without getting onto that learning curve. Many thanks for the kind offer, though.
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

If you don't want a linear regulator because of power, you are essentially talking about is a switched mode power supply.

After a few years of designing them, I can say it's not trivial. Don't attempt building you own system. Companies like Maxim, Texas Instruments, Linear Technology, National Semiconductor, Fairchild etc do ICs that soak up all of the control circuitry. You'll more than likely find one that will have an integrated switch too.
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

[ QUOTE ]
Don't attempt building you own system.

[/ QUOTE ] /forums/images/graemlins/ooo.gif
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

Hi Lemain for a charger chip just google SLA battery charger circuit.

Re switch mode supply I found the SG 3525 driving a pair of FET seemed very uncritical regarding inductor no saturation etc. I used a ferrite pot core and also an old computer power supply core with like 20 turns centre tapped for primary and 15 turns for secondary. olewill
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

Olewill, nice one.. !!! excellent stuff

David, pics really are superb, its a thing, if you ever have time lol, to get into. its a risc core, huge programs in little space.. you can buy a pic for your use here for around 2 pounds, contains harare PWM circuits, A to D on board, 16 bit math etc.
I use a compiler called Proton, its a BASIC type language, I originally used raw code though. In critical timing areas you can insert assembler direct.
they are very nice toys.
Have fun with the circuits, sounds good.
Joe.
 
Re: Electronic circuit design for battery charger - PWM/PLL? - help pl

Ah, brilliant! That looks perfect. Many thanks.
 
You must log in or register to reply here.

Other threads that may be of interest

P
Old AGM battery.
Replies
9
Views
307
samfieldhouse
samfieldhouse
webcraft
Resetting SmartShunt to 100%
Replies
7
Views
342
webcraft
webcraft
Skipper Felice
Optimisation of the battery charger infrastructure
Replies
1
Views
152
PaulRainbow
PaulRainbow

Share this page

Top