Caravan Charger/P.S.U.

I'm looking for idea's if you can help. I have recently bought a caravan and need to make a 12V Charger/P.S.U to run the low voltage appliances.

My problem is that I can build a charger to charge my leisure batteries easy enough but I also want the same transformer & rectifier to run all 12v appliances at the same time. I have thought about this for days and looked at all sorts of circuits on the net for ideas but I've filled my head with so many idea's now that I can't figure out what is my best move.

I have 2x 12volt 85Ah lead acid batteries wired in parallel giving 170Ah. I expect to pull approx 10 - 15 Amps from the P.S.U. running the radio, fridge & lights etc. I had considered simply making a constant current 13.8volt P.S.U. and connecting that directly accross the batteries and running the appliances direct from the batteries.

Will doing the above risk over charging the batteries? As I have read, most lead acid chargers charge in stages and reduce the charge voltage as the battery becomes full, Bearing in mind that the P.S.U. will be connected for long periods of time and obviously will not be disconnected when the battery reaches the 'Full charge' state.

I am currently building a control unit based around a 16F877 microcontroller which will (Using an LCD display) control water heater, Radio power, light power etc etc etc. It would be nice if I could incorporate this into an already designed PIC controlled charger circuit of some sort.

Ideally I would like this thing to be 'intelligent' so as not to overcook my batteries but at the same time it must be able to supply the relevant voltage & current to run the other electrical appliances. If I end up making the constant current P.S.U. I plan to regulate it using a simple LM317 regulator or similar with 4 or more bypass transistors to boost the current to somewhere near 20A. I have already made up the start of a P.S.U. using a 12v 20A transformer, full wave rectifier and 39,000uF capacitor. This P.S.U. gives me 16.5 volts with no load, and approx 15.5volts with a load of approx 30W. If I were to use 6 X TIP31A transistors as 'bypass' transistors (because I have a dozen of them sitting spare here) what size/value current limit resistors would I need for each one.

Sorry far all the questions in one post but I though I might aswell get them all (hopefully) out of the way.

Thanks for any replies...

Paul.

I think that the idea of reducing the voltage at the end is so that the voltage before that can be high to get a really full charge reasonably quickly.

Car alternators just run at 14 to 14.4 V all the time.

with that much capacity, just cut the voltage down to 13.5 to 13.8 V, make sure the current limit works, and it should be fine. It might not charge to maximum capacity, but it should be OK to leave it connected all the time.

hackableFM said:

I had considered simply making a constant current 13.8volt P.S.U. and connecting that directly accross the batteries and running the appliances direct from the batteries. Will doing the above risk over charging the batteries?

Click to expand...

You mean a constant voltage, current limited, power supply. Yes, this will work and won't over charge the batteries if you set it to apx 13.5V-13.8V depending on temperature.

As I have read, most lead acid chargers charge in stages and reduce the charge voltage as the battery becomes full, Bearing in mind that the P.S.U. will be connected for long periods of time and obviously will not be disconnected when the battery reaches the 'Full charge' state.

Click to expand...

Yes, this is how they normally work. The charger I mentioned above will still charge the batteries to apx 80-90% capacity but will be slower than a 3 stage charger.

If I were to use 6 X TIP31A transistors as 'bypass' transistors (because I have a dozen of them sitting spare here) what size/value current limit resistors would I need for each one.

Click to expand...

I would opt for the switching regulator option due to the power levels involved. You could quite easily dissipate 20-50 watts of power in the pass transistors if you are charging at 15Amps while trying to keep up with the load from the other electrical appliances.

hackableFM said:

I'm looking for idea's if you can help. I have recently bought a caravan and need to make a 12V Charger/P.S.U to run the low voltage appliances.





Ideally I would like this thing to be 'intelligent' so as not to overcook my batteries but at the same time it must be able to supply the relevant voltage & current to run the other electrical appliances. If I end up making the constant current P.S.U. I plan to regulate it using a simple LM317 regulator or similar with 4 or more bypass transistors to boost the current to somewhere near 20A. I have already made up the start of a P.S.U. using a 12v 20A transformer, full wave rectifier and 39,000uF capacitor. This P.S.U. gives me 16.5 volts with no load, and approx 15.5volts with a load of approx 30W. If I were to use 6 X TIP31A transistors as 'bypass' transistors (because I have a dozen of them sitting spare here) what size/value current limit resistors would I need for each one.

Sorry far all the questions in one post but I though I might aswell get them all (hopefully) out of the way.

Thanks for any replies...

Paul.

Click to expand...

With the transformer you are using don't expect to get more than about 12 amps DC on a continual basis, because of charging current for the filter capacitor. With brute force filtering, bridge rectifier feeding the filter capacitor the maximum load current on can realize is secondary amps x 0.6. Drawing more than that on continual basis will cause the transformer to overheat. So I would use current limiting resistors for about 12 amps. Going one step further, I don't think the 12 volts from the transformer is enough voltage considering the drop across the rectifiers and the regulators. I think you will need very close to a 15 volt secondary

OK, thanks to all of you for the info, I've done a bit of looking around and found THIS circuit which is an automatic charger of the battery. Can anyone think of a reason why the charger may get confused and go into a cycle of "switching on/off" repeatedly because I intend to use it to power accessories up at the same time as charging the battery?

Thanks again for your input on this as I would like this one to work right from the word go.

Paul...

That circuit will decide on every AC cycle whether you need a boost. With your huge batteries, this cycle time can be several minutes, and could be hours. Since it's only designed for 4A, it still might not keep up with your load.

I would consider using a full wave rectifier in place of D1 to be nicer to the transformer. You might need a resistor across the unfiltered supply to help turn off the SCRs. I would try 200-270 ohms, 2W. It will get very warm.

I would be concerned that the BT136 is rated at just 4A and its actual current is limited only by the transformer resistance. The heatsink will need to keep Q1 tab below 100c which will be difficult in any enclosed space. I would normally use a 25A SCR or TRIAC in its place.

I don't understand why he used a TRIAC for Q1. If Q1 was an SCR then D2 would not be needed, and R4, 5, 7, 8 could be higher values for decreased standby consumption. (In this case I would add a reverse diode from the gate to cathode of Q1.)

I should have said earlier but I clean forgot to mention my intentions of 'upgrading' this supply so it will produce somewhere in the range of 20 - 30 amps. I intended on piggybacking 2 BT139 (16Amps each I believe) Triacs for Q1.

I have already made up a 12Volt full wave rectified & smoothed P.S.U. as mentioned above, so I planned to replace the existing transformer & D1 in this circuit with the ready made P.S.U. I now assume from your above info that the smoothing cap "39,000µf 30wv" will intefere with the operation of this circuit since it uses the 'unsmoothed DC' to decide when to go into 'Charge' mode. Will it work ok if I simply replace the transformer & D1 with my ready made P.S.U. but remove the smoothing caps and place them on the output of the final circuit?

mneary said:

I don't understand why he used a TRIAC for Q1. If Q1 was an SCR then D2 would not be needed, and R4, 5, 7, 8 could be higher values for decreased standby consumption. (In this case I would add a reverse diode from the gate to cathode of Q1.)

Click to expand...

So what you are saying is that if I replace the triac with a 2 or 3 SCR's piggybacked such as a TIC126 rated at 12A each then I could omit D2 and insert resistors such as 150Ω or are you thinking something higher like a 1k or more? Why may I ask would you use a diode from gate to cathode of the scr, I'm interested as to why you would add this.

I attach a modified diagram, can you please have a look and see if my mods are suitable, As for the diode you suggest from gate to cathode of the scr, can I ask your suggestion for a part number or type?

Another silly question I suppose but I might aswell ask it to go with the rest of them, Would a TIP126 or a C106D work fine for Q2 as a replacement for the listed BRX49?

Thanks for your help.

Paul.

New diagram is a very good start.

Correct you need to provide pulsed DC without smoothing. This supply is probably not suitable for use without a battery, it's pretty noisy.

I added the diode from gate to cathode because you don't want reverse voltage on the gate. (General Electric SCR manual, first edition, 1961, page 33). Looks a lot like a base-emitter junction, max -5 to -7 volts. Any 1N400x is good.

Unless very closely matched, thyristors are not happy in parallel. The hotter one would tend to conduct more current, which would make it still hotter. You can force current sharing with resistance or reactance (same manual, section 6.)

150 ohms for R4, 5, 7, and 8 sound good. Don't bother trying 1K, not enough trigger current.

The C106 is a lot more sensitive, but I don't see a problem with it in this case. NO on the TIP126, it's a PNP transistor.

I should remind you that the fuse is NOT optional. You may also need to include some means of inrush current limiting, because this design is not current limited at all.