Double output battery charger

Posted by: smsboit on 02/04/2014

Hi, I have to design a new battery charger with two outputs 27.5V 0.5Amps. In the attached scheme resulting from PI Expert 9, the feedback voltage is connected to only one section, while I need control on both sections because, having to charge the batteries, I might have a section that requires more load and the other less. With the proposed circuit by PI Expert, it seems to me that if on the 'uncontrolled' section there is little current consumption, while in the controlled section there is a greater absorption, the voltage on the 'uncontrolled' section goes up improperly. I wonder if it is possible and recommended to create another equal feedback for the uncontrolled section and put the optocoupler output in series with the first optocoupler. There is a better alternative ? Can you give me a little help ? Thank you very much in advance, Emanuele Castagnini - SMS Sistemi e Microsistemi - Tech. Dept.

Comments

Some questions for you.

Do you need your two outputs to be isolated from each other as schematic is shown? Can your two outputs share a common return? If you can have same return for both outputs then it may be easier to have a better cross regulation

How important is the efficiency of the power supply? Can you allow to have some preloads at the outputs?

When I know your answer I may something to suggest

Thanks for your prompt answer. My application is a battery charger for a device with n.2 12V or n.4 12V lead acid batteries series connected, a sort of UPS for elevator market. The common return could be possible if the '0' of the first charger is connected to the +27.5V of the second charger. The efficiency is quiet important because the charger will be closed in a box with a low air flow and not forced. What do you mean exactly with output preloads? Consider that batteries will be connected to the device. I'm evaluating to use 2 separated sections with 2 PI devices and 2 transformers sharing a common input filter and rectifier...maybe the cost will increase a little, but the transformers will be smaller and the final charger thinner. What do you think about it ? Best regards, Emanuele.

Emanuele:

Please forgive my ignorance. I do not know what you mean “n.2 and n.4”

Talking about your comment, when you said “common return “0” of first charger is connected to the +27.5V of the second charger”

I think what you mean is to stack the output windings. In other words, the outputs will be in series as shown in the link below.

http://www.powerint.com/sites/default/files/PDFFiles/der28.pdf

if the DER28 example is suitable for your application, then there are some tricks to improve the cross regulation. Please let me know if you like to try this solution.

Thanks a lot, my ignorance is surely higher than yours :-)... When I wrote n.2 and n.4 I mean 2 or 4 12V batteries series connected as indicated in the attachment 4Batteries.I think that DER28 could be good if the output voltage of my stages remains stable even if the 2 load conditions are different. I see that this application uses feedback on primary side, maybe this is one of the tricks you mean ? We have to consider also that a battery charger have to limit the output current with a circuit as in the attacment CB_ILim. The devive can't be switched off in case of overload like a normal power supply. Maybe we will change our solution to 2 separated stages with 2 Tiny switch or one single stage to limit costs, but I'm interested technically to the solution for the first idea. Thanks in advance, best regards, Emanuele.

Ok Emanuele, talking about the first idea, an architecture for the secondary side as DER28 is acceptable for you right?

So one output (V1 from OUTRET to V1+) will be 27.5V. The second output (V2, from V1+ to V2+) will be 27.5 volts too. If this is the case, then the total VOUT from OUTRET to V2+ will be 55V

If this is the case, you can use your actual prototype and connect the lower leg of the floated output to the +vout of your regulated output. Then you can use two resistors feedback to sense your outputs as shown in schematic report of DER28 (R10 senses one output and R11 senses the other output.

You may need to add a preload resistors as (R13) in one or both outputs to avoid your output voltages to go too high when one of the outputs is fully loaded and the other output is unloaded. The preload resistor will conduct current and will keep the voltage down. Hope this recommendations help you in your design

Dear PI-Worf, your solution is more than acceptable and this recommendation helps me very much, many thanks. I have only one 'generic' doubt. What is the criteria to select a feedback from the primary side or from the secondary side of the transformer? Normally PI Expert indicates secondary feedback, while DER28 uses primary feedback. It is due to the multiple controlled output? Best regards, Emanuele.

DER28 uses secondary feedback. The DC voltage at the cathode of D6 is used to bias U2 transistor. The feedback signal comes from U2 Diode.( It is wrong label as U3 in schematic). The feedback comes from sensing the output voltage. I do not understand why you said “DER28 used primary feedback”

Some circuits (other than DER28) used primary feedback. You used it when you do not care about your output voltage regulation and want a very economical power supply design.

Thanks for your support. I've supposed the feedback comes from the primary because it is drawn at primary side, but this doesn't mean 'feedback from the primary', and it is a usual drawing mode. My best regards, Emanuele Castagnini.

My pleasure to talk to you