Isolated DC-DC schematics

[DanD3]

I'm looking for schematics for DIY isolated DC-DC with 12V input , 4.2V output and 5-10A capable .
Application is to build a Li-xx charger with 6 ch (= 6 lipo's) with each channel individualy µP controlled .
It should be low cost , so I exclude commercial modules for so far .

any links to subject ??? (or schematics of course welcome)

Thanks,
Dan

[z-matrix]

use a toroid core, wind 2:1:1 turns ratio (primary secondary, sense)
an irf7455
schottky at secondary,
you can use synchronous rectificaion at secondary side to increase current capability and efficiency.
wind a flyback sense and monitor the secondary voltage using an adc,
drive the 7455 on a port, use PWM at 20-200khz switching frequency, this is the simplest, or you can implement a flyback in software.

though this isolated topology might not give engough accuracy, load regulation can be a few tmes 10 millivolts, so this is a too cheap method.

either use a microcontroller per cell, or somehow galvanically isolate adcs and sense the cell voltages to accomplish stable CV output mode, and use seperate transformers at the secondary, or use isolated mosfet drive at output and use 1 transformer with n windings.

What do you think about using a balancer and a charger?

[DanD3]

Thanks for this quick reply z-matrix.
I have done the Charger-balancer way (both commercial and DIY ) but I think the future is definitivly with individual channels.

A main RC manufacturer (ROBBE) goes already that direction with his low-cost 5 channel lipo charger (4A current) . The LIPOMAT 5-4000 use dc/dc small toroids with 2 windings ? (4wires) , control is still 100% analogic .

Most balancers are dummy units which works at end of charge.
With intelligence , you could balance at any level of charge , i.e not waiting to end of full charge to use the batteries. New generation balancers do it that way.

I don't have too much experience with isolated sense or ADC's but I'll will try it .
Using a low cost µC per channel is a new challenging idea I din't thought about!

[z-matrix]

you can isolate adcs at the outputs using 6n136, or 6n137 optocpuplers, if you do not need very fast sampling, then 4n35 will be enough.

i do not thnk parallel charging is the future, hope balancing will not be needed for any cells in the future, this needs the cells be created equal.

sure, the balance chargers used to be a dummy, using a 4.2V shunt regulation at 100-200mA max. current,

btw.: i'm just testing my balancer, that balances at any voltage level.
completely analog, using 2 operational amplifiers per balance point (2cell).

of course C/10 should be an overkill for a balancer that is not just balancing during the near CV stage.

[DanD3]

Hi z-matrix,

Thankx for the opto's information.
6N136 is indeed popular and used in AN's from MAXIM.

The DIY balancer I used so far is 1-2A shunt capable.
I have to say that I play with Lipo's (Eng. samples ) from LG 10.000 mA/h and (claimed) 50 C .
Now experimenting with Kokam 4800 x2p = 9600mA/h 20C.

But you're right in saying that the best would be a rugged battery with no specific need's .
A123 batteries are my next test object (not available directly in Europe so far).

I will start a single channel / µC controled experiment .
Winter is coming , so timing is right.

Thanks,