analog mppt controller to charge battery

[mmgall]

Hello.

Can I use this simple analog mppt circuit to charge a lead-acid battery?




Thanks in advance.

 

[mmgall]

What do you mean not obvious? Actually the reference is here **broken link removed**

So it is not possible to charge a battery? Thanks

 

[mmgall]

Thanks for your response. I just wondered if this simple circuit can do mppt because most mppt controllers use digital(Mcu/fpga) implementation. Well, I may experiment this to see if this works.

 

[FvM]

I believe it's not obvious at first sight if and how the circuit is operating as MPPT controller.

- - - Updated - - -

"Not obvious" means that the supplemented explanation is urgently needed. I understand your question in the way, that you don't understand how the circuit operates? Otherwise you would answer the question yourself.

So it is not possible to charge a battery?

I hope the question is answered in the link.

- - - Updated - - -

It seems to me that the controller can only provide a kind of hysteretic control, switching the pwm on and off. Or it must rely on a well calculated amount of pwm residuals in input voltage and current feed through to the FF input. This looks like a rather unsure method to me. I guess, the authors tuned circuit parameters untik they achieved a "good-looking" result.

Unfortunately there are no quantitative data of the analog signal processing chain, so you have to start almost from the scratch when designing the circuit.

Needless to say, that the circuit doesn't yet include battery voltage or current control. It's apparently assuming that the maximal output current can be handled by the battery. End-of-charge cut-off must be implemented though.

All-in-all, it's an interesting attempt to make a purely analog MPPT controller. It can be most likely improved. I assume that a dedicated PP modulation oscillator and true PWM should be considered for a good controller. Obviously, it can be more easily implemented digitally these days.

 

[Orson Cart]

This circuit is in fact quite clever, it uses a change in voltage and a change in the seen power (V*I) to change the operating point of the buck converter. It should not be dismissed because it is hard to see at first glance how it works, if anything it may need a min freq circuit to keep it switching above a certain min freq, but apart from that it looks to provide a pretty good basic MPPT...!

Oh I see the clock provides a static Sw freq...

 

[Anna Conda]

Yes you can charge a battery with it, however there is no provision for current limit, or voltage limiting.

 

[schmitt trigger]

In addition to what others have said, the schematic itself is incomplete, and you have to design the following functions.

- The flip-flop itself does not have the proper capability to drive a high side switch (Mosfet). There must be a high side gate driver in between the two.
- The analog multiplier block is shown as...well, only as a block, and there are several connections and passive components that must be taken into account as explained on the data sheet.
- It is not obvious, but the analog portion requires bipolar power supplies. Which BTW, both must of a higher voltage than 4 volts.
- care must be taken when interfacing the opamps to the logic gates to avoid pulling them below ground.

All in all, it appears an interesting circuit. But as it stands right now, it is only conceptual; significant work is required to make a real and functional circuit.

 

[Anna Conda]

I think the sketch is intended to show a quasi block diagram only and not a finished design.... else pin numbers etc would be shown...

 

[schmitt trigger]

Anna Conda; that is precisely my point. This is only a conceptual circuit.
And also to reinforce FvM's statement that: "Unfortunately there are no quantitative data of the analog signal processing chain, so you have to start almost from the scratch when designing the circuit."

 

[Orson Cart]

I think Anna's point was that it is obviously a slightly developed conceptual circuit (although proven in the paper cited), and therefore to criticise it as if it were a finished design is a little over the top.
However it is fairly easy to see how this circuit can be inplemented (and modelled) to useful effect. The component calulations are easy and no software required, just a bit of thought towards controlling a suitable converter/inverter from the output signal (not too hard either...)

 

[schmitt trigger]

I wasn't making a criticism of Annaconda's reply. Not at all, and I hope my meaning wasn't misconstrued.

If any -and there wasn't any- it would be towards the OP, who asked FvM what wasn't obvious about the circuit.

The answer is: many things as it is not a fully developed circuit.

 

[mmgall]

In addition to what others have said, the schematic itself is incomplete, and you have to design the following functions.

- The flip-flop itself does not have the proper capability to drive a high side switch (Mosfet). There must be a high side gate driver in between the two.

Hello, if i use irf540 or irf3250 as mosfet switch, what high side gate driver circuit could you recommend. Thank you.