Parallel Switching Regulators

[dknguyen]

Hi. I have two switching regulators with 4.5-9V and 9-18V input. I want to put them in parallel in order to make a system that can accept 4.5-18V, where only only one regulator is ever active and the the input voltage stays well away from 9V will never cross it for a given input voltage.

I'm not sure if all I have to do is use the disable pin on the non-functioning regulator to protect the low voltage regulator from higher voltage inputs (the company also won't seem to tell me). The two regulators are from the same series (they share the same datasheet), and just have different input ranges, but I am not sure what physical (not functional) input voltage a device can tolerate.

Also, I am not sure what will happen on the outputs when one regulator is working and the other isn't (will it behave as a sink? unexpected capacitive load?).

Does anyone have any idea? These are buck-boost converters I think (they work on an input that is both higher and lower than the output voltage).

 

[mneary]

The buck converter will pass the output voltage to the input through the MOSFET's intrinsic diode, even when it's turned off.

And the boost converter will pass the input voltage through to the output via the series diode.

Better use a SEPIC design.

 

[dknguyen]

Maybe I'll just use blocking MOSFET on the input and output then.

 

[mneary]

If you really must block input and output, given your original conditions, then you might consider a relay.

 

[Hero999]

You can connect6 MOSFETs back-to-back to prevent the current from flowing in the reverse direction.

 

[dknguyen]

The company said it won't work so I probably have to have a physically disconnected switch or relay. It's more likely I'll just use a socket so I can exchange the SIP regulator.

 

[mneary]

If you can give more of the requirements (or the chip part numbers) we could help a lot more.

 

[dknguyen]

It's the RS or RS3 from Recom Power (or equivelant from other companies.)
**broken link removed**

 

[mneary]

If you can arrange to have your 5V and 12V inputs come in on separate rails, then this isolates the inputs. You can then isolate the outputs from one another with P-channel MOSFETs connected as "diodes". (one module to each MOSFET Source; Drains connected together [this is your combined output]; and both Gates to 0V.)

 

[dknguyen]

The outputs are already isolated from each other due to the fact their sole purpose is to feed into a circuit that requires a diode. SO I guess I could just have a dedicated diode for each output where the cathodes are a single node(rather than connect the converters together then feed through the diode. WHy do you suggest use a backwards connected PMOS though?

The different power rails won't work because the power feeding the regulators is also the power source feeding the motor and they take up a lot of space. Perhaps I should just use PMOS pass transistors for the outputs.

There is a regulator with a much higher input voltage range powering the logic, and the purpose of these switching regulators is soley to drive the motor MOSFET gates. So I can have the PMOS transistor's gates being pulled high, with only the correct one being pulled low after the logic has decided what battery voltage has been connected.

Great help thanks. Wouldnt have thought about it wihtout you.

 

[mneary]

WHy do you suggest use a backwards connected PMOS though?

A backwards connected PMOS with the Gate to 0V can be seen as an "active diode". Although the intrinsic diode conducts like an ordinary diode, with the Gate grounded, the channel turns on for zero voltage drop when the Source is positive. (Respecting voltage ratings of course).

When I have more time, I'll try to draw a circuit.