Can I Use Parallel Buck Converters to Achieve Higher Amperage?

[Puglife]

I have a ton of 1A buck converters, and have a 5A application in which i need to run. I do not want to order any new buck converters, because I already have so many 1A ones, and do not have money for any new ones. Each one can only handle 1A before it over heats.

I was wondering if I could use multiple 1A buck converters in parallel with each other in order to achieve 5A? if not, could i some how modify the buck converter i have to handle 5 times more amperage?

 

[Hesch]

I was wondering if I could use multiple 1A buck converters in parallel with each other in order to achieve 5A?

Yes, you can.
But you must have a common controller, sensing five currents, that distributes the total load current equally between them.
Otherwise one of them will be over heated.
The controller must generate various pulses to each of the converters.

 

[Puglife]

Yes, you can.
But you must have a common controller, sensing five currents, that distributes the total load current equally between them.
Otherwise one of them will be over heated.
The controller must generate various pulses to each of the converters.

Is is at all possible to do it without one, if not how would i make one?

 

[Hesch]

Is it at all possible to do it without one, if not how would i make one?

I don't think so. There will always be small differences between the converts, and if one of them yields just a little more current than the others, it could be a little more heated up, increasing the problem/difference. It could make a thermal run away.

I will suggest five VCO's ( voltage controlled oscillators ) that counts up five counters, therby forming an integrating A/D-converter that smoothes the noisy currents.
A μ-processor could read/sample the counters and be programmed to output suitable pulses for the buck-converters.

 

[Puglife]

I don't think so. There will always be small differences between the converts, and if one of them yields just a little more current than the others, it could be a little more heated up, increasing the problem/difference. It could make a thermal run away.

I will suggest five VCO's ( voltage controlled oscillators ) that counts up five counters, therby forming an integrating A/D-converter that smoothes the noisy currents.
A μ-processor could read/sample the counters and be programmed to output suitable pulses for the buck-converters.

so what exactly would happen if I where to just put them in parallel?

 

[Hesch]

so what exactly would happen if I where to just put them in parallel?

Try! You have five fingers, so you can sense the temperature for each of them with one hand.

Your other hand could be used to increase voltage/current/load slowly.

If one of your sensing fingers starts smoking, you have a problem with your parallel converters.

 

[Puglife]

Try! You have five fingers, so you can sense the temperature for each of them with one hand.

Your other hand could be used to increase voltage/current/load slowly.

If one of your sensing fingers starts smoking, you have a problem with your parallel converters.

I don't want to burn any out if their is a risk, I would like to know first, exactly how to solve the problem. Could i distribute it between 6 or 7 in parallel instead? would that work?

 

[Hesch]

I don't want to burn any out if their is a risk, I would like to know first, exactly how to solve the problem. Could i distribute it between 6 or 7 in parallel instead? would that work?

Your fingers will be the ones to be burned as the first. Your transistors will give up at 200 C°.

Using 6 or 7 converters, one of them may still make a thermal run away.

 

[Puglife]

Your fingers will be the ones to be burned as the first. Your transistors will give up at 200 C°.

Using 6 or 7 converters, one of them may still make a thermal run away.

so what exactly causes a thermal runaway?

 

[Hesch]

what exactly causes a thermal runaway?

E.g. some transistor that decreases its V_BE voltage due to a higher temperature, thereby conducting more current than the other transistors.
So this transistor will be even more heated up, . . . . ., and so on.

Also the kernel of the inductor will change characteristics due to temperature.

 

[Puglife]

E.g. some transistor that decreases its V_BE voltage due to a higher temperature, thereby conducting more current than the other transistors.
So this transistor will be even more heated up, . . . . ., and so on.

they each have potientiometers on them, in which i can control the voltage, is it possible for me to adjust each one individually, and test them out, and change the POT for each one, to prevent that from happening, sort of like calibrating them?

 

[Hesch]

they each have potientiometers on them, in which i can control the voltage

You need a potentiometer that can control the temperature/current on the fly.

That's what my suggested controller will do: Adjust on the fly.

 

[Puglife]

You need a potentiometer that can control the temperature/current on the fly.

That's what my suggested controller will do.

so you can't just calibrate them once, and avoid any sort of thermal runaway?

 

[Hesch]

so you can't just calibrate them once, and avoid any sort of thermal runaway?

The term "thermal run away" means that something is completely out of control. It won't just increase the temperature by some predictable 10 degrees. It will increase the temperature forever until something evaporates.

It's the butterfly that starts a tornado.

 

[Puglife]

The term "thermal run away" means that something is completely out of control. It won't just increase the temperature by some predictable 10 degrees. It will increase the temperature forever until something evaporates.

i know, but the reason that happens, is because one of the buck converters gets a change in voltage, which causes it to draw more current, causing a chain reaction right, so why can't you just limit the voltage that can come into your buck converters to 5v to prevent that?

 

[Hesch]

why can't you just limit the voltage that can come into your buck converters to 5v

You cannot limit some voltage to 5.00000000000000V, but to say 5.0001V

These extra 0.0001V is the butterfly to start the tornado.

 

[Puglife]

You cannot limit some voltage to 5.00000000000000V, but to say 5.0001V

These extra 0.0001V is the butterfly to start the tornado.

so what is the absolute cheapest way to stop this from happening, I need to make a ton of these, and need it to be as cheap as humanly possible. Thank you

 

[Hesch]

so what is the absolute cheapest way to stop this from happening, I need to make a ton of these

Make a big buck-converter that can handle 5A.
No distribution problem.
You need a ton of these.

 

[Puglife]

Make a big buck-converter that can handle 5A.
No distribution problem.
You need a ton of these.

is their any way of doing it asside from that?

 

[Hesch]

is their any way of doing it asside from that?

Yes, of course.

Example:
You can connect a current limiter at the output of every buck-converter, say 6 converters + 6 current limiters, each limited to 0,9A.
Then you will have a total current = 5.4A.

But I don't think that will be the cheapest way.

 

[Puglife]

Yes, of course.

Example:
You can connect a current limiter at the output of every buck-converter, say 6 converters + 6 current limiters, each limited to 0,9A.
Then you will have a total current = 5.4A.

But I don't think that will be the cheapest way.

would i buy current limiters, if so what are they called, can i make them, if so how?

 

[Hesch]

would i buy current limiters, if so what are they called, can i make them, if so how?

Well, there is a simple one here:

You can choose other circuits here:

https://www.google.dk/search?q=curr...-JbLAhUzb5oKHWmBDS0Q_AUIBygB&biw=1920&bih=946

 

[Puglife]

Well, there is a simple one here:

You can choose other circuits here:

https://www.google.dk/search?q=curr...-JbLAhUzb5oKHWmBDS0Q_AUIBygB&biw=1920&bih=946

can you buy them cheaper than you can make them?

 

[Hesch]

I don't know.

 

[Puglife]

I don't know.

ok, that's fine, do you know where i could buy one, so that i may compair the prices, Thanks!

 

[Hesch]

No, I'm not staying in the USA.

 

[Puglife]

No, I'm not staying in the USA.

ok, that's fine, you have been of great help to me, just one last question, do i put the limiters on the inputs, or outputs of the buck converters?

 

[Hesch]

do i put the limiters on the inputs, or outputs of the buck

On the outputs.
You must calculate with a voltage drop across the suggested current limiter of about 1V, using silicon transistors.

 

[Puglife]

On the outputs.
You must calculate with a voltage drop across the suggested current limiter of about 1V, using silicon transistors.

what do you mean?

 

[Hesch]

Say you need 10V on the output of the current limiter, the buck converter must yield 11V on the output.
11V - 1V = 10V.