Evening all,

I'm trying to pick out a suitable DC-DC Converter for our low power gate driver circuitry and I'm unsure about a number of different things.

First some background on the circuitry and what the DC-DC Converter is needed for.

To date, we've designed, built and tested a low power gate driver circuit similar (almost identical) to the one shown in the figure below. The end goal is to use such a circuit to drive a Power MOSFET Module for an electric motorcycle.

Anyways, we built and tested this circuit on a lab bread board that has its own variable 15V and fixed 5V supplies, and everything seems to work fine. However, the final product must have its own low voltage supplied via a DC-DC converter fed by battery bank of the motorcycle which is ~55V.

Thus we need a DC-DC converter capable of supplying us with an isolated 12V(or 15V) and 5V supplies. The 12V(or 15V) supply is used to power the gate driver IC (IR2110) as well as the power buffer circuitry,(i.e. VCC = 12V or 15V) while the 5V supply will act as the logic supply for the chip. (i.e. VDD = 5V)

One simple option would be to find a DC-DC converter that will provide the 55V to 12V(or 15V) needed and simply feed the 12V(or 15V) into a regulator such as the LM7805 and obtain our 5V logic supply from there.

It seems straightforward enough to find a suitable DC-DC converter that will meet our input/output voltage requirements, but I am uncertain as to what our output current requirements are.

According to the application note for our gate driver, the power buffers in the schematic shown above (i.e. the circuitry shown to the right of the IR2110) are needed to provide a higher gate drive current and lower gate drive impedance than what a typical gate driver IC can provide. (Application note is attached below)

They mention that the one shown in the figure above is capable of providing 8A peak output current, which is sufficient for driving Power Modules. (We are using a MOSFET power module, I've attached its data sheet below)

So does this imply that the 12V(or 15V) output from DC-DC Converter (i.e. VCC = 12V or 15V) must be capable of supplying 8A peak output current? The current needed to the drive the gate must be supplied by VCC, right? Or does the power buffer somehow amplify or boost the current?

Thanks for reading!

 

That's what the decoupling caps are for...to provide the 8 amps for less than a millisecond for switching purposes.

 

Hi #12!

That's what the decoupling caps are for...to provide the 8 amps for less than a millisecond for switching purposes.

This wasn't obvious for me, as the only context I've seen decoupling caps are in textbook amplifiers where you want allow the small ac signal to pass in order to be amplified, but block the DC by using a capacitor.

What I'm referring to above, are they called coupling capacitors or decoupling capacitors? (Maybe I'm mixing up some terminology here)

On the bottom of page 12 of the application note, they suggest the use of good quality 10mF tantalum or 10mF electrolytic and 0.1mF ceramic capacitors at the output of the buffer, but this is confusing because the values shown in the schematic are 10μF and 0.1μF across the rails of the buffer.

How do I figure out which is the correct value they are referring to?

Also, are both the 10μF and the 0.1μF referred to as decoupling capacitors?

Is the 10μF the decoupling capacitor that provides the necessary current drive while the 0.1μF decoupling capacitor is used for filtering undesired noise?

I realize a lot of what I've said above may sound a little goofy, but I am genuinely confused.

Hopefully you can help me clear things up!

Edit: After reading a little more, I can see that both the 10μF and 0.1μF are referred to as decoupling capacitors and the both serve to act as a charge resevoir and filtering of unwanted noise.

Are the 10μF decoupling capacitors are referred to as the bulk capacitors, while the 0.1μF decoupling capacitors are referred to as the local capacitors?

Shouldn't the higher capacitance bulk capacitors be placed closer to the power supply (i.e. the input of the power buffer), while the local capacitors closer to the actual load? (i.e. the output of the power buffer)

 

To me coupling capacitor are when you want to allow the transfer of the AC portion component but block the DC signal.
De-coupling capacitors from my old tube days are the ones used across a DC power supply, across the cathode resistor of a tube, or across the emitter resistor of a transistor etc.
And the function is to "de-Couple" the signal from the function path, IOW in the above cases, the signal at the decoupled point will posses the same potential as the +supply or chassis or other common.
Max.

 

That's what the decoupling caps are for...to provide the 8 amps for less than a millisecond for switching purposes.

Correct. Some people call them local bypass or load support caps, but there job is to be placed very close to whatever needs the high peak current to supply it.

 

It seems part of your problem is a misspelling. mF means millifarad and it is wrong. Both capacitors should be placed as close to the chip as you can get them. At these speeds, an inch makes a difference.

 

Just a couple of questions/observations:

You probably don't need the 4 additional fet drivers. Your modules would turn on and off in less than 500 ns with just the gate driver and a 4.7 ohm series gate resistor.

If you keep the extra FETs you are missing a wire from the drain of the IRF0110 to the source of the 9110.

You don't show the motor connection, but I assume it is from the positive supply to the line marked load. If so a clamp diode may be in order even though you can turn on the upper fet to clamp the motor voltage.
You might also want to check how much current you get as "shoot thru" when the bottom transistor is turning on before the top one is completely off.
What motor is it?
Does it only go in the forward direction?

 

Just a couple of questions/observations:

You probably don't need the 4 additional fet drivers. Your modules would turn on and off in less than 500 ns with just the gate driver and a 4.7 ohm series gate resistor.

If you keep the extra FETs you are missing a wire from the drain of the IRF0110 to the source of the 9110.

You don't show the motor connection, but I assume it is from the positive supply to the line marked load. If so a clamp diode may be in order even though you can turn on the upper fet to clamp the motor voltage.
You might also want to check how much current you get as "shoot thru" when the bottom transistor is turning on before the top one is completely off.
What motor is it?
Does it only go in the forward direction?

I've attached the specs of our motor, sorry if the data sheets provided aren't the greatest.

We are designing the converter to have regenerative braking capabilities, but due to time constraints we are more focused on simply achieving unidirectional operation.

Anyways, I had another question regarding the decoupling capacitors. The application note says that they should be "good quality 10mF tantalum or 10mF electrolytic and 0.1mF ceramic capacitors".

What do they mean by the term good quality? What kind of capacitors should I be looking at for this purpose?

 

Anyways, I had another question regarding the decoupling capacitors. The application note says that they should be "good quality 10mF tantalum or 10mF electrolytic and 0.1mF ceramic capacitors".

What do they mean by the term good quality? What kind of capacitors should I be looking at for this purpose?

Low ESR so they can supply the high peak currents. Al and Tant caps are both made in low ESR types. The ceramic will have low enough ESR for bypass use.