Why Haven't You Applied for RoadTest: Gate Driver EVM with Truly Differential Inputs?

Randall,

The RoadTest looks very interesting, especially with the flexibility of the MC2GO controller board and PC software, but I'm going to be busy with another road test in the same timeframe.  The learning possibilities with the evaluation kit would be worth the time, but to split my time between 2 road tests at the same time wouldn't do justice to either.  If only the review completion date was for the Infineon kit was out another month.

Regards, Gordon

I just don't have a good idea of what kind of tests would be interesting for a product like this. I haven't yet built any kind of power electronics circuit (outside of school), so it doesn't really seem reasonable to assert that I could perform a reasonable RoadTest

I wouldn't know what to do with it, or how to test it.

I would like to test it but I don't have a differential probe. Although I have just realised that I could probably manage with a multimeter on JP1.

EDIT: I suppose the main question really is what would I do with the output, I have a 48V DC motor, but It's not really conventional to drive them from the output of a hefty low pass filter. Would it make more sense to use it, with current feedback, to charge a battery?

I just don't feel I'd do a good job of road testing this. It's complicated and not in my area of expertise. I don't apply for road test for free stuff. I apply if I think I can contribute something worthwhile. I don't think I'd have any use for the device after a road test either.

'm also busy with other things - including a road test in progress.

In my case, the main problem is lack of time. I have other projects which are in stand by and which need less time than roadtesting the Infineon gate driver, but they have to wait until I could free myself for them.

As I am not familiar with building systems using this kind of gate drivers, applying for this road test would imply a higher effort from my part and would certainly take more time than the usually allocated for a road test.

Probably, for those who are not familiar with this kind of device, but are willing to try a roadtest, browsing the roadtest archive for similar roadtests would be a good start. They could find an idea on how to deal with this device, what kind of tools they would need and maybe what other components would be necessary to build an appropriate roadtest bench.

I voted "it's too complicated" but the real reason is more complicated than that.

The Infineon part is a really interesting and potentially useful chip. As far as I'm concerned playing with their demo board is waste of time - they have some good material on the web - I'm fully convinced that the part actually works.

To go to the next step of using it involves a lot of time - I have a potential application but the board design for that is done already using a different chip (and that project is stalled due to lack of time).

baldengineer

The critical feature appears to be the differential inputs. I guess that these target applications where the designer wants to isolate the digital electronics from the power electronics. An interesting application, but a difficult one to find a widespread audience.

The special advantage of the chip is it's ability to deal with what you might call ground bounce. In any real implementation of your circuit there will be inductance and resistance in the source path of the NMOS switch. So when you try and turn the MOSFET on and the current builds up the source voltage rises and the MOSFET either current limits (slowing it down) or even oscillates. The Infineon driver is good enough at being differential to reduce or eliminate this effect. The second benefit of the driver chip is that it works with large differential voltages so you can use it to do high side driving with an NMOS power device.

MK

It is a pretty nice driver chip with an interesting differential input. It could solve a nasty issue, but I would tend to solve those issues in other ways during design. It would be good to study this chip so that when you come across an application that needs this functionality, it could save a lot of design effort to implement an equivalent function. Of course it makes a pretty good gate driver even if you don't need differential inputs. I'm not surprised that there aren't  many members who have a need for this chip although the high-side driver capability might be a more common requirement. Maybe you can throw down a gauntlet to postulate interesting applications for this chip.

A gate driver is used to turn-on (and off) a MOSFET from a microcontroller input. The simplest example of a gate driver is using another transistor to turn on the FET. For example, a standard method is to use a small BJT to turn on the MOSFET. So why does that get done?

MOSFETs turn-on (or off) depending on the voltage difference between the GATE and SOURCE pin. In the case of an N-Channel MOSFET, the SOURCE is generally tied directly to GND. So as long as the voltage on the GATE is higher than the MOSFET's Vgs-th voltage, the MOSFET will turn on.

BUT, things are not as simple as that explanation. While it is true, it lacks a few details. The Vgs-th voltage for a MOSFET is actually the threshold that the MOSFET turns-off! In other words, its the minimum voltage necessary to turn on the FET. Anything below it and the FET stays off. Take the popular FQP30N06L. It is called a "logic level" MOSFET because the Vgs threshold is only 2.5 volts. With either a Raspberry Pi's 3.3 volt or Arduino's 5 volt I/O pin, there is enough voltage to go past the Vgs-th value.

However, at 3.3 volts or even 5 volts, the MOSFET is JUST starting to turn on. The resistance between the drain and source (Rds-on) is very high with such small voltages. Even on this "logic level" MOSFET to get the rated "35 mΩ" on-resistance, you need to apply at least 10 volts to the gate.

So where do you get that 10 volts? Well, what if your load was a 12 volt motor? You could use it's power supply to turn on the FET, giving 12 volts across Vgs, and fully turning on the MOSFET. You "just" need to use a gate driver.

The microcontroller controls the gate driver (or NPN's base) which then controls the MOSFET's gate, with a higher voltage. This method does a little bit of electronics irony. When your I/O pin goes high, that turns on the NPN, which means its collector (pin 2) is now at 0 volt. R2 is dropping the entire 12 volts. So the N-Channel FET turns off. To turn it on, you have to use a logic LOW.

---

SOOOO, where does that leave us on the Infineon part? Well, it is a gate driver designed for high voltage, aka "power electronics," circuits. These are MOSFETs operating anywhere between 48 and several hundred volts DC. In those cases, every milliohm of Rds-on wastes a lot of power. So an integrated gate driver like this one simplifies driving those MOSFETs from a PWM or Microcontroller.

The critical feature appears to be the differential inputs. I guess that these target applications where the designer wants to isolate the digital electronics from the power electronics. An interesting application, but a difficult one to find a widespread audience.

For me it's usually a lack of time, but in general for RoadTests I'll consider it if it's in line with a project I have in mind.

I'm not sure what this dev board is for, or what a Gate Driver is in general, which then means I also have no idea of how I'd incorporate it into any of my projects

(The description is kind of above my head too)

-Nico