Getting Ready to RoadTest a Low-Side Gate Driver with Truly Differential Inputs That Prevents False Triggering. Are You Interested?

Question

I'm getting ready to roadtest a 1-channel low-side gate driver EVM that has truly differential inputs to prevent false triggering. What the truly differential inputs do is to overcome driving issues when DC offset (DC GND shift) or AC voltage drops (AC GND shift) are existing between the driver and the MCU ground potentials.

Common situations characterized by DC and AC GND shifts are, respectively, the driving of high-side devices and the driving in noisy environments with significant stray inductance on the ground path.

Some Features

48V half-bridge buck converter tested up to 60W
Integrated section to generate and regulate the DC GND shift
Possibility to easily regulate the AC GND shift
Possibility to compare the 1EDN TDI and standard low side driver in GND shift situation

A Word about Low side gate drivers │ EiceDRIVER 1EDN TDI

The input signal levels of conventional low-side gate driver ICs are referenced to the ground potential of the gate driver IC. If in the application the ground potential of the gate driver IC shifts excessively false triggering of the gate driver IC can occur.

The 1EDN7550 and 1EDN8550 gate driver ICs have truly differential inputs. Their control signal inputs are largely independent from the ground potential. Only the voltage difference between its input contacts is relevant. This prevents false triggering of power MOSFETs.

Here's a schematic block:

Here's some documentation

Product Brief / Datasheet

App Note

I'm getting 10 units to roadtest. I'd like to get some roadtesters that both have the test equipment and a background in this type of circuitry to test this out. If you are interested, please send me a message at rscasny

Thanks.

Randall Scasny

RoadTest Program Manager

Jan Cumps · Answer

I happen to have a GaN half-bridge on the desk at the moment. It's a next step in a blog series I'm writing here on the community: https://www.element14.com/community/groups/power-management/blog/2017/05/14/gan-point-of-load-converter-48v-to-1v-50a-pa… I'd certainly be interested in putting the Infineon design. Here's a board with a GaN half-bridge I made earlier. Blog on element14: Hercules LaunchPad and GaN FETs - Part 3b: BoosterPack Layout - my version

michaelkellett · Answer

The chips are very interesting parts of which I was already aware of from Infineon direct marketing. The EV board is very much designed for evaluation in a a fairly abstract (ie not directly related to a specific product design) environment and the app note that goes with it is very comprehensive. So much so, that there seems little point in actually getting the EV board oneself - I would go straight to a product prototype design - using the EV board and the documentation of the chips and board to inform that design. I'd like to see the parts used to make a 4 switch buck/boost controller (see LTC3780 for example) since this topology, although superficially attractive, seems always to result in truly disgusting noise levels with the power devices utterly failing to turn on and off cleanly. I suspect that the ground bounce immunity of these drivers might be able to make that topology work much better. So an evaluation board, maybe using an FPGA or CPLD to control 4 switches and designed for maybe 6 - 30V in and 12 or 24 V out at 100W (or more) would be a very interesting thing to see. (Of course if it doesn't manage better noise than existing solutions it won't be much use, but I think it could.) MK

hlipka · Answer

You can find a picture on the product homepage: https://www.infineon.com/cms/en/product/evaluation-boards/eval_hb_bc_1edn8550b/

rscasny · Answer

I fixed it

Getting Ready to RoadTest a Low-Side Gate Driver with Truly Differential Inputs That Prevents False Triggering. Are You Interested?

Top Replies