Can the ground B (GNDB ) of ADuM4221-1 be driven at -4V?

Hello,

The ADuM4221-1 is able to operate SiC devices that require a negative voltage, but there is a special way that this is achieved. The ADuM4221-1 itself cannot provide a negative voltage relative to it's GNDA or GNDB pins, but it can provide a negative voltage with respect to the source of the SiC device. The way that we can achieve this is by offsetting the GNDA/B of the ADuM4221-1 relative to the source of the SiC device. Here is an article that talks about how this can be achieved. If you have a single output isolated power supply, you can get a negative supply by using a Zener diode. If you have a bipolar isolated supply, it's just a matter of hooking up the bipolar power supply connections to the correct points.

https://www.analog.com/en/analog-dialogue/raqs/raq-issue-158.html

Please let me know if you have any other questions.

Thanks,

RSchnell

Hi,

Thank you for your response.

1. This is helpful. I wush to know if ADuM4221-1 has a short circuit protection feature like DESAT. This was not mentioned in the datasheet.

2. I can see that ADuM4221-1 can be oprated in a half bridge configuration. Is there a similar driver operable in full bridge configuration (four chanbels) capable of delivering 4A/channel output with DESAT and miller protection features.

Thanks

Hello,

The ADuM4221-1 does not have DESAT protection. Our current offerings that include DESAT are the single channel ADuM4135 and ADuM4136 devices. The ADuM4137 and ADuM4138 also include DESAT, but they are much more complex parts, and include many more protection features including DESAT.

One benefit of using the ADuM4221-1 in a half-bridge is that since it has two channels, and controls those channels, it can provide deadtime control between the channels. It is also possible to operate a full-bridge with two ADuM4221-1 devices, or even four single channel devices. ADI does not currently have an isolated full-bridge driver. If you need DESAT protection in a full-bridge, I think using four ADuM4135 drivers would be the best fit.

RSchnell

Hello,

The ADuM4221-1 does not have DESAT protection. Our current offerings that include DESAT are the single channel ADuM4135 and ADuM4136 devices. The ADuM4137 and ADuM4138 also include DESAT, but they are much more complex parts, and include many more protection features including DESAT.

One benefit of using the ADuM4221-1 in a half-bridge is that since it has two channels, and controls those channels, it can provide deadtime control between the channels. It is also possible to operate a full-bridge with two ADuM4221-1 devices, or even four single channel devices. ADI does not currently have an isolated full-bridge driver. If you need DESAT protection in a full-bridge, I think using four ADuM4135 drivers would be the best fit.

RSchnell

Hello,

Quite useful. Please accept my appreciation. Again,

With four ADuM4135 for full bridge configuration, I've got a challenge. I have six full bridges to control independently. Therefore I'll require twenty four ADuM4135. The challenge now is:

1. I'll need a microcontroller with at least 24 independently programmable channels to control the 24 PWM (Such isn't common. I don't know anyone yet).

2. Would they (ADuM4135) require special shielding from each other?

Kindly advise.

Hello,

The ADuM4135 drivers do not need to be shielded from one another. The ADuM4135 drivers are also natively bipolar, so you can hook up the -4 V between VSS2 and GND2, and the part will operate as expected.

Multi-phase systems so get a bit crowded with PWM signals. You might be able to save some GPIO's by creating discrete deadtime controllers out of glue logic between the controller and the gate driver pairs. This would reduce your GPIO count in half, but would require external components. I haven't looked for discrete deadtime controllers lately, and there might be one already in existence.

One way I know of doing this is with NAND gates and controlling the RC on their inputs. It's simple, but there are more robust methods available. The ADuM4135 has VI+ and VI- inputs which you can use to prevent static shoot-through errors by routing both highside and lowside signals to each ADuM4135.

Let me know if you'd like more details and we can start brainstorming some designs.

RSchnell

Hello 

Thank you for your advice. I'll explore this approach but seems I'll still need a microcontroller with at least 12 independently programmable channels. Is this feasible please?

Regards