UCC21320-Q1: UCC21320 VDDA&VDDB fail

Eric Liu(SZ-FAE)

Part Number: UCC21320-Q1
Other Parts Discussed in Thread: UCC21550-Q1

Tool/software:

Hi team,

For this device, our customer need cover ICT test ,when unpowered up , they will inject a external voltage into UCC21320 VDDA& VDDB pin, they found that this device output abnormal when do the functions test. input a pwm into the IN pin ,there is no output.

they need know,

1. can they inject a voltage to the VDDA and VDDB pin when unpowered up?

2, when powered up ,if VDDA failed, will the output abnormal with no PWM wave?

3.When unpowered up, What 's the VDDA state? and the VDDA can withstand -0.3v-30v voltage too?

28 days ago

0 Hiroki Honda 27 days ago

TI__Expert 3631 points

Hi Eric,

Yes, they can inject a voltage to VDDx when the device is not powered up on the primary side
Correct, if the output is not showing a PWM signal when inputting a PWM signal and VDD is powered on, that channel is most likely not functional.
Powering up the primary side will have no affect on the secondary side due to the isolation.

As long as the injected voltage is not surpassing the limitations of the driver, there should not be any failures. Some things to consider would be the edge rate (dV/dt) of the injected test voltage. For this device, there is a limitation of 5V/us on the VDD pin. The newer gen devices such as UCC21550-Q1 has a much more robust design on the VDD pin.

Kindly refer to this FAQ for this topic: https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1338508/faq-why-is-ucc2155x-more-robust-than-ucc215xx-against-vdd-supply-noise?tisearch=e2e-sitesearch&keymatch=UCC215xx%20faq#

Regards,

Hiroki

0 Eric Liu(SZ-FAE) 27 days ago in reply to Hiroki Honda

TI__Intellectual 1350 points

Hi Hiroki,

Thanks for your sharing, so the UCC2155X ESD cell will be triggered by the voltage level rather than the voltage edge, is this right? so for the UCC21320 , its ESD cell will be triggered by the voltage edge, so we need to control the VDD noise. is this right?

BTW, what actions will be taken by UCC21320 when we inject a external noise source and the device is unpowered.

0 Eric Liu(SZ-FAE) 27 days ago in reply to Eric Liu(SZ-FAE)

TI__Intellectual 1350 points

Do we have a description about this picture?

what's the withstand voltage of those ESD cell? if ucc21320 is unpowered, we inject a 1KHZ, 200mV voltage source into the VDDA pin, will this hurt the ESD cell and the VDDA pin?

0 Hiroki Honda 24 days ago in reply to Eric Liu(SZ-FAE)

TI__Expert 3631 points

Hi Eric,

Correct, UCC21320 is edge triggered so VDD noise slew rates are what triggers the ESD cell here. With the device unpowered on the secondary side, the ESD cell would not be clamping current as it would when the device is powered. The voltage across the internal ESD MOSFET clamp would be too low to induce high enough current to damage the driver.

A 1kV 200mV noise signal injected into VDDA with VDDA unpowered should not cause damage to the driver.

The robustness of the driver against noise also highly depends on the layout and decoupling capacitor placement and size. Even with low injected noise, if the decoupling capacitors and layout are not ideal it can induce more noise.

Hope this helps.

Regards,

Hiroki

0 Eric Liu(SZ-FAE) 11 days ago in reply to Hiroki Honda

TI__Intellectual 1350 points

Hi Hiroki,

when VDDA and VSSA unpowered, the ESD cell won't works? is this right?

and if the VDD short to the OUT, How can we check whether the ESD cell failed or the FET between the VDD and OUT failed

0 Hiroki Honda 10 days ago in reply to Eric Liu(SZ-FAE)

TI__Expert 3631 points

Hi Eric,

For UCC21320, the ESD may be triggered if VDDA-VSSA sees a sharp transient.

For VDD-OUT shorts, this is typically the result of failure in the HS FET of the output stage. This can occur when the abs max specifications are being exceeded on VDD/OUT pins. When the OUT pin is biased higher than the VDD, this will cause the internal body diode to start conducting which can cause thermal overstress to the body diode.

When running these tests, please ensure that the absolute maximum specifications of the pins are being met!

Please let me know if you have additional questions.

Regards,

Hiroki

0 Eric Liu(SZ-FAE) 9 days ago in reply to Hiroki Honda

TI__Intellectual 1350 points

Hi Hiroki，

you means the VDD-OUT shorts can only occur when the HS FET failed？ how about the ESD cell？ if the ESD cell failed， the VDD will also short to OUT， is this right？

for the internal ESD forward voltage , it is 0.3V, is this right?