TMS320C28341: I/O pin protection

Larry Richardson37

Part Number: TMS320C28341
Other Parts Discussed in Thread: TMS320F2808,

Hello,

I am designing lightning protection circuitry around the I/O pins for the TMS320F2808 processor. I couldn’t find this in the input menu, so I chose one close. The test waveforms to be injected on the input connector pins conform to the airborne standard RTCA/DO160G Section 22, waveforms 3/3 and 4/1 at level 3.

My approach is to augment the internal clamp diodes with external schottky clamps, as illustrated below:

My question is; what is the nominal forward drop across the internal clamp diodes for the GPIO I/O pins for this processor?

I have simulated this with a relatively high drop schottky, the MBR21011CT, but I am just guessing.

Also, I assume that the maximum spec for input clamp current (±20ma) refers to the maximum current through the internal clamp diodes.

Thanks,

Larry Richardson

over 6 years ago

0 Hareesh Janakiraman over 6 years ago

TI__Guru* 93755 points

Larry,
I am unable to see the image. Could you attach it as a file? On a different note, the "Part number" on the title is TMS320C28341, but the body of the post refers to TMS320F2808. Which is the correct part number?

0 Jason Whiles over 6 years ago

TI__Expert 4810 points

Larry,

Can't see the image and I'm not personally familiar with the RTCA standard you mention but I assume this is protection for pins which could see a system level ESD event cause high pin voltages.

Nominally the clamp diodes are about 0.7v, but they can begin conducting weakly at as little as 0.3v (will vary by device and temperature). The 0.3v level is more of a consideration to avoid for continuous over-voltage conditions, however we don't have a specified limit for when the clamp diodes turn on.

The 20mA maximum is for all components in the circuit, this shouldn't be exceeded whether it is through the ESD clamp diode or the Output buffer. The approach of external schottky's is good, as well as a series resistor to the F2808 device to limit the current to below 20mA.

You might check out the IBIS models as well, you can find them here:

Best regards,

Jason

0 Larry Richardson37 over 6 years ago in reply to Hareesh Janakiraman

Prodigy 40 points

Jason,

This reply resolves my issue. I already have 50 ohms in series with the GPIO pins, but can't go much higher because of rise/fall time requirements downstream. The DO160G requires a 300V sustained transient to be applied to the input pin, which results in some pretty hefty currents internally. I have two parallel low-drop schottkys on the two pins of concern, and am modeling the internal clamps with higher drop diodes. I just wanted to make sure the internals were indeed higher drop than my externals.

I have attached an image of my LTSpice simulation since it didn't make it originally.