Due's ARM Processor, ESD Protection Diodes or What?

Hi,

I looked at the data sheet for the SAM 3x8E chip and could not find data on the type of ESD protection on the input/output lines.

On the chip used in the Uno for example, the 328P, the data sheet clearly shows the two protection diodes. On the ARM sheet i could not find that information. The data sheets typically give the current rating of the diodes also.

So the question is, does this ARM chip have protection diodes, and if so, what is the current rating of the diodes?

Thanks :slight_smile:

ALL chips these days have protection diodes, or they wouldn't last 5 minutes in the real world. The maximum allowable voltage on any SAM input pin is ~3.6V, and is limited by the ESD diode maximum current. If you are connecting higher voltage signals to the Due, use series resistors that will limit the maximum voltage to less than 3.6V, and/or limit the maximum current INTO the pin to less than 3mA, and you''l have no problems with the diodes. If you search the Atmel forums, you'll find a long discussion on this topic from a few years ago.

Regards, Ray L.

Hello,

Hey thank you Ray for that info. I had looked around quite a bit and could not find anything that specifically shows that information. I dont like guess work, because that is what blows chips unnecessarily. I dont think i can replace that chip on the Due if it blows out, i'd have to buy a whole new Due board.

On many chips they have it right in the data sheet so there is no guess work. For example with many of the PIC chips they spec 20ma for the diodes. This allows us to use some means of protection like a series resistor that you suggested. This is somewhat like how i did it using the PIC line of chips...just make sure the path current can never go above about 1/2 the rating of the diode.

I had guessed that it was lower for the ARM chip because of the lower i/o pin ratings, but didnt want to take any chances with this one :-)

Thanks again, and i'll see if i can find that on the other site too just for reference.

[LATER] I checked 'one' of the forums and they dont have any information, as other people have asked this same question and no one is able to give a direct answer. Do you have a link to that discussion on hand?

I don't know if I'd consider the "input protection" diodes on AVRs (and nearly all CMOS ICs) to be "ESD" protection. ESD is usually considered momentary conditions of quite significant over-voltage relatively high-current ... discharges. The input protection diodes are more for smaller, longer, in-circuit transients and noise...

westfw:
I don’t know if I’d consider the “input protection” diodes on AVRs (and nearly all CMOS ICs) to be “ESD” protection.
ESD is usually considered momentary conditions of quite significant over-voltage relatively high-current … discharges.
The input protection diodes are more for smaller, longer, in-circuit transients and noise…

The ESD diodes ARE the protection for the I/Os. There is nothing else. This I know very well, because I have designed, and managed the design of, dozens of chips up to 6M gates over the last 25 years. The diodes are there specifically to provide a current path to dissipate over-voltage and under-voltage on the pin, and the limiting factor is nearly always the amount of energy required to fuse the diodes, or the wires connecting them. Virtually all commercial chips are designed targeting the same industry-standard specs and validation tests.

Regards,
Ray L.

Yeah, the diodes are all there is. But actual "ESD Protection", IMO, would have claims like you see on rs232 transceivers:

ESD Protection to IEC 1000-4-2 Level 4 ±15kV Air Gap, ±8kV Contact

Which is quite a bit more than the protection diodes in a "typical" chip.

westfw: Yeah, the diodes are all there is. But actual "ESD Protection", IMO, would have claims like you see on rs232 transceivers:Which is quite a bit more than the protection diodes in a "typical" chip.

And how many chips have you personally designed? You're welcome to your own opinion, but it doesn't change the facts. It is the diodes and the physical layout of the I/O pad cells that provide that protection.

Regards, Ray L.

Hello there Ray,

Thanks for the info there.

Why do you suppose they did not put any info on the data sheet for the diodes like they do on most other controller chips i have used in the past?

Also, i guess you are sure that there are ESD diodes on each i/o pin then, one to ground and one to the +3.3v supply i presume.

I am looking into this because the upper ESD diode is the one that is usually used for small currents that arise from an over voltage and would be limited by the external circuit impedance (such as a voltage divider). i've used this technique in the past many times and it seems to work well.

What would you suggest for a test for the upper ESD diode properties? How about a 1k to 10k series resistor and potentiometer, the potentiometer connected to +5v and ground so that the wiper terminal has a range of 0 to +5v. The current into the pin (and thus diode) would be about: (5-3.3)/R=1.7/R amps, so using a 1k would limit current to 1.7ma. Does that sound like a reasonable test for the ARM chip? Note i would be watching the pin voltage itself and the current through the series resistor. If i see the voltage rise to 3.7v with no increase in current flow through the resistor, i would have to stop the test and assume that there was no upper ESD diode. Sound reasonable, or do you have another, better recommendation for performing this kind of test?

Many thanks :-)

I spent over 30 years as an IC circuit designer and have designed many output drivers. The ESD protect devices we used were much more complicated than simple diodes. They were custom devices designed by the process development physics guys. Some companies may use simple diodes. We did not. One of the guys I worked with wrote the book on ESD protection. I cannot speak for ARM.

charliesixpack: I spent over 30 years as an IC circuit designer and have designed many output drivers. The ESD protect devices we used were much more complicated than simple diodes. They were custom devices designed by the process development physics guys. Some companies may use simple diodes. We did not. One of the guys I worked with wrote the book on ESD protection. I cannot speak for ARM.

Hello there,

Well thanks, that's nice to know. But my argument was never about how complicated or simple the diodes were, or how they were made. Just what the continuous current rating is for the Due's ARM processor.