Can I use serial resistors to have "some" ESD protection on IO pins? Would 100ohms help at all or should I just remove them?

[Bageraz]

I am using STM32F207 on the main board for an industrial remote control. The radio connector is connected to the MCU.

[Danish1]

Everything helps.

100 Ohms isn't much, but sometimes that's all you can have in series with the signals - depending what the signals are carrying.

Where possible I would want higher series resistance (won't work on the I2C lines given your choice of pull-up). I would also consider diode clamps to Vdd and Vss on the connector-side of the resistors.

I think the "gold standard" is opto-isolation, but I don't think that can be used with I2C and it could well be overkill - depending on your application.

How "far" is it from the radio to the MCU - physically (both in terms of actual distance and length-of-wires), electrically (do they share the same power-supply, ground)?

Is there the possibility of inductive-coupling onto those wires (is it a high-power radio; are there any relays or strong magnetic-fields)?

Incidentally, have you checked what the stray capacitance of the entire signal run will be; that might affect how fast you can run I2C and what pull-up resistors you need.

Also consider cross-talk between the signals. Quite often connectors e.g. for disk drives would have every second cable grounded to prevent crosstalk. I note that you have a couple of grounds on the connector - on your schematic they're shown to be close to the I2C lines, but given the pin numbering of the connector I can't see if they help.

Regards,

Danish

[Bageraz]

Opto-isolation would be overkill for sure i think.

The actual distance is about 7cm, and the trace distance of the SDA and SCL lines are about 6-7cm. The other traces are about 10cm long

They share the same power supply and ground, yes.

The radio is 2,4GHz 5mW. No high power.

There are no relays. However, there are magnetic fields coming from magnets (strong pot magnets 2cm apart, facing away from the connector)

The board is already tested without any disfunction on the radio communication/ i2c lines.

The pins on the connector symbol is actually placed as the real pins are. So what you see on the symbol is what you will see on the physical connector.

The series resistors package size and power rating is a little tricky to determine. The purpose is ESD protection which means short duration of a current.

In my case the resistors are 0402 and 1/16W, quite small, but are they too small regarding airgap between resistor pins/ pads, thermal considerations regarding current due to duration of an ESD event, voltage rating and the resistor value, and also, should/ can I go bigger in resistor value for better protection (pullups should be higher perhaps)?

[LCE]

Ah, so this "radio connector" will be inside the final product, so that the end user will not be able to access / contact these pins? And it will be assembled in ESD-protected surroundings?

If yes to both, then the 100 ohms might not be necessary, but I tend to uses such series resistors anyway - if space, cost, signals allow.

If no, then add the above mentioned diodes to VCC/ GND, or some TVS diodes, "strength" depending on accessability of these pins in the final product.

And if you have the space and money, use gas discharge tubes! ;)

[Bageraz]

Yes that is true. The end user will not access the connector. However, its not only the end user this protection should be aimed against. As the assembly house should be an ESD-friendly place, its not always the case. A survey from the EOS industry council reviles that assembly house does not always follow the ESD rules; see the last link bellow.

Nice research on ESD protection:

https://www.st.com/content/ccc/resource/technical/document/technical_note/group0/6e/dd/e8/bb/62/67/43/80/DM00670774/files/DM00670774.pdf/jcr:content/translations/en.DM00670774.pdf

A survey from the EOS industry council (found in the link above):

https://www.esdindustrycouncil.org/ic/docs/Industry%20Council%20White%20Paper%204%20rev1.2.pdf

[LCE]

Interesting! And terrifying...

Good to know that one should check and advise external manufacturers.

I expect an ESD-safe environment to be standard for electronic manufacturers at all stages of production.