Static Electricity Protection

We’re trying to relay status of a charging base to a computer - that part is working. However, when the user walks around their body builds up static electricity and the static electricity gets transferred into the charging base when the device is returned to the cradle. Our Wixel connects to an undocumented header on the base, which provides 3.1v, ground, and charging LED status. Though we are using an undocumented “feature”, power spikes like this were unexpected - we felt the power from the base would have been regulated better.

I was curious if anyone has had a similar problem, and what kind of protection circuit was used. We’re looking into a few different methods, but thought I’d ask around here as well.

Thanks!

For ESD protection, TVS diodes are great! For power, you can use high-power-rated TVS-es (1500W, 5 kW even.) For serial communication, you want lower ratings to provide less capacitive load on the line.
A TVS diode is basically a form of Zener diode – if you have some 3.3V or 5V Zeners laying around the toolbox, you could start by trying those, although they may not be up to the abuse of significant static dissipation.

Ah - thanks! I was reading up on TVS diodes this morning, and they sounded like what we need, but wasn’t sure.

We ended up choosing this: katalog.we-online.de/pbs/datasheet/824013.pdf

Soldered a chip onto a prototyping board, and hooked it up to the data lines. We were consistently able to crash the Wixel with a static charge without the TVS - and with the TVS we could not get the Wixel to crash.

One thing that concerns me - and I don’t know enough about static electricity to fully understand what is happening - is whether this higher voltage will shorten the operating life of the Wixel.

For both protected and unprotected circuits, the duration above 5 V lasted for no longer than 20 micro seconds. Unprotected we were able to measure voltages as high as 13 V - protected, the voltages were clamped around 9 V. However, with the oscilloscope we used, I’m not sure we were reading the full voltage in the unprotected state. Another oscilloscope (an Oscium) captured a much higher voltage peak that went past the threshold for how much it could measure, but we didn’t use that this time.

The question I have is: with a voltage of 9v for such a short duration - will the microcontroller have a shorter lifespan?

Also, as part of the final circuit, we are putting in a fuse. From what I’ve read online, TVS diodes can fail as a short, and if that happens, 1.5A will flow unimpeded.

Thanks!

John

Oh - we’re using pins: P0_4, P0_5, P1_6, P1_7, 3.3V.

I am not sure if applying 5V or 9V for short durations will negatively impact the lifespan of the microcontroller. According to the “Absolute Maximum Ratings” section of the CC2511F32 datasheet, the maximum voltage on any digital pin is VDD + 0.3, which is significantly lower than 5V. However, it also mentions ESD ratings of 750V and 500V for the human-body model and the charged-device model, respectively. You might find additional useful information in the datasheet.

- Jeremy

I was wrong about the 5 V - should have written 3.3 V.

Regarding the data sheet… it is very likely without the TVS we were going over 500/750+ V, just weren’t able to measure it. With the TVS, the maximum ESD discharge voltage seems to be around 9 V, well within those limits.