ESD Protection

[schlaray]

Can anyone cite some references for ESD protection? I have a home application using a PIC and I want to insure it is not destroyed by a spark discharge after walking on a rug. I assume that all external inputs should have ESD protection along with the power terminals.

Schlaray

[mister_e]

all unused i/o have to be tied somewhere (Vdd, Vss), via resistor or not.

[Russ Kincaid]

For an active input, I use a 10K resistor in series. If the input can tolerate it, I also use a .01uF cap from input to gnd. Not much you can do for an output except diodes to VSS and VDD.

[T.Jackson]

I think problems with ESD Electro Static Discharge tend to be overly exaggerated. I've done all sort of projects with CMOS IC's and never had one go on me. As Mister_e pointed out - tie all unused inputs to ground and this should be enough. PIC's have built in diode clamps too, you stand a better chance at winning the lottery than damaging your PIC through ESD.

[skimask]

Can anyone cite some references for ESD protection? I have a home application using a PIC and I want to insure it is not destroyed by a spark discharge after walking on a rug. I assume that all external inputs should have ESD protection along with the power terminals.

Schlaray

I've zapped various PICs in my carpeted lab with audible, visible, tangible sparks numerous times. I think I've only actually fried one or two pins on one or two PICs over the last 7-8 years.
But at the same time, I also just got done with a Class 3 Micro-Mini soldering school which put special emphasis on ESD, grounding, etc for about a week.
We got to see really good pictures of the damage that can be caused by sparks that can't be seen, heard or felt.
So, I know it's entirely possible to fry something without even noticing that you just zapped it, but I also know that it's also possible NOT to fry something by zapping it and feeling that lightning bolt course thru your arm and into that silicon.
If the part is installed on a PCB, not much to worry about, plenty of ground paths... If the part is laying on a piece of something with good insulation properties and you zap it without a halfway decent ground, you might do some damage...
I keep a few chunks of copper on my workspace and ground myself regularly (like every change I feel like it) to help out, I also have a cheapo wrist strap, don't know if it actually helps but it makes me feel good.
My two cents...

[Charles Linquis]

PICs are pretty tough, but I design a lot of military stuff that has to pass ESD testing. This is my checklist of things to do - and it has always proven successful. I have drawn sparks to the PCB several times, and the PIC didn't even reset.


On all inputs where signals come from "off the board" either buffer the input
with a logic gate or put a 2K resistor in series with the input pin and a .01uf cap from the pin to GND.

Buffer all outputs that go "off the board" with logic or transistors.

Make sure MCLR is pulled up to VCC with a 4.7K (or smaller) resistor. Keep this line short.

Use a 4 layer (or more) PCB with power and GND planes.

Use 4 .1uF and 2 10uF low-ESR tantalums across the power pins.

Put a reverse-biased diode across the power pins.

Use a 7V TRANSZORB across the power pins.

Keep the lines to the crystal very short.

[precision]

and keep your running project in microwawe
high microwawe
pause 1000
low microwawe
if pic is safe THEN
your project is safe from ESD
ELSE
Keep more prtection From ESD
ENDIF

[schlaray]

I thank everyone for their input. Along with ESD protection I should have included any and all other sources of distruction. Lightning strikes and an industrial type enviorment should have been included. Many years ago I lost an IBM Aptiva computer when there was a strike close to my home. All other electronic equipment survived.