ESD protection for programming pins

[rt_joerg]

Hello everybody,

for a small application I want to use a PIC16F917. The pcb will be put into a box which will be filled with potting compound.
Since it might be needed to update the software later on I need access to the programming pins and VPP to connect them with the PICkit. Therefore those pins will be routed to a plug.
My question is about ESD protection. Does anyone know a proper IC to protect those pins against ESD discharges?
Or does anyone know a cheap and easy solution without an expensive “special IC”?

[treitmey]

As you have read in the spec sheet, there is some protection provided by clamping diodes,..ect. I have never used more than that.
Check this appnote.
http://support2.microchip.com/KBSearch/KB_StdProb.aspx?ID=SQ6UJ9A001VUZ
http://ww1.microchip.com/downloads/en/AppNotes/00595.pdf

On other pins I've use the sp720.

Also I always thought that if I potted a circuit I wanted to try a Infrared interface. only thing exposed is power/gnd, and LED receiver.

Has anyone done that before?? I suppose it would have to be a lvp (low voltage programming) setup since I wasn't thinking about bringing the mclr up to the high programming voltage.

[newguy]

I too have used the SP720 and recently the SP721 (ROHS compliant SP720). Just a word of warning regarding the SP721....

If your application will never see temperature extremes, this chip works very well. However, I recently had big issues with a product of mine and traced the problem to the SP721.

This chip is rated for operation down to -40C. It works fine above 0C, and it also works fine below approximately -5C to -10C. Between 0C and -5C to -10C or so, it acts funny. The protect inputs of this chip simply clamp the voltage on them to lie between the Vcc and ground potentials connected to it. They're not supposed to load external circuitry at all.

Well, in that narrow temperature range just below the freezing point, pin 1 on the SP721 in particular, and one other as well would load the line it was connected to. My power supply is 5V, and in that narrow temperature range, pin 1 wouldn't allow the voltage on it to rise above 3.0V. Took me 3 days to isolate and identify the strange behaviour I was seeing.

I've also used these parts for ESD protection: http://www.littelfuse.com/data/en/Data_Sheets/ML.pdf and they work very well (at all temperatures too).

[treitmey]

Newguy, you saved my butt.
I was just about to use this chip again,.. in an outdoor product.
I will rethink this chip now.
THANKS ! ! !

[newguy]

Glad I could help. When I first found the problem I had trouble believing what I was seeing, but time & again when I blasted that chip with cold spray it would do the same thing. I took it out and the problem went away.

I emailed the manufacturer to let them know but no reply so far. I imagine that if they can duplicate the problem that they're probably sh***ing a brick right now.

[treitmey]

I use sp720 normally at 5V 10k pull-up and with 2 pins in parallel.
I'll see if it is reproducible with sp720.
I'll see if I can get my hands on some 721's from digikey.

[PCM programmer]

I have a board right here on my bench that has SP721 chips on the
input side of a 16-bit parallel port.

The circuit looks like this:

1. A 50-pin shrouded 3M header has 16 digital input pins.
2. Each input signal first has a 10K pullup to +5v.
3. Each input signal then goes through a 100 ohm series resistor.
4. After the series resistor, each signal goes to a pin on an SP721.
The chips are marked with "721A". The V+ and V- pins are
connected to +5v and Ground.
5. Then each input signal goes to the "B" side of a SN74LVT245B buffer.
6. The buffer is then read by a micro-controller (non-PIC).

What input voltage should I apply to the input pins, in order to test
if I can duplicate your problem ? I have a can of freeze spray to
do this test.

[newguy]

My circuit is a bit different from yours....

Mine, in a nutshell, has 100k pullups to 5V. These pullups go to the inputs of a PIC and to the SP721. External (to the PCB) switches ground these lines, which the PIC detects (user input). Pin 1 on this device was the worst, and one other pin was also affected. The rest of the inputs on it weren't affected at all. There's a little more to the circuit, but it's not relevant.

If I measured the voltage at pin 1 of the SP721, I'd see 5V at room temperature and immediately after I sprayed it with freeze spray. I know the chip got down to around -40C, and it worked fine at that temperature. After a couple of minutes, as the temperature of the chip rose, the voltage would dramatically fall to 3.0V and came back up as it warmed some more. The temperature this happened at was just below freezing, based half on measurement and half on educated guess. [My temperature measurement devices are rather rudimentary.]

If I froze the SP721 alone, I'd see the behaviour I mentioned. With the PIC removed, I still saw the behaviour. With the SP721 removed, the behaviour stopped.

Let me know if does the same thing with your setup. 10k vs 100k pullups may make a difference.

[PCM programmer]

I tested it and I didn't get the problem. I froze the SP721 with chill
spray with 3 or 4 blasts. I don't know how cold it gets, but on the can
it says it will freeze things down to -51 degrees C. Then I put my
scope probe on pin 1 of the SP721 and waited 2 or 3 minutes while
it warmed up. It always stayed at +5v. I tested two different SP721
chips and got the same results.

The data sheet for the SP721 claims the leakage current is 20 na, max.
So you would think a 100K pullup would work. But maybe that is the
problem, as you suggested. My board has SOIC R-packs soldered in,
so it's not easy for me to switch to 100K pull-ups. Maybe you should
test it on your end.

[newguy]

The product is out of my hands at the moment, but even if it was here I'd have difficulty testing with a 10k pullup because everything is surface mount.

Whoops, just noticed the SP721 is the 6 input device and the SP720 is the 14. I had problems with the 721 (6 input), but I did notice that the SP720 did act similarly, though not as bad. It's tough to compare because I have the SP720 in a different board and it is connected to 12V for Vcc. However, it is used similarly, protecting pullups to 12V. When I sprayed the 720 with freeze spray, the voltage on pin 1 would fall a little (like from 12V to 11.7V). I didn't notice the same problem on other pins. This happened at the same temperature too - just below freezing.

[newguy]

A support engineer from Littelfuse just contacted me and has requested that I send them samples of the chips that were acting up. I'll report back when/if I hear anything.

[rt_joerg]

I am looking forward to the answer from littlefuse.

But nevertheless I will try the SP720 / SP721, because I don’t need the programming pins at those temperatures.

So far thanks to all.