Capacitive Touch Sense

[chrisstmk]

Hello.

I assembled the circuit for cap. touch sens with a pcb. It works very fine. But now i would connect a 2m long wire as antenna. So wath happens? I destroy one 12F615 after the other! I think the problem is a to high (induced) voltage. Could that be? I try to save the PIC with resistor, with coil, with zenerdiode....don't work! After a couple of seconds (from 2sec to max. 1 min) the PIC is dead!

I gave my promise to a freind that he could have it this weekend. He include the antenna in his stairs and want to controll a LED lighting with this sensor. The electronic for controlling the LEDs is no Problem for me. The registration of the touch is the big problem! Please help me!

thx
Chris

[kam]

WOWWWW!!! What an interesting post!
First, show me which circuit you've used, i suppose the PIC Frequency-Change Capacitance Touch Button

If you check this page (How Touch Buttons work)  i have a video for the ac hum touch sensor in which  show how much voltage can be induced from AC power sources.
1- Do you have such power sources near the antenna?

Also, do not forget that PIC are sensitive to high ESD
2- Do you touch the antenna bare handed?
3- Is there any other sort of physical interaction with he metal of the antenna?
4- Is the wire of the antenna sealed with plastic or bare metal?

5-Is it possible to use an oscilloscope and probe the antenna?

You mention:

After a couple of seconds (from 2sec to max. 1 min) the PIC is dead!

6- This happens even if the antenna is not touched? For example, if you just leave the circuit running without being close, the pic will fry?

Finally
7- what sort of power supply you use? Is it isolated from AC???


Edit:
8- Some photos from the antenna and the circuit overall would help

[chrisstmk]

THX for your fast reply!

Yes, I've used the linked Circuit.

1- there is no power source near the antenna
2- No, the wire is isolated, but doesn't matter - it works bare handed without antenna
3- No
4- it is a 0,6mm² isolatet wire (e.g. for breadboardwork)
5- I use a oscilloscope, i think i don't see a differnce, but the time to the fry was everytime very short.
6- i don't have any Pics anymore, so i can't try
7- It's a galvanically isolated power source
8- There are 5 components (PIC, 100nF, 10pF, 47K, BAT46) in a compact design (25mm x 25mm)

[chrisstmk]

I can't understand. I use a fix capacitance of 10pF and try different sizes of Sensors. It works fine. I try it with alumionium foil, copper boards, ... Everything works fine.

I also tryed it with a wire antenna in the past, it doesn't work good with the wire but i never fry a PIC. And now: Same circuit, same workshop, same kind of wire, I doesn't change anything, but every PIC fry.... I bild a new hardware, for 3 or 4 times, but everytime the same problem.

Sometimes (after the fry) the PIC makes a shortcircuit so that it get hot and take 0,5-1A. Sometime the GP1 is high after fry sometimes it's low. But i don't think that this information is very helpful..

On monday i'll post some detail images, but you can see the whole project here:

http://www.facebook.com/media/set/?set=a.544569918902790.140989.349926295033821&type=3

[kam]

hey! noch einer Freund aus Deutschland!   

As i see from the photos, you're preparing a very cool job! Should you finish this job, let us know - i will share it in my facebook page as well. Now i have the full picture of what you're about to make.

- One question: How many wires are in each antenna? Do you have also a ground wire in the antenna? If yes, then make sure that the ground wire and the pad wire are not connected to each other

I totally do not believe in ghosts whatsoever, BUT your case gives me the creeps. It so happens to have a degraded or burned I/O port, if it has been electrostatically damaged or if more than 20mA have gone through for extended period of time. Your case drives me to a dead-end. The only thing that i can think of, is that the pic is already degraded prior your test due to electrostatic discharge. During assembly for example... It has happen to me many times. If your pic is already installed on the board, it is harder to fry it with electrostacit discharge because of the grounding layers. But when you receive the PIC and touch it bare-handed to place it on the board, this may mean sudden death. Do you take all ESD precautions during installation? Like for example, ESD wrist-wrap and isolated soldering iron.


One last thought: In my opinion, one wire will not work properly. The wooden step seems to be very thick (seems thicker than 30mm to me). Add to this the shoes... We're talking about a thickness of 35mm or more. You definitely need more sensing area underneath. My approach would have been different: A thin 3mm plywood would be the base, on which i would experiment with different pad sizes made of aluminum or copper to ensure 100% reliability even for a kid with small feet.

[chrisstmk]

I'm not from German, I'm from Austria 

The antenna is just one wire. The wood is 40mm. I try it befor we install it at the stair, and it worked. The problem is that the antenna is now fix mounted in the first and in the last step. There is no way to change it anymore.

ESD is no Problem i think, i don't use any ESD prodection, but the PIC allways run for a short time, and then fry...

I think I try it a 100times befor with aluminium foil, and everytime it worked (perfect), and now it fry also with the foil. There must be witches in my workshop!!!!

Tomorrow I get new PICs and try it again. It must work!

[kam]

The problem with ESD (and overheating during soldering) is that, sometimes it does not immediately kill the chip, but it degrades it causing curious behavior. Some time ago, i had such a curious problem with a chip that worked perfectly fine, only that the current it was drawing during operation was abnormally high - if i remember well about 20mA with no load, and around 2 mA during sleep. After tons of senseless search, i turned to microchip and reported the fact, only to tell me that ESD or overheating during soldering can cause such behavior.

But you did not answer - the antenna has one or two wires inside?


PS I was in Austria before summer for training, in Absam near Innsbruck, Tirol, and went into a restaurant for eating, and noticed that despite being thursday night the restaurant was full of people. And then, suddenly, a crazy mob wearing costumes and making noise went into the restaurant. Some where dancing, others dressed like witches were screaming, and then there was this guy dressed like a bear, chained to a brutal fat dude with a club.... And everyone were going around the people, hitting them in the back and giving them a liqueur to drink. Definitely the best day of the best trip I've ever made so far!!! I think it was called Absam Matchgerer

[chrisstmk]

I don't think it's ESD. I never have problems with it in the past... I got new PICs today, the first I fry in 30sec (it was the last PIC from last week). Now i replaced it and disconnect the antenna. It runs and runs and runs... I connect the antenna again and it runs and runs and runs... I disconnect the osci, so it works much better! Maybe I really wear very bad-ESD-clothes last week? I'll take a video for you.

PS: Where are you from?

[chrisstmk]

Here is the video. Now it runs! I don't change anything... Maybe the last turn of Pics have a Problem? But I never have dead new PICs befor..

You can't see it in the video but i don't have to touch the tabel on the bottomside. A disdance of 5mm is enough to switch.

[kam]

Probably the PICs had indeed problem. This makes much more sense than anything else. I really like your approach with the wire antenna. Works even under a 40mm chipboard... Very interesting. Frankly speaking, i did not expect to work at that thickness with only wire.... Which gives me some ideas as well....

So, i'm looking forward to see the video with the circuit mounted on the stairs!!!!!

[cheerio]

i would always prefer aluminum tape for a smth like you are building.

[chrisstmk]

I'm so sorry that I can't post new photos or videos. I'm so busy: other projects (for a change from paying customers ), arrangements for my attendance at electronica12 in munich, ... I hope I can make it until weekend ...

PS: Do anyone else visit electronica?