Asking the Ancients about capacitive sensors


Hi, I am working on a project that needs capacitives sensors. I saw the really cool tutorial on making a capacitive sensor with a piece a wire and a 10M resistor but it says that for long distance it's better to use a qt.

I've got a qt140 from when I was in school but I remember that it was really slow, like 1s delay between the moment the copper is touched and the info is send to the computer. Must come from the chipset we used, we didn't have the arduino at the time :'(.

To describe the thing. It's a 1 meter high "bag" filled with foam. I need to have long capacitives sensors running along it ( top to bottom ), spaced approximately by 15cm. I need those sensors so I know approximately where the hand is placed on the bag.

A guy on the forum says that the qt are not that complicated to used if you are using the simple one. But is here really a difference between a 4 and 10 inputs? Is it complicated to "plug" the qt to the arduino ? If anyone have any experience with those and could put me on the track, that would really be great.

Thanks for you time!

Sorry for my bad english.


I know a guy who built a floor sensor using the QT products. He had some panels engineered, about 25Cm square each, with 100 touch-sensitive spots on each. Worked great. Had them manufactured. Looked great, top quality and excellent engineering… or so it appeared.

It turned out that when he put a grid of these panels together ( it was a floor sensor) the fields of QT chips interacted with each other. Many of the panels ended up in the trash, they were unusable for the original purpose. So you should be aware that the QT products are pretty carefully engineered; if you try to make them do things they weren’t specifically engineered for, like work together in a grid, you will have problems. (…The engineer responsible for the above design couldn’t fix it.)

Have you considered using conductive foam sensors? You could make a sort of grid. Also there is a company in Victoria, BC, Canada called Tactex that makes conductive fabric. The conductive foam solution works great for touch sensing. See the Playground exhibition section, the one called “customers design with their fists”.

Finally, a good rule of thumb in designing this kind of thing is to ask if you have ever seen such an interface available commercially. this will tell you if the technology is stable enough and cheap enough to manufacture and have it actually work. I haven’t seen anything bigger than maybe 100 sq cm that is really touch sensitive, have you? Experimental technologies are sometimes trouble because the “experimenting” take $$$ and lots of time. Even for a hobbyist thtis can mean $500 and fifty days… So you have to be smart about your design from the beginning, and realize your capabilities, and ask how much time and money you have to spare.

This is one of the interesting side-effects of the Arduino project: it makes electronic design so modular that something that even two years ago would have been beyond the reach of a hobbyist now appears to be simple. A lot of things are simple with Arduino, but for really complicated things like a touch-sensitive grid array, you are still going to need the $$$ and time, time and more time. You just have to remember, in the case of things like the QTProx sensors, orsay, an Ipod, that many hours of fine-tuning and preparation have gone into making it look simple. If you want to change that simplicity to another design, it’s probably goign to be very complicated.


PS: I feel ancient, but I hope i don’t look ancient :slight_smile:

Thanks Daniel.

I forgot about that interactive bag. I was going to do the classic home-made touch sensor made out of foil ( or by cutting a dance dance carpet ), but theirs are force sensitive so I am defenetly going to try that. Even if I don't know how their thing is force sensitive compare to the foil technique.

I though about textile sensors but I assumed it was going to cost us to much money. We have to much surface and volume to do. For the qt sensors, I feel really sorry for the guy. He must have been really upset against Qprox! If I use them it won't be has complicated as this guy so it might still be a way of doing some of it. I don't need a precise grid for that, just a 4 or 5 lines. The installation is composed of several elements that are using differents type of sensors ( like non mercury vibration sensors ), so this is just a element, people won't focus only on it.

I am not going to try to challenge people who have studied their products for months or years. The main focus of our project if not only on the interactivity but on the environment shaped with sound and volume made with fabrics. Although it would be a real shame if the interactivity was rubbish. It's really a short description, we hope is going to look good. The volumes looks already great. We will propose the installation to the exhibition page, just in case ;) We are exhibiting it in Brighton ( UK ).

Thank your for your answer.


the guy I'm talking about used a commercial engineer, not a QTprox one. It was expeeeeensive. :( that stuff is such a headache, as you're talking about moving electromagnetic fields essentially... Anyway, good luck.