Force/Contact and temperature sensor for robotic hand

[umerwahla]

Hi all,
I am working on a project where im making a robotis hand. The basic structure, skin and servo motors are all set up. But I am having trouble in finding good tactile sensors.  I am looking for the following sensors:-
- Force + Contact Sensor
I have tested QTC as the force sensor but I am not satisfied with its performance. I am using a very simply potential divider circuit with the 4 QTC Sensors in wheatstone bridge configuration. The sensitivity is very low and it requires too much force to show the required change in resistance. Is it possible to change the circuit in any way to improve the sensitivity? If not, can you suggest any other force/pressure + contact sensor that is responsive in the 0.1N to 5N range. A sensor with simple circuit will be great since I dont have much time left.
- Temperature Sensor
I have tried a few temperature sensors and they have the same problem of having a very slow response. Can you suggest a temperature sensor that can give me an analog value of the temperature within 1 sec of the object touching it, or remotely measure the temperature.

I would really appreciate you input on this. Thanks a lot

[Qubix]

Have you seen these, they are pretty easy to use

http://www.trossenrobotics.com/store/c/2670-Force-Sensors-Pressure-Touch-Sensors.aspx

David

[umerwahla]

Thanks for the reply Qubix. Yes I have seen these and have also ordered 1. Waiting for it to arrive. Do you have any recommendation for the temperature sensor aswell? I was thinking of http://sgbotic.com/index.php?target=products&product_id=250
Have you used that before?

[Qubix]

One of the guys in college is using those force sensors for small forces and he is pretty happy with them.
Temperature is really my field. I haven't used that sensor before. How about using a thermistor, they have fast response times I think. 

[Soeren]

Hi,

I am using a very simply potential divider circuit with the 4 QTC Sensors in wheatstone bridge configuration. The sensitivity is very low and it requires too much force to show the required change in resistance. Is it possible to change the circuit in any way to improve the sensitivity?

Hard to tell when we don't know the original schematic, so please post it.
It sounds weird that it shouldn't be sensitive, but if you have the same amount of force on all four of them, it would explain it though.

I have tried a few temperature sensors and they have the same problem of having a very slow response. Can you suggest a temperature sensor that can give me an analog value of the temperature within 1 sec of the object touching it

Low mass equals fast response times.
Low mass NTCs should be quite easy to get hold of, they were used extensively in the health sector before the IR ear thermometer outdated them and they are very fast responding.

[umerwahla]

One of the guys in college is using those force sensors for small forces and he is pretty happy with them.
Temperature is really my field. I haven't used that sensor before. How about using a thermistor, they have fast response times I think. 

Hi qubix,
I am testing NTCRL3004-291-97-D1 Thermistor. Its just a normal NTC Thermistor. I was having some trouble and since temperature is your field, I thought I might be able to help me out.
I want to use the thermistor to know when a human hand makes contact with the sensor. I am using a normal potential divider circuit right now but the results arent that good:
 - The change in resistance/P.D upon human contact is low
 - After the contact ends, the Resistance/P.D takes quite a long time to go back to the original value.
Can you please suggest a better circuit so as to overcome these problems or any thermistor you know that woould not pose these issue. I would really appreciate your help.
I am using te folliwing circuits:-

[umerwahla]

Hard to tell when we don't know the original schematic, so please post it.
It sounds weird that it shouldn't be sensitive, but if you have the same amount of force on all four of them, it would explain it though.

Actually, QTC are not supposed to be that sensitive. According to the graph in their datasheet, they only start responding after 5N of force has been applied and since I am dealing with forces upto 5N only (handshakes, grabbing objects) so it is not sensitive enough for me.

Low mass equals fast response times.
Low mass NTCs should be quite easy to get hold of, they were used extensively in the health sector before the IR ear thermometer outdated them and they are very fast responding.

Do you have any low mass ntc in mind? The one I have tested and am having trouble with is NTC RL3004-291-97-D1

[madchimp]

Could part of the problem with the temp sensor being slow to return to room temp be the material surrounding it? It's probably not just something that popped in my head and thought I would share in the off chance it might be.

[Soeren]

Hi,

I am dealing with forces upto 5N only (handshakes, grabbing objects) so it is not sensitive enough for me.

How much precision do you need?

Do you have any low mass ntc in mind? The one I have tested and am having trouble with is NTC RL3004-291-97-D1

None in particular, anything used with a body thermometer should be useable and since they are disposable sensors (for the sake of hygiene), they're comparatively cheap - perhaps check with a local hospital or a doctor - they may have some leftover that you could get for free.

The only thing that I could find (in a shallow search) was that your NTC is 500 Ohm and has a diameter around 8mm (Remember, linking to datasheets or providing full info of the component in question yields a much better chance of a useable answer).
If it's indeed 500 Ohm (@25°C), it's a bit too low, since it will get lower when heated. I'd go with something around the several kOhm mark, say 10k to 100k, for your app.).
Also, 8mm (and probably a few mm thickness?) is quite a thermal mass, so go with something way smaller.
That said, with a well designed bridge circuit, even small changes should register.

Please provide full data and a schematic of your circuit if you want further help - none of us here are psychics, so we need the input