Group Actions
Engagement
  • Author Author: AE7HD
  • Date Created: Date Created
  • Views 4594 views
  • Likes 7 likes
  • Comments 34 comments
Related
Recommended

555 timer Remote Capacitance Sensor - fluid level, body detector, touch sensor

AE7HD
If you want to measure capacitance for a sensor some distance away, you can't do it at the end of long wires. Just temperature changes and nearby objects will make larger changes that can swamp what you are trying to measure.

This was originally for my first car in the late '70s. I put sensors on the windshield washer tank, the radiator hose, even inside the gas tank. I had already designed and built my own precision capacitance meter using two 555 timers. This triggered each sensor separately, then sorted which one was active with a decade counter and an LM3914 linear dot/display driver.
Two wire twisted pair is a lot cheaper than 3 wire. And when you aren't flush with cash, that can make a difference. This one is designed to send the signal back to an Arduino, which is then programmed to measure just the On time.

I've used this to measure liquid levels in water tanks, on roofs, even inkjet tanks. The sensor itself might be metal tape on the outside of a plastic or glass tank, or insulated wire in the liquid. Make sure the insulation is compatible with the liquid.

One 555 timer goes onto the remote sensor. The signal and power travel on the same line similar to how I2C and several other methods work, although in this case there are no addresses.
image
  • in reply to AE7HD

    Attaching the pictures of the Sensor I have made again.

    Full Length.     Close Up.  Aluminium Wire

    You are right. This is a PVC insulated Aluminium wire normally used for drying cloths here. This is a strong wire and can't be bent easily...Will check on Teflon insulated Single stand wire. The looping back till the top is a good idea. I had done that some time back with 1.5 core, Insulated Copper wire (Multi-Stands). I need to check how this would work, will give it a try.

    W.r.t 75ft CAT5 cable, I have a single LAN cable upto from Sump to the OHT. The connections on either side (Arduino on one side and sensor on the other side)  of this wire is taken with the help of connector as shown below...Connections are stiff and conducting properly

    Connector

  • in reply to Gururaj

    I can't find the picture you posted before.

    What insulation is on the wire? PVC insulation absorbs water. I notice you put some kind of sealer on the ends of the wires, are you sure that isn't leaking? I like to use teflon insulated wire, and rather than one wire down, I like to do a loop down and back up so there is complete insulation only in contact with the water.

    The behavior with a scope attached points to something very odd going on. A bad connection, perhaps? The scope should only add a few pF if you are using 10x probes and that should not change. What length are the pulses?

    How are you connecting 75ft of twisted pair? Are you replacing the 30ft, or adding 45ft? Soldered or just twisting the wires together?

  • in reply to AE7HD

    @AE7HD : As you suggested I changed R3 100E and used Twisted pair. But no luck. I tired the setup with R3=1K, with 30ft Twisted CAT5 cable. It worked fine. But with 75ft cable its not working.. Infact when the CL (Sensor) is outside the water tank, It shows 0.11 nF. But when I start inserting it into the water (OHT), Capacitance reduces, like it become 0.07nF and goes as low as 0.01nF when fully inserted into water.

    I even used Digital Oscilloscope to test the signal. Weirdly, as soon as I connect Scopes probes, Capacitance starts varying like 0.11nF, 0.08, 0.05,  0.12, even with 2ft Cable and Sensor not in Water. If I remove the scope's probes..It stays constant at 0.11nF. The timing of high and low signals on Scope also shows varying results..

    I though of increasing R3 to something like 4.7K/10K  so that Sensor can detect the "Low" at this length of cable. Any other thoughts will be helpful.

  • in reply to Gururaj

    If it works with 2ft of wire, then perhaps lowering R3 to 100 ohms and using both wires in a twisted pair will fix the numbers..

    I've seen nearby noise sources cause problems when not using twisted pair or shielded wire. Brushed motors, motor start relays, LED/fluorescent lights.

  • in reply to AE7HD

    As it says, One Picture is worth thousand words, I am attaching pictures of my sensors.

    image  image image

    Pic 1  is complete view of 4ft Sensor.
    Pic 2:  Closer look at the green Insulated Aluminium wires on either side of the pipe. Also the end of this Aluminium wire is insulated with rubber tape and Araldite (Adhesive).
    Pic 3: Shows how 555 timer is connected with this Insulated Aluminium wire.  

    I'll surely use a single twisted pair and test it tomorrow. Borrowing a scope on Sunday to check timings.

    W.r.t Grounding the circuit at the tank, I don't have any additional grounding provided to circuit. Just the power supply wires (The Red-Black twisted wire in Pic -3)

    I'll even try to lower the R3 to 100 Ohms and give it a try. 

    IMO, Since this setup is working fine with 2ft of wire, we can zero down our analyse around stray capacitance, Noise, EMI, timing issue or other.