Sensor Hub

Hi Everyone!

I recieved my first ever arduino board yesterday (uno R3), and really happy with it! I've messed around with microcontrollers in university so it's good to finally get one of my own.

I'm currently looking at doing a project as just a proof of concept for my dissertation in university. It's aim is to structurally monitor a pylon, this will hopefully incorporate 5 or 6 sensors:

  • An Anemometer to measure wind speed
  • 2 Strain Gauges
  • An Accelerometer to measure vibration
  • 2 thermocouples to read temperature at ground level and at a high level.

I'm not quite up to speed on all of the associated circuits for these sensors yet. But I'm just wondering if off the top of an experienced arduino users head, the main issues this sort of system would present that I'd need to iron out? whether I'm being a bit too ambitious? etc...

I'm also planning on doing datalogging to an ardulog microSD logger.

Thanks

the main issues this sort of system would present

The thermocouples will be the biggest problem. Typically thermocouples have either a very wide range but limited accuracy or a very limited range with high accuracy. Neither is likely suitable for your purposes, not to mention far more complicated and expensive than a simple(r) thermister. What range of temperatures are you expecting to measure?

The strain gauges will need amplification, too, which introduces noise. What are you measuring?

PaulS:

the main issues this sort of system would present

The thermocouples will be the biggest problem. Typically thermocouples have either a very wide range but limited accuracy or a very limited range with high accuracy. Neither is likely suitable for your purposes, not to mention far more complicated and expensive than a simple(r) thermister. What range of temperatures are you expecting to measure?

The strain gauges will need amplification, too, which introduces noise. What are you measuring?

Thanks for the reply, I’m looking to measure deformation of structural steel members with the strain gauges under certain loadings, this will be a very slight change but hopefully enough to indicate stress.

I’m looking to measure ambient temperature, so anywhere between -20 and +40 (about the maximum temperatures in the UK).

A thermistor ey? I’ve only ever used thermocouples before now, so will definitely look into them!

How would a thermistor work? just one end to the 5V and the other to an analog input?

How would a thermistor work? just one end to the 5V and the other to an analog input?

A thermistor is a resistor whose resistance varies with temperature. So, they need to be used in a voltage diivider circuit. In your scenario, the changing resistance would change the current flowing into the pin, not the voltage.

If you're talking about an electricity pylon, there will be another couple of issues - the size means you'd need to run a lot of wire to your sensors, which would impact the accuracy of any analog read. The other things is that HT voltages will give you all sorts of issues with interference. You might consider stand along processors for each sensor and commnicate the data via radio. Take a look at Jeenodes for example.

Heimdallofasgard:
A thermistor ey? I've only ever used thermocouples before now, so will definitely look into them!

How would a thermistor work? just one end to the 5V and the other to an analog input?

That's possible, but not the best way to use them. Like a lot of other sensors thermistors are resistive elements that change resistance due to specific physical conditions. Because a microcontroller's ADC is based on voltage, the best way to wire them is as part of a voltage divider. Here is a link to a simple schematic and some sample Arduino code.

wildbill:
If you're talking about an electricity pylon, there will be another couple of issues - the size means you'd need to run a lot of wire to your sensors, which would impact the accuracy of any analog read. The other things is that HT voltages will give you all sorts of issues with interference. You might consider stand along processors for each sensor and commnicate the data via radio. Take a look at Jeenodes for example.

I had a discussion with my project supervisor and we decided that mitigating interference from the HV conductors wouldn't be in the scope of the project (thank god!). As for the issue of the long wires causing innacurate readings, haven't really thought about this. Is it just a power issue?

It's the resistance of the wire causing a voltage drop. Worst case, you might be running 200 feet of wire to your thermister at the top of the tower, which I would expect to have an unfortunate effect on your accuracy.

Heimdallofasgard:
I'm also planning on doing datalogging to an ardulog microSD logger.

Any luck with this device? Did you succeed in using it? Did you reprogram it or just use default firmware?