MicroControllers at competitions

[Ether]

Originally Posted by yash101

.025 ohms was just a practical example

It's not practical. .025 ohms will have a significant effect on motor torque.

In real time, I would use a 25mOhm resistor

25mOhm is .025 ohms.

Also, about the inductance-based current sensors, don't you need a circuit to constantly monitor it if you are using DC?

A clamp-on DC ammeter uses a Hall Effect device to sense the static magnetic field created by the DC current.

[yash101]

Originally Posted by Ether

It's not practical. .025 ohms will have a significant effect on motor torque.

25mOhm is .025 ohms.


A clamp-on DC ammeter uses a Hall Effect device to sense the static magnetic field created by the DC current.

Sorry. I confused mOhms with uOhms!

I think it would be better to skip the resistor and bring along a foot of gauge 8 wire. That aught to do it without over-dropping the voltage.

Using a hall-effect sensor is a different way of thinking about it! I was thinking about measuring current by creating a transformer, with the second winding as a coil around the main wire!

[gpetilli]

If you are truly looking at shunt resistors <<25mohm you absolutely should use the voltage across the wire from the PDB to the speed controller. The resistance of the wire per foot is easily found online and will be just as accurate as a resistor you could afford to add - plus its FREE. Just put a volt meter across the wire (or add your micro-controller). If you are looking for a failure, the difference between 133 amps and 125 amps is not really relevant anyway.

[Joe Ross]

Originally Posted by yash101

Sorry. I confused mOhms with uOhms!

Assume you use a 25µΩ resistor. At 100 amps, how much voltage drop would there be? How many useable bits of resolution would there be between 0 amps and 100 amps with your ADC?

[yash101]

To increase the 12 bit resolution, I want to decrease the vRef voltage to something like 1v or .1 v, etc.

[gpetilli]

Originally Posted by yash101

To increase the 12 bit resolution, I want to decrease the vRef voltage to something like 1v or .1 v, etc.

As someone who designs ADCs for a living, I strongly recommend against using a Vref less than 1v. At 1volt each count of yor ADC is already 25uv and noise of a CMOS amplifier is about 10uV. The best answer is to use a resistor that gives you a voltage that matches the input range of your ADC. I still think the suppy wire is in the right range. The key is to use a high impedance sense circuit with a Kelvin connection in parallel with the supply routing to the speed controller so a not to influence the system performance.

[Al Skierkiewicz]

Dev,
Your original statement about DC is true for transformers. You only get an output when the input (primary) is a varying voltage. In a transformer, when you switch on a DC current, the secondary will produce an output during the transition and then settle back to zero. That is why Hall effect devices are used for current monitoring in DC applications.

[yash101]

That makes sense. Does anyone have a good place for me to grasp info on how hall-effect sensors work? I am interested in these neat devices!

[techhelpbb]

Originally Posted by yash101

That makes sense. Does anyone have a good place for me to grasp info on how hall-effect sensors work? I am interested in these neat devices!

2 for one - tutorials and something specific for hall effect sensors.

http://www.electronics-tutorials.ws/...ll-effect.html

So has anyone used a length of wire, resistor or hall effect sensor on a fielded robot as a current sensor?

[Alan Anderson]

Originally Posted by techhelpbb

So has anyone used a length of wire, resistor or hall effect sensor on a fielded robot as a current sensor?

At IRI a few years ago, I know a few robots had self-contained logging hall-effect current sensors strapped against the battery leads. I don't remember finding out who was responsible for them.

[Joe Ross]

Originally Posted by Alan Anderson

At IRI a few years ago, I know a few robots had self-contained logging hall-effect current sensors strapped against the battery leads. I don't remember finding out who was responsible for them.

FIRST was trying to determine current draw in order to set the requirements for the 2009 PDB.

[Al Skierkiewicz]

WildStang manufactured a current monitor called StangSense a number of years ago. We used a one foot piece of #10 wire and a current monitor circuit from Maxim. We then ported the data to the robot data link and recorded the data at the driver's station. I believe the file is still on CD someplace.
The current monitor that was added to robots at IRI was from HQ. They wanted to gather data to confirm the high current spikes I had mentioned in conversation. The data recorders were supervised by Wayne Penn as I remember.

[Dad1279]

http://www.andymark.com/product-p/am-2709.htm

New product from AndyMark, using a Hall-Effect, but needs to be inserted in the circuit.