SpudFiles

I will be going to a robotics competition in the fall. There are two classes of competition, autonomous and remote controlled. My team will most likely be in the autonomous division. The goal is for the robot to drive around in the sand and detect a magnet that is buried 6-12 inches deep. Once it detects the magnet, it has to plant a flag over it to indicate that it found the magnet.

Now to my question for some of you braniacs out there, how would you go about trying to detect a magnet buried 6-12 inches? I do not know how powerful the magnet will be. I have several ideas, but I would also like some more suggestions so I don't leave out any possibilities.

Presumably a conventional metal detector will suffice - unless there are other metallic components strewn about the sand to confuse the robot.

I don't know yet. That is one of the guidelines they haven't given us, so I think we are supposed to assume that it is a regular beach with litter and such.

You could try a hall sensor, but I'm not sure about the range at which they can detect a magnet.

Ya that's kinda the way we are leaning right now. They're cheap, so it's worth a shot even if it doesn't work out. I have no clue how sensitive they are since I have never used them.

the most common hall sensor arnt that sensitive. you need to figure out how a metal detector works.

but i know you can make metal detectors bij yourself. dont know where ive seen the electric diagram of it.

Detecting a piece of iron at a depth of 6" to 12" is really pretty hard. A magnet will be easier but still not all that easy.

As others have said, a Hall effect detector is probably not sensitive enough. The magnetic strength drops off as 1/r². So a sensor designed to work at a few mm (which I believe is what most are) will only be getting ~1/10,000 as strong a signal at 6".

Here a couple of homebrew metal detector designs using coils to detect metals;
A pretty simple one.
Using a micro controller to do most of the work. (Need to reverse engineer what the microcontroller is doing.)

Also check out www.metal-detector-schematic.com

Since you are looking for a magnet you might be able to just use a very large number of turns in the coil and a high gain amplifier to boost the voltage you'll get when the coil is moved across the magnet's field.

Or, use a circuit specifically designed to detect magnetic fields. Here is one and a two that can detect the Earth's magnetic field.

As jimmy said, at 6", you've got a challenge on your hands. Even assuming it's a neodymium magnet, at that distance, the flux density will have fallen to a less than 100 gauss at best, and more likely less than 10.
At 12", you'll have immense difficulty.
At that point, the magnet's field probably barely different to the earth's magnetic field, and if you've got motors onboard the robot, the magnets on those will easily obscure any field from the magnet.
Adding in the field from moving currents, it sounds next to impossible.

There are some possibilities. But any sensor you have will have to be magnetically shielded, or at least seperated by some distance from your motors and electronic workings.

Mount a DR tiller on the robot and have it scan the dirt/sand it's kicking out in the back

But really, having sensors that can be partially buried would probably increase the chance of this working.

Have you ever seen how they run telephone lines underground in the country? There is a vehicle with a large disk that cuts into the dirt while cable is being fed around it. It you could use a disk with a sensor in it, you would have a lot more luck than sensing from above ground.

I don't think it would be too hard, sand isn't extremely tough to cut through.

_Fnord wrote:Mount a DR tiller on the robot and have it scan the dirt/sand it's kicking out in the back

But really, having sensors that can be partially buried would probably increase the chance of this working.

Have you ever seen how they run telephone lines underground in the country? There is a vehical with a large disk that cuts into the dirt while cable is being fed around it. It you could use a disk with a sensor in it, you would have a lot more luck than sensing from above ground.

I don't think it would be too hard, sand isn't extremely tough to cut through.

Pneumatic spear like thing that jabs down into the sand a certain distance, with the sensor on it. Pull it up and move somewhere else and stick it back into the sand.

Would that work or am I talking out my arse?

Modderxtrordanare wrote: Pneumatic spear like thing that jabs down into the sand a certain distance, with the sensor on it. Pull it up and move somewhere else and stick it back into the sand.

Would that work or am I talking out my arse?

Well it is a fairly large area, I think 20' x 20', and it is a race against the clock, so I'm not sure how efficient that would be.

Maybe you could have the sensor on the end of a rod that drops nine inches below the wheels of the robot, that way the sensor is in the best place to detect a magnet that is at 6-12". That would only work if the robot is powerful enough to pull the rod through the sand, but I think it would be possible.

Depends on the strength of the buried magnet really.

If it's a relatively powerful one, you could have a strip of metal suspended over a contact from a weak spring. When it approaches the magnet, it will be sealed shut, thus creating your signal.

robert23 wrote:

Modderxtrordanare wrote: Pneumatic spear like thing that jabs down into the sand a certain distance, with the sensor on it. Pull it up and move somewhere else and stick it back into the sand.

Would that work or am I talking out my arse?

Well it is a fairly large area, I think 20' x 20', and it is a race against the clock, so I'm not sure how efficient that would be.

I think this is actually a good idea. Not only would this mean your sensor doesn't need to be as sensitive, but you could aldo rig it to where the flag is on the sensor probe, then when it finds the magnet, just have it disconnect the whole apparatus and go.

paaiyan wrote:

robert23 wrote:

Modderxtrordanare wrote: Pneumatic spear like thing that jabs down into the sand a certain distance, with the sensor on it. Pull it up and move somewhere else and stick it back into the sand.

Would that work or am I talking out my arse?

Well it is a fairly large area, I think 20' x 20', and it is a race against the clock, so I'm not sure how efficient that would be.

I think this is actually a good idea. Not only would this mean your sensor doesn't need to be as sensitive, but you could aldo rig it to where the flag is on the sensor probe, then when it finds the magnet, just have it disconnect the whole apparatus and go.

Only problem is you'll have to be jabbing into the sand every 6" or so, otherwise you could just use the detector up at the surface of the sand. For a 20'x20' area and jabbing the probe in every 6" would mean 1,600 probings. (-- Insert joke about alien abduction here --)