Beach Detecting Rover project

[Emio052]

Thanks for update....your are doing nice work...

[ApBerg]

Thanks for the schematics. I had seen those, but I assumed they were an earlier version. Has anyone done a CAD version, with the same component IDs and circuit? All the CAD drawings I saw don't match up to that or Silverdog's PCB. I'm guessing that's why the component IDs were left off the PCB - too many schematics floating around with different IDs.

Anyways. I'm gonna hold off on getting too deep into the PI circuit for now on this thread. I will start a new thread specifically going over my questions so that it doesn't clutter this one up, which I just want to be the build details.

http://www.geotech1.com/forums/showt...t=17396&page=2 # 49

There you find the Sprint Layout files from the Baracuda pcb, made after the hand-drawn schematics made by Alexis.
In the Sprint File, turn layer 0 on.. now you have the component ID's that match with Alexis schematics.


Ap


What targets are you want to find ? small surface ... or deep large ?

[fixstuff]

Thanks for the link. That will be a big help.

I will be mainly looking to start with small targets, shallow depth. I won't have access to heavy digging tools where I will be using this mainly. I'm looking at small nugget or smaller deposits. I'm hoping the detectors I have now will work for this. Would a Gamma 6000 or the Baracuda handle that requirement, or should I look further?

The boom detector arm will hold the smaller pinpointing coil. Even though a coil like that is extremely inefficient to use manually, I think an automatic sweep with fine grain movements, it will be just fine.

For each sweep of the boom arm, I will extend the arm 0.5 inch or so. This will allow really good coverage. Every 12 inches I will stop. Reset the arm to the home, and drive the rover forward 1 foot. Everytime I finish my rover move, I will shut down the motors, so that noise is minimized, with only the boom moving.

I'm planning on using this to maintain boom height to the ground, but it isn't released yet.

http://www.maxbotix.com/articles/024.htm

It has a 1mm accuracy, so I should be able to have almost perfect sweep technique for the coil. Nice thing about the coil, is the target will be very localized, and I can mark using my mounted paintball gun on the boom arm.

[ApBerg]

Thanks for the link. That will be a big help.

I will be mainly looking to start with small targets, shallow depth. I won't have access to heavy digging tools where I will be using this mainly. I'm looking at small nugget or smaller deposits. I'm hoping the detectors I have now will work for this. Would a Gamma 6000 or the Baracuda handle that requirement, or should I look further?

The boom detector arm will hold the smaller pinpointing coil. Even though a coil like that is extremely inefficient to use manually, I think an automatic sweep with fine grain movements, it will be just fine.

For each sweep of the boom arm, I will extend the arm 0.5 inch or so. This will allow really good coverage. Every 12 inches I will stop. Reset the arm to the home, and drive the rover forward 1 foot. Everytime I finish my rover move, I will shut down the motors, so that noise is minimized, with only the boom moving.

I'm planning on using this to maintain boom height to the ground, but it isn't released yet.

http://www.maxbotix.com/articles/024.htm

It has a 1mm accuracy, so I should be able to have almost perfect sweep technique for the coil. Nice thing about the coil, is the target will be very localized, and I can mark using my mounted paintball gun on the boom arm.

For nuggets you need a 'fast coil' look at http://www.geotech1.com/forums/showthread.php?t=16322 think the "coax" coil is a good coil to use with the baracuda also read the pdf from BBsailor.

To use more coils look at the jpg.... do not now if it will work..

To log the found nuggets is very hard ?? you use dgps ?
Never found nuggets... are not here in The Netherlands

[fixstuff]

Wow, that was a great response. That is really good info for me. I really like the fact that it would allow me to use the Baracuda with the spiral coax coil. Since I'm sweeping in very small increments, this should be an excellent way to pick up the small gold.

At this point, I just might be tempted to do a rev to the Baracuda, and add the ability to remote control everything from my control system, or be able to use a Prop Quickstart board for the onboard processing.

If I added remote delay and threshold adjustment, I could throw in some nice improvements, such as terminal headers for everything. I could also incorporate some of the various mod circuits folks have introduced as well, and incorporate them into a single board. Make the board able to handle 4 coils maybe.

Any suggestions on the best mod circuits and improvements to the Baracuda?

on the dgps: I'm looking at my own mapping system on the rover.

[fixstuff]

Now we are getting serious. I've measured the size of my enclosures, and it looks like I'll be able to fit in a full size ATX motherboard rather than use the Zbox HTPC. This opens up other options as well such as a CUDA graphics card... Hmmm...

[fixstuff]

re: Differential GPS

I still haven't quite figured out how to do a true dgps. There are some techniques I will use though to hopefully improve my positional accuracy by quite a bit.

During scanning I will be sweeping back and forth using a boom arm. This arm will be capable of indexing out in small increments. As I complete each pass, I will index the boom arm out. I will have a 12" stroke available initially. While I'm in sweep mode the rover is stationary.

I will have 3 or more GPS sensors in the Rover, and I will read all three GPS units and average them over a period of about 2 minutes. This will increase my accuracy by quite a bit.

Each scan will increment the boom index by 0.5inches, and with a 12inch stroke, that will take appox. 2 minutes. Just before moving I will update my position. Since I will know my rover distance upon moving, I should be able to have a pretty good idea where I'm at. Add in the boom index and rotation position from there, and I will know the target location.

I've done some research on multiple GPS, and it looks like I can improve my accuracy over a single one by a factor of 3 or 4 at least by averaging multiples over a longer period of time. With a GPS sensor capable of WAAS and EGNOS decoding, I'm hoping for a 1/2 meter or better. Again, I'll have to see what it ends up. This is one reason why I'm marking target IDs with a paintball.

[Farside]

It might pay to consider the inertial mass of the coil assembly with respect to using a boom arm to do the sweep...

A gantry setup could be a better option.

[fixstuff]

Yeah, thats why I'm using PVC pipe to keep it lightweight, and I'm using a BIG motor with a lot of gear reduction so that I don't have any issues with reflected inertia. Hopefully, I'll be ok, but I'm not going to swing real fast, just enough for the motion pickup to work.

[fixstuff]

more parts have arrived

Well. I got in my motors and linear actuators today. Man those motors are freaking HUGE. The ebay pix didn't do them justice. 13.5" x 3" tires. Motor looks like a car starter. Weight about 30+lbs each. That will put the weight from the wheels and motors alone at almost 150lbs. I powered the brakes, and ran the motors, and they ran pretty good. Not bad for $50. I've got servos the size of my thumb that cost more than that.

Powered up the linear actuators also. They worked well, power is 12V though. I'll run those from my smaller motor controllers.

National Instruments stuff came in too. cRIO rack didn't power up . Not the fuse either. Oh well. Good thing I only spent $50 on it. I'll use the boards with my other rack.

Heres some pix on the motors and actuators. Note my foot for scale. Thats a 13 shoe...

[Paul23]

Fixstuff,

Just thinking of your gps problems.

Do you really need on going GPS readings? If you have a start point and encoders on your buggy it will always know where it is from the start point.

You can then just take one reading of the start point (Doesn't even have to be on the buggy.) and then mark this with a disc and go from there. Could be a good power saver.

In the UK accident investigators do something similar and measure all points from an initial set point, this in turn gives them their 3d pointing/measuring system. May be worth a little look to see if anything fits your ideas.

Paul....

[UFO]

Interesting project.

Now if you could just get that darn Robot to dig up your targets, you really have something then!

[Tepco]

Yeah, thats why I'm using PVC pipe to keep it lightweight, and I'm using a BIG motor with a lot of gear reduction so that I don't have any issues with reflected inertia. Hopefully, I'll be ok, but I'm not going to swing real fast, just enough for the motion pickup to work.

Beach detecting rover!?! come on, what on Earth you actually want to build? Some sort of autonomous automated minesweeper or what? These things are in high demand and i suppose you can find some funding for this easily. If im right, only thing wrong whit this project is MD itself, somewhat useful for finding unexploded ordinances, but almost useless for demining. Instead, your front-end may look like this: (sorry, file too long, this is the link)

http://www.s2is.org/Issues/v1/n3/papers/paper8.pdf

PS: For LRL guys, sympathetic resonance at its best, but as far as i know, this trick will not work whit gold. After all, even whit detector working on subatomic level, someone will allways complain about discrimination!

(Hope this can be useful, best regards)

[Midas]

Fixstuff,

Just thinking of your gps problems.

Do you really need on going GPS readings? If you have a start point and encoders on your buggy it will always know where it is from the start point.

You can then just take one reading of the start point (Doesn't even have to be on the buggy.) and then mark this with a disc and go from there. Could be a good power saver.

In the UK accident investigators do something similar and measure all points from an initial set point, this in turn gives them their 3d pointing/measuring system. May be worth a little look to see if anything fits your ideas.

Paul....

What you describe is called dead reckoning (it can also be done with inertial measurement). It has excellent short term accuracy but suffers from long term drift (in the example you gave it works fine because they only do short measurements, always coming back to a known point). GPS is just the opposite it has good long term accuracy but large amounts of drift in the short term. So the best strategy is actually to use both and come up with an algorithm that mixes the two data streams together to remove the weaknesses of each.

Midas

[Davor]

I have some experience with GPS in (surface) radio propagation drive tests, and it is true that drift makes it, well, not that much accurate. In radio propagation it is not that much of a problem since the clutter size is seldom smaller than 25m.
Many of the GPS errors can be fixed in post-processing but you don't have that luxury in the field.
I expect your GPS readings to be as accurate as can be because it is a beach with perfect reception conditions and no scattering at all. If you aim for even better accuracy go for a pro antenna that emphasizes satellites near zenith, thus eliminating problems of near horizon satellites.