Probe for Garrett Sea Hunter

[rgecy]

Will this one at Radio Shack do?

http://www.radioshack.com/product/in...&tab=techSpecs

How is the MOSFET wired in?

Robert in SC

[bbsailor]

PI Probe

RGECY,

You do not need to buy a MOSFET. Go to RS and only purchase a 10 ohm metal film, 1 watt resistor. The resistor comes in a two pack. Where you solder the coil leads to the coax wire, just put a resistor in between the coax center conductor and one coil wire. This will raise the resistance of the probe to approximately match the resistance of the coil. You could even put it in the switch box.

The MOSFET is inside the PI control box and is what drives the coil. It will sometimes get hot when the coil resistance is too low. You do not need to open the control box and monitor the MOSFET heat if you just put the resistor in series with the probe coil. I was suggesting that you might not need the resistor if the MOSFET (that looks like the RS IRF510) is not getting hot without it. The MOSFET would be located near the coil connection on the circuit board and have three leads and a metal tab on top.

bbsailor

[bbsailor]

More PI Probe

RGECY,

Since the Tesoro coil is 11 ohms, you should plan to use the 10 ohm resistor in series with that probe coil.

Since the Garrett Sea Hunter coils are bewteen 1.18 ohms and 1.38 ohms, you should plan on making the probe have between 1.2 and 1.4 ohms of resistance. AWG 30 wire (purchase a 50 ft spool of Kynar insulated wire wrap from Radio Shack) has .1 ohms per foot resistance so your coil should contain about 15 ft of wire, 14 ft for the coil with 6" for each coil lead. Start out with this amount of wire and see if the detection range is acceptable. You should be able to make three probes with one spool of the Kynar insulated wire.

If you use thicker wire, you will be putting on many more turns to match the coil resistance but the inductance will probably be too high. So stay with AWG 30 wire to keep the probe small and thin. The AWG 30 Kynar wire is .019 Outside Diameter. You will be making a single layer of wire about 2" long in the center of the core to obtain a 1.4 ohm coil (107 turns on a .5" diameter rod core). You can go as low as 1.2 ohms with 90 turns.

There is a posibility that you may need to add a small damping resistor inside the probe housing (in parallel with the probe coil where the coil wires are soldered to the coaxial wire) if it does not work at the lower end of the delay range. (Assuming that your PI has an adjustable delay.) This is not the same as the series resistor mentioned before. This is because the probe inductance may be higher than the damping resistor in the PI control box is set for. So try adding values of about 5000 ohms, 4000 ohms, 3000 ohms, 2000 ohms to see which value allows you to operate over the full delay range. Pick the highest value that allows operation over the full delay range.

bbsailor

[rgecy]

After widing the coil for the SeaHunter, using 14' of AWG30 magnet wire and 98 turns, I measured the resistance of the coil without the resistor and got a reading of 1.9 ohms. This is significantly higher than expected. Should I unwind some of the turns to get it down to say 1.3ohms.

With this, should I still use the 10 ohm 1 watt resistor? How will I be able to tell if I need the damping resistor? Do you have a diagram of how the two should be wired? One in series the other in parrallel?

Thanks,

Robert in SC

[bbsailor]

PI Probe

RGECY,

Give it a try and see how it works as is. Then, if the low delay does not work just remove about 3 ft. This will make the inductance less and make the probe closer to the coil inductance. You would only need to put a series resistor on higher resistance coils like the Sand Shark.

To see how a damping resistor is connected to a coil, print out the Hammerhead schematic and look at R11 (the damping resistor). You probe is essentially another L1. R11 is on the circuit board. Your external switch box is just replacing the coil with the probe when you flip the switch.

Probes design is not too critical. My only problem came when I tried a probe with very high inductance on my PI that samples at minimum of 7.5uS. At the low end of the delay, the probe locked up. See my article about the coil locking up or stopping to respond to targets. Too much inductance would couse this but I think you are close enough to have it work as it is now. The simple fix is to remove coil turns until the probe works through the full range of delay.

I'm out for a week on vacation tonight.

Good luck.

bbsailor

[rgecy]

OK, I hooked the probe up at 1.9 ohms and it would hardly register an object and would only detect it after a second or two. And then would continue the tone for a few seconds after the object was removed. I removed about 5' of wire and got the resistance down to 1.3 ohms with coil and 6' RG8X cable. The same thing happens. It will only detect an object at about 1" and only after a second or two.

I added damping resistor and it just got worse.

I am going to rewind the coil. I think I may have rubbed some of the enamel coating off the wire and it could be grounding to the ferrite rod. I am going to insulate the rod this time and see if it makes a difference.

Any suggestions?

Robert in SC

[Carl-NC]

I made a PI probe with a ferrite rod I had on hand. It was 3/16" x 1-3/8". I wound 102 turns of 28AWG, 2 layers, and got 300.3uH (my target was 300uH). Resistance ended up 7.1 ohms.

The PI circuit I happen to have on my bench is a very crude design that detects my gold ring at about 4-5 inches with an 8-inch coil. With the ferrite probe, it gets the ring at almost 3 inches, so that's not too bad. Next is to try it with my Sandshark. Actually, next is to get the Sandshark back together.

I kept track of inductance as I wound:

25T = 22.1uH
50T = 69.9uH
75T = 165.3uH
85T = 217.5uH
100T = 293.1uH

This comes out roughly to

L = 0.03 * T^2 (uH)

Your mileage will vary.

- Carl

[rgecy]

Carl,

Where can I find the connector for the SandShark? Darren wants me to make up a probe for his.

Thanks, and let me know how the coil works on the Garrett as well. I ordered the LCR Meter and should have it Friday.

Robert in SC

[rgecy]

OK. After talking with Carl we decided that a two coil solution may be needed with the Garret Sea Hunter MkII since there appears to be two coils in the 8" mono coil. Basically a TX and an RX.

Following guidelines provided earlier of 98 turns of awg30 I wasn't able to get a stable reading just using a single coil over the ferrite core.

What I did, after talking to Carl, is wind a second coil of 98 turns over the first and connected them to the TX and RX pins on the unit. And bingo, I am getting a fairly good response.

But I guess the settling characteristics need improving since the tone will ring for a second or so after removing the coin. I am only getting two to three inches on a quarter, but I think once I get my LCR Meter and OScope, I should be able to resolve this and increase the distance.

I did not connect the desoldering braid yet. Should I connect it as a ground to the negative pins on both coils?

Thanks Carl and Joe for getting me on the right track. I am not done yet, so I am sure will still need you guys for additional input.

Robert in SC

[Carl-NC]

Where can I find the connector for the SandShark? Darren wants me to make up a probe for his.

Ah, now comes the hard part. On the SS, the connector and coil are a molded assembly. The easiest way to get a connector for the SS is to buy a coil, cut the cable off, and throw the coil away. Now you have a connector/cable assembly.

You would think that ordering a cable assembly from Tesoro would be easier. But, for some bizarre reason, Tesoro refuses to sell the cable assembly. OK, you could also discover that the cable is manufactured by Methode. But it is a custom cable made only for Tesoro, and Methode will not sell any unless Tesoro approves. For some bizarre reason, Tesoro refuses to approve.

I really like the Methode connector, but I'm thinking of switching my SS over to a Bulgin connector. I want to make some custom coils for it anyway.

- Carl

[rgecy]

Yeah,

Darren is a Tesoro Dealer and they wouldn't even sell him one. He is willing to let me chop one for him. Is there any good data that can be gotten from the 8" coil?

Could it be adapted to another detector? I believe its a printed circuit board coil right?

Maybe someone could give Darren a few bucks for it to offset the cost.

Robert in SC

[Carl-NC]

Yes, the SS coils are printed circuit coils. They might work with other detectors, but they generally don't perform as well as a regular wire coil. That's why I wanted to build some custom coils.

I've already swapped out the crystal in my SS to a higher frequency; this reduces the sample delay and improves performance on gold targets, about 2 inches on a man's gold band. I've suggested both of these improvements to Tesoro, but they don't seem to be interested.

One option on the Tesoro is to build a shaft-mount switch box... all the connectors on the switch box can be Bulgin or Switchcraft or whatever you like... then when you clip the connector off the coil, leave it a couple of feet and use that piece to feed the switch box... then add a new connector to the castrated coil. That way you don't lose the coil, and don't have to change the bulkhead connector.

Below is my latest probe winding... it uses a smaller ferrite rod, and has about 135 turns for 300uH. Diameter ended up larger because it has 5 layers of winding. What I would like to make is a skinny coil that fits in a hollow fibreglass rod, like is used in arrows. Then I could literally probe INTO the soil.

- Carl

[bbsailor]

Probe Design Considerations

Robert, Carl,

Probes should be grounded to minimize noise. I use rubber grommets to slip over the desoldering brad which holds it tight and flat against the ferrite material. This makes a good ground connection. Just stretch a grommet to tightly wedge the ground and the start coil wire lead under the grommet.

Then, I use a piece of Scotch 24 wire mesh shield that I almost make one complete turn so as not to make a shorted turn. I solder the desoldering wick to one edge of the Scotch 24 wire mesh used as the outer shield, then take that lead back to the coax outer braid shield along with the one coil wire to make the ground connection. The other coil wire goes to the coax center conductor. When shielding a balanced coil probe coil, ground the outer RX and TX coil portions separately to the respectve coax grounds.

If a separate RX and TX coil are needed, I would try to make it like a stacked balanced coil. I would divide the ferrite rod into three winding sections by using a two thin nylon washers in the center sections with the rubber gromments on the ends. Drill a small hole in one nylon washer to hold the start winding lead. Then, I would try to make either a center TX coil with two out of phase RX coils or vica versa to see which one balances better. Just wind the same number of turns on both end coils that are out of phase. Add a few more turns on one outer coil and unwind a turn at a time to obtain a null balance. This can be done with a signal generator on the center TX coil and the scope looking at the outer RX coils to obtain the best null balance. AWG 30 Kynar wire wrap or Teflon insulated wire should work well for this application.

Here is where a larger .375" to .5 " diameter ferrite rod, 3" long or greater may be helpful. These larger ferrite rods should give you 5" to 6" range on coins. The smaller ferrite rods will typicall give 2" to 3" on coins, based on my early mono coil probe designs.

If someone measures the resistance of the damping resistors across the coil TX and RX coil connections and the TX pulse width, I can try to reverse engineer some recommended probe coil winding data.

bbsailor

[Carl-NC]

If a separate RX and TX coil are needed, I would try to make it like a stacked balanced coil.

For an ordinary PI, induction balance is not necessary. Just wrap the TX winding right on top of the RX.

- Carl

[rgecy]

The resistance across the TX and RX pins on the unit are 1k ohm. I dont have the ability to measure the pulse width yet.

Carl posted it as 750pps with a pulse width of 70us.

I have made several different coils, but didn't have the ability to measure the inductance until this weekend. All the coils I had wound were on the order of 450-580uH.

The latest one, I wound a coil over a single .281" ferrite core with 52 turns of AWG26 magnet wire. This gave me an inductance of 140uH with a resistance of .65 ohms. I then wound another coil that matched over the first. The results are the best yet, but still needs some work. I am picking up a quarter at 4", but I am getting a ringing noise when there is no object present.

Could this be a ground or shielding issue? I am just using twisted pair for bench testing.

Also, is there a way to make the sides of the coil less sensitive and keep the tip sensitive?

Robert in SC