• Please join our new sister site dedicated to discussion of gold, silver, platinum, copper and palladium bar, coin, jewelry collecting/investing/storing/selling/buying. It would be greatly appreciated if you joined and help add a few new topics for new people to engage in.

    Bullion.Forum

DIY Pulse Induction Detector (Impulse III)

Prospecting Australia

Help Support Prospecting Australia:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Looks like the homework has been done on larger coils for the sdc.
i was out of the picture for some time . ...
Target Signal vs Amplitude in Hyperphysics.
Oh ! Well ! ... Spare working coils is not so bad.
I`ll just re-wind the windings into within original 8`` coil spec with what i have.
The 4500 might like the larger ones.
Closing Ceremonies i`d say.

With one exceptional thought.
200 x 200 piece of tin foil air tested with Minelab sdc OEM 8`` repaired coil .
Max reading of 800mm.
The modded 12`` NF air test reading maxed out at 950mm but is it sdc stable ... .
 

Attachments

  • Screenshot (1818).png
    Screenshot (1818).png
    48 KB
Last edited:
Coil Works in Progress.
1. 12`` coil foam insert.
2. New coil top & base. Armstrong bucket base & 20 liter oil tub.
3. Base & coil workings
4. Conductive painted shields.
5. 13`` Shield plate Ohms resistance test.
6. 13`` Shield 1.5 AAA battery DC volt test.
7. 8`` shield Ohms test.
8. 1940`s 8`` shield Ohms test.
Conductive Paint.
New Coil Faraday Shield Coating Formula.
Drying Agent - Automotive Acrylic Thinners 40 %.
Pigment - Graphite Powder 55 %.
Binder - Acrylic Lacquer 5 %.
4 to 5 thin coats brushed on.
Advantage over acrylic water base paints as the thinners cleans the dried surface skin of the previous coat creating better conductivity
contact and adhesion.
Required Ohms per inch is achievable.
65 - 85 ohms thereabouts.
 

Attachments

  • 20241221_073112.jpg
    20241221_073112.jpg
    979.9 KB
  • 20241221_073618.jpg
    20241221_073618.jpg
    2.6 MB
Last edited:
Conductive Paint.
New Coil Faraday Shield Coating Formula.
Drying Agent - Automotive Acrylic Thinners 40 %.
Pigment - Graphite Powder 55 %.
Binder - Acrylic Lacquer 5 %.
4 to 5 thin coats brushed on.
Advantage over acrylic water base paints as the thinners cleans the dried surface skin of the previous coat creating better conductivity
contact and adhesion.
Required Ohms per inch is achievable.
65 - 85 ohms thereabouts.
Hi Bush are you painting it on paper card? how does it go painted directly on to Styrofoam and I have not forgotten about the pictures just a bit hectic over Christmas here.

Regards, Ian.
 
Hi Mate,
I have not tried it on foam i think it will react with the thinners though.
I`m favoring full disc shields for the coils just my feeling that target eddys may pass through the spoke arrangement but i have no idea really.
For me to create a spoke coil in foam to test i would find most difficult regardless.

Just now finishing off a 14`` Elite rebuild i bugged some of the winding cable whilst dismantling so have decided to twin wrap it in flat wound dual cable.
Doubt it will work on the sdc but positive it will on the 4500.

Flat wound a F1A4 8`` twin layer into 9 1/2` `results so far air testing looking good & sensitive.
Another F1A4 8`` i`ve finished into 9`` standard layer wrap not tested yet.
Decide to use that white foam as the coil frame partly because i had some laying around about half the weight of the denser polystyrene stuff.
Plastic dinner plates for the casements but trying to get plastic angle for the shaft connection is proving most difficult to acquire so far best i can get is off those angles that protect hardware products in their packaging.

The cardboard / paper shield base was difficult to get the right gauge as it either is too thin prone to breakage or too ridged to bend around the foam rim.
Art supplies had a variety so picked up some various sheets including very thick blotting paper $1.50 a pop.
Pizza cardboard packaging works well but restricted to 9`` dia.

Question in my head i`m dealing with is the high & low end of the graphite coating continuity & what variables could occur.
I`m guessing excessive conductivity will result as a targeted & less will be susceptible to false ground signals.
The other idea is potentially some variables there may suit different types of ground.

Hopefully i`ll have them all completed and functioning soon and actually get out into the field.
Time slips by with great speed this i do know.
cheers
 
The Dinner Plate Coil Top.
Plastic type seems a little brittle and difficult to work with.
The two separate packing protection angle fitment idea for the shaft attachment works well.
It will do for a while to test the new 2 layer 9 1/2 `` flat wound coil.
 

Attachments

  • 20241227_082648.jpg
    20241227_082648.jpg
    2.2 MB
  • 20241227_082842.jpg
    20241227_082842.jpg
    1.7 MB
The Dinner Plate Coil Top.
Plastic type seems a little brittle and difficult to work with.
The two separate packing protection angle fitment idea for the shaft attachment works well.
It will do for a while to test the new 2 layer 9 1/2 `` flat wound coil.
It'll be melamine, which is a thermosetting material (like bakelite), not a conventional thermoplastic. Be careful of dust and fumes when you're working with it:
https://en.wikipedia.org/wiki/Melamine
 
So awesome coming back and seeing the discussion live and well. So much to go over! I put a pin in the project for a while after we had a baby and recently we decided to book in a trip with some family so I think its a good time to continue the work and make an improved coil. So much amazing information in this thread
 
The main issues I faced with my original coil that I am currently trying to fix with my rebuild are:

- No shielding of the wires or coil
- Coil was loose and i noticed when I could swing too aggressively or bump the coil head it would sound off

The second issue is easily fixed when I wrap the coil with insulation and then the shield etc. My biggest concern is the faraday shield and how to ground it, since the PCB I am using wasn't built to have a shield for the coil I am left with trying to figure out where to connect the shield with my limited electrical knowledge... so I asked AI a quick question to see if it would offer some sort of insight and this is what it spat out. While it sounds pretty good for an answer, I want to see what others think before I try this and potentially damage my unit.

My question was: I have a battery powered metal detector and I am making a faraday shield for the mono coil and the wires connecting to the PCB from the coil. The drain wire from the faraday shield I believe should be grounded, but I do not know where I should ground it. Is it problematic if I solder it to the negative terminal that the battery connects to the PCB? Will this cause any issues and will this correctly ground the shield?

The answer below in a spoiler to lower clutter in this post.

Grounding the drain wire of your Faraday shield to the **negative terminal of the battery** is generally a good approach and should work effectively in most cases. However, there are a few considerations to ensure it doesn't cause issues and provides proper grounding for the shield:

---

### 1. **Why Ground the Faraday Shield?**
- The Faraday shield is designed to block external electromagnetic interference (EMI) from affecting the coil and its wiring.
- The drain wire provides a path for any induced currents (from EMI) to be safely diverted to ground, preventing them from interfering with the metal detector's operation.

---

### 2. **Grounding to the Battery Negative Terminal**
- Soldering the drain wire to the **negative terminal of the battery** (or the PCB's ground plane connected to the battery negative) is a common and effective method.
- This ensures the shield is at the same potential as the circuit ground, which is necessary for proper operation.

---

### 3. **Potential Issues to Consider**
- **Ground Loops**: If the Faraday shield is grounded at multiple points, it can create ground loops, which may introduce noise. Ensure the drain wire is grounded at only one point.
- **High-Frequency Noise**: If your metal detector operates at high frequencies, the grounding point should be as close as possible to the coil's connection to the PCB to minimize inductance in the ground path.
- **Battery Isolation**: If the battery is not isolated from the rest of the circuit (e.g., in a floating system), grounding to the battery negative should not cause issues. However, if the system has a separate ground plane, ensure the battery negative is properly connected to that ground plane.

---

### 4. **Best Practices for Grounding the Faraday Shield**
- **Single Ground Point**: Connect the drain wire to a single ground point on the PCB, preferably near the coil's connection point.
- **Short and Direct Connection**: Keep the drain wire as short as possible and route it directly to the ground point to minimize inductance and resistance.
- **Use the PCB Ground Plane**: If the PCB has a ground plane, solder the drain wire to the ground plane near the coil's input. This is often better than connecting directly to the battery terminal.
- **Avoid Floating Grounds**: Ensure the ground point you choose is part of the circuit's reference ground (e.g., battery negative or PCB ground plane).

---

### 5. **Testing the Grounding**
- After grounding the shield, test the metal detector for proper operation.
- Check for any increased noise or interference, which could indicate a grounding issue.
- If necessary, experiment with different grounding points to find the one that provides the best performance.

---

### Summary
Soldering the drain wire to the **negative terminal of the battery** (or the PCB ground connected to it) is a valid and effective way to ground the Faraday shield. Just ensure the connection is secure, the wire is as short as possible, and there are no ground loops. If the PCB has a ground plane, grounding the shield there (near the coil's connection) is often the best approach.
 

Latest posts

Top