Hi BBsailor,
I would love to be able to give you specific techniques to build the ideal DD coil, but I am still learning, so I can't since I still do not know just what works the best for a particular situation.
One thing I have noticed is a DD coil will generally be easier to get to work at shorter delays than a mono coil. It will also be much quieter. So, it should be a little easier to get the DD to work once the damping is correct.
One also has to remember that I am working with one of Eric's PI's, so there may be differences when trying to use similar techniques on a different PI. If you are working with a PI that has a higher pulse on current, most likely you will not be able to get the delay down to anywhere near the delays obtained with a PI having a much smaller pulse current.
What I have found is I have had better luck with a round housing when trying to build a stable DD so I would start with a round housing or design. My smaller elliptical still will produce some falsing which I bang it against a rock, even when the windings are epoxied in place. Also, it appears I have better luck with stability with larger housings than the 11" round one I built.
My 11" round coil is really quite stable. In fact, I have one that just has the windings held in place with foam strips which generates very few false signals and works at the short delay. So, at the present time, that is the one I am working with.
As for building the coil itself, I prefer to make the windings for the xmit and receive the same inductance and size. The reason for my preference is I am changing enough things all the time now. At some point I will try to increase the receive windings but not not now.
Also, if I ever settle down to a minimum delay, then I hope to be able to really refine my coil building techniques, but so far, that hasn't happened. This is one of the big reasons I can't say just what works best.
A good example is the most recent change to try to detect the "invisible" nuggets. My previous DD's wouldn't tune to a short enough delay to allow their detection. So, I had to build a new one.
Ok, now what I have had to do is to reduce the inductance to get mine to work right at the shorter delays. I also have had to adjust the damping accordingly. I slightly underdamp the xmit winding for the best results.
To be honest, since reducing the delay to less than 10 usec, each complete coil I make is individually damped to match the detector. I have a damping resistor in the detector but the value is high enough that all coils are underdamped. Then I will add a damping resistor in the coil or in the plug for final damping the xmit winding. The receive damping is done in a similar fashion.
All coil winding leads are taken back to the plug. I do not tie the receive and xmit common together in the coil housing. Instead, each coil winding is isolated from the other even at the crossover points. At the present time, each coil winding is individually shielded and that shield is tied to the common lead of that particular winding. The shielding is taped with plastic tape at the crossover points to assure there is no electrical connection between the two coils.
At some point, I plan on experimenting more with the shielding connections but not now. Once I settle in on a minimum delay, then I will try different things but until then, I will use what seems to work best for me. However, it is something that a person might want to try just to see what happens.
Generally, I will adjust the windings looking for the "null" point. By observing the signal at the output of the preamp while adjusting the width of the overlap, one can see the signal change as the crossover point is approached. Generally, this will occur when the windings are about an inch or less apart.
Now, I have built coils where I have deliberately left the overlap area wider than normal. This seems to give a little better, or maybe better stated, wider response to very small targets like real small nuggets in the couple of grain range. Again, it is something for a person to experiment with.
In the case of Frank, the detector he is working with is designed to work at about 15 usec. Building a DD to work at that delay is much easier but will still be a challenge at the beginning.
A scope is really a necessity to assure there is no excessive oscillation and to see the nulling. One could build a coil without a scope but if things are not correct, strange results could occur. They may not be noticeable on coin size objects, but will show up when trying to detect when trying to detect the real small non-ferrous objects.
I guess the bottom line is, don't try anything fancy to begin with. Stick with a coil housing like a 11" round one to start. Use careful shielding techniques. Keep all windings tight. Eric mentioned he uses spirawrap on the windings. I have found it works great also. Don't be afraid of reducing the inductance if trying to build a coil to work at the shorter delays.
As for false signals, I still get some of them when operating at the extremely short delays. I am not sure if they are because of coil winding movement, winding to shielding movement, or if some of the problem is simply static discharge. At less than 10 usec, even the movement of the coil lead becomes very critical.
I hope this helps get you started.
Reg
