need coil guru help

[Anonymous]

i came across this site last week and i am impressed with the knowledge here. i am playing with a design but i'm not sure about my coil. it is 25 turns of 24awg(same as found in cat5 cable) in a 10.5" loop. it seems fast enough but is power hungry even at 40us/12v transmit time. is there a way to compensate? series resistance, more turns, shorter pulses? should i use a smaller wire size? the cat5 wire attracted me because of the thicker insulation, which i believe will keep the capacitance down. any help or pointers in the right direction greatly appreciated <img src="/metal/html/smile.gif" border=0 width=15 height=15 alt="">

 

[Anonymous]

Eric and Reg seem better qualified to comment on your question.
bbsailor

 

[Anonymous]

Hi John, try to turn to the theory. The sensitivity of a metal detector is proportional to the intensity of magnetic field
H = nI/D = nU/DR = nU/Dnr = U/Dr
In other words, intensity of magentic field in a coil is determined by diameter D, number of turns n and current I. The current is given by the supplied voltage U and coil resistance R which is just a summ of resistances of single turn r.
True, it is just a state for ideal transistor in static regime - long pulse. We must count with real elements and pay attention to real properties. So I think your coil will be optimal with forty turns, using a basket or Lorenz winding, fifty turns. Then the pulse should not exceed 160 microseconds, not to break your FET, with bipolar transistor,three hundred.
Nice day Sid.

 

[Anonymous]

Hi John,
I am not sure what PI design you are using, but the principle is basically the same on all of them. Also, it helps to know just what you want to find with your PI. It does make a difference.
My guess is your coils is around 450 uh or so. This should allow for a fairly fast sample time.
Now, as for your question about what to do about the detector being power hungry, you have basically answered your own question with your suggestions. Again, part of the answer to your question lies in what you want to find.
As for the easiest adjustment to reduce the current, I would use a resistor in series with the FET. Don't be afraid of reducing the current significantly. I think you will find it will not reduce the sensitivity that much to most targets, but the battery life will show a dramatic increase.
Lets say you primarily want to detect small gold with your PI. If that is the case, then you can shorten the pulse time also. However, a shorter pulse will reduce the sensitivity to copper, silver, and other objects.
Reducing the pulse on time and the current in the coil will generally allow for a shorter delay to be used. By sampling sooner, one will see a greater sensitivity to all objects, but especially gold, since a typical gold object will have a very short decay time. Strange as it might seem, this reduction in both current and pulse on time combined with a shorter delay time just might have a net effect of an increase in sensitivity to small gold obects.
There are a lot of misconceptions about what is necessary to do on a PI to increase the depth or sensitivity. The most common one is brute force (increased current) increases depth. Well, it can increase the signal coming back from a target, but if one cannot sample soon enough, that increased signal will be gone before the sample is taken. So in some cases, brute force is of little or no value.
Everything is a tradeoff when trying to design a PI. One just has to know the final objective and go from there.
I hope this helps a little.
Reg

 

[Anonymous]

thanks for the reply. the design i'm working with is a collection of circiuts from variou designs i have seen on the net. what i am after is a machine with a fairly balanced sensitivity to coins and rings, current consumption under 100ma, and air test to 10+ inches. my concern is that my coil uses wire too thick to achieve these goals, as the wire i usually see used is 26 to 32awg. my current coil will allow me to sample at < 10us. the circuit and coil will detect a nickle at only about 9" and a small ring at 7" but is consuming 210 ma. i guess what i really need to know is if my goal is realistic with only 100ma. thank you. <img src="/metal/html/smile.gif" border=0 width=15 height=15 alt="">

 

[Anonymous]

Hi John,
Your goal of having a detector work well while only consuming 100 ma is very realistic. In fact, Eric's GQ does that now. The GQ consumes less than 100 ma and the sensitivity is extremely good. A typical air test on a nickel will usually be 13" to 15", with the variation do to the noise level.
It is hard to say what it would do on a ring without knowing the ring's composition and actual physical size and shape, but I would think it would easily detect it more than 7".
As for the wire used in the coil, I have used various sizes and seen very little difference if any in sensitivity because of it. However, on my PI, there is a resistor in series with the FET that limits the current more than the small changes in resistance do to different wire size.
I really need to say the final sensitivity of a machine is not just dependent upon the coil current. The rest of the circuitry makes a big difference.
I suspect once you reduce your coil current, you will have to concentrate on increasing the sensitiivty through circuitry design and reducing the noise level to obtain your desired results.
Noise is a very big factor in determining the final sensitivity of a PI. One has to reduce the internal noise as much as possible through filtering, board layout, proper choice of parts, and specific circuit design aimed at reducing internal signal interactions. Good power supply regulation is a must.
External noise needs to be reduced as much as possible also. Obviously, one can't shut off all external noises, but good shielding of the windings will greatly minimize the problem.
Reg

 

[Anonymous]

thanks for your help <img src="/metal/html/smile.gif" border=0 width=15 height=15 alt=""> , and a couple questions if you don't mind. what is your experience in using al foil tape for shielding?(the standard hardware store variety)i know the ends can't touch but from what i've read so far maybe its too thick? also what kind of cable is normally used? rg58,59,62? i am currently using a regular computer power cable for experiments and to prevent line loss. anyway, it looks like i'll have lots eo experimenting to do, and lots of time to do it. the water up here won't be wet untill late april. <img src="/metal/html/frown.gif" border=0 width=15 height=15 alt="">

 

[Anonymous]

Hi John,
I haven't tried the foil tape you mentioned in a very long time, so I am not the one to give advice on how it works. If it were me, I would try it to see how well it works and just what were the negative side effects.
A good way is to try the coil before shielding and carefully observe the response, then try the same coil with the shielding. One should see a reduction in the noise level but not much of a change in the overall response curve at the output of the preamp. Don't bother looking at the xmit signal. Normally, it is too difficult to see what happens there.
I have tried lead foil such as what Eric uses, but it is difficult to find, so I switched to something different that seems to work well. Unfortunately, it is fairly expensive to try or build a coil or two. I use a product made by 3M. The part number is 1091-1. I get it from Contact East. The link to their site is:
http://www.contacteast.com/product/group.asp?parent_id=9427
Their part number is:141-317. The info they list for the product is: CONDUCTIVE ACRYLIC 5.5 MIL COPPER-PLATED POLYESTER, FABRIC
This is a fabric coated with a thin layer of copper, not the foil normally found.
Now, I have used different types of coax including the ones you mentioned. However, I generally prefer a DD coil and use a special S video cable made by Mogami. It is one of the few that lists its capacitance. I have tried a few off brands but they generally don't work as well do to their characteristics.
As for mono coils, I will still use the Mogami simply for the low capacitance of the wire and the fact it is shielded a little differently. The shield is a served type, not the typical braid. This makes it less likely to be detected.
Whatever you decide to use, try to avoid any coax using steel either in the core or shielding. Yes, some of them have it.
I like your description of your water. First impression is you have dry water. Do you just add hot instant coffee to make it wet? I just had to see if Canadians have a sense of humor.
Reg

 

[Anonymous]

Hi John,
The cable from the coil needs to be coaxial, or at least with an outer grounded screen, otherwise you will get strange effects. The AN19/2 PI mine detector has an unscreened cable, and it detects the aluminium shaft just from the proximity of the cable. Not good if the cable moves. Also you will get additional electromagnetic noise pickup. I use RG58 cable for most PI's that use mono coils.
Eric.