Menu
What's new
Find's Treasure Forums

Welcome to Find's Treasure Forums, Guests!

You are viewing this forums as a guest which limits you to read only status.

Only registered members may post stories, questions, classifieds, reply to other posts, contact other members using built in messaging and use many other features found on these forums.

Why not register and join us today? It's free! (We don't share your email addresses with anyone.) We keep email addresses of our users to protect them and others from bad people posting things they shouldn't.

Click here to register!

Need Support Help?

Cannot log in?, click here to have new password emailed to you

Open transmit coil

A

Anonymous

Guest
I know this is far fetched but it seems to me that if the transmit coil was open circuited more of the magnetic field would colapse into the targets. The coil doesn't represent a shorted turn because of the damping resistor but absorbes a lot of energy as evidenced by the flyback voltage imposed on the damping resistor. A shorted coil really eats energy.
I don't kid myself by what this means in the way of huge voltages on the open coil and the difficulty of doing this. But has this ever been done? Or anything like this.
JC
 
Hi JC,
The magnetic field from the coil is determined by the number of turns and the current flowing through it while the pulse is on. It is this field, or a portion of it depending on the relative size and range, which interacts with the object. How you switch off, fast or slow, affects the signal generated in the object for a different reason. Also the energy stored in the field has got to be dissipated somewhere. If a coil were truly open circuited instantaneously, then a vast voltage would cause a spark which would result in breakdown of insulation and destroy transistors etc in its path. The damping resistance across a coil allows the energy to be dissipated in a more controlled, slower, manner so that the coil voltages pose no danger to surrounding components. Provided the collapse of the field is fast (see my post
 
Eric,
I have just finished measuring my coils. I used the lowQ formula, an impedance meter built to measure lowR cross-over inductors, and an oscillator with accurate readout. I measured at around 1khtz. The Tx coil (5x paralled Kynar) is Rdc = 1.6ohm and L = 200uH. My damping R = 207ohm. The Rx coil (5x 7/41 Litz) is Rdc = 2.3ohm and L = 185uh. My damping R = 470 ohm.
Please prompt me how to calculate time constant.
g
 
Hi Eric
I did read/print your It even gets worse posting. Thanks for the response. I know opening the coil is insane as far as practical electronics but I have seen some pretty clever does done before (switching with high voltage diodes). Maybe a single turn coil (lowest C low L). Run fast (filters out better) with smaller current. You answered my question though, it probably doesn't do any good.
JC
 
Hi Eric
That's interesting about the first order response on non ferrous objects. So Candy doesn't have it quite right. The complexity of the subject grows larger. That's why I felt they made two periods short and long, and sampled the long one at two places, got three signals which reacted differently to different objects and then tested it until they could kinda discriminate close to the coil where the signals are stronger. But that most of the tweaking of the look up table was made by testing. Don't blame them a bit.
Once more on the litz wire subject. If I remember correctly Candy talked about using it because of the decay of eddy currents in DD coils being especially bad where the coils overlap. This transformer action wouldn't be so bad except that the coupling of the transformer changes as the coil is moved closer and further from the ground, so the deal was to use fine litz wire and get the currents to die quickly. Sorry the coil fell apart but glad minelab is practicing what they preach. Especially for the money, you should get some litz wire whether you need it or not.
I agree with the audio hype. Too many friends asking too many questions about the stuff that is for sale. Seems any gimik will sale stuff. Thought about one of my own, but seems all gimiks have been taken. Problem is driving a reactive load like speakers with a fixed impedance transmission line. Until they solve that things like skin effect are way down in the mud.
Large coils produce better signals/depth for large objects (buried gold and silver bars), while better ignoring small surface objects, like 22 rifle brass. Like you said since the energy is spread out over a larger area the brass doesn't get very much.
Large coils love to pick up AM radio stations better than small ones. The entire planet is full of RF radiation not a quiet spot even out in the middle of nowhere in New Mexico.
JC
 
Hi JC,
If you had read WHITLOCK'S paper you would have learned one of the principal advantages of his balanced amplier circuit is RFI suppression. Despite his address being Jensen Transformers, the paper does not prescribe using them!!
Been quite busy - trying to speed up my sampling times as Eric suggested - not easy as the time-base is a 32768 watch crystal. Rembered late in the afternoon that I had some FM stereo crystals - 38khtz. It works, though I havn't had time to fully explore - should take my short sampling periods to around 12uS delay. If Candy uses the 32768htz crystal, then it may be a quick hot-rod retro!!!
Don't know about you, but I am a hobbiest only. Very pleased to rat some computer power supplies - very nice low impedance electros, and to my great delight 600v 0.6ohm IRF mosfets. Guess where that is going in the morning - current time here is 3:07am.
g
 
Hi Eric,
There is no technical argument presented in Candy's paper. He states the Litz DD coils were tried with a detector with ground-balancing. Raising or lowering the head "did not significantly alter the ground balance, and across ground a "very significant improvemwent in maintaining stable ground balance - hence significant improvement in discrimination sensitivity" He mentions a single strand coil, but dismisses it on grounds of driver complexity.
Commenting on previous postings re static and your comment re conductive plastic, I recall someone patenting making the search head out of carbon-fibre. They think of everything!!
Thankyou for your comments re time-bases. I started by graphing Candy's Tx and Rx patterns. I soon realised that the intervals divide by eight - so the device I used was the4022 - then And and Orred.
You are most probably right with your comments re common-mode. Whitlock boot-straps the input Rs and parallels this with bootstrapped C filters to deal with RFI. I still like two two coil, balanced input though, And like Candy I like switching the Rx - initially I was just going to reverse the Tx pattern, but plumbed for a very slight delay - from memory under 1uS. Still playing on the bench, neither of my coils are shielded!!
For those who havn't enough reading I list all Candy's patents (excluding litze previously mentioned)
04894618 04942360 o5506506 05537041 05576624
JC!! You seem stuck on 05506506!! Forget "look-up" - the latter two papers are far more interesting!
g
 
I reread that one patent. Most of these I read years ago, and I would have to reread them all to get everything just right for you. Too busy building and working. Moving the search coil up and down changes the coupling of two coils. Even without mineralization there is a change in dielectric constant of the air and earth which distorts the magnetic field, adding salt water or black sand makes this much worse. Therefore to keep good ground balancing correct for one spot when lifting the search coil up and down, you have to consider this effect (least as far as Candy is concerned). He states this made a significant improvement (litz wire)(has to for patent) could be all wet. For the purposes of maintaining ground balance.
White's did the patent on carbon fibre and getting the length of the fibres just right to form a shield without absorbing all the transmit energy.
JC
 
Hi Eric
Two receive coils used or wound antiphase certainly does cancel external noise very well. Better than almost anything else you can do. Modified the coil structure for a walk through metal detector company once to include two coils instead of one for each of 12 zones, wound differential and common mode noise went away.
Great way to go.
Speaking of noise, would it be possible to electrically dither (change back and forth) the timing of the transmit and receive times to prevent beating with external signals.
Another scheme I haven't had time to try would be to continuously sample across the receive waveform with every other sample subracted (in the diff integrator) in order to better cancel frequencies instead of two samples which will certainly beat with even harmonic signals which match the spacing of the samples. The first sample would be positive as now, the second also positive, next negative, next positive, and so on till the end with two negative samples. Now we have some samples across the interfering signal to try to integrate to zero. The width of the samples could also be varied to better sample various frequencies, but total positive and negative time still needs to add to zero with exception of the first and last sample. This is hard to discribe but hopefully you understand.
JC
 
Actually I wasn't very clear on dithering the time between the two differential samples. You want the sample times to be the same or the integrator drifts. I use the same timing generator to make both sample times, so even if it changes a small amount the sample times are the same. The timing between them is not so critical, so if this time could be varied somewhat with say a ramp generator reset at the integrator time constant (say 1 or two seconds) then this between sample time could vary in a linear manner an amount (say 100 microseconds) over the 1 or two second period. Then fewer samples would beat hard with the interference and high order frequencies would be cancelled better. This would only offer some improvement and would not be perfect, but would be easier to implement then other schemes.
JC
 
Hi Graeme
Please tell what is so great in Candy's more recent patents.
JC
 
Hi JC,
I am trying to follow your reasoning - and when I see mention of positive and negative signals, then I am inevitably drawn back to Candy. I am not a mathmetitian - even a clever mate with a Phd in maths sat back in silence.
Candy transmits a 150uS pulse, then follows with four equi-spaced 60uS pulses, shuts up for 300uS, then repeats the pattern. He uses a total of 12 sampling periods in Rx - 4 of which interogate the longer Tx pulse at differing durations. The four short Tx pulses are sampled at 15uS and 45uS.
The later sampling periods are inverted and subtracted from the earlier sampling. Is this your positive and negative signals thinking??
g
 
Yes the positive and negative is the first sample going to the positive and the later subtracted (negative) sample. Candy does it different than Eric and most in so much as he has and inverting amplifier for making the negative and others go to a differential amplifier, with positive and negative inputs. I prefer this otherwise you get to deal with offset drift of the second amplifier.
JC
 
Hi JC,
I'll buy that - not up to speed, but I appreciate where you"re blowing.
g
 
Are you sure? Which ones are they. All are analog to a point.
JC
 
You must log in or register to reply here.
Top