Litz wire size for coils

[Darin]

Thanks guys for all the reply's, it really help.
I am still trying to find that list I read some where and will post it if I ever find it.
I feel better about the 330/46 litz wire I have coming after reading fatbob's post, thanks.

[green]

You can wind a 300uH coil(short lead length) with a 2Mz resonance not using litz wire. What difference in target response, decay time or anything else should I expect with a PI detector with a litz wire coil if both type coils with lead wire have the same resonant frequency? Is there a reason other than maybe resonant frequency to use a litz wire coil for a PI detector?

[bbsailor]

You can wind a 300uH coil(short lead length) with a 2Mz resonance not using litz wire. What difference in target response, decay time or anything else should I expect with a PI detector with a litz wire coil if both type coils with lead wire have the same resonant frequency? Is there a reason other than maybe resonant frequency to use a litz wire coil for a PI detector?

The two factors that restrict target detection are: 1. coil circuit capacitance which affects the damping resistor value and, 2. the residual eddy current decay time in the coil wire itself which is affected by the delay setting.

The coil resonance is related to 1 above, however when detecting very small targets like gold nuggets, you need to quickly switch the RX circuit on to catch the quickly falling eddy currents in a small gold nugget or the links of a fine link gold chain. Using Litz wire prevents eddy currents from being generated in the coil wire. The issue is: at what delay are you sampling? The lower the delay, the more care must be used to allow the fastest RX turn on time (lowest delay).

Another issue is what is the peak current per pulse? Lower current pulses can be damped quicker.

Typical delays of 10 uS to 15 uS can be used on beach PI machines looking mostly for coins and large jewelry like rings at higher delays. However, when you get near and below 10 uS delay, the ability to reduce the delay and not lock up the RX circuit can be a challenge. You will be detecting the residual signal in the coil caused by less than optimal coil damping and the residual eddy currents decaying in the coil wire. Remember, we are only dealing with a few microseconds here.

The PI machine needs to have the ability to operate at low delays to see if the chosen search coil can operate at that PI machines lowest delay. One trick is to use a variety of small target sizes and materials and then reduce the delay to see if the PI machine locks up. If it does then you have found the limit of your PI machine and that coil! Eric Foster reported making a PI machine and coil that went down to 5 uS. I believe he used Litz wire on this 5 uS coil. All of the above issues are interrelated.

I hope this helps?

Joseph J. Rogowski

[green]

The two factors that restrict target detection are: 1. coil circuit capacitance which affects the damping resistor value and, 2. the residual eddy current decay time in the coil wire itself which is affected by the delay setting.

The coil resonance is related to 1 above, however when detecting very small targets like gold nuggets, you need to quickly switch the RX circuit on to catch the quickly falling eddy currents in a small gold nugget or the links of a fine link gold chain. Using Litz wire prevents eddy currents from being generated in the coil wire. The issue is: at what delay are you sampling? The lower the delay, the more care must be used to allow the fastest RX turn on time (lowest delay).

Another issue is what is the peak current per pulse? Lower current pulses can be damped quicker.

Typical delays of 10 uS to 15 uS can be used on beach PI machines looking mostly for coins and large jewelry like rings at higher delays. However, when you get near and below 10 uS delay, the ability to reduce the delay and not lock up the RX circuit can be a challenge. You will be detecting the residual signal in the coil caused by less than optimal coil damping and the residual eddy currents decaying in the coil wire. Remember, we are only dealing with a few microseconds here.

The PI machine needs to have the ability to operate at low delays to see if the chosen search coil can operate at that PI machines lowest delay. One trick is to use a variety of small target sizes and materials and then reduce the delay to see if the PI machine locks up. If it does then you have found the limit of your PI machine and that coil! Eric Foster reported making a PI machine and coil that went down to 5 uS. I believe he used Litz wire on this 5 uS coil. All of the above issues are interrelated.

I hope this helps?

Joseph J. Rogowski

Thanks for the reply. Charted the target decay for a 2 inch piece of awg19 magnet wire. A 1 inch piece charted the same TC at lower amplitude. Tried a piece of awg28 magnet wire the coil is wound with but didn't get a response. Wondering if there is a way to predict time constant knowing wire size. Maybe the decay of the no target recording between 4 and 5usec is the awg28 coil wire decaying. Maybe not since the target decays from a plus level.

[green]

Thanks for the reply. Charted the target decay for a 2 inch piece of awg19 magnet wire. A 1 inch piece charted the same TC at lower amplitude. Tried a piece of awg28 magnet wire the coil is wound with but didn't get a response. Wondering if there is a way to predict time constant knowing wire size. Maybe the decay of the no target recording between 4 and 5usec is the awg28 coil wire decaying. Maybe not since the target decays from a plus level.

Coil: Rx two round coils side by side, opposite phase surrounded by a race track Tx coil. Maybe Rx coils not quite cancelling and decay between 4 and 5usec is coil wire.

[Monolith]

Thanks for the reply. Charted the target decay for a 2 inch piece of awg19 magnet wire. A 1 inch piece charted the same TC at lower amplitude. Tried a piece of awg28 magnet wire the coil is wound with but didn't get a response. Wondering if there is a way to predict time constant knowing wire size. Maybe the decay of the no target recording between 4 and 5usec is the awg28 coil wire decaying. Maybe not since the target decays from a plus level.

A valid alternative to Litz wire, is fine strand tin plated wire.
It would be very interesting to actually see the difference in coils, wound with different wires.
All the coils would need to be as equal as possible, in inductance, resistance, winding Technic, diameter etc.

[WM6]

330/30 / fast coil / makes good coil for hunting gold nuggets

660/46 / really fast coil / makes good coil for hunting gold nuggets

It is pretty nonsense to hang on such tinniest Litz wire till you still have a lot of much bigger solder joints in your coil.

Minelab patented means to overcome this problem.

[godigit1]

Here is a fast coil made with 28 awg solid tin plated copper
4.5"x 9" 3dss

.

It runs 325 uh at 3.3 ohms resistance.
Here is a scope shot on the minipulse at 5us sample delay.
I have also now tested this coil on my TDI Pro and its a keeper. I wont say it beats my folded mono all around but it does have a better nose and tail depth.
As far as sensitivity it will get a 1/4 x 1/4" piece of a cola can at about 1.5 " on the minipulse in air.
This coil has not seen the ground yet but GB on the tdi seems fine with it so I think it will be fine.



Here is a shot of a 12 dual field coil also on the minipulse with the same settings.
I think the dual field is a litz coil but I may be wrong I cant find any pictures inside.
Its is 367 uh at 2.9 ohms.



As you can see from the shots the solid wire coil has a slightly faster rise time and is clearing as fast. But the solid wire coil is charging to a higher current level.
Both coils are matched closely enough that I can switch from the dual field to my coil without changing the damping resistor. Both scope shots are the same damping resistor..
As far as I can tell about Skin effect And I'm no expert by any means at a 1000hz its so thick it penetrates the whole conductor on smaller awg wires.
That being said you would think a 30 awg wire would be effected very little. But 30 awg poses other problems resistance being the main one.
I think 28 awg would be a good choice but I have not been able to find any locally yet.
Best
digit

[green]

A valid alternative to Litz wire, is fine strand tin plated wire.
It would be very interesting to actually see the difference in coils, wound with different wires.
All the coils would need to be as equal as possible, in inductance, resistance, winding Technic, diameter etc.

The two factors that restrict target detection are: 1. coil circuit capacitance which affects the damping resistor value and, 2. the residual eddy current decay time in the coil wire itself which is affected by the delay setting.[bbsailor reply #18]

I agree with inductance, resistance and coil diameter not winding technic. A bundle wound coil with Teflon coated wire is going to be better than a bundle wound coil with magnet wire. A spider wound coil with magnet wire might be better than a bundle wound coil with Teflon coated wire. Other than cost and maybe diameter for the same resistance Litz is probably better. Is it necessary? I ran the TC for a 2 inch piece of awg12 cu wire(12.5usec after 10usec) and a 2 inch piece of awg24 cu wire(.67usec). I'm guessing the awg28 magnet wire I use probably shows up with short delays. Guessing awg32 or34 wouldn't. I agree (It would be very interesting to actually see the difference in coils, wound with different wires.) To see if wire diameter has any effect, coil resonance with lead needs to be the same. Another question, does stranded wire act like Litz wire in a PI coil?

[Monolith]

Here is a fast coil made with 28 awg solid tin plated copper
4.5"x 9" 3dss

[ATTACH]34834[/ATTACH].

It runs 325 uh at 3.3 ohms resistance.
Here is a scope shot on the minipulse at 5us sample delay.
I have also now tested this coil on my TDI Pro and its a keeper. I wont say it beats my folded mono all around but it does have a better nose and tail depth.
As far as sensitivity it will get a 1/4 x 1/4" piece of a cola can at about 1.5 " on the minipulse in air.
This coil has not seen the ground yet but GB on the tdi seems fine with it so I think it will be fine.

[ATTACH]34835[/ATTACH]

Here is a shot of a 12 dual field coil also on the minipulse with the same settings.
I think the dual field is a litz coil but I may be wrong I cant find any pictures inside.
Its is 367 uh at 2.9 ohms.

[ATTACH]34837[/ATTACH]

As you can see from the shots the solid wire coil has a slightly faster rise time and is clearing as fast. But the solid wire coil is charging to a higher current level.
Both coils are matched closely enough that I can switch from the dual field to my coil without changing the damping resistor. Both scope shots are the same damping resistor..
As far as I can tell about Skin effect And I'm no expert by any means at a 1000hz its so thick it penetrates the whole conductor on smaller awg wires.
That being said you would think a 30 awg wire would be effected very little. But 30 awg poses other problems resistance being the main one.
I think 28 awg would be a good choice but I have not been able to find any locally yet.
Best
digit

The skin effect on the coil wire is of no importance.
The thickness of the copper wire has influence in the coil DC resistance and coil inter wire capacitance and the coil to shield capacitance.
The way your coil is wound, there is very little influence in the inter-wire capacitance as it is very low.
The copper of the coil has it's own eddy currents. Eddy currents in thin copper wire are of shorter duration than in thick wire.

This site has a calculator for calculating the capacitance between two parallel wires. Play with wire thickness and distance between the wires, it is quite educative.
http://www.mogami.com/e/cad/electrical.html

[Monolith]

The two factors that restrict target detection are: 1. coil circuit capacitance which affects the damping resistor value and, 2. the residual eddy current decay time in the coil wire itself which is affected by the delay setting.[bbsailor reply #18]

I agree with inductance, resistance and coil diameter not winding technic. A bundle wound coil with Teflon coated wire is going to be better than a bundle wound coil with magnet wire. A spider wound coil with magnet wire might be better than a bundle wound coil with Teflon coated wire. Other than cost and maybe diameter for the same resistance Litz is probably better. Is it necessary? I ran the TC for a 2 inch piece of awg12 cu wire(12.5usec after 10usec) and a 2 inch piece of awg24 cu wire(.67usec). I'm guessing the awg28 magnet wire I use probably shows up with short delays. Guessing awg32 or34 wouldn't. I agree (It would be very interesting to actually see the difference in coils, wound with different wires.) To see if wire diameter has any effect, coil resonance with lead needs to be the same.

This wire should be nearly as good as real Litz wire and very much cheaper.
http://www.aliexpress.com/item/High-...8.3.355.5MS74U

A bit more expensive is the silicon coated high temperature wire with similar stranding.

[baum7154]

It is pretty nonsense to hang on such tinniest Litz wire till you still have a lot of much bigger solder joints in your coil.

Minelab patented means to overcome this problem.

----------------

I agree completely that all solder joints should be eliminated in fast coils for small gold detection. This is why I recommend the use of a continuation of the coil wire to form a twisted pair feedline of low capacitance and keep it as short as practical to minimize capacitance. This is especially true when sampling below 8us. Every technique must be used to push the sample time shorter.

Smaller wire gauge minimizes coil capacitance but at the expense of lower coil current. I think 24 through 28 gauge stranded 600 volt PTFE insulated wire is a very good choice for coil/feed construction. I'd like to build a litz 3DSS coil to see how it compares to the ones I've already built. However I am very pleased with the performance of the 4 X 12.5 3:1 328uh 3DSS coil running at 6.98us in my overclocked Chance PI detector. Perhaps a Litz 3DSS coil would be a bit better but only construction of one can tell. The 3DSS coil I built with 23awg solid copper 200 volt PTFE insulation was a dissappointment as it would not operate sub 10us as I recall.

Regards,

Dan

[WM6]

----------------

I'd like to build a litz 3DSS coil to see how it compares to the ones I've already built. However I am very pleased with the performance of the 4 X 12.5 3:1 328uh 3DSS coil running at 6.98us in my overclocked Chance PI detector.

Regards,

Dan

Yes or very likely.

There is simple test.

Good built Litz wire coil could not be (in its coil circle) detected by using par example small (5") coil of one sensitive VLF detector as Gold Bug.

If something (like solder joints, contact wire for shielding etc.) in such test is detected, then coil design need some more improvement.

"Litz" properties of regular multi-strand wire could be refined by strands surface oxidation which can be forced by connection (~30m wire per coil) to output of electric welding machine for short (seconds or so) muffle heating. Surface oxidated strands are more joint-isolated than bare. Procedure suitable for PTFE insulated wire.

[bbsailor]

----------------

I agree completely that all solder joints should be eliminated in fast coils for small gold detection. This is why I recommend the use of a continuation of the coil wire to form a twisted pair feedline of low capacitance and keep it as short as practical to minimize capacitance. This is especially true when sampling below 8us. Every technique must be used to push the sample time shorter.

Smaller wire gauge minimizes coil capacitance but at the expense of lower coil current. I think 24 through 28 gauge stranded 600 volt PTFE insulated wire is a very good choice for coil/feed construction. I'd like to build a litz 3DSS coil to see how it compares to the ones I've already built. However I am very pleased with the performance of the 4 X 12.5 3:1 328uh 3DSS coil running at 6.98us in my overclocked Chance PI detector. Perhaps a Litz 3DSS coil would be a bit better but only construction of one can tell. The 3DSS coil I built with 23awg solid copper 200 volt PTFE insulation was a dissappointment as it would not operate sub 10us as I recall.

Regards,

Dan

Dan, and all interested

Go to this old 2004 web post http://www.findmall.com/read.php?34,...197#msg-135197 to see more information about Litz wire. Also, use the findmall search feature to search on the word "Litz" under the name Eric Foster for some more good Litz wire background information.

Joseph J. Rogowski

[bbsailor]

----------------

I agree completely that all solder joints should be eliminated in fast coils for small gold detection. This is why I recommend the use of a continuation of the coil wire to form a twisted pair feedline of low capacitance and keep it as short as practical to minimize capacitance. This is especially true when sampling below 8us. Every technique must be used to push the sample time shorter.

Smaller wire gauge minimizes coil capacitance but at the expense of lower coil current. I think 24 through 28 gauge stranded 600 volt PTFE insulated wire is a very good choice for coil/feed construction. I'd like to build a litz 3DSS coil to see how it compares to the ones I've already built. However I am very pleased with the performance of the 4 X 12.5 3:1 328uh 3DSS coil running at 6.98us in my overclocked Chance PI detector. Perhaps a Litz 3DSS coil would be a bit better but only construction of one can tell. The 3DSS coil I built with 23awg solid copper 200 volt PTFE insulation was a dissappointment as it would not operate sub 10us as I recall.

Regards,

Dan

I found another findmall link to read about Litz wire: http://www.findmall.com/read.php?34,...780#msg-136780

I hope this helps.

Joseph J. Rogowski