Differential Coils ... Making

[WM6]

Hi all,

if you want a very low capacitance differential coil, just use a 300 Ohm twin lead antenna cable. ( http://en.wikipedia.org/wiki/Twin-lead )
Capacitance: 11.8 pF/m (meter - not foot!)

Wind the coil spiral like with polystyrene tape between windings (winding spacer, 1-2 mm thick). The voltage between each winding will be low so the energy stored in the interwire capacitance will therefore be low too. Only the two coil halves will see the full half flyback voltage. But the distance between them forming a low capacitance.

Connect the coil halves to a center-tapped coil.

Aziz

Very good advice, Aziz. Probably it is hard to get such old-art cable now.

[moodz]

This coil aint like your mother used to make.

The expression for the current decay across an Inductor is given by:

iL(t) = i0 e-(R/L)t t

≥ 0

where,
i0 is the initial current stored in the inductor at t = 0 ( ie current at switch off )
L/R = τ is time constant of coil.


Note the differential coil is wound with twisted pair .... which is a transmission line ... lumped LRC analysis will only be accurate for frequencies less than 10 Khz ... There is around 40 - 50 meters of twisted pair in my coils this is significant from an analysis point of view. Reducing the capacitance between the twisted pair will raise the characteristic impedance of the twisted pair and the "self sheilding" effect will be reduced .... ie near the ground.

moodz.

[moodz]

iL(t) = i0 e<SUP>-(R/L)t </SUP> t ≥ 0

[moodz]

Thank you for the great project. Can't wait to see it working.
I have some doubts regarding twisted pair coil design:

1. There is a huge capacitance between twisted wires.
If good coaxial cable has 16 pF/ft. then twsted pair has 30-40pF/ft or so.
Coil will have additional approx. 35pF x 60ft. = 2160 pF
You can find twisted-pair capacitance calculation here:
http://www.signalintegrity.com/lib/htm/TWIST.htm
It may help to reduce capacitance if wires will not be twisted
or coated thick insulation with low dielectric constant (teflon,polyethylene)

2. If flyback voltage can reach 1100V then you can expect 550V between
wires in twisted pair according diagramm below.
Regular magnet wire can survive 130V only.
So it requires much stronger insulation. At least 600V teflone coated wire.
http://www.bulkwire.com/wire-cable/p...nded-wire.html
Unfortunately this wire is silverplated so it inducing more eddy currents in the
wire then tinned wires. Teflon coated Litz wire will be best if it is possible to find.

Some clarification must be made here .....

1. Breakdown voltage of polyureathane insulated magnet wire is better than 2700 volts .
Wire to wire is better than this.
magnet_wire_8051.pdf

2 Capacitance of twisted pair is intrinsic ... lowering the capacitance only increases the characteristic impedance ...must not forget about the mutual inductance between two closely spaced conductors. My ADSL link runs across 7000 meters of twisted pair ... but the signal is not shorted out by the huge "capacitance" across the pair.

Moodz.

[Waikiki_Sweep]

Just want to remind why low capacitance is so important for PI coils.

May be ADSL modem can work with high capacitance cable but not a good pulse induction detector.

Just want to remind why low capacitance is so important for PI coils. Coil with lower capacitance will decay faster so you can sample early on most sensitive period just after flyback and also rise sensitivity to small low conductance targets like small gold nuggets or jewelry. This theory has been established on the Geotech forum as well as other forums for many years. Most of members knows it.

Some quotes:

Normally when a coil has less capacitance the damping resistance is increased. Conversely, higher coil capacitance requires lower values of damping resistors. Think of it as coil circuit capaitance stores energy and must use lower values of damping resistance to acheve the best damping for early sampling.
http://www.geotech1.com/forums/showp...56&postcount=3

To understand basic principals of design PI coils it is strongly recommendes to read:

Making a Fast Pulse Induction Mono Coil
A Practical How-to Guide and Tutorial
By: Joseph J. Rogowski
Also known as: bbsailor
http://www.geotech1.com/pages/metdet...s/FastCoil.pdf

Quote from the bbsailor document:
...The magnet wire 384uH coil with a 733KHz resonance has 123pF of capacitance while the Teflon wire insulated coil of 317uH resonates at 1.25MHz and has 51pF of capacitance...

( And it is not a twisted pair. You can see that for good coil capacitance for whole coil is expected close to 51pF. That is equal of approx. 1ft. of twisted pair wire capacitance. So capacitance of "around 40 - 50 meters of twisted pair" is 100 times greater. That will make a very slow coil with thousands pF capacitance.)

Also I am still not sure that it is good to make twisted pair with 550V between wires with using enameled magnet wire. I belive that most electricians will agree that it is mistake. Brake any transformer even for such low voltage as 110V and you will see that construction avoids direct contact even between layers of magnet wire not only ends. If anybody will twist 110V enameled ends from transformer he will be fired from the job. :-)

With my experience coils from wire with poor insulation (like enameled magnet wire) are unstable especially in presence even a little amount of moisture. Static discharges between windings and stray capacitance makes noises. I have throwed out several coils for my underwater detector that I have made from magnet wire because of such noises and unstable results. And now I am using teflon wire because no such problems.

Also a great coil has been explored by Apberg. Core of coaxial cable with shield stripped out. Great insulation and spacing makes it low capacitance and sensitive to small gold. http://www.geotech1.com/forums/showthread.php?t=16322

[moodz]

Firstly thanks for your comments which are all perfectly correct ... I have made some commentary below.

May be ADSL modem can work with high capacitance cable but not a good pulse induction detector.

That example was probably a poor comparison on my part however I was trying to explain that frequency response / bandwidth of electronic systems is dependant on more than just capacitance and PI is a wideband system.

Just want to remind why low capacitance is so important for PI coils. Coil with lower capacitance will decay faster so you can sample early on most sensitive period just after flyback and also rise sensitivity to small low conductance targets like small gold nuggets or jewelry. This theory has been established on the Geotech forum as well as other forums for many years. Most of members knows it.

Some quotes:

Normally when a coil has less capacitance the damping resistance is increased. Conversely, higher coil capacitance requires lower values of damping resistors. Think of it as coil circuit capaitance stores energy and must use lower values of damping resistance to acheve the best damping for early sampling.
http://www.geotech1.com/forums/showp...56&postcount=3

To understand basic principals of design PI coils it is strongly recommendes to read:

Making a Fast Pulse Induction Mono Coil
A Practical How-to Guide and Tutorial
By: Joseph J. Rogowski
Also known as: bbsailor
http://www.geotech1.com/pages/metdet...s/FastCoil.pdf

Quote from the bbsailor document:
...The magnet wire 384uH coil with a 733KHz resonance has 123pF of capacitance while the Teflon wire insulated coil of 317uH resonates at 1.25MHz and has 51pF of capacitance...

( And it is not a twisted pair. You can see that for good coil capacitance for whole coil is expected close to 51pF. That is equal of approx. 1ft. of twisted pair wire capacitance. So capacitance of "around 40 - 50 meters of twisted pair" is 100 times greater. That will make a very slow coil with thousands pF capacitance.)

All of the above is well researched and proven for "conventional" coils. I am trying here to promote the differential PI monocoil which is not "conventional" ..... A primary demonstration of this is that the Faraday sheilding required on a good "conventional" coil is NOT required on a differential coil.

The conventional coils you decribe above effectively have one side grounded and the other runs "Hot" or high impedance. This means that the flyback will try to capacitively couple back via any path to get back to ground ( the other end of the coil ) and it will do this by whatever conductor is nearby .... like the ground etc. I describe these coils as VOLTAGE coils.

On the differential coil there are two "hot" ends of opposite polarity so the capacitive leakage to ground is equal but opposite so it balances. The signal difference is only measured across the two "hot" ends NOT to ground ...this again is a fundamental advantage over the single ended coil as described above. The differential coil is a CURRENT coil.

The principles of the traditional PI coil are not necessarily applicable to the differential coil .... the laws of physics have not changed ... only the analysis.

Also I am still not sure that it is good to make twisted pair with 550V between wires with using enameled magnet wire. I belive that most electricians will agree that it is mistake. Brake any transformer even for such low voltage as 110V and you will see that construction avoids direct contact even between layers of magnet wire not only ends. If anybody will twist 110V enameled ends from transformer he will be fired from the job. :-)

No argument here ... however high voltage baluns / pulse transformers similiar in construction to the differential coil and different to power transformers are reqularly wound with twisted pair so long as peak specified insulation voltages are not exceeded so noone will be fired :-).
Of course Teflon insulation will provide even higher voltage breakdown limits. The main issue with polyureathane insulated magnet wire is poor mechanical handling during construction that introduces scratching or other damage to the wire coating.
Teflon coated wires will result in a slightly thicker coating which will result in higher characteristic impedance for twisted pair over polyureathane coating.

With my experience coils from wire with poor insulation (like enameled magnet wire) are unstable especially in presence even a little amount of moisture. Static discharges between windings and stray capacitance makes noises. I have throwed out several coils for my underwater detector that I have made from magnet wire because of such noises and unstable results. And now I am using teflon wire because no such problems.

Also a great coil has been explored by Apberg. Core of coaxial cable with shield stripped out. Great insulation and spacing makes it low capacitance and sensitive to small gold. http://www.geotech1.com/forums/showthread.php?t=16322

Moisture in any circuit at these voltages would not be good. I am definitely not making any claims that the differential coil should be wound only with magnet wire .... it should work with any sort of twisted pair wire ( taking voltage ratings into account of course. )

moodz

[Stefan]

Hi Moodz,

Have a question for you which has been bugging me for some. So here is my long winded sort of a query.

With a differential coil the common mode signals cancel out. This is fundamental in all instrumentation circuits using twisted wire with a shield. In order for the canceling to occur, the current in the two wires must flow in the opposite direction. This is true for differential transformers as well and equally applies to differential coils.

With your coil on power up you are using half of the coil, and on sampling you are measuring the voltage of the two coils in series in which case the current flowing through the coils are both in phase and not directly opposite. So from my understanding you cannot cancel out common mode signals with this arrangement.

Regards,

Stefan

[moodz]

Hi Moodz,

Have a question for you which has been bugging me for some. So here is my long winded sort of a query.

With a differential coil the common mode signals cancel out. This is fundamental in all instrumentation circuits using twisted wire with a shield. In order for the canceling to occur, the current in the two wires must flow in the opposite direction. This is true for differential transformers as well and equally applies to differential coils.

With your coil on power up you are using half of the coil, and on sampling you are measuring the voltage of the two coils in series in which case the current flowing through the coils are both in phase and not directly opposite. So from my understanding you cannot cancel out common mode signals with this arrangement.



Regards,

Stefan

Thanks Stefan I think your description above is partly correct however you are not taking the "centre tap" connection which is effectively at Analogue ground. With respect to this point the current in one coil is negative and the current in the other coil is positive ....or another way of visualising it is in one coil the current B flows OUT of the common point C and IN from A the other coil. So the current is not in phase as you describe. Check the dot notation of the connected coils and you will see that the current is opposite which is correct as the coils are closely coupled via mutual inductance. See diagram below.

The reason for sheilding on twisted pair is more to do with imperfect common mode rejection due to mechanical imperfection of the coil and the lack of "perfect" common mode rejection at the amplifier. If you talk to radio guys many old timers will swear by unshielded balanced lines over "shielded" coax for performance and low noise.

[moodz]

...or another way of thinking about it.

In the diagram above ... if you have a common mode voltage at A and B ... no current will flow because the voltage is the same ! Only a differential signal will result in current flow like a target decay field across L1 and L2.

moodz.

[Monolith]

I have a question about the ground wire.
Do you twist the A and B leads together and leave the ground wire straight, or twist all 3 wires together?

How about using a shielded twisted pair? with the shield being the ground?

I also tried using a CAT 5 cable, using one twisted pair for the A and B RX and another twisted pair for ground.

Each one of the 4 different ways give somewhat different results.

Another question is about the damping resistor(s), under certain conditions the A signal ad the B signal look like a mirrored signal on the scope.

At other times one of the signals "folds over" so that both signal peaks are on the same side.

Have you observed similar things?

Monolith

[moodz]

I have a question about the ground wire.
Do you twist the A and B leads together and leave the ground wire straight, or twist all 3 wires together?

I dont use a "ground wire" as such ... there is only a twisted pair .... there is a ground/common connection which is the centre tap ... the cross connection between the two coils ... this connection is very short though and thus magnetically "invisible".
It would be interesting to see the scope shots for a third wire ( ground ) twisted in.
Post some pics of your results.

How about using a shielded twisted pair? with the shield being the ground?

I think this would be similiar to a three wire twist ....post some pics .... maybe the results are better than two wire twisted.

I also tried using a CAT 5 cable, using one twisted pair for the A and B RX and another twisted pair for ground.

Each one of the 4 different ways give somewhat different results.

Another question is about the damping resistor(s), under certain conditions the A signal ad the B signal look like a mirrored signal on the scope.

This will depend on whether coils are in series or parallel etc. However some configurations will give less noise immunity.

At other times one of the signals "folds over" so that both signal peaks are on the same side.

ditto

Have you observed similar things?

Yes ... however I am pretty much sticking to the monocoil wound with twisted pair and cross connnected as in diagram in previous post above. As time permits I am getting the electronics and DSP ready now. I am have prepared a 100 turn twisted pair coil ( ie 200 turns in series ) to investigate the capacitance issue that seems to dominate some discussions.

moodz

[kenbarls]

You are a real guru moodz. You have questions and then answered by yourself! Amazing questions= Amazing answers!

[Aziz]

You are a real guru moodz. You have questions and then answered by yourself! Amazing questions= Amazing answers!

Have a look one page before (posting number 25):
http://www.geotech1.com/forums/showp...3&postcount=25

He is answering to someones questions.

Look before you think. Think before you post.

Aziz

[taliesin]

Hi J_P

You mean using litz wire only?
Using litz wire in differential coil construction suppress EMI.
It is worth to try litz wire as proposed instead of say (untwisted) speaker cable (see Tesla diagram) to minimize received noise.

Letter marks of connection wires on Tesla diagram do not fully suit to coil connection letters on schematic.
Maybe this for clarification:

this connection looks like a impedance matching balun .interesting

[moodz]

this connection looks like a impedance matching balun .interesting

It is ... an air core balun.