The step by step build of the Delta Pulse Rev. 5

[Infamy]

So the first image is my board with stage three complete minus IRF840 and Resistor 680.
Image 2 are the materials used to build my first coil, a 36 inch mono. The board underneath rotates, it's for the wire.
Image 3 & 4 is the coil built with 4 rows 4 columns of wire 16 turns target inductance 750 uH.
The dowel pegs that held the coil in place will be knocked out and the coil will be tied down with zip ties using the dowel peg holes



I used tape to keep the wires in place. Using the spiral wrap was interesting. To achieve the tightest packing of the wires the tape need to be removed as I progressed
so the wires could move and I had to stop fighting the wrap & and let the wires move naturally to find the best fit. The length of spiral wrap used was 10m. I had a bit left over, but not much.

I used a an LCR Meter DE-5000 to measure inductance and resistance.
Inductance LS = 731 uH with the spiral wrap. Next stage will be to wrapped with a water resistant tape, readings will change a bit, we will see.
Resistance DCR = 0.6 Ω

Now capacitance hmmm. If your a beginner and looking for help with this measurement, the maths part can be found on page 135 of ITMD 3rd edition.
The diagram on page 134 of the test setup will leave you wondering, missing labeling and explanation of the test process.

Its for finding the self resonant frequency of the coil which is used in a formula to calculate the capacitance.

'Making a Fast Pulse Induction Mono Coi'l by bbsailor has an understandable diagram of SRF testing on page 8.
A 'Surf Detector' video gives a method for testing SRF https://www.youtube.com/watch?v=fD57KGaGZBM
But by far the most comprehensive explanation is gven by Woody.
Metal detector coils testing self resonance on nugget finder coils. https://www.youtube.com/watch?v=DuU_CJbT3EQ

Now I have my large coil values I will make a small test coil with an inductance of about 730 - 750 uH. Might as well have a test coil
as close to what I'm actually going to use.

One final bit in the comments on Woody's video, some one mentions using a 'Nano VNA' to measure SRF. These are very cheap,
any opinions on using one and what else can they do which is relevant to building a detector?

PS The setup for building and using the 36 inch coil can also be used to make and use a 22 inch coil.

[Carl-NC]

One final bit in the comments on Woody's video, some one mentions using a 'Nano VNA' to measure SRF. These are very cheap,
any opinions on using one and what else can they do which is relevant to building a detector?

Not applicable to detectors. A network analyzer is used in high-frequency circuits to measure s-parameters, which characterizes impedance, gain, and signal reflection. You could use it to measure the SRF of a coil but it's very easy to do with just a sig gen, and the network analyzer would have little use beyond that. The last time I used one was 30 years ago to measure the input characteristics of a cell phone chip I designed.

[Infamy]

The last set of images I posted are some how messed up so I'm posting them again.
The paint on the spider web coil support is to protect the plywood base from water splashes. Eventually it will
be sprayed black with a white outer edge, partly stealth partly cooling.





So its just a signal generator Carl!
Can you give a clear explanation of how this is done.
All examples I have found using a signal generator also have a an oscilloscope attached.
I tried using just a signal generator counter but no dice.
On the 'Surf Detector' video using just oscope I could not see a resistor in play, is it necessary?

OK so in the ITMD 3rd edition diagram there needs to be two cables from the signal generator on the left.
The first cable is output and the frequency is changed until the second cable which is input, (cable dc input) shows
an unstable frequency when you press counter.

20.4 MHz is where it collapses on this coil, does that look like its in the ball park?
Now to work out the units in the equation.

[Carl-NC]

Sorry, sig gen + oscope. Per the drawing in Fig 8.3, you use an oscope to measure the coil signal and vary the frequency until you get a peak. You can also reverse the position of the coil and resistor (as in Fig 8.2), then use the oscope to measure the resistor signal and vary the frequency until you find the null trough. I find the peak measurement to be easier.

The Surf video uses coil ringing to measure SRF. For this you need an active PI transmitter with no damping resistor. The coil will ring and you just measure the ringing frequency. This is mentioned in the last paragraph of 8.1.1 and illustrated in Fig 19.11.

I expect 20.4 MHz is way too high. I would expect below 1MHz.

[Infamy]

Well that begs the question what was I measuring.
Based on the coil design I was able to find an online calculator to work out the theoretical SRF figure.

https://www.teslascientific.com/prod...cy-calculator/

SRF = 760 kHz

Decided to get out the oscilloscope and check this, o'dear what I dreaded, 5 years not used.
Power supply caps probably need changing as it did not coming out to play.
That was the Philips PM3055 60MHz full of blue Philips caps, the worst type of cap I came across when repairing CD players.
I have an Hitachi V525 50MHz I picked up not working, I guess now is a good time to fix them both.

Will be a few week before this thread continues as I am unable to find a component list, so can't order caps in advance.

[Infamy]

This my setup for finding the SRF of the coil showing the collapse of the sine wave.

Image 1 is 2.6 MHz Rise
Image 2 is 3.1 MHz Apex SRF
Image 3 is 3.3 MHz Fall



Image 4 is how the cables are connected, the resistor is 100 ohm.
Image 5 is the power supply board of the Oscilloscope I had to repair, full of lovely philips blue low value caps.

The paper in the background of images 1-3 have the six bad blues also replaced 1 of the 2200 uf browns.
Channel 1 is not as it should be out by x10 which is odd, means when probe is set to x10 you don't need to make any
adjustments. Any suggestions as to whats going on?

This site has a set of calculators for coils
http://radiantarc.com/tools/tools_index.htm

This link is to the calculator for capacitance when you have the SRF and Inductance values.
Is usefull as it gives a explanation of the equation and gives units. Why are units important, read about the Mars Climate Orbiter.
Units are missed so often mistakes are going to happen so use them, big problem for the unfamiliar.
.
http://radiantarc.com/tools/calculat...nk_circuit.php

So these are the values for my coil.

Inductance = 731 uH
SRF = 3.1 MHz
Capacitance = 3.61 pF

Woody's video that explains this method for working out SRF has a comment about water in the wires impacting SRF.
Thinking of wrapping the coil in a cut tire inner tube, need to experiment with glue to see if I can make the coil water tight.

Now what?
Are yes back to the Delta Pulse. With the values I have I must be able to work out what damping resistor I need?

[Carl-NC]

Your connections are not correct. This is what you have:



Also, depending on how the sig gen and oscope are designed, they may have an inherently shared ground (dashed line) through the AC power ground.

This is what you want:



Also, use the 10x probe setting to minimize the probe's capacitance on the measurement. If you look at Table 8-1 in ITMD3 you will see the measured SRF for some ~1mH coils. None of them are even close to 3MHz so you should immediately question your result. The best scramble-wound coil (Teflon) is 667kHz which yields a self-C of 60pF. Expect an SRF of 500-1000 kHz and a self-C of 50-150pF.

[Infamy]

If ITMD3 and Woody have measuring SRF wrong is this delibrate or are folks in a Hubble needs glasses loop?
Their measuring setup is wrong.

In the last example by Carl I isolated the function generator and ran it off a battery, it was the only way to
achieve results which were reliable?, meeting the error criteria for x1 and x10 probe measurements 30-100kHz
from page 8 of 'Making a Fast PI Mono Coil' by bbsailor.
& close to the predicted value of 760 kHz from post #80.

What I had on my scope was a sine wave 2 volts peak to peak which drop slowly when SRF is reach?
I assume the point of change is the SRF but this does seem odd to me.
I expected a rise and fall, not just a fall.
This is a bit of a pigs ear as I still don't know if this is correct.

Averaging it out I have 693 kHz off a spread fall of 675-710 kHz.

This probably all irrelvantfor what I'm building but maybe be a usful exercise for future projects.

[bbsailor]

If ITMD3 and Woody have measuring SRF wrong is this delibrate or are folks in a Hubble needs glasses loop?
Their measuring setup is wrong.

In the last example by Carl I isolated the function generator and ran it off a battery, it was the only way to
achieve results which were reliable?, meeting the error criteria for x1 and x10 probe measurements 30-100kHz
from page 8 of 'Making a Fast PI Mono Coil' by bbsailor.
& close to the predicted value of 760 kHz from post #80.

What I had on my scope was a sine wave 2 volts peak to peak which drop slowly when SRF is reach?
I assume the point of change is the SRF but this does seem odd to me.
I expected a rise and fall, not just a fall.
This is a bit of a pigs ear as I still don't know if this is correct.

Averaging it out I have 693 kHz off a spread fall of 675-710 kHz.

This probably all irrelvantfor what I'm building but maybe be a usful exercise for future projects.

[bbsailor]

I used a 1 meg ohm resistor to drive the coil from the signal generator and a 1 pf capacitance, both to isolate the coil from the loading of the oscilloscope and signal generator from loading down the coils resonance. This setup allows the coil resonance peak to be more easily seen to get a better view of the resonant peak frequency. This resonance measurement allows you to calculate the coil capacitance when you build many coils to experiment with coil seen capacitance to make a faster mono coil.

My fast coil document has a picture on the top showing many coils each of which was resonant frequency measured and noted that motivated me to write the article in 2006. There was minimal documentation about this issue at that time.

Those making coils should lookup the Skin Effect for each gauge of wire they use and note the frequency of their detector coil.

Bottom line- reducing coil seen capacitance allows you to make a faster coil that transitions from TX to RX faster thus allowing a higher value of damping resistor. I devised the variable damping resistor jig to find the best damping resistor value
for the coil
being used. I used Scotch24, a wire mesh to shield the coil. This mesh has less surface area and adds less capacitance than a solid shield like foil.

using the setup in my fast coil document allows very accurate coil resonance measurements to be documented.

bbsailor
Joseph J. Rogowski

[Detectorist#1]

For these measurements I use 1Mohm and 4.7pF. With 1pF and a 10:1 probe with 10MoHm and an input capacitance of 18pF, the error is very small. With a capacitance of 1pF, the 50/60Hz electromagnetic field in the house makes a large unpleasant amplitude modulation on the oscilloscope's screen.

[bbsailor]

For these measurements I use 1Mohm and 4.7pF. With 1pF and a 10:1 probe with 10MoHm and an input capacitance of 18pF, the error is very small. With a capacitance of 1pF, the 50/60Hz electromagnetic field in the house makes a large unpleasant amplitude modulation on the oscilloscope's screen.

Every measurement environment is different. You were aware enough of your environment to do the proper thing to get consistently stable measurements.

Thanks for your input on this to get measurement results that can show detection improvements.

bbsailor
Joseph J. Rogowski

[Infamy]

Thanks for the comments, I was aware of the 1Mohm resistor in bbsailor method which I tried after using 100 ohm. It produced a rise and fall and a reading I was not expecting so given the problems I had, I looked for another way to measure SRF. Its very simple as can be seen in image 1. bbsailor method with x10 probe no cap. gave 415 kHz. Second method using x10 probe gave 420 kHz, close enough. So now finally the true coil figures?

SRF = 420 kHz
Capacitance = 196.44 pF
Inductance = 731 uH
Resistance DCR = 0.6 Ω

[Carl-NC]

If ITMD3 and Woody have measuring SRF wrong is this delibrate or are folks in a Hubble needs glasses loop?
Their measuring setup is wrong.

The set-up in Fig 8.3 is identical to that in post #82 and also the same as what Joe used, except that Joe adds a 1pF cap to the probe to minimize probe loading.

What I had on my scope was a sine wave 2 volts peak to peak which drop slowly when SRF is reach?
I assume the point of change is the SRF but this does seem odd to me.
I expected a rise and fall, not just a fall.

Yes, it should rise, reach a peak, and then fall. How fast this happens depends on the resistor value, which sets the "Q" of the coil. 100 ohms produces a very low Q so that the amplitude varies slightly with frequency, 1M produces a high Q so the amplitude varies quickly. A high Q makes it easier to find the peak frequency.

I suggest you keep on this until you really know what is going on. Try different resistor values, plot the amplitude vs frequency on graph paper. Try the probe in 1x and 10x, try adding a series 1p (or 10p) cap, see what happens. Debugging PI circuits only gets harder, and it sounds like you need more practice with the oscope.

[Infamy]

I suggest you keep on this until you really know what is going on.

I've always known whats going on!

Thicker wire gives a lower SRF.
Thicker wire gives a higher Q value.
Thicker wire gives lower RESISTANCE.
1.5mm litz is about AWG 15 SRF 420 kHz appox.
0.5mm litz is about AWG 24 SRF 550kHz approx.
Looks about right?

Debugging PI circuits only gets harder, and it sounds like you need more practice with the oscope.

Odd how you fail to mention Woody's deliberate XXXX up.
https://www.youtube.com/watch?v=DuU_CJbT3EQ

Hard to over look the obvious
https://www.youtube.com/watch?v=c1Omuu78CyA

100 ohms produces a very low Q so that the amplitude varies slightly with frequency, 1M produces a high Q so the amplitude varies quickly. A high Q makes it easier to find the peak frequenc

Looks like upside down world to me.