New advanced PI project

[Tinkerer]

This project will discuss a PI project with a high, variable repetition rate, high power, bi-polar, square wave TX.
A high sensitivity, very short delay, fast response RX.
Low power consumption.

[Altra]

Will follow your project with interest. Also I am interested in your thoughts on the benefits of a square wave TX. Thanks

[Tinkerer]

Will follow your project with interest. Also I am interested in your thoughts on the benefits of a square wave TX. Thanks

Thank you for the feedback. I am looking forward to your input and ideas.

I have this half finished coil that gives me the opportunity to show the inside of it. So this is the coil I will use for this project.

TX = outer diameter about 190mm, 3.12 mH, 3.35 Ohm
RX = about 115mm diameter, 325uH, 6.63 Ohm
Coil cable 4x coax,

This will be the coil for the initial tests. Then I want to build an improved coil.
Suggestions are welcome.

[Altra]

Is your coil balanced?

[Tinkerer]

Is your coil balanced?

This coil is not balanced. Just simple TX with 100 turns outside and simple RX with 40 turns inside.
Since the TX is a continuous current square wave, a separate RX coil is needed.

[Tinkerer]

On the picture we see a rather crude coil built. It works. For the next coil I plan to make a serious effort to build the best performing coil.
Suggestions are welcome of how to improve the coil # 9B.
Now let?s look at the coil specifications:
This coil should work at a pulse rate of 20,000 to 5000. That is, from 50us to 200us between each Flyback.
Flyback voltage around 1000V.
One positive pulse, then one negative pulse.
About 300mA continuous coil current, x 100 turns = 30 Ampere-turns.

Any questions? Suggestions? Anybody?

I would much appreciate help for the following:

Pointing out any mistakes in the coil build and my reasoning.
Translating my plain language into technical language.
Translating the technical language into mathematical language.

[eclipse]

If square wave pulse induction is used /Deemon/ you can have 10kHz @ 200uS or 20kHz @ 100uS, it's bipolar. See Altra's topic for the schematic if interested. Drawbacks /at least in my book/ somewhat less sensitivity with a regular DD coil /Gpx/Sd/ type.

[Altra]

It's hard to evaluate your coil until you hook it up and see the results. Some people may be interested the SRF of the coil. It looks well constructed.

Concerning high pps, I ran into a problem trying to ground balance. What I found in my case was to get enough ground signal to subtract from the target signal, the tx pulses needed to be greater than 90us. Even widening the the sample and gain in the gb channel didn't work. This also increased the noise in the overall signal. I would like to hear anyone's thoughts on this phenomenon.

One solution to the above is use a train of narrow tx pulses for the target balanced against some wide pulses for the gb. ala sd2000.

[Tinkerer]

eclipse, thank you for the feedback.
Yes indeed, the original idea square current wave came from Deemon.
Altra's and King's designs are quite different from mine.
Recently Moodz also posted a square wave TX design. Different too.

My TX design has been running for a few years. As I am 79 years old now I decided to post it now, before it is too late.

I plan to explain in detail how it works so that everybody can use it.

[Tinkerer]

Altra, thank you for the feedback.

I have not looked at your design in great detail, but it seems to be quite different from mine. It will be interesting to compare step by step.
As you have probably noticed, the first big difference from all other designs (that I know) is the high inductance due to the many turns of the TX coil. With about 3120uH, roughly 10 times more inductance than a traditional PI TX coil. The reasons for this I will explain in detail.

[Altra]

I should have said that the problem with GB was with a normal BP PI and not the constant current square wave. I ran that around 1200 pps.

Your high inductance coil may work out well since the current is stored there between pole reversals.

[Tinkerer]

I should have said that the problem with GB was with a normal BP PI and not the constant current square wave. I ran that around 1200 pps.

Your high inductance coil may work out well since the current is stored there between pole reversals.

Yes, the storing of the energy or Joules or whatever we want to call it, in the TX coil, is the most important lesson I learned from Deemon.
The initial charge of the coil takes some time, but once charged, it can be switched fast.

[eclipse]

Can't wait to see what's this all about you got my interest.

[Tinkerer]

The SFR you ask. To measure the real, in-circuit self resonant frequency, I need to finish the coil, shield it and connect it to the circuit and the switching Mosfets, because between shielding, cable and Mosfets a lot of capacitance is added. As I have built many coils, I know that the SFR is going to be close to 156.25 KHz.
However, I have a similar coil already connected to the circuit, so I can show you the waveform.
At this point I recommend you to download a very useful little calculating program that I use for many calculations for PI design. http://www.miscel.dk/MiscEl/miscel.html
[IMG]file:///C:/Users/pirat/AppData/Local/Temp/msohtmlclip1/01/clip_image002.jpg[/IMG]
Looking at the screenshot of the misc_el, we see that with a TX voltage of 1V, the coil will take about 5.4ms to charge to about 285mA.
The energy in the coil will then be about 126 micro Joules.
Discharging the coil through a 1k damping resister will raise the Flyback to about 1000V and take about 30us.
These numbers are only for the 3120uH inductance. In reality there is also a lot of capacitance, so the times would be greatly extended.
How we are solving this problem I will try to explain in the next post.
I have this manic need to add some philosophy in everything, so here it goes:
?Every problem gives you an opportunity?

[Tinkerer]

Still part of the answer to the SFR question:
Calculating the capacitance with the misc_El calculator for an inductance of about 3120uH and frequency of 156250Hz gives about 330pf.