Realization of 22Hz transmitter by using magnetic coupling

[Mohammad_Pirzadi]

Some of us are getting just a little bit tired of your continual DEMANDS.

yes, I know that you have tired by me. excuse me!
but it's vital for me.

- - - Updated - - -

That's a great example of lateral thinking! It would certainly allow a small battery powered motor to spin the magnet and have a reasonable life expectancy. I'm thinking the magnet would have to be aligned so maximum N and S fields radiated at 90 degrees to the pipe. I'm no expert in 'power' magnetism but I wonder if the pipe might form a shorted magnetic loop around the field and counter-drive the motor, making it harder to rotate. There is also the issue of the body of the Pig counter-rotating as well but there might be a solution in using two magnets, presumably spaced as far apart as possible but rotating in the opposite directions. Please don't ask me to calculate the resulting field pattern, it would be a nightmare to work out.

Brian.

hi Mr. Brian. excuse me for persistent question. but if you want the fact, I am under excessive pressure by my boss. he gave me an ultimatum to January 1 for realizing this project. I am so tired by my PhD thesis and this project disturbed my life. I earned very useful information by you but help me to finalize this project.
what is your final&practical approach? (economic design should be cared).
to build an in-pipe transmitter, should we use 22Hz frequency or we can use higher frequencies?
thanks to all who helped me and excuse me for continuous questioning. best regards
Mohammad

 

[Warpspeed]

Many sensitive devices sample low amplitude dc signals at 33.3 Hz because most of the noise picked up will originate from 50Hz or 60Hz mains frequencies and the resultant harmonics and sub harmonics.

By sampling at a non sub harmonically related frequency, any received noise becomes non synchronous and can more readily be rejected by the receiver.

I can see a case for using 22Hz or even 33Hz for any system where extreme sensitivity and maximal rejection of mains interference is desired. Frequencies such as 20Hz, 25Hz, 30Hz would be far less desirable.

All this has nothing to do with pipes or attenuation or soil effects, its about very carefully selecting an operating frequency that should potentially provide the most noise resistant receiving system.

 

[FvM]

The attenuation by steel pipe tubes as well as the field propagation along the tube are strongly affected by the operation frequency of a magnetic pigging system.

Á newly designed system should be able to work over the commonly used 10 to 30 Hz frequency rage. For possible avoidance frequencies, electric traction using 16.7 Hz (Europe) and 25 Hz (parts of US) should be considered.

 

[Warpspeed]

Á newly designed system should be able to work over the commonly used 10 to 30 Hz frequency rage.

Yes, but some spot frequencies in that range will be less desirable than others.

A couple of AA batteries in a thick steel pipe might have a tough time competing with these:

 

[kthackst]

Jumping in a little late, I skipped a the second and third page, but in case the idea of a WPT system is still being tossed around, I have a good amount of experience designing WPT inductive systems, and I agree with the consensus here that even optimally matched magnetic resonant coils are going to have virtually zero coupling with one another if there is a steel pipe in the way. Even if there was coupling, another consideration is that inductive systems are inherently directional. If the coil is not facing the right direction in the pipe, no coupling, period.

Its a pretty cool idea but I'd be surprised if you could make it work. The first way to test this would be to specify the dimensions of the pipe and perform a simulation or estimation to determine the magnetic coupling coefficient one can expect between a coil in the pipe and and a coil above the soil. After that, with some estimates on the Q factors of your coils, you can come up with an upper bound on the S21 between the two coils (Witricity has a white paper with the right equation in it somewhere). If that corresponding signal is strong enough to detect, there might be hope, but I suspect the simulated k will be < 0.001.

 

[Warpspeed]

There are three huge hurdles to overcome, the first is the mild steel pipe which typically might have a wall thickness of 25mm. Achieving significant magnetic penetration of that will be difficult.

Available transmission power will probably be in the tens of milliwatts region, if a couple of AA batteries are going to last several days. Some of these pipelines are hundreds of Km long.

The potentially high magnetic noise level due to mains power transmission, and phantom earth currents travelling through the pipeline. That is going to set some practical limit to receiver sensitivity.

 

[Mohammad_Pirzadi]

Jumping in a little late, I skipped a the second and third page, but in case the idea of a WPT system is still being tossed around, I have a good amount of experience designing WPT inductive systems, and I agree with the consensus here that even optimally matched magnetic resonant coils are going to have virtually zero coupling with one another if there is a steel pipe in the way. Even if there was coupling, another consideration is that inductive systems are inherently directional. If the coil is not facing the right direction in the pipe, no coupling, period.

Its a pretty cool idea but I'd be surprised if you could make it work. The first way to test this would be to specify the dimensions of the pipe and perform a simulation or estimation to determine the magnetic coupling coefficient one can expect between a coil in the pipe and and a coil above the soil. After that, with some estimates on the Q factors of your coils, you can come up with an upper bound on the S21 between the two coils (Witricity has a white paper with the right equation in it somewhere). If that corresponding signal is strong enough to detect, there might be hope, but I suspect the simulated k will be < 0.001.

thanks for your comments. do you have any practical idea???

- - - Updated - - -

There are three huge hurdles to overcome, the first is the mild steel pipe which typically might have a wall thickness of 25mm. Achieving significant magnetic penetration of that will be difficult.

Available transmission power will probably be in the tens of milliwatts region, if a couple of AA batteries are going to last several days. Some of these pipelines are hundreds of Km long.

The potentially high magnetic noise level due to mains power transmission, and phantom earth currents travelling through the pipeline. That is going to set some practical limit to receiver sensitivity.

thanks. what is your proposal for finalizing this nightmare project?

 

[mtwieg]

hi Mr. Brian. excuse me for persistent question. but if you want the fact, I am under excessive pressure by my boss. he gave me an ultimatum to January 1 for realizing this project. I am so tired by my PhD thesis and this project disturbed my life. I earned very useful information by you but help me to finalize this project.

Well then you had best start building prototypes now, as I don't think anyone here is going to ship you one or post exact schematics.

 

[CataM]

what is your proposal for finalizing this nightmare project?

There are several proposals in this thread, why don't you try this one?

I quote:

mtwieg

Instead the receiver should have resonant filter > low noise amplifier > antialiasing filter > ADC.

FvM (referring to transmitter)

A class-D amplifier driven by a precise analog or preferably digital oscillator.

 

[betwixt]

Well then you had best start building prototypes now, as I don't think anyone here is going to ship you one or post exact schematics.

That about sums it up.

Dr_Mohammad, you are asking for the almost impossible and in a highly specialized field. I think we have gone through all the possible solutions and it is now up to you to see if they can be implemented. We are not being obstructive, we will help in any way we can but there comes a point where the only way to evaluate the ideas is to build some hardware and conduct experiments. I hope you appreciate that is something we do not have the time or money to do for you.

The easiest idea to construct is the spinning bar magnet, all you need is a small motor, a battery and the magnet. 1320 RPM is roughly the speed of a small DC motor from a scrapped VCR or cassette player and even if you cant salvage one they cost very little. The exact speed doesn't matter to prove the principle. Try building it and see if you can induce a voltage in a pick-up coil nearby. Then place it inside a metal pipe and observe the change in output and them move 4m away from it and check again. With some figures from experiments we can assess how much amplification you will need.

Brian.

 

[Warpspeed]

The above is exactly how I would begin with a spinning magnet at a distance, first fully exposed, then within a pipe.

The next step would probably be the detecting loop antenna. I cannot say exactly why, but experience leads me to suggest that a reasonably large screened multi turn open loop might be the place to begin. Something like this perhaps, its not perfect, but just illustrates the general idea.



That will very likely beat a really physically small coil with a flux concentrator at such a low frequency. Aperture is what you need for receiving low level "anything", and there is no real substitute for physical size.

My own thoughts lead me in the direction of suggesting a open loop antenna connected to a resonant transformer with a high voltage step up ratio, tuned with a capacitor.The Q of the tuned circuit could be further increased with some positive feedback, to just below the threshold of oscillation.

Its much easier to get a very high inductance with a closed magnetic circuit such as a transformer, but that will pick up nothing externally. So you probably need the open loop pickup, connected to a very high inductance transformer. I might suggest a silicon iron tape wound toroid for that.

Resonant tuning over some frequency range will be very necessary, and the most practical way to do that would be with a capacitance multiplier circuit that can readily be potentiometer tuned over a useful range.

I could be completely on the wrong track about all this, but that is at least how I would begin. And then gradually try to improve performance by trying various alternative ideas.

 

[c_mitra]

I can see a case for using 22Hz or even 33Hz for any system where extreme sensitivity and maximal rejection of mains interference is desired. Frequencies such as 20Hz, 25Hz, 30Hz would be far less desirable...

I would also think 31 or 37 to be good candidates; is 31 too close to 30? or 37 to 60?

I do not know why 22 Hz was chosen in the first place but 23 Hz would have been better; it (23) is a prime!!

 

[BradtheRad]

Eddy currents were brought up (post #80), caused by moving a magnet near metal. The effect is that greater physical force is necessary to maintain magnet speed.

If the pipe is ferrous (steel, iron) the magnet will want to stick to the pipe. It might help to have 2 or more magnets spinning equidistant from each other, so that forces are balanced, and centered.

Is it known whether the pipe is ferrous? This brings to mind the fascinating trick of dropping a magnet down a copper pipe. (The pipe should be thick-walled, with a diameter just larger than the magnet.) The magnet automatically slows its fall, taking a second or two to travel through the pipe. As soon as it reaches open air, it drops quickly. Eddy currents are the cause.

 

[Warpspeed]

Yes definitely thick wall welded steel pipe.

Typical smallest pipe might be only 50mm diameter with 6mm wall.
Mid range 300mm diameter with 20mm wall.
And much larger still 1000mm+ diameter in some cases.

If we had tens or hundreds of watts to do this it would be nice.
What we actually have is milliwatts.

 

[FvM]

The next step would probably be the detecting loop antenna. I cannot say exactly why, but experience leads me to suggest that a reasonably large screened multi turn open loop might be the place to begin. Something like this perhaps, its not perfect, but just illustrates the general idea.

I expect a low quality factor (not much above or even below unity) for the 22 Hz loop and thus no advantage of a resonant circuit. A low noise preamplifier with electronic band pass filter seems more realistic. Optimal number of turns has to be evaluated.

Passive resonant circuits tend to need inconveniently large capacitors, as seen with some quantitative examples before. Their possible advantage is the elimination of reactive currents for the transmitter coil and a possible increase of receiver input voltage relative to amplifier noise. It does not reduce the contribution of coil resistance to total receiver noise.

A good preamplifier and noise-matched number of coil turns in a non-resonant input circuit is my first approach.

 

[betwixt]

I agree a loop is a good choice to start evaluation but if one is available, a lot of turns along a ferrite rod might also be worth trying. As mentioned, resonance would be nice but probably not achievable but the ferrite would increase the inductance without losing directivity. I'm thinking of the 60KHz/77KHz tuned circuits used in radio controlled clocks which use a short ferrite in a resonant circuit. I appreaciate there is a huge difference between 22Hz and 60KHz but there may be some merit in trying it.

Brian.

 

[FvM]

Yes, a ferrite rod with a respectively smaller coil achieves a larger reactance to winding resistance ratio so that a certain quality factor can be achieved at 22 Hz and a resonant circuit is feasible. I didn't yet try to figure out which configuration has better signal-to-noise ratio for a given field strength.

 

[mtwieg]

A well-made multilayer solenoid could probably get Q>1 at 22Hz, but would probably be limited to Q<5 (for something handheld, anyways). Starting with non-resonance circuits is probably best for a prototype.

 

[Mohammad_Pirzadi]

thanks to all for their valuable feedbacks. the proposed approaches by you dears pushed me into a meandering labyrinth and now I am completely baffled, because no schematic has presented for a better decision. but the most important note that I have noticed is that utilizing resonance inductive coupling method is almost useless due to absorption by a steel pipe. consider the following text:

1- please aware me if you think using the resonance inductive coupling is still a useful method.

2- since I have no previous experience in design of transceiver systems, plz help me step by step (accept my apologizes).

3-firstly I picked 22Hz frequency because some commercial PIG locator systems use that for passing steel pipe and soil. but which frequency is the best?

4-what is the best method for realizing this nightmare project?

best regards
Mohammad

 

[Warpspeed]

Resonance coupling is not really possible at 22Hz because in order to transfer power the magnetic circuit needs to be open (so it can radiate and receive).

Any inductor which has the necessary large external air path cannot be made an efficient enough resonator. So forget resonant coupling at 22Hz its just not possible.

However a large open multiturn non resonant coil can be coupled into a reasonably efficient tuned circuit, without the pickup coil itself needing to be made self resonant.

Its done all the time in radio receivers, where any old crappy non resonant antenna is fed into a very carefully tuned resonant circuit ahead of the first gain stage.

Thinking about this, your pickup loop is going to be alive with mains induced hum, especially within a building.
Use of a screened pickup loop should help, as your transmitter is not going to radiate any electric field, only a magnetic field.
So screen out the electric interference with a Faraday screen around the pickup loop.
If done correctly, your loop should then only be sensitive to magnetic fields.

But I expect the 50/60Hz hum pickup will still be horrible. Probably orders of magnitude above your wanted signal. This is where a frequency selective tuned circuit can really help if placed ahead of the first gain stage.

Building a tuned circuit of reasonably high Q on a tape wound steel toriod that has very high permiability and low loss at 22Hz should be fairly straightforward, and it need not be very large.

Ferrite is not really suitable because of its low permiability, especially with a large external air gap. If you wish to try a flux concentrator, a block of iron laminations should be more effective than a ferrite rod of equal size at these low frequencies.

Ferrite rods are much more effective at radio frequencies than at power line frequencies and below.