Fisher Patent & MF

[Anonymous]

US Patent number 4,868,504 claims a way of returning energy from the collapse of the magnetic field to the power supply. The machine they use this on is not a PI but is an IB or induction balance machine which transmits square waves in place of a sinusoidal signal. Here is the fun part. The claims clearly indicate that this is done by alternate switching of one side of a transmit coil to +V and -V and that the other side of the coil is connected to storage devices (capacitors). This means that all one has to do is to drive both sides of the coil together with opposite voltage signals to break the patent! Think of this as a push pull amplifier. There are DC motor driver H bridges available today from the FET manufacturers. These consist of four FET's which are arranged so as to switch the motors winding +/- or -/+. Substitute the motor winding for a transmit coil and you are Fisher proof. The rest of the patent makes interesting reading as the method used is the same idea as Minelabs Explorer. The Fisher transmits a single square wave at say 5KHz and makes use of the third harmonic at 15KHz. The received signals are seperated by two pairs of synchronous demodulators which demodulate the resistive (R) and reactive (X) components for both frequencies. A ratio of the R and X signals from the 15KHz signal is then subtracted from the 5KHz signal to cancel the ground signal. Minelab use a multi period digital signal to produce more than one fundamental and harmonic signal. My guess is that the Explorer uses three sets of three different fundamental frequencies. The receiver processes the fundamental frequency, the third harmonic, and the fifth harmonic. This gives you three frequencies per fundamental signal. This provides nine frequencies per set. three different sets make this 27. Does this number ring any bells?

 

[Anonymous]

Hi Dave,
The Fisher Impulse is a PI that uses the energy recovery system. It uses alternate positive and negative TX pulses. Draws about 15mA total.
Eric.

 

[Anonymous]

Eric, Unless they have another patent for their energy recovery than US4868504 then you can use an H bridge to do the same thing without fear of infringement. Patents are only as good as their claims. This guy claimed a method of switching one end of a coil to alternate +/- supply rails with the other side of the coil going to capacitors. The H bridge switching is different and is NOT covered by the patent. Go to the IBM patent site and download the patent to read the claims. http://patent.womplex.ibm.com/ The difference is that an H bridge is in reality a push pull pulse, power amplifier. They really messed up by not including it as a method as they are now limited by their claims. Both sides of the coil are driven with an H bridge while only one end is driven in the patent. In my humble opinion, the world is free to use an H bridge as an energy recovery system with a PI or otherwise. Correct me if I have misread it, but I dont think that this is the case. Smaller batteries anyone?

 

[Anonymous]

Hi Dave,
One reservation about energy recovery circuits is whether they will degrade the clean switch off of the current and field that you can achieve with conventional MosFet switches where the energy is dissipated resistively. In the days before MosFets you had to be careful which type of bipolar you used otherwise the back emf popped it on the first pulse. To prevent the back emf rising above the maximum voltage of the device, a diode was connected to the collector which diverted the back emf into a capacitor. Effectively, the diode connected the capacitor across the coil for one quarter cycle in which all the energy was transferred to the capacitor. This worked fine provided you did not want to use short sampling delays. Short at that time being 50uS. Today we sample at 10uS as in the Aquastar, and considerably shorter for special applications. Because the receiver is looking at microvolts of signal, low level spurious signals, particularly from the diode caused lumps and bumps to appear on the decay curve. Maybe a MosFet bridge will be fine; we will just have to try it and see, but I have found that you have to be very careful what you hang on the coil if you want a smooth fast recovery.
Eric.

 

[Anonymous]

Eric, It is great for me to be able to draw on your vast experience. The point about the clean switch off is well taken. I guess that the Fisher PI may not have too short a sample delay. As you say, testing is in order. I saw an H bridge in the Digikey catalog. It was a single chip like device and sported very good transistor matching, super low Rds-on and a very high current ability.
This might be a good candidate to test with. I will try to locate it in the catalog and will forward the info to you. It would be nice to be able to take advantage of the lower current requirement without having to worry about Fisher. I must admit that until recently I thought that the high voltage flyback pulse was the signal used to induce the eddy currents in the target. I had no idea that it did not play a role in the detection process! Most of my efforts have been with sinusoidal machines. I built up a few PI's and played with them but the lack of discrimination made them of little value to me at the time. Did you read my posting on an iron reject idea I had? I was wondering what if anything you thought of it? Dave. * * *