PI Timing (a bit long)

[Anonymous]

Hi Folks
I'm new to this forum, and new to detector design. I have been on the Geotech Tech forum for a couple weeks. I am trying to understand the Pulse timing for PI detectors. From my reading I gather that earlier designs used a quite low Pulse Repetion Frequency on the order of 20 Hz up to 100 Hz. with the sampling pulse occuring somewhere between 20 and 50 microseconds after the TX pulse ended. On newer machines I understand that the PRF may be as high as almost 10KHz.with sampling occuring between 5 and 90 microseconds after the TX pulse ends. I did some quick calculations on the circuit Carl posted and it seems to have the ability to range from 1.6 KHz to 2.5 KHz for the PRF with a Tx on time of from 35 to 55 microseconds. Am I on the right track with these numbers? What is typical for the decay time of the flyback pulse? I assume that it is around 5 to 15 microseconds based on the above mentioned sample pulse times. I would sure appreciate it if someone could elaborate on typical PI timing a bit.
Thanks
Russ

 

[Anonymous]

Hi Russ,
There is quite a variation in PI timing; it not only depends on what you want to achieve in detection sensitivity, but how it is generated in the timing circuits. A typical PI for beach use, where one wants good sensitivity to rings, coins etc, would have a TX pulse width of 50

 

[Anonymous]

Thanks Eric,
I have been trying to put together a general timing diagram for PI units. Nothing written in stone but showing a possible range of timing signals for different objectives. It really helps me in the learning process to be able to visualize the sequence of events. Your response will go a long way in helping me achieve that diagram.
Thanks Again
Russ_NY

 

[Anonymous]

Hi Russ,
Most PI's have a second sample pulse occuring much later than the first; maybe 100uS after the first one. The second signal sample is then subtracted from the first to eliminate low frequency noise. The sources of this noise is firstly the movement of the coil in the earth's magnetic field, and secondly 50/60Hz noise from power lines. The two sample pulse widths must be equal and the individual sample widths are usually the same as the initial sample pulse delay i.e. for a 15uS delay, each sample pulse is 15uS in width. Others are experimenting with much narrower widths and using sample/hold circuits rather than the usual integrator, so you pays your money and takes your choice.
Eric.

 

[Anonymous]

Thanks Eric
This is very interesting. Is the 100uSec between the two sampling pulses sufficient to allow for settling and return to true ambient noise conditions? On the surface it would seem that a sampling just prior the TX pulse would be more representative of ambient noise, particularly at slower TX frequencies. Does this make any sense, or am I splitting hairs?
Thanks again
Russ

 

[Anonymous]

Hi Russ,
For rings, small gold items and many coins the decaying eddy current signal has substantially decayed to zero by the time the second sample is taken. But even if there is a bit of object signal left, it is much lower in amplitude due to the exponential nature of the decay. Sometimes the second sample is just before the start of the next TX pulse - depends on the TX frequency. The dual sampling method is most effective for cancelling low frequency noise (power line and lower) where the noise amplitude is similar for both the first and second samples. When subtracted, the two equal noise amplitudes subtract out, whereas the unequal signal amplitudes don't. Higher frequency noise is reduced by integration and low pass filtering in the receiver after sampling. Sometimes a shift in TX frequency is required to avoid a beat note effect with an interfering signal.
Eric.

 

[Anonymous]

Eric,
I really appreciate you taking the time from what must be a busy schedule to help a newcomer like myself better understand this technology. You have been very helpful. This is a great Forum. You and the other contributors to it are providing an extremely valuable service. I hope that as my experience and understanding grows, I will be able to contribute also.
Thanks Again
Russ