PI Max Frequency?

[chudster]

Dear Forum,

Looking at PI detectors, they seem to all max out at a pulse frequency of about 1KHz- presumably due to technical limits. If a TX pulse time is 100uS and allowing 100uS for sampling and processing the pulse, then 5KHz would seem to be possible in theory.

1KHz is pretty fast already. Would there be any advantage in a 2-5KHz PI?

Is it technically possible?

Chudster

[6666]

I have a whites PI2000 , the clock runs at 30kHz, the tx pps is 3,000, the tx pulse width is 30uS, the sample pulse width is 30uS.
http://www.geotech1.com/forums/attac...1&d=1215509798

[chudster]

I have a whites PI2000 , the clock runs at 30kHz, the tx pps is 3,000, the tx pulse width is 30uS, the sample pulse width is 30uS.
http://www.geotech1.com/forums/attac...1&d=1215509798

OK. I see it is 9V operation with a 5.6 ohm resistor so this is probably about 1.5A current max.

I am wondering if the limit is power dissipation and battery. I can see you would get away with 3KHz at 1.5A current with a 30uS pulse but what about a hefty 3A with a 100uS. Maybe just a bigger battery is needed?

Chudster

[Carl-NC]

Most monopolar PI designs us a Very Late Sample for Earth field cancellation, which limits the max frequency. If you use bipolar pulsing then the late sample isn't needed and you can run much faster. Also, 100us pulse width isn't necessary, in fact at high pulse rates you'll probably want a smaller PW to keep power consumption reasonable. Eric Foster runs the GoldQuest at 10kHz with around a 20-30us PW (don't recall exactly).

[chudster]

Most monopolar PI designs us a Very Late Sample for Earth field cancellation, which limits the max frequency. If you use bipolar pulsing then the late sample isn't needed and you can run much faster. Also, 100us pulse width isn't necessary, in fact at high pulse rates you'll probably want a smaller PW to keep power consumption reasonable. Eric Foster runs the GoldQuest at 10kHz with around a 20-30us PW (don't recall exactly).

Thanks Carl. Your answer seems on the money. That pretty much describes the trade-offs.

Chudster

[Tinkerer]

Thanks Carl. Your answer seems on the money. That pretty much describes the trade-offs.

Chudster

Why run at high pulse repetition rate?

The S/N improves very much if you can integrate 100 samples instead of 10 samples. However, for large, deep targets we would like about 1 target TC between the early and late samples, so, with a target like a silver dollar, that has a TC of several hundred us, we want several hundred us between the early and late sample.

Of course, if we look outside the box, we might think of recycling most of the TX power and so reduce the power consumption. Let's say we recycle 90% of the power and multiply the cycles by 10, to keep the power consumption similar?
And then we could think of a way to accelerate the target signal decay, like speeding up the decay time by a factor of 5, then we need only one fifth of the Target TC between the early and late sample for the same signal amplitude.

It all depends if we want to continue with 40 years old technology or if we want to step into innovation.

Tinkerer

[Davor]

speeding up the decay time by a factor of 5

... will definitely ruin every possibility of successful discrimination

[Tinkerer]

... will definitely ruin every possibility of successful discrimination

WRONG

Tinkerer

[Davor]

I'm all ears. I'm thinking of a speeded up target response as a device that adds resistive component to the target, hence all targets with long tau will be packed tightly one next to the other, and all of them will be recognised as "good conductors". OK, that doesn't have to be bad at all.

[Tinkerer]

I'm all ears. I'm thinking of a speeded up target response as a device that adds resistive component to the target, hence all targets with long tau will be packed tightly one next to the other, and all of them will be recognised as "good conductors". OK, that doesn't have to be bad at all.

Ha, you got me there. I don't have all the answers yet.The problem with the decay time of long TC targets is the exponential decay. A deep, long TC target gives little difference in amplitude between a first sample and a last sample. 1TC = about 67%I never looked at this problem, using a log amplifier. Will the log amplifier exasperate the problem or remedy the problem? How can we influence the TC of a distant target? In recent real circuit experiments, I noticed that the TC of targets are reduced by about a factor of 5. However, I did not make a full test target series of tests, to see if all the TC's were reduced by the same factor. If this holds true, it would mean we can increase the PRR by the same factor. For short TC targets it is of little importance, however, for large deep gold nuggets it can make all the difference. Tinkerer

[Carl-NC]

Why run at high pulse repetition rate?

The S/N improves very much if you can integrate 100 samples instead of 10 samples. However, for large, deep targets we would like about 1 target TC between the early and late samples, so, with a target like a silver dollar, that has a TC of several hundred us, we want several hundred us between the early and late sample.

Of course, if we look outside the box, we might think of recycling most of the TX power and so reduce the power consumption. Let's say we recycle 90% of the power and multiply the cycles by 10, to keep the power consumption similar?
And then we could think of a way to accelerate the target signal decay, like speeding up the decay time by a factor of 5, then we need only one fifth of the Target TC between the early and late sample for the same signal amplitude.

It all depends if we want to continue with 40 years old technology or if we want to step into innovation.

Tinkerer

Not everyone is looking for large & deep. The GoldQuest is designed for smaller nuggets. Also, with bipolar pulsing no late sample @ several hundred us is needed.

Target signal decay depends only on the characteristics of the target. You can speed up the coil decay (sans target), but not the target decay.

[Tinkerer]

Not everyone is looking for large & deep. The GoldQuest is designed for smaller nuggets. Also, with bipolar pulsing no late sample @ several hundred us is needed.

Target signal decay depends only on the characteristics of the target. You can speed up the coil decay (sans target), but not the target decay.

I have no doubt that the GoldQuest is an excellent design for small targets. At 100us per cycle, it is fine for a target of the size of a US penny, that has a TC of about 70us. A quarter has a TC of over 100us (out of memory, I hope the TC's are about right). Gold nuggets would be of comparative size, certainly good size nuggets.

So we are back to the same subject: Do we want to be satisfied with the limits of PI technology, or do we want to strive to push the limits further? Go deeper, go more sensitive, improve discrimination, improve anything and everything possible, even if it is just for the satisfaction of knowing.

Bipolar pulsing eliminates the need for an "Earth field sample" but, if the target sample and the baseline sample are close together, any target with a TC significantly longer than the sample interval, will produce a much diminished signal amplitude when we take the difference between the two.

I agree that we can not change the TC of the target. But, yes, we can speed up the target signal decay. I can prove that.

Tinkerer

[Carl-NC]

So we are back to the same subject: Do we want to be satisfied with the limits of PI technology, or do we want to strive to push the limits further? Go deeper, go more sensitive, improve discrimination, improve anything and everything possible, even if it is just for the satisfaction of knowing.

What are the limits of PI technology? I don't know, I've only begun to explore.

Bipolar pulsing eliminates the need for an "Earth field sample" but, if the target sample and the baseline sample are close together, any target with a TC significantly longer than the sample interval, will produce a much diminished signal amplitude when we take the difference between the two.

Then don't take a baseline sample.

I agree that we can not change the TC of the target. But, yes, we can speed up the target signal decay. I can prove that.

Well then, maybe we're talking different "speeds". A target's step response is t*e{-t/tau}; the decay is fundamentally tied to the tau. I'll have to ask for more details, which you may not be ready to divulge at this time.

- Carl

[Orbit]

Detector is given whites tdi pro, bipolar pulsing or Mono pulsing ?

[Orbit]

After the course tdi whites monopulsing !