Questions on PI hardware

[massimo]

Hi all
I have two question on PI machine
1)what kind of climping diode is better in frontend pi schema 1n4148, 1n400x,fast or superfast? And there are different performance in base the kind of diode?
2)nedd the PI coil to be shielded like vlf or bfo or is a detriment of performance?
Thanks in advant at all
Max

[Carl]

Re: Questions on PI hardware

I recently did a quick experiment with the 4148, a schottky (don't have the number at my fingertips), and a superfast UF1002. The schottky was actually slower, and the UF no faster than the 4148. The 4148 seems to be what everyone is using.

Yes, you need to shield the coil.

- Carl

[massimo]

Re: Questions on PI hardware

Thanks Carl
Even in my test i dont see difference between 1n 4148 and fast diode.
In your opinion which si the best time set for a PI ?
Time pulse, delay and sample wimdows?
Thanks in advant
Massimo

[Reg]

Re: Questions on PI hardware

Hi Massimo,

The answers as to the best settings for the pulse length, delay, and sample window depend upon several things, but primarily on what you are looking for.

When I was experimenting with the Stuart PI, I found I liked a pulse length of about 100 usec. This seemed to have about the same depth as a longer one, but didn't draw as much current from the batteries. I also found that since the Stuart PI didn't take two samples it would detect hotrocks with a loud signal until I reduce the pulse length to 100 usec. Actually, I think the pulse length was about 85 usec.

Again, the delay determines what objects you can detect. I prefer a delay less than 15 usec so I can detect small gold objects. With a delay of about 30 usec, most small gold items will not be seen.

The sample window can be very short depending upon how the sample is taken and what type of circuit is used. You probably will not be able to get less than 10 usec delay. It takes some doing to reduce it that low.

I set the sample time on the Stuart to be just a few usec. I think it was about 4 usec or 5 usec. When you have a short delay and then you shorten the sample time, you increase the sensitivity even more to gold objects.

The pulse repetition rate again can vay alot depending upon the sample hold circuit you use. I had a sample rate of about 4 msec which seems quite slow but it worked quite well. The sample rate will also determine the about of interferrence you might encounter. Picking the wrong rate can cause increased noise levels, so it is probably best to experiment with this some.

On my Beachscan type PI, I have a minimum pulse delay of 10 usec, which then gives me a pulse length of only 30 usec due to the circuitry of the detector. Even this short of a pulse length seems to work quite well and is quite sensitive. With a delay of 10 usec, the Beachscan sample time is also 10 usec. If the circuitry would allow it, I would prefer a shorter sample time.

I hope this helps a little.

Reg

[massimo]

Re: Questions on PI hardware

Hi Reg
Thanks for your explain, now i am playng whit a circuit very similar at Twin loop(http://www3.telus.net/chemelec/Projects/Metal-1.htm)because i think is thery simple change every paremeter; all times are given from 555.
I'm not sure i can do so little delay and sample whit this circuit bat i will try because now i have a 40 microS of dalay and 55of sample so now i'm very distant from time you said and in actual pic up very well all little iron but it's hard pic up copper, silver and gold.

Thanks a lot for yours input

Massimo

[Reg]

Re: Questions on PI hardware

Hi Massimo,

The website you posted didn't work for me so I couldn't look at the schematic.

Based on the settings you have, I would expect the detector to pick up iron well but you will probably not be able to pick up most typical gold objects.

You are right, it will not do as well on copper and silver due to both the long delay and the long sample time, but I would think it should pick up coin size objects.

I would think it shouldn't be too difficult to reduce the delay and the sample time if they are using 555 timers.

Reg

[massimo]

Re: Questions on PI hardware

Thanks Reg