C.scope SC4PI Mod

[geroms]

Hello,

I create a new post after exchanges from here :
http://www.geotech1.com/forums/showt...194#post165194

I want to increase the sensitivity of CS4PI to detect small gold nuggets in highly mineralized ground.

A coil smaller than 6.5 inches (154mm) with 0.5 mm wire gives better results than the original coil.
j'ai augmenté

I increased the original 4 Mz quartz frequency with a 6 Mz, with the original coil the results are very good in the air but very disappointing on the ground, we must excessive reduce the power to find correct ground effects management .
What's happening, why this problem with the original coil? Is it an capacitor effect?
So I will make a faster coil.

To understand this detector, I drew the 4pi schematic, there may be some minor errors, it remains to put the components value, but you can see the principle of operation.

[WM6]

Hello,

To understand this detector, I drew the 4pi schematic, there may be some minor errors, it remains to put the components value, but you can see the principle of operation.

Eagerly awaiting for your schematic.

[geroms]

SC4PI Schematic

[geroms]

pdf file

CS4PI Schema 29 01 2013.pdf
here pdf file

[WM6]

Great work, geroms. Thanks.
I have one CS4PI in past.
Hard to get stable working frequency but excellent depth.

[geroms]

The diode D13 is not installed on my cscope.
In some pictures I saw that this diode was added, can you tell me if you have this diode or not and if your CS4PI stable?

[WM6]

The diode D13 is not installed on my cscope.
In some pictures I saw that this diode was added, can you tell me if you have this diode or not and if your CS4PI stable?

Sorry, I sell my CS4PI couple months ago. I didn't open it. So I cannot say D13 was there or not.

But my CS4PI was pretty unstable on terrain (phantom chirps and frequency "drift" quick after tuning).

[porkluvr]

geroms,
I think you need to have a faster settling coil. From what you wrote earlier, your new coil would measure about 1mH. What was the original coil inductance?

A lower inductance would help to obtain a faster settling time but at the same time, you need to make sure that impedance does not drop to a level that is dangerous to your FET (you knew that already). You have a series R, which when combined with the coil R, should be made equal or maybe a little higher, than in the original detector.

I think Eric Foster migrated to using PVC insulated tin plated 10 strands of 0.1mm (which is equavalent to ~30SWG, or between 28.5AWG ~ 28AWG). Not all PVC is the same. You want the lowest dielectric constant (low dc) insulation you can find. Unfortunately, dc is not one of the most commonly published specifications when you are looking at wire.

There are probably some better insulation materials but my mind is a big blank right now. More later.

[porkluvr]

For a 300uH coil on your 156mm form, you would use about 26t of #28AWG. Series resistance will be only about 1.7r. Add resistance to the series resistor R40 to make up for that. Be sure and use some decent hookup wire.

You might think that increasing the tx PRF (pulse repetition frequency) would reduce pulsewidth (PW) proportionally thus maintainining the same average current through the coil, but that ain't necessarily so. Sure, as F goes up PW goes down... but the original circuit's coil series resistance combined with R40 serve to flat-top the charging current, or very nearly so. That means that coil charging current may have already reached near-peak several microseconds before switch cut-off. Shortening the PW may not appreciably reduce peak current in that case. Just my thoughts, I know there will be some disagreement.

What you might have would be a current pulse that reaches about the same value as before, only doing so more often. That would mean increased average current.

If you then reduce series resistance (changing to a coil with fewer windings) that means that your higher frequency tx current will reach higher peak and average values. Your poor little IRFD110 might decide to pop one day soon.

What can you do about that? If you have increased operating frequency and thereby want to use a lower Z coil, you should limit coil current by slightly increasing R40.

You can monitor coil current indirectly by scoping the voltage waveform across R40. Current then becomes i=v/R40. Look for your peak current to not exceed the original value (that with the original coil). To maintain reliability you should reduce the new peak current value by an amount proportional to the frequency change. Or, you might instead get by with installing a small adhesive-backed DIP heat sink to dissipate the higher average power but you're rolling dice with that approach.

But, do not let your peak coil charging current get above the value established in the original design.

Whatever you choose, you should keep some extra IRFD110 on hand for repairs, just in case.

I do not have details of what is happening in your machine. I am guessing about what may be happening, with no certainty. I have to assume that the design guidelines for the CS4 would be mostly in line with what I have learned about the CS6, but they are substantially different machines so that might be an invalid assumption on my part.

[geroms]

Thank you for your help. Your analysis have good sense for me, I'll follow your advice and order some IRFD110.

If I understand the battery will drain faster.

The original coil is 15.6 R and R40 39 R, I think first make a new coil with a variable resistor to test.
Do you think it's better to replacing R40 with a Metal Oxide R27 and a resistor with low noise (Metal Film)?

[porkluvr]

Do you think it's better to replacing R40 with a Metal Oxide R27 and a resistor with low noise (Metal Film)?

Could you rephrase that question?

**********************************
I forgot to say, you will also need to try and find another damping resistor to fit whatever coil you make.

Damping resistor value depends a lot on your coil L and R. You know how to use a scope and find the best damping resistor value?

[6666]

Check your connections on U5

[porkluvr]

If I understand the battery will drain faster.

You've got that right (but it does depend on what coil/series resistance combination you use).

Your series limiting resistor r40 might need be rated for 2W! Metal oxide would be a good idea.
I think 1/4 or 1/2W metal film should be OK for the damping and preamp input resistors.

I need to know the ranges of your tx PPS and PW for a simulation on your tx circuit or else I can't do much more for you. It would also be good to have some numbers on the coil you intend to use. I would recommend something in the range of 300uH~500uH as a start.

[porkluvr]

Check your connections on U5

I'm more worried about all those R values that are left "?", but yessir that regulator is reversed.

[Sean_Goddard]

YUP, Porkluvr is spot on! Increasing the XTAL frequency to 6MHz has a two fold effect, yes, sure the Tx frequency goes up and your sample delay drops, but what else is going to drop?? PULSE WIDTH and probably by 25% or MORE and I suspect that the 4PI doesn't "hard drive" the coil anyhow (top out on Imax) in order to save battery power .

Unless you can re-write the code in the PIC (or replace the PIC with your own version) you are going to get nowhere doing what you did as what you gain in one area, you lose in another as all the parameters are inter-dependent.

I know the guys who worked on the 4PI and it is a very good design, but none the less a compromise between price, performance, complexity (as are all mass produced detectors). The 4PI is designed to do a job, and it does - VERY well too. Could be better? YES but you're looking at spending nearly double to gain a few inches depth if you want a PI to experiment with, start with the Hammer Head, lots of things to mess about with on that, but as Carl designed it, it's already going to be very good. Bear in mind people like Aziz (and there ae MANY more on here too) have made some VERY major steps forward in PI performance in the last few months. Something VERY wonderful is happening on this forum -REAL, tangible progress. It's in the public domain and the big boys are worried!

Sorry to rain on your parade, but I'm just trying to save you from wasting your time.