Whites Surf PI schematic

[kuraja]

ApBerg...

Could you please make one layer PCB?

[nelson]

Hi porkluvr. Yes i think you are right, cause i have checked everyting and is fine.So now i´ll try to play with the damping resistor.
I ´ll post my results.
Thanks and take care.
Nelson

Hello Nelson,

I have a feeling that you should play around with the damping resistor. If you used 390 ohms; that is tailored to the IRF9640. I don't know if it should be larger or smaller but you should not assume that 390 is good for IRF9630. GOOD thing you have a scope!

I think the best value would be higher but I might have it backwards. I don't have a good spice model for IRF9630 to put into my stupid simulator.

update: I found a model and I'll compare IRF9630 and IRF9640 side-by-side. I think best: LOWER damping R (I changed my mind) but I want to run some tests.

[cordius]

I' don't know where is the problem on surfmaster??? it is very very simple to built!!!

[porkluvr]

I' don't know where is the problem on surfmaster??? it is very very simple to built!!!

Is it easy to adjust? I have problem with Surfmaster because I can't get LTspice simulation of the sample timing one-shots (built around 4093 nand gates) to run correctly. I'm not saying there is anything wrong, but it is impossible for me to analyze or confirm proper operation of the timing.

I do not believe that the timing circuits are particularly flexible, but that is only my opinion. If the RC components are not kept within a certain range, then the 4093s may fail to trigger.

I do not want to badmouth Surfmaster. It has a very low parts count for what it does. An alternative might be CD4538 circuits (or HC221 etc.), but when you use CD4538 one-shots; you may get more flexibility and straght-forward setup, but the parts count goes way up.

Surfmaster is certainly functional. It does what it does, but until I can flog it I could not possibly feel safe modifying downstream from the LM555 timer in the Surfmaster circuit. That does not mean that YOU should not do it, but I won't. That's just me.

This morning I have pulled out my instruction book and a copy of PSpice Schematics (which I once loved but now thoroughly despise). I will attempt to setup and play with the Surfmaster timing circuits. It will take some time... let me get going. MAYBE I can get Surfmaster working on my computer.

[cordius]

I never believed in the simulator , the SURFMASTER is to be tested on the ground and not on the PC, look at the scheme, if you want to change the pulse delay put a trimmer in place of R1

[kuraja]

Surf Pi is a nice project. I wonder if somehow you could put an analog meter just for estetic reasons...

[Unregistered]

... if you want to change the pulse delay put a trimmer in place of R1

Ummh, not sure which schematic you're looking at, but I think changing R1 near the LM555-7 won't change delay time very much. It will change pulse frequency.

[cordius]

yes, pulse frequency

[nelson]

HI
Last weekend i have retune my surfmaster and now is working very very stable. But sensitivity is still strong btw 14 and 15 cms for a 24 mm coin, and very weak btw 16 and 18 cms.
So does anyone now how to get this machime more sensitive?
Does the dumping resistor help with sensitivity?
Oh i forgot, i used two coils, one is 360 uH, 3 ohms and the other 407 uF, 3,6 ohms. Both are 21 cms in diameter.
I hope anyone can help me , cause my knolegments of electronics are very poor.
Regards
Nelson

My experience with TIP32 it is very very hot and surfmaster dont' work.
IRF9640 is good for this project.
No good 14-15 cm for 24 mm coins.
my surfmaster detect 20 cent euro at 23 cm in air and my mobile phone at 37 cm.

[Unregistered]

Use for coil wire 0.50 / set offset null (or in midposition) / adjust delay pot. for best sens.

[porkluvr]

>....snip

So does anyone now how to get this machime more sensitive?
Does the dumping resistor help with sensitivity?
Regards
Nelson

The damping resistor affects sensitivity because too low a value can load down the return signal. Strong returns get weakened and weak returns might get lost entirely.
Too high a value will cause overshoot and ringing. Coil flyback won't die down as quickly as with the correct value. Since you want the coil flyback to be essentially dead before you start sampling, this would mean you must sample later than otherwise.

If you are still using IRF9630 I would suggest that you try again to find IRF9640. The difference is not so great, but the higher IRF9630 drain resistance means less power to the coil. It also means that your coil decay curve will be longer. Both factors are detrimental to sensitivity.

[bbsailor]

[quote=kuraja;85487]ApBerg...Today I think I've learned that different damping resistors affect the decay curve differently depending on the value of coil and cable capacitance. Since this capacitance is an unknown quantity, I am blowing smoke if I try to pin down a "best" damping R value.


[quote]

Yes, this is exactly true. Anything that contributes coil capacitance or capacitance in the TX circuit will lower the coil's self resonant frequency that is stimuated when the back EMF pulse is being damped by the damping resistor (Rd). The value of Rd will be lower with higher TX coil circuit capacitance. This lower value of Rd will change the coil discharge curve, extending sampling at the earliest time, causing lower sensitivity to some metals.

You have a good mental model of what is happening so you can appreciate that the fastest sampling occurs with coils that have low capacitance and TX circuits with low capacitance.

bbsailor

[kuraja]

Can the Surf PI have a flat wound type of coil? And how would it benefit in depth field?

[nelson]

Hi porkluvr and bbsailor, and thanks for such clear explanation about damping resistor.
So first i ll go experimenting with it and after that if nothing happends i ll try to get IRF9640 instead of IRF9630 that i m using now.
Kind regards
Nelson

The damping resistor affects sensitivity because too low a value can load down the return signal. Strong returns get weakened and weak returns might get lost entirely.
Too high a value will cause overshoot and ringing. Coil flyback won't die down as quickly as with the correct value. Since you want the coil flyback to be essentially dead before you start sampling, this would mean you must sample later than otherwise.

If you are still using IRF9630 I would suggest that you try again to find IRF9640. The difference is not so great, but the higher IRF9630 drain resistance means less power to the coil. It also means that your coil decay curve will be longer. Both factors are detrimental to sensitivity.

[porkluvr]

HI
So does anyone now how to get this machime more sensitive?
Regards
Nelson

Here's a trick you could try. Needless to say I have not done it myself because I don't own a Surfmaster. But if I ever get my PI finished it will have this feature. (Maybe if my design weren't so darn complicated... .)

You are probably aware of my ongoing recommendation to not try any extreme modifications without benefit of an oscilloscope to make sure everything is properly broken. (That's an extension of my military training. I loved the military.)

Now that I got that out of the way: take a look at Carl's Hammerhead II.
http://www.geotech1.com/forums/showthread.php?t=14940
Notice the silicon rectifier 1N4004 between the FET and coil. This should knock about two micoseconds off a coil's decay time. It depends on the coil. I would have reservation about using 1N4004 because coil current in the Surfmaster -well- exceeds the 1 amp rating of the 1N4004. Of course it's not continuous current, but I would still use a diode rated for 3amps. Minelabs uses a BYF28-200 in their SD2000 (with IRF640 FETs). That should work here. BYF28-200 is a 200PIV@3A fast recovery diode. There are plenty of others.

Aziz explained that the improvement comes from the fact that interelectrode capacitance of the diode is in series with the FET capacitance. This reduces the capacitance seen by the coil and reduces coil current decay time. You can then use a higher value damping resistor and your receiver will see a stronger signal.

You may also be able to take your primary sample sooner but there are risks involved in changing the timing. Hence the scope recommendation.
The scope would also be useful for finding the best damping resistor.

Here might be only two microseconds improvement, but even though I can't run experiments to prove this - I think that it would mean an improvement in target illumination even if you don't move the timing forward.
If the FET runs much warmer than before you might want to reduce the 3.3k 1% resistor a little.
Please don't blame me if something goes terribly bad, it's not my fault.