new MineLab Machine

[Chiv]

So did all multi in all modes , looks like gold uses all frequency's to I was told that it did when I picked up the detector.









Hope this helps

[Skippy]

That worked well. What you did when you chose 680R is you matched your resistor value to the Inductive (and capacitive) impedance at the resonant frequency. So 2 * pi * f_res * L = 680 Ohms. F_res was 450kHz, so I calculate your pickup coil has an inductance of about 240 microHenry. Matching R value to L impedance gives you 'critical damping' , where ringing is minimal.
The reason your 5kHz square waves in particular have drooping flat sections is because a pickup coil is a differential sensor, it cannot respond to static, continuous magnetic fields. So those flat sections would eventually droop to zero.

So .. we've confirmed the 5/10/15/20/40 modes do exactly what they should. By not transmitting the 'unused' frequencies, more energy can be put into the wanted one. This should give better resistance to EMI, and electronic circuit noise will also have less effect.

All the 'multi' modes transmit the same signal, which presumable makes the hardware/software simpler. The performance differences between the modes is all down to what frequencies are processed, and how. You only need two different frequencies to produce an effective multifreq machine, so it's not unreasonable to suggest some modes only 'look at' two frequencies, even if more are available.

Your next challenge is to get the Spectrum Analyser function of your USB scope working. It seems pretty certain the 'multi' waveform contains a frequency of 39kHz, and some lower ones, eg 1/2, or 1/3rd or 1/5th of 39, hard to tell, but your analyser should make easy work of it.

[Chiv]

Will I will have to look at the spectrum side of it and learn what that's about, good bit is, learning something new all the time on this forum. All that calculation stuff dose my head in as I'm dyslexic with number as will read them one way the wright them back wards more often than not, have to do a lot of checking or by the tools to do the job, the inductance meter gave me 241 mh on my coil so given a bit here and there

[green]

That worked well. What you did when you chose 680R is you matched your resistor value to the Inductive (and capacitive) impedance at the resonant frequency. So 2 * pi * f_res * L = 680 Ohms. F_res was 450kHz, so I calculate your pickup coil has an inductance of about 240 microHenry. Matching R value to L impedance gives you 'critical damping' , where ringing is minimal.
The reason your 5kHz square waves in particular have drooping flat sections is because a pickup coil is a differential sensor, it cannot respond to static, continuous magnetic fields. So those flat sections would eventually droop to zero.

So .. we've confirmed the 5/10/15/20/40 modes do exactly what they should. By not transmitting the 'unused' frequencies, more energy can be put into the wanted one. This should give better resistance to EMI, and electronic circuit noise will also have less effect.

All the 'multi' modes transmit the same signal, which presumable makes the hardware/software simpler. The performance differences between the modes is all down to what frequencies are processed, and how. You only need two different frequencies to produce an effective multifreq machine, so it's not unreasonable to suggest some modes only 'look at' two frequencies, even if more are available.

Your next challenge is to get the Spectrum Analyser function of your USB scope working. It seems pretty certain the 'multi' waveform contains a frequency of 39kHz, and some lower ones, eg 1/2, or 1/3rd or 1/5th of 39, hard to tell, but your analyser should make easy work of it.

I've been using pi * f_res* L instead of 2 * pi * f_res* L to calculate R for critical damping. Have I been doing it wrong?

[Skippy]

Carls "Waveform Challenge" thread does raise an interesting possibility:
If the Nox can happily generate any old frequency, from 5K to 40K, and beyond these limits, probably; and any arbitrary mash-up that fits this range, then it's possible they could do software upgrades that give totally different operating frequencies, modes etc.

[Carl-NC]

Green is correct. Critical damping is when R = 0.5*sqrt(L/C). Replacing C with f_res, R = pi*f_res*L.

[Skippy]

Apologies, bad maths by me. Not paying attention enough.
It's critical damping when
2R = 2*pi*f*L = wL.

[Skippy]

Bah, another example of not paying enough attention:
Take a closer look at Chiv's (tiny) screengrabs of Beach Mode, b1,b2 in Post 106. Clearly two frequencies, I think 3 cycles and 9 cycles in 385 usec, giving freqs of 7.8kHz and 23.4kHz.
Worth examining in more detail, perhaps when Chiv's got the Spectrum Analyser 'tuned up'.

[Chiv]

I had ago at spectrum analyser, tried a few different coils and make shift probes, but where I see a sig on the screen its not clean and this lap top or maybe 205b cant handle it but it really laggy on the screen running about frame a second. Next time I'm in town I get some solid coax cable and make a tool like 6666 posted as that seems to me the norm type that you see on the net. and I make pics a bit bigger too

[Skippy]

Your existing pickup loop should be fine. You need to fix it it position when doing a test, as a wobbly signal level is going to cause problems.

And you can low-pass filter the signal from your loop, to reduce noise.
We know that there's no signals of interest over 50kHz, so filtering at this sort of freq should be OK.
Leave the 680R damping resistor fitted across the pickup coil output.
Connect a 2700 Ohm resistor in series with a 1 nF capacitor. Join this pair across the pickup coil output.
Connect the scope input across the 1 nF capacitor.
2k7 & 1n0 give cutoff freq = 59kHz. If you're stuck for parts, 5k6 and 470 pF would work, etc.

And general noise-reduction tips include keeping wires short, twisting together the leads coming out the pickup coil, keeping away from noise sources like PC's, WiFi gear, routers, TV's etc.

[Chiv]

ok I use your little hint guide Skippy, Blue ch 2 was coil from before and yellow is a tool made from the net EQ800 was set on p1 @5khz so attached two pic from each oscilloscope and spectrum, but real means nothing to me hope it in the right direction, emi are very high in my area

[Chiv]

Still trying to nut out this spectrum side of the scope still think its to slow for it, but cover coil and pick up in a lose sheet of ally for bit of a shield and seemed to help as have readings on the scope now. 1 pic park 1, second beach 1. now Am I heading in the right direction or just flogging a dead horse?

[Skippy]

Not heading in the right direction.
I guess I'm going to have to find a manual for your scope online and study it... I see you've now got some extra controls on the right-hand side, to adjust 'centre frequency'. That looks to be where your problem may be. We're interested in frequencies of below 50kHz, specifically we think the highest freq is likely to be 40kHz, and the lowest probably 5kHz. So the 'centre freq' needs adjusting so we see those frequencies on screen, not MHz stuff.
Try 5kHz single freq mode again. Then you would expect the spec analyser to show signals at 5kHz, and weaker ones at 15kHz (the third harmonic) and 25kHz (5th harm). If you can get that to look OK, then the multifreq modes should display OK, too.

[Chiv]

I have had another crack at this,
park
park
park
park
park
park


beach

[Skippy]

That's better, looks like you've tamed the beast now.
Your 'field multi' clearly matches the analysis that we did on the other 'challenge' thread, showing the three strongest frequencies of 7.8kHz, 18.2kHz and 39kHz. And residual bits of (n*2.6)kHz. I'm not sure why the signal level is lower at 39kHz than it is at the other two main freqs. If you look at the raw waveform, it clearly has a lot of 39KHz.

The 'beach multi' is obviously different. Could you re-do the raw square-wave beach-multi screengrab? (last time it was tiny in size)
It looks to have less 39k and more of the lower freqs.