CTX3030 noise cancel vs detection

[bklein]

I noticed in air testing (Beach mode) of a bobbypin that the ability to detect it linearly follows the noise cancel setting: setting "1" can't hear it but as you set towards 11 it detects better and better. I didn't try other types of targets - I was wondering why some days I'd find them and others not. it is recommended by Andy Sabisch to noise cancel periodically as you move along. Now I am wondering about NC and its effect on detection. When you set it you arent necessarily maximizing the ability to detect a target(especially if you want the damn bobbypins) - you are minimizing whatever EMI there may be nearby (or put more broadly- affecting tx or rx frequency range). This could be a flawed concept as target detection should be the priority. If you are in an area with little EMI would the task of minimizing it affect target detection?
What do you think? I really wonder now if "11" is wide open detection and as you reduce from there you may risk detection of good targets - or perhaps just ferrous ones?. Would all EMI be audible on the CTX if I were to move from an auto-NC chosen number? When I am hunting and come across an area with EMI sometimes a number other than the Auto number cancels it better. Seems many settings default to a number of 6 or so - never seen it at 11.
I'd like to better understand what is tweaked here...

[Sean_Goddard]

IMHO Andy has this totally wrong. If you look at the way NC works it sets the FUNDAMENTAL of the FBS so everything will be a harmonic or sub harmonic. NC should ONLY be used when another FBS machine is in the area. Noise cancel has very little to do with reducing EMI from other sources, and where ever possible should be set to 11.

[bklein]

I tend to agree but need to prove it. I don't really understand what NC does - whether it happens on the TX or RX side. Interesting this is more important than ground balance....

[Carl-NC]

CTX (like all BBS/FBS designs) transmits a single cycle 3.125kHz wave followed by 8 cycles at 25kHz, a 1:8 frequency ratio. Noise cancel settings slightly alter these up or down a few percent (don't recall how much) but the frequency ratio remains 1:8. The receiver is synchronized to the transmitter so it follows. Usually the total range of NC is a few percent so you wouldn't notice a performance difference.

[bklein]

Carl, how does moving up and down a few percent help eliminate EMI?

[Carl-NC]

On most detectors the RX baseband bandwidth is only 10-20 Hz, so if there is in-band EMI you only need to move it a small amount to get it out-of-band.

[bklein]

What signifies to the code that the signal is emi and not a target?

[Funfinder]

https://en.wikipedia.org/wiki/Attenuate

[bklein]

I don't see an answer in there... My question centers around the detection of the bobbypin in my test. It was only detectable under higher settings of NC. I didn't notice any EMI interference at any NC setting. So did NC hurt detection and we have to think about it even though we don't have any apparent EMI present? Is the NC code "looking" at the RX signal in a broad sense and shifting variables to minimize amplitude then moving to a detect mode on the result? If you don't have EMI then it would pick the lowest amplitude target response - not what we want. Perhaps also it has to do with the target material or size. I only tried the bobbypin as I was wondering why some days I detect them and others not. I haven't had anyone come back with their personal results - just opinions like "you're nuts - the Minelab engineers know what they are doing and it doesn't work that way."

[Skippy]

You must be aware that different targets are more easily detected with different frequencies? Eg. big silver coins like U.S Dollars, half-dollars would be ideally picked out with a machine running at 2 -> 5 KHz, whereas small cupro-nickel coins (U.S 5 cent, British 5 pence) would favour 15KHz -> 20KHz, and other items may favour 40+KHz (tiny gold nuggets, gold micro-jewellery, certain ancient coins).
So choosing a NC channel that has the highest freq may just give a slightly improved response to tiny 'difficult' targets.
For Info: On the E-Trac and CTX, NC = 1 is the lowest freq, NC = 11 is the highest. The difference is about 23%, ie. centre freq (NC=6) can be adjusted +/- 11% approx.
Of course you have to consider other factors. The machine may be somewhat optimised for operation at the central NC=6 frequency, and operating at either NC1 or NC11 may be less than optimal. Machines may vary slightly from each other (component tolerances etc) so one may work well at NC11, the other may be a bit worse.
And regarding "What signifies to the code that the signal is emi and not a target?"
You perform 'Noise Cancel' when there is NO target signal, eg. coil held in the air away from the ground, away from nearby vehicles, spades, fences etc. So it's only EMI and self-generated noise it sees.

[Koala]

Reliably testing a metal detector is not easy. Small difference in ground balance can affect not only the depth but also discrimination. Adjusting one setting often have a knock on effect. Improvements on one type of target can have a detrimental effect on another.

Some steels can be tricky to detect. Have you tried in all metal mode and then try adjusting NC again.

[Sean_Goddard]

Out the other day on a beach with 10 or so ships anchored offshore loads of noise on the etrac on noise cancel auto (ship radars I presume) changed channel to 11, all nice and quiet and loads more finds. Go figure

[Carl-NC]

What signifies to the code that the signal is emi and not a target?

As I recall, when you do a noise cancel you hold the coil still and the machine cycles through all 11 channels. I assume for each channel it looks at a bunch of samples and sees how noisy the signal is. Then it selects the quietest channel. If the coil is not in motion then the signal won't have target info. If you move the coil during NC then all bets are off.

[bklein]

Skippy-thanks for the confirmation of Minelab NC behavior.
Sorry for my question about EMI vs target. I had some kind of brain fart as it made sense to me when I asked it but sure makes me look foolish now. If I figure out what I really wanted to ask I'll give it another try.
Koala - on the CTX I was running without any target masking and not in Ferrous/Coin which seems to bump ID's around.
Most detectors I've used (desert and beach) like to see small iron/steel rods regardless of their operational frequency. I find it interesting that the NC can mask bobbypins though. I need to try things like nuggets, chains, and earrings... Coins, rings, and bottle caps seem impervious. I haven't messed with Ground Balance much though. The manual says to leave it on Auto for the beach so I just do...
I have very little park/inland experience with the CTX. A whole nuther world.

[Funfinder]

> The manual says to leave it on Auto for the beach
This can't be cause beach (especially if soaked with salt water) can be the worst "high mineralized" situation for a MD ever!

The next and even more important thing is that you must not cancel the noise completly because
still good and valuable find-signals "modulate the noise"!
It may disturb you but if you walk around with a completly quiet detector you will miss alot stuff.

Of course I hope your coil fits good enough to the objects you wanna find.

This by itself already noise creating "frequency-cycling" only is of value if its good enough in resonance with the coil
and if the frequencies have a great band-width. Something from 500 Hz to 100 kHz would be useful
and not form 5 to 15 kHz.

Of course it also counts how good matches the RX receiver the frequency itself -
per instance with usuals walkie-talkie radio the noise starts already if the receiver is
just 5kHz away from the real frequency.
For MDs this means that they RX has to listen around 0,5 Hz exactly to the TX or the noise will start,
especially if there is mineralized ground or all kind of outside EM-Field disturbances.
The stability of the used circuit and its parts of course also counts.