Noise Cancel Question

[Iron Buzz]

I don't understand noise cancellation. Actually, there are two things that I don't understand about it.

First... the manual says that it finds the frequency with the least noise. But how does that work with multi-Q? How can it select one frequency when it runs on multiple frequencies?

Second... if I run noise cancel multiple times without moving, or even moving the coil, it seems to select a different frequency every time. Seems random. How does that work?

 

Amazon Forum Fav 👍

Minelab PRO-FIND 35 Waterproof Pinpointer Metal Detector for Experts with Tone ID & Adjustable Sensitivity - Grab it through Amazon!

[vferrari]

I don't understand noise cancellation. Actually, there are two things that I don't understand about it.

First... the manual says that it finds the frequency with the least noise. But how does that work with multi-Q? How can it select one frequency when it runs on multiple frequencies?

Second... if I run noise cancel multiple times without moving, or even moving the coil, it seems to select a different frequency every time. Seems random. How does that work?

Those are good questions, and I have wondered about the first one myself. I can't fully answer the question, but I can say that Multi-IQ is more than just multifrequency though, it determines the frequency weighting which distinguishes each search profile mode but ML has not seen fit to give away the secret sauce on this is accomplished. The other thing it does is determine how the target return signals are processed. I can't really tell you how noise cancel works but the key to the manual description is that is SHIFTS the frequency channel. In multi IQ I have interpreted this to mean that it has set up 19 discrete frequency shift channels around the base frequencies. For simplicity let's assume the base channel is "0" and that means that each of the 5 frequencies is at 5, 10, 15, 20, and 40 khz. If noise cancel then selects channel "1" as the quiet channel, I am assuming all the base frequencies are just shifted by a small, set amount say 0.1 khz, so channel "1" corresponds to 5.1, 10.1, 15.1 20.1 and 40.1 khz. I am making up the shift number but you get the idea. Similarly, Channel "-2" might be -.2 khz so everything is shifted -.2, or 4.8, 9.8, 14.8, 19.8, 39.8 etc. for each of the 19 "channels" so you are basically getting the same multifrequency performance around the base frequencies but a slight frequency shift is potentially eliminating some interference without detracting from the multi IQ performance. So it is not a single frequency change.

Regarding your second question, if there is basically no EMI, then I can see the auto noise cancel feature just randomly picking any of the channels because they are all about the same or maybe 4 or 5 are good enough, so it may pick any one of those 4 or 5 good channels each time you run it. Same thing could be said for HEAVY EMI , where none of the 19 channels appreciably lowers the EMI. Also, I have no idea how good the auto noise cancel algorithm really is, especially if the EMI intensity and frequency is variable. It may pick what it thinks is the best channel but I have found under power lines that the auto picked channel is not so good and I manually pick another channel that is better, which is a feature I like on the 800 since the 600 only has auto noise cancel.

Finally, since all the base frequencies are multiples 5 khz, you may find a situation where Multi IQ is just not cutting it EMI wise regardless of auto or manual noise cancel and lowering of sensitivity. In that case, your only option may be biting the bullet and going to single frequency and picking the frequency that gives you the least interference. Better than having to dial sensitivity down 10, but you do loose the advantages that Multi IQ brings to the table in terms of salt water balancing (beach mode does not have a single frequency mode), iron bias (does not work in single frequency mode), and all the usual target optimization disadvantages that single frequency brings to the table. HTH

 

[tnsharpshooter]

I don't understand noise cancellation. Actually, there are two things that I don't understand about it.

First... the manual says that it finds the frequency with the least noise. But how does that work with multi-Q? How can it select one frequency when it runs on multiple frequencies?

Second... if I run noise cancel multiple times without moving, or even moving the coil, it seems to select a different frequency every time. Seems random. How does that work?

No need to think much about the how behind.
Just remember for peak perormamce in each detect mode using multi freq, or using single freq ops, do a noise cancel.
A noise cancel done in a site, doesn’t mean Emi hasn’t changed and hence another noise cancel can enhance preformance.
Depth is really the biggie affected, granted even mid depth could be effected with a lot of Emi.

And noise cancel doens’t necessarily mean ALL Emi is mitigated. Just the most at a given time (ideally). It is possible to manually select a channel and detector perform better.
Don’t forget aircraft overhead, this can definitely change Emi levels depending when the aircraft passes overhead.
Things like nav systems p, crypto systems, radar, etc,,all the emitters on board.

 

[Iron Buzz]

No need to think much about the how behind.
Just remember for peak perormamce in each detect mode using multi freq, or using single freq ops, do a noise cancel.
A noise cancel done in a site, doesn’t mean Emi hasn’t changed and hence another noise cancel can enhance preformance.
Depth is really the biggie affected, granted even mid depth could be effected with a lot of Emi.

And noise cancel doens’t necessarily mean ALL Emi is mitigated. Just the most at a given time (ideally). It is possible to manually select a channel and detector perform better.
Don’t forget aircraft overhead, this can definitely change Emi levels depending when the aircraft passes overhead.
Things like nav systems p, crypto systems, radar, etc,,all the emitters on board.

Thanks for the reply, TN, but I always wonder about the why behind things. That's just how I'm made.

I'm not talking about noise mitigation. That really hasn't been a problem for me. And I understand EMI. I'm just trying to understand their noise cancel feature.

 

[cudamark]

If there is no apparent EMI, do you still need to hit noise cancel? I've tried it both ways on the Nox800 and my Etrac and don't see any performance difference.

 

[vferrari]

If there is no noise to cancel out then you can skip the noise cancel step and save the 5 seconds.

Regarding noise versus detector performance with the Equinox, my experience has been that raw performance like depth and response will not suffer from not doing a noise cancel. But one key to getting the Equinox working optimally is to maximize the signal-to-noise ratio. That means more than increasing sensitivity or choosing the optimal search profile corresponding to the targets of interest but also minimizing EMI, if present, through noise cancelling, and doing a ground balance, if needed. Also, avoiding overdriving the machine with too much sensitivity or creating too much ground feedback by lowering recovery speed too far which people sometimes do when they are trying to maximize raw depth. These approaches to maximize depth performance become counterproductive at the point where increased noise starts to overcome the smaller increase in signal.

HTH