Sources of EMI

I have posted this on another forum, but am interested in feedback from readers here. Hope other forum does not mind.

I came across an article in the Minelab web site about how to tell what is causing emi interference, that I found interesting, because there are so many confusing theories on forums about this. I checked the details with an engineer friend who is with me now looking over my shoulder to ensure I don’t write something real dumb, so this is a combined contribution.

Because the “Cancel” setting (on the front panel) of a 4000/4500 subtracts the magnetic signal of the transmit coil from the magnetic signal of the receive coil (using a DD), the net result is a cancellation of magnetic signals from magnetic sources far away, so that the field strengths at the transmit and receive coil are more or less identical.

The following have been stated for causing EMI interference on forums:

Charged dust particles generated by wind.
Sun spots.
Radio stations/TV stations.
Radars.
Mains fields.
Lightning (sferics).

If the coil is very close to mains equipment, even the “Cancel” setting may not work 100% because the fields may not be the same in the transmit and receive coil (because the field is changing with distance from the source). But “Cancel” still should at least reduce the intensity of even this interference, unless it is simply too strong.

Radio waves from radio/TV stations and radars will not be cancelled in the “Cancel” setting because apparently the coils no longer act as magnetic receivers at these frequencies and the “phases” end up being random, or the electronics does not behave properly to allow for cancellation.

Sunspots are radio waves, so the same applies, but apparently these are tiny signals compared to the radio transmission strength anywhere on earth, so these cannot be a source.

Charged particles will strike the coil at random locations, or at least different intensities at different locations, either more near the transmit coil, or more near the receive coil, so this source too will not be cancelled in the “Cancel” setting.

So, if one is not close to mains, and the “Cancel” setting does not get rid of the EMI almost completely, then the source must be some sort of radio waves (radio/TV stations or radar) or dust particles.

My experience is that Cancel does its job 100% of the time (except very close to mains), so wherever I have been, neither radio waves nor changed particles have ever been a source.

But I certainly do suffer from interference when not in “Cancel” on some days in the field!!

This is apparently from lightning signals (sferics) which transmit at all frequencies within the receive frequency range of metal detectors, unfortunately, so changing the detector frequency will make no difference. These signals are transmitted right around the earth, so can be of significant strength even if the lightning is 1000s of km away.

Obviously the sferic field will be uniform at the coil, so the signal from the receive coil will be identical to that of the transmit coil and thus “Cancel” should work more or less perfectly regardless of the orientation of the coil.

However, when NOT in the “Cancel” setting, far from the lightning source, the magnetic field is horizontal, so if the coil is also horizontal, the sferic signal will more or less be cancelled, but not when the coil is tilted away from the horizontal. Some hills/ground conditions may tilt the sferics signal away from the horizontal, so one will need to tilt one’s coil similarly to cancel this signal (again when not in the “Cancel” setting). Nearby lightning signal field are not horizontal, so holding the coil horizontal will not cancel out these, but will in the “Cancel” setting.

On trap are fences. These apparently act as excellent antennas for sferic signals and retransmit their signal vertically! So that if one is near a fence (or any long conductor such as a telephone line), one will detect this amplified sferic signal even when the coil is horizontal. Very annoying!

Hope this is of interest to some as it was to me.

I am interested in the experience of others using "Cancel" in the field; does it work 100% of the time for you when not next to mains electricity or a fence or any other sort of long conductor like a telephone line? I have heard that some non-Minelab DDs do not have the same number of turns in the coil windings of the transmit and receive coils, so these will not work properly in the "Cancel" setting, so when checking, better use a Minelab coil just in case this is true

Excellent information, thanks Peter.

Hi Pete,
Thanks for the info mate, I'll see if I can put some of your suggestions into practice when I head up to NQ later this year. The main problem I come across up there is toward the end of the season when the odd storm is about further North, the interference can drive you batty If simple adjustment of the angle of the coil will help then it's certainly worth a try.
Cheers,
Andy

Thanks Goldfish. But most of it came from my friend reading the Minelab article, not me!

Unfortunately Andy, close lightning (I think "close" means less than several hundred km), produces interference fields with fairly random directions, so a mere tilt in the coil angle may not help too much. The main weapon is Cancel if things get too bad. This is what I learnt from this research; adding various sorts of screens to your coil or control box and so on suggested by others will produce no improvement unfortunately, because sfeics interfere at the same frequencies as metal detectors.

JP has offered good tips about not running gains too high. Very important!

Hi Pete,
Yes, I was referring to distant storms - made all the more frustrating because usually there is no other indication of what might be causing the interference - it's only if it happens to be mentioned on the radio or if you have access to online weather info & can see what's happening 500 or a 1000 K's distant.

Unfortunately cancel extracts too much of a depth penalty for many of the places I go to bother with.

Cheers
Andy

I must say, a brilliant topic thanyou!
Plenty of food for thought and much experimentation to be done.

I haven't used a DD in years but will dig one out for sure!

cheers!

Interesting topic!!

shandeemax wrote:
Unfortunately cancel extracts too much of a depth penalty for many of the places I go to bother with.

I feel a bit the same Andy, Unless near mains power or electric fence etc. I tend to shy from using cancel for that reason. Personally I would prefer to put up with the ferics or just pack it up!

Can anyone give an indication of the depth loss when using 'cancel' or better yet how much depth do we actually get when in that mode, ballpark figures of course, conditions pending.
Cheers

Excellent post Peter. Thank you and your friend.

I have found that I have learnt to live with EMI provided it's not too extreme.
After countless hours detecting with spikes, EMI or interference of some kind, I believe have developed an "ear"
for discerning the signals I'm getting.
I'm not about to say it's foolproof but if you spend enough time detecting under these conditions you can learn to overcome them.Unfortunately the newbie or "once a month" detectorist will need to persevere until they become more adept at it.
Having your threshhold turned down so as you can just barely hear it is, I believe, one of the secrets. You are not going to miss signals even with it turned down.
This is just my findings and they work for me. They may not for you.
Btw, I use a 4500 with a 12" or 16" NF Advantage mono exclusively.

Robert

The subject of whether SETA works came up else where, which I hope may be of interest here too.

I thought that it did not work perfectly until I realized a mistake I was making, which clearly others are making too:

From the original post, it should be clear that if the plane of the coil is vertical, you will normally clearly notice an increase in threshold instability because of sferics. When I originally did the testing, I flipped the plane of the coil from horizontal to vertical, and noticed a signal, but soon realized that this was only because the threshold had become noisy. To check, I then monitored the difference in threshold between the threshold when the coil was horizontal and vertical. Taking this into account, it was clear that this explained the "apparent" movement signal, and that SETA thus seemed to be doing its job well.
I would imagine SETA will not work perfectly if one waves a magnet way stronger than the earth’s field near the coil, for obvious reasons! Of course it will have to be one of those black ferrite magnets, not a metal magnet, but maybe even the the "ferrite" in the ferrite magnet will be detected slightly too, if too close to the coil, because the ferrite may not behave like a "normal" ferrite??
　
P.S. remember that the earth’s field is not parallel to the surface of the earth except near the equator and that it has a more and more vertical component as one moves from the equator to the poles.

I too was scepitacl about the "Cancel," but after measurement, I was not what I assumed.

For large targets just detectable at large depths, the decrease using Cancel is roughly 20% loss of range compared to Normal (using a DD).
This loss steadily reduces with decreasing maximum detection depth until about 80% of the coil diameter at which point the detection ranges of Cancel and Normal are about the same (both using a DD).

For targets with a maximum detection depth less than 80% of the coil diameter, Cancel produces GREATER detection depths than Normal (again both using a DD).

However, Cancel does produce more ground signals typically than Normal using a DD, so this in effect reduces the 80% equal detection range mark to maybe 70% of a coil diameter, but this can be combated using "Enhance," which is better for detecting small/medium sized nuggets than "Normal" timing mode.

In conclusion, one should be able to find more small nuggets using Cancel than Normal using a DD (obviously especially if EMI is really bad!) So maybe it is worth detecting on those really bad days so long as one is aware that large nuggets may be missed. I tell myself that these are rare, so am probably not missing any!

(Obviously using Enhance + a mono or Normal timing setting + DD is better than Cancel for the usual days.)

I was out detecting the other week and came across interference that I could not work out the source, and being in the bush with no power lines or fences close by.
I eventually came across a marker for underground telstra lines.
I wouldn't have thought they would have caused a problem especially a hundred or so meters away and being buried at I assume would be a reasonable depth.
Mark

MS wrote:I was out detecting the other week and came across interference that I could not work out the source, and being in the bush with no power lines or fences close by.
I eventually came across a marker for underground telstra lines.
I wouldn't have thought they would have caused a problem especially a hundred or so meters away and being buried at I assume would be a reasonable depth.
Mark

The telephone lines channel the sferics just like old fence lines do.

JP

Hi Johnathan
Yes your right there, I didn't think that they could effect the detector that much, but now know better.
Cheers Mark

A summary is posted at
http://arizonaoutback.ipbhost.com/index.php?s=89c5360885b2f7427c5c09a483e33099&showtopic=8544&st=140
of a difference between my views and detecting buddys' observations and that of Doug's from GoldProspectingInOz.

Note that if wind is a cause of EMI then whenever it is windy, it will always be a bad emi day, because the cause is wind. Or to put it another way: if it is windy and there is little or no EMI, then wind cannot be a cause of EMI because it cannot be a cause one day and not the next: Nature does not work like that.

Also, that weather systems are huge, typically 1000s km across, so often when it is windly in some places, this is often associated with lightning = sferics which is the true cause of any EMI.

Apparently the older 4500's had worse problems with EMI. The later machines handles it better. Please correct me if I'm wrong but I think due to the limitations of EMR the machines can put out, Minelab have made the recievers so damn sensitive, you can hear someone fart 2000 KM away. For me cancel mode is a complete waste of time and only good for finding your car door in a fog... but only if you accidentally walk into it. If EMI becomes a problem for me, I reach for the old 2200d.
Ismael at Minelab mods claims to improve EMI by replacing a component that is used in the later 4500's.

Cheers

Undertaker wrote:

Apparently the older 4500's had worse problems with EMI. The later machines handles it better.

This is NOT correct, there was a recall but it had nothing to do with EMI.

For me cancel mode is a complete waste of time and only good for finding your car door in a fog... but only if you accidentally walk into it. If EMI becomes a problem for me, I reach for the old 2200d.

Sounds like you have been trying to use the 4500 with a Monoloop coil and Cancel mode together, Cancel mode only works effectively with an in-spec DD coil. There is a section in the owners manual on how to use Cancel mode.

Ismael at Minelab mods claims to improve EMI by replacing a component that is used in the later 4500's.

I seriously doubt if this is true, all detectors from the first release were subject to the re-call and as such should have been repaired by Minelab.

Sorry to seem like I'm jumping to Minelabs defense again, but this needed to be cleared up before it becomes gospel.

JP

Hi Jon,

It's your job to defend minelab, I don't blame you for it. If there was an improvement made in EMI, and I strongly suspect there was, I doubt if they'd tell you about it anyway. To do so would be an admission that the early 4500's were faulty and therefore covered by warranty. From memory my first 4500 was no 1520 and was pretty frustrating to use at times. Now i'm using my 7th 4500 which is less than 12 months old and I can honestly say it's a big improvement on my 1st one. I wonder if anybody else on the forum has noticed any improvement?

Regards

Undertaker wrote:Hi Jon,

It's your job to defend minelab, I don't blame you for it. If there was an improvement made in EMI, and I strongly suspect there was, I doubt if they'd tell you about it anyway. To do so would be an admission that the early 4500's were faulty and therefore covered by warranty. From memory my first 4500 was no 1520 and was pretty frustrating to use at times. Now i'm using my 7th 4500 which is less than 12 months old and I can honestly say it's a big improvement on my 1st one. I wonder if anybody else on the forum has noticed any improvement?

Regards

7th?!!! Why in heavens name would you be on your seventh 4500 in a bit over 12 months? Undertaker just for a second think about this, could it be the interference was really bad around the same time you were experiencing problems with your original unit and now conditions are more stable? To my knowledge nothing has been changed on the 4500 other than correcting the original issues via recall with the first batch.

BTW you haven't come back to me on why you were having problems with Cancel mode?

JP

PS

It's your job to defend minelab, I don't blame you for it.

It is not my job to defend Minelab, I consider it my civic duty to defend them when I see public comment that is incorrect.