SDC Vs 6000 & Geosense

[Aureous]

38 minutes ago, phrunt said:

So, while it does appear the GPX 6000 is just a pumped up SDC with nothing overly new it does have some new technology in it that was good enough to patent, although Minelab would patent their lunch ladies home-made tomato sauce for their canteen pies.

Can just imagine the ML patents attorney trying to invent buzzwords to obfuscate the real ingredients and use thereof for the canteen pie sauce....the paragraphs have to be 20 sentences long...minimum 😁

[Norvic]

Aye them ML buzzwords, but the 6K used in either of the auto modes.... call them Geosense or whatever... tis a wee rattly but a SDC on steroids.

[phrunt]

I like reading the patents, although they go way over my head some of it can be translated to English.  The hardest part for them seems to be proving their "invention" isn't a prior art and justifying the fact is in indeed new, they cover themselves often too for future detectors by saying things like although this method is used in PI detectors, it is possible we could adapt it in the future for VLF detectors or something along those lines, so if someone in the future adapts it for another purpose, they can fall back on their patent to prevent it being done.  They've even got their patent for the GPX fine gold technique patented to cover things like walk through and detectors for kitchens to recover tableware dumped into the bin by mentioning they could adapt it for that in the patent application.  They probably have no intetions of ever doing so, but don't want anyone else to be able to either.

It's probably someone's job just to word these documents to give them the best protection of their patent possible to prevent future products by other manufacturers.

[RONS DETECTORS MINELAB]

The three refined GPX-6000 channels would explain its great performance abilities on a wide range of target sizes. And the CTX Auto Sensitivity would explain the ability to work with the varying ground conditions automatically. Guessing SETA is integrated into these complex channel loops also, guess we will never know the secret.

Below is from a couple of articles about the MPF’s refined enhance/fine gold and salt timings which seems about right for the performance we see in the 6000. Also the auto sensitivity from the CTX.

 

From Minelab’s SDC’s/MPF  knowledge based article:

To improve sensitivity to small targets, the transmit frequency is about 3kHz, which is much faster than a conventional PI detector. This means more averaged signal per second (3000 signals per second) of the weak small gold signals. The faint metal target signals are the same following each transmit pulse, whilst the electronic noise is random. This means that the noise averages to a low level compared to the faint metal target signals , making the target signal ‘stand out’ from the noise.
Reduced Ground Noise.

The SDC 2300 incorporates Minelab’s proven “Enhance/Fine Gold” type of receive processing that minimises mineralised soil signals and false detection of hot rocks, unlike traditional PI detectors. This greatly assists with nugget detection in most gold fields, especially in highly mineralised gold fields. This substantial reduction in ground noise makes prospecting less confusing and more time efficient, It incorporates Automatic Ground Tracking and
Fast Ground Balance so that the detector is always optimally ground balanced.

Performance in Salty Conditions
The SDC 2300 incorporates a Salt Mode for which the transmit and receive timing signals have been optimised for saline conditions found in some gold field areas.

Nenad’s CTX Auto Sensitivity write-up:

 I’d like to explain briefly how Auto works. As you are sweeping, the CTX 3030 automatically measures the ground, and uses this information to set the Sensitivity to the maximum stable setting. Now it is important to know, that the detector can accurately analyse the signal from positive or magnetic mineralisation, but does a poor job on conductive ground, which saltwater beaches are. What happens is that the detector will continue trying to gather ground information, but a primarily conductive ground response can cause the Sensitivity setting to be inconsistent, leading to an incorrect setting. So as the detector struggles to read the conductive ground accurately, the operator ends up with a noisy detector, due to a Sensitivity setting that is too high.

The reverse can also happen, and due to incorrect ground information being read, the detector sets the Auto Sensitivity conservatively low, i.e. 16, whereas if set manually, the operator would find that the detector is still working well on a Sensitivity setting of say, 21. In this instance, the operator who is working in Manual Sensitivity would be getting better performance. 

[phrunt]

yep, little has changed, that's why they put an Auto + on there. 🙂

What I find intriguing is they clearly state using a threshold can make finding weaker target signals easier which means more chance of small gold or deeper targets than no threshold, yet for those that use Auto and Auto + the threshold is off by default, and to make it worse, even if you turn it on which you have to do in each Auto mode individually, each time you power cycle your detector you have to turn the threshold on again.  It's like they think the people that are likely to use auto are the same type of people that would prefer to hunt with no threshold, make of that what you will.

Fortunately for me I don't use the Auto modes, or I'd be nagging Minelab to release a firmware that saves your last used threshold configuration setting.

[RONS DETECTORS MINELAB]

On 4/1/2024 at 8:26 PM, phrunt said:

Fortunately for me I don't use the Auto modes, or I'd be nagging Minelab to release a firmware that saves your last used threshold configuration setting.

I also use the manual modes. Interesting that the owner manual does not mention what the difference is from auto+ verses high manual settings for overall performance differences. In my soils which are more challenging I find a small edge on smaller targets using the manual settings. Not sure if benign soils make any difference here but from Nenad’s auto sensitivity description it would make sense.

On the manual gain settings I am also surprised that you can go from 1 to 10 and really do not loose very much depth to targets as you would expect from all the previous legacy models. It seems like the manual sensitivity levels act more like a stabilizer noise filter than an actual gain adjustment, If this was the case then the sensitivity levels are really all automatically set on the fly and also dependent on ground conditions.

The motion speed is somewhat interesting as you mentioned, it can go from a dead to a supersonic signal. This supersonic signal sounds like a modded 5000 or what people refer to as it being on steroids.

[WesD]

"each time you power cycle your detector you have to turn the threshold on again"

For someone like myself who hunts mostly in auto plus, that's got to be the most annoying feature on the 6000!!!  Wishing Minelab could give us a reprogram update to change that option.

 

[jasong]

2 hours ago, RONS DETECTORS MINELAB said:

It seems like the manual sensitivity levels act more like a stabilizer noise filter than an actual gain adjustment, If this was the case then the sensitivity levels are really all automatically set on the fly and also dependent on ground conditions.

I think the last part is definitely true, maybe the first part too. I did some experiments a while back and found that both large iron targets as well as high EMI levels caused the 6000 to adjust itself in manual, not just in auto. Further, by memory, I believe it stayed dumbed down even after the iron and EMI were removed to some degree, requiring a reset to really trust it was back to normal. I noticed this happening in the field at first but was unsure, but when I posted here a while back about this effect, other people didn't seem to see it, so I tested in my shop and found enough to change my trust in the 6000 personally. I've posted about this a few times in the past here. 

6000 is a prospecting machine only to me. I toss it into auto+, and use it for what it is. What happens under the hood is too opaque for me to ever trust it over the 7000 for cleaning applications because there is no way to verify if I am running sub-optimally or not.

This is why they need to at least provide some bar graph readouts to show an operator what the detector has adjusted, if controls aren't provided. Not knowing is not acceptable for pro-sumer grade machines IMO.  I do like having the auto option though, and I believe it will improve in future releases.

[phrunt]

I think what you were seeing was the auto ground tracking going out of whack Jason, I can replicate it by going over a big bolt or something a few times then attempting to find a small nugget, it takes a while to recover, and yes, a quick way about it is a factory reset.

You can confuse the detector, I don't think its sensitivity is changing, I think it's all to do with the ground balance.  I guess high EMI is like a target to it or even some ground, so it's trying to balance to the EMI and going all funky from trying to do that.

I wish it had a way to fix the balance, a feature it is lacking that it would greatly benefit from.

It's funny though, the 6000 doesn't have fixed and people want it, the Algo has fixed and people complain it's not auto 😛

I think it's best to have both.

[Gold Catcher]

11 hours ago, RONS DETECTORS MINELAB said:

On the manual gain settings I am also surprised that you can go from 1 to 10 and really do not loose sensitivity to targets as you would expect from all the previous legacy models.

You don't loose that much sensitivity for shallow targets when you reduce gain, but the loss in depth is much more profound.

GC