I didn't need another project -

[ivconic]

...But without schematic it will be again numerous textual posts diluted by various distractive questions and opinions from others.
What we need here is solid ground. The schematic...

You must give me a credit for this Eric.
Just as i finished my latest post; it happened again.
See what happens when we don't have schematic to focus on?
Next 124 pages we'll talk here about pins and needles and "i did this ... i did that."
I am afraid i will not live long enough to finally see the schematic of your upgrades.
The heck with it!
I am not getting any younger here!
There must be better things to do in life, than this.

[WM6]

To detect small things, do this mental model.

1. Identify the time constant of the target.
2. Assume it takes 5 time constants of that targets time constant to decay to zero.
3. Any Pulse Induction detector delay beyond the full decay of the target will not allow that target to be detected.

What are the things that will allow me to attempt to detect this small target.
1. Let us assume that the needle time constant is 0.25 micro seconds (uS).
2. Now we know that the energy in this target will decay to zero in 1.25uS.
3. Can I make a Pulse Induction detector that can sample a little faster than 1.25 uS so I can detect this low TC target?
4. What size coil would be optimum for detecting this small target?
5. What distance should my coil be from the desired target?

Once you think about this detecting challenge in the above terms, you may begin to find a solution.

Joseph J. Rogowski

Thanks for interesting insight bbsailor.

According your calculated specification, I cannot see such available technology in hobby sphere.

Our devices sampling capability end up at 5μS at best.

In my view this remain unsolved problem, not only with PI but with VLF/IB tech also. I tested a lot of
prospecting VLF detector, but no one is able to detect hypodermic needle. Empty hand yes, but not needle.

Nupi started a thread on hypodermic needles a few years ago. I tried recording decay back then without much success.
https://www.geotech1.com/forums/show...564#post196564

Thanks green, it is known to me, but I was in hope, that maybe technology advanced somewhat in meantime.

As I see state of the art, it is only PI technology capable to solve this problem. So I ask Eric, first authority in
field of PI solutions.

[6666]

I still have a way to go with all the updates that I would like to see, and when it is done then the full information will be made available. I'm not sure how much interest there is, but since 1983 and 1991 a lot of XL500 and XL200's must have been sold.

Hello Eric I am interested in your work thank you.

[bbsailor]

To detect small things, do this mental model.

1. Identify the time constant of the target.
2. Assume it takes 5 time constants of that targets time constant to decay to zero.
3. Any Pulse Induction detector delay beyond the full decay of the target will not allow that target to be detected.

What are the things that will allow me to attempt to detect this small target.
1. Let us assume that the needle time constant is 0.25 micro seconds (uS).
2. Now we know that the energy in this target will decay to zero in 1.25uS.
3. Can I make a Pulse Induction detector that can sample a little faster than 1.25 uS so I can detect this low TC target?
4. What size coil would be optimum for detecting this small target?
5. What distance should my coil be from the desired target?

Once you think about this detecting challenge in the above terms, you may begin to find a solution.

Joseph J. Rogowski

[Vadim18]

Nice progress !
What about STP14NK60ZFP for RFP8N18 replacement?

Rds 0.45 Ohm (vs 0.5 Ohm)
Vds 600V (vs 200+v)
Coss 240pf at 25v (vs 250pf)
Gate treshold 4.5 V max (vs 4V)

Price around $1.5
I dont know how, but I was lucky to get it for free as it came from salvaged tv board. Fate)

If Coss after installing isolation diode play no role, SPA17N80C3 may be superior in Vds and Rds

[green]

According your experience, Eric, is this delay range still sufficient to detect sub-gram gold nuggets?

Or from another point of view, what would be most adequate delay range (using fast 18-22cm coil),
to detect sub gram gold nuggets?

Asking this I hope to correlation between sub gram gold nuggets and syringe injection needle lost in
parks and playgrounds of schools and kindergarten, made from medical grade stainless steel which
is here my main goal of interest.

Nupi started a thread on hypodermic needles a few years ago. I tried recording decay back then without much success. https://www.geotech1.com/forums/show...564#post196564

[WM6]

Delay range 20 - 118uS

Eric.

According your experience, Eric, is this delay range still sufficient to detect sub-gram gold nuggets?

Or from another point of view, what would be most adequate delay range (using fast 18-22cm coil),
to detect sub gram gold nuggets?

Asking this I hope to correlation between sub gram gold nuggets and syringe injection needle lost in
parks and playgrounds of schools and kindergarten, made from medical grade stainless steel which
is here my main goal of interest.

[ivconic]

It does say on the sheet 'Proprietory Notice - This information is the property of Garrett Electronics Inc

---
give me this sch and i will delete it in PHOTOSHOP. a job on 1 minute.
while cool story and too pathetic. i just reminder you Eric now... Colonel Phil Corso HAD
gave all info before he gone. so Remember about The Death. do our hobby best.
you should garrett nothing btw... because sir Garrett had gone.

[ivconic]

"...That is a tad cynical..."

Supposed to be.

"...I have rarely made public any complete circuits..."

That was my point indeed.

"...Why would I want to hand it out on a plate?..."

It is always much better when it comes from author than from a third party. One (even hand drawn) schematic speaks much more than 100 posts.
Instead of "i did this... i did that.." is much simpler, easier and better to post schematic.
You will spare many words like that.
Later on you can answer the questions if you like and if there are any.
So far i never saw author to do that on these forums. Except Carl. He is indeed unique exception (Hammerhead and few other articles that he did).

"...I will disclose information as I do the various stages so that others can comment and offer advice and suggestions..."

Alright. But without schematic it will be again numerous textual posts diluted by various distractive questions and opinions from others.
What we need here is solid ground. The schematic. Of all what you wrote about in your previous post.
Hand drawn, no problem, later taken photo of it by camera on your phone.
...
Be second in history of MD forums to do something like that (counting Carl as first one).
Overall i mean nothing derogatory about you and your work, on contrary!
Take this as small advice. If i may dare to offer it to you.
Cheers!

P.S.
Take my word on this; reverse engineering i can do.
But i never enjoy it. What would make me happy is to see author to do all the work for forum instead.

[kt315]

It does say on the sheet 'Proprietory Notice - This information is the property of Garrett Electronics Inc

---
give me this sch and i will delete it in PHOTOSHOP. a job on 1 minute.
while cool story and too pathetic. i just reminder you Eric now... Colonel Phil Corso HAD
gave all info before he gone. so Remember about The Death. do our hobby best.
you should garrett nothing btw... because sir Garrett had gone.

[Ferric Toes]

If you don't eventually offer a schematic with all your upgrades... than "...I didn't need another project..." can be translated to "...we didn't need another project here..." with much more sense.
Decide on this, would it be project to share among members here or something that will just fulfill your days against boredom and kept undisclosed.
Sad to notice; i haven't seen much from your work completely disclosed and documented well, on this forum, done by yourself.
Usually is reverse engineering done by somebody else.

That is a tad cynical in my view! For well over 20 years I have shared information by the means of forums on the technology side of metal detectors. This was because I was not happy about the secrecy that was general at that time in the detector industry as to how their products worked, particularly PI as that was my speciality. However, copying a product by means of reverse engineering was always a problem, not so much with hobbyists, but with other manufacturers. What is the point of developing something and putting it to market, only to find that a copy is on the market a short time later and eroding one's own sales. This comment does not include reputable manufacturers such as the likes of Garrett, Whites, CScope, Detectopro.That negative aspect has always been there, and always will be there unless you spend a fortune and patent as much as possible as some do today. As reverse engineering is a fact of life, I have rarely made public any complete circuits. Why would I want to hand it out on a plate? Some people love the challenge of reverse engineering, so why should I deprive them of the opportunity to do that. It is alive and well today and with genuine hobbyists, I don't have a problem.

Regarding the Garrett XL500, the circuit as I developed it is already on this thread,so there is no point in repeating it.(post 3 attach 2,3. The latter being an update) Garrett's official circuit, dated 28th Jan 1983 I have a copy of, but it is not the same as the 1989 unit I have and also departs further from my original design. It does say on the sheet 'Proprietory Notice - This information is the property of Garrett Electronics Inc, and shall not be copied or reproduced for the purpose of sale or manufacturing of apparatus, except by written permission of Garrett Electronics'. You may say that it is now an old design and doesn't matter. I could write to Garrett's for permission but the people I dealt with are no longer there. They may say it is OK but there is little point as it is not the latest version with Mosfet Tx and better opamp at the Rx front end. The board layout has not changed, as far as I can see, since 1983, so the modifications that I am doing will be relevant.

I still have a way to go with all the updates that I would like to see, and when it is done then the full information will be made available. I'm not sure how much interest there is, but since 1983 and 1991 a lot of XL500 and XL200's must have been sold. These detectors are very rugged and ideal for both shallow and deep water hunting. With uprated performance; longer lasting and lighter batteries, they should be good for a few years yet.

Meanwhile I will disclose information as I do the various stages so that others can comment and offer advice and suggestions if they wish. I would like to find a better logic level Mosfet than the RFP8N18 with higher avalanche volts and lower Rds.on. There is no boredom in my neck of the woods. I already had two projects going which are suffering a bit as a result of acquiring the XL500. But it was hard to resist getting the detector when it apeared to be in good shape and I knew it would be relatively easy to upgrade.

Eric.

[kt315]

garrett deposited. whole pocket of files. no secret files. if we are back to matter ok we are back.

http://depositfiles.com/files/9ru2lph3g

[ivconic]

If you don't eventually offer a schematic with all your upgrades... than "...I didn't need another project..." can be translated to "...we didn't need another project here..." with much more sense.
Decide on this, would it be project to share among members here or something that will just fulfill your days against boredom and kept undisclosed.
Sad to notice; i haven't seen much from your work completely disclosed and documented well, on this forum, done by yourself.
Usually is reverse engineering done by somebody else.

[Ferric Toes]

I have made some progress on updating the electronics while keeping the circuit board the same. First, I measured the various pulses, which were as follows.
Tx width 215uS
Tx rate 106pps
Delay range 20 - 118uS
Sample widths 20uS
Delay to second sample pulse 230uS measured from start of first sample pulse.
Inverter frequency for +5V rail 84.6kHz. A different circuit was used to my original.

Also noted was the the Rx amp had not fully reached zero until 50uS, so the first sample was on the slope prior to full recovery. The detector still worked OK at 20uS by adjusting the threshold control. This was the case both with the Garrett coil and one of my latest fast shielded coils. Also no different when the OPA37 was changed to a NE5534. This was found to be partly due to the 150 ohm resistor used for damping. EF performance was not the best when tested with a ferrite magnet waved over the coil. This is due to the capacitor ahead of the integrator only being 0.1uF I used a 22uF tantalum capacitor in the prototype. This was pointed out to Garrett at the time but they stuck with the 0.1uF for some reason not disclosed.

Another interesting feature is that the Tx Mosfet gate is coupled to the 555 timer o/p directly via a 10nF capacitor and that the gate/source resistor is whopping 150K. Trying a couple of other Mosfets did not work. I then found that the RFP8N18L is a Logic Level device that switches at lower gate/source voltages. Looking at the Tx waveforms it appeared that the drive was still not switching the Mosfet fully. Swapping the 10nF cap for 330nF cured the problem and I reduced the gate/source resistor to 100K which gives very clean Tx waveforms with a sharp rectangular 5V gate drive pulse. The RFP8N18 avalanches at 225V which flat tops at that level for 4uS before decaying via the damping resistor which has been increased to 560ohms. Also a HER208 diode now isolates the coil from the Mosfet Coss.

The gain of the Rx amplifier has been halved by using a 2K input resistor, and as been suggested, the gain can be made up later in the receiver chain. At this stage I have not incorporated a two stage preamp, but comparing the OPA37 and a NE5534 in this circuit, the 5534 recovers about 5uS quicker.

More to come later, but the upgrading of the XL500 is going well.

Eric.

[Qiaozhi]

Those flexible shaft coupler's are also used on 3D printers if you search on ebay on
"3d printer couuplings" you will find lots of all sorts and cheap, you can drill them out to match the pot shaft.

I thought I'd seen those somewhere before. Here's one on my Geeetech i3 Prusa.