Alternating pulse width

[Kev]

Hi Wirechief,
You surely brightened up my day
I'm West of you alright, try 90 degrees West, and 75 degrees South, ruffly.
Isn't the internet great. I'm one of the few oddballs in this city that goes beach detecting, I say oddball cause that's how people look at me. Many give me a wide berth, except kids of course, I've got some sort of magnetic attraction for them, anyway every serious metal hunter knows what I mean. The beauty is that I can connect to the internet, and be among like minded oddballs
I mean people.

Bye for now.
Kev.

 

[Eric Foster]

Hi Kev,

I don't think so, but ML machines are a bit of a mystery. I took a standard ML mono coil and fixed it up to one of my earlier Deepstars, which had been modified, so that the transmitter arrangement was about the same, except for the short pulses. The earliest I could sample was 15uS, and that was a bit close to the edge. It appears that the ML long pulse generates the ground signal and the three short pulses are for the target. It could be that the three short pulses and their associated samples are to equalised the integrator time constants for the signal and ground balance channels. In the SD2100 and 2200 The short pulses are of a much smaller amplitude than the long pulse, due to the current being inductance limited. Summing the signal samples from the three short pulses, would also make up some of the lost amplitude. In the ML machines that use "dual voltage technology", the TX supply on the short pulses is higher, so as to give a similar peak current amplitude to the long TX pulse.

Eric.

 

[Wirechief]

Why would people be afraid of a guy with a metal detector on a beach? Wirechief.

 

[Kev]

Hi Wirechief,
It's probably not as bad in the States, since detecting has been a known hobby for decades, but down here many people don't even know what a detector looks like!. I've had many Maoris asking me "findin many pipis mate?"(shellfish). We've also had a large number of Asian immigrants arrive recently who perhaps associate detectors with UXOs, something they thought they left behind in their old country, then they see a dude on their local beach mine sweeping

Then again it could be me, concentrating, with the distant look in my eye, could look a little menacing to timid people I suppose?
Or maybe I am the only oddball in the world?
Anyway, by your response I must add, I'm certainly not calling anyone else here an oddball, it's just my own perspective. I'm sorry if I've offended you or anyone else.

This is a PI forum so I'd better stop waffling about nothing.

I'm building a new DD coil this weekend for my GQ, using a 15" Sovereign WOT scuff cover. The Explorer scuff plate makes an excellent light 10" coil.

Cheers
Kev.

 

[Wirechief]

No problem Kev and my misunderstanding. I thought for some reason you were here in the states but when you say cheers that should have told me. To tell you the truth I live in a little town of 300 and when everyone else is farming, ranching, playing cards at a party or dancing I'm at home building a ham rig or metal detecting or gazing at the stars so I know I'm an oddball right here where I live. Well like you said this is a PI detector forum and I hope your DD coil works well and thanks again for the link to the riser card and I saved it to my favorites.
That's kinda funny about the mine detecting but really it is not for the people that have lived through that all their lives. Ok Kev let us know how your coil turns out and I bet it is lightweight. Cu later Kev cheers! Wirechief.

 

[Stef]

Hi Eric,

I own a SD2000 and also a SD2200 and have been reading this thread with interest.The way the ML works is covered in Candy's patents but he obfuscates and its hard to read the patent. So I'll put in my 2 bob's worth.

Basically Candy talks about long time constants and short time constants and corresponding ground balance channels. The long time constant channel infers large targets and the short time constants to smaller targets closer to the surface. In the SD2000 the pulse train consists of a long pulse followed by 4 short pulses. The on and off times are equal. The SD2200 has a long pulse followed by 3 short pulses.

The ground channel is derived by taking a sample after the flyback voltage goes to zero ( off time ) and then taking a sample towards the end of the sample period. These two samples are subtracted and averaged with previous historical samples to give a ground balance channel. THis calculation is performed on both long and short pulses to give corresponding long and short ground balance channels. The time constant of the ground balance channel is approx 10 seconds.

This part gets a bit fuzzy but I think the long and short time constant channels have significantly shorter time constants as compared to its associated ground balance channel. So for example the signal derived from the short time constant channel is compared with it's ground balance channel. In the absence of a target, the channel is balanced either manually (SD2000) or automatically (SD2200) to give a null output. When a target is detected the null point will go out of balance and a signal will be detected. With this type of ground balancing the detector must always be moving. If you detect a
target and hold the coil stationary over the target for at least 10 seconds the ground balance channel will null out the signal.

The ML runs with two ground balance channels with the greatest signal routed to the ouput. With Dual Voltage technology the long pulse is approx 5 volts with the short pulses approx 30 volts. I noticed in the SD3500 the ground balance time constant can be varied between short medium and long to give better response on highly mineralised ground.

I hope this note helps somewhat. I'm trying to build a prototype using Candy's patent but it has been a slow process.

Regards,

Stefan