Vallon VMH3CS Mine Detector

[Ferric Toes]

Metalmorph ready for test run

I got the waterproof connectors in and fitted an extra nimh battery. I figured that four 1.2V C batteries are little different in voltage to three 1.5V D alkalines in the original detector, and the former looks neater. The home made pack disconnects so that a spare is easily fitted. A semiconductor fuse is incorporated as a protection. The phone is the one supplied, but removed from its headstrap which isn't very elegant. I shall put it under a head sweatband just off the ear. It was best with the phone reversed, then it operated like a bone conductor phone. The coiled cable from the coil shaft is what resides in the telescopic shaft, but with most of it pulled out and anchored at the top of the handle. I have left about 6in of telescopic adjustment which is sufficient for me. All external metal fittings are stainless steel, so no corrosion problems. I just have to make a gasket between box and lid, although I believe that Hammond possibly make them.

Everything fitted neatly into the box with no room to spare. The green T shape pcb is a distribution board from where most of the cable runs diverge. The sealed grey box underneath houses the main electronics. There is a small piezo speaker on the opposite side to the single control and this cuts out when the phone is plugged in.

Weather permitting I hope to spend some time at the beach over the weekend.


Eric.

[Ferric Toes]

Thank´s for your help, but it is not possible for me to activate the automatic channel adjustment. Maybe it is an other limitation of tis older model. I can switch to manual mode (hold c and then adjust with + or -). If I press + and - the same time I adjust only the volumen to more louder.
Is there a way to adjust delay or the filters that you mentioned?

I haven't tried the channel adjustment on mine, so I don't know if my one works. I have tried the filters but no luck there so far. However, in it's stock state the performance that I see so far is very good. The VMH3CS is still in production, so maybe later improvements, or better firmware, has been incorporated in later years. I would have thought that channel adjustment, so that several detectors could operate in the same area, would have been necessary since the detector first came out.

Eric.

[Skippy]

The T-shaped board looks to have SMPSU on it, based on the 6-pin SOT23 packaged part, the small inductor, and maybe the odd diode. This is obviously low powered ( <500 mW ?) and will not be for driving any of the main power-hungry stuff.

I'm surprised by your use of 4 cells, I'm sure the design is competent enough that it will work fine with either 3 x dry or NiCd/NiMH.

I'm not familiar with the connectors you used, are they 2-pin? I was thinking it would be an idea to bring out 2 pins as a charging circuit, with a protection diode and small fuse etc added. Then you won't have to dismantle it to recharge the pack.

[Ferric Toes]

The T-shaped board looks to have SMPSU on it, based on the 6-pin SOT23 packaged part, the small inductor, and maybe the odd diode. This is obviously low powered ( <500 mW ?) and will not be for driving any of the main power-hungry stuff.

I'm surprised by your use of 4 cells, I'm sure the design is competent enough that it will work fine with either 3 x dry or NiCd/NiMH.

I'm not familiar with the connectors you used, are they 2-pin? I was thinking it would be an idea to bring out 2 pins as a charging circuit, with a protection diode and small fuse etc added. Then you won't have to dismantle it to recharge the pack.

Yes, the detector works fine with no change in performance down to 3.2V supply. The interesting thing is that as the supply voltage falls, so the standing current goes up.
3.2V gives 455mA
4.0V gives 350mA
5.0V gives 287mA
5.1V gives 282mA

The demining literature states that the detection performance stays the same as battery volts fall, so TX current must have a stabilised supply. Perhaps this is why the current goes up at low voltages.
Some of my testing is done using a bench power supply, so to avoid a lapse of attention that results in my connecting the supply the wrong way round, I have incorporated a series diode which loses me nearly 1V, which is one reason for the extra battery. I do not know if there is any reverse protection on the T shape board in addition to the fact that in the original machine using D batteries, putting them in the wrong way gives no contact. If I fry something, particularly on the main circuit module, then that is curtains for the unit, hence I am ultra careful.
The connectors are Souriau UTS series which are very good and reasonably priced. Although plastic, they mate with MIL spec. metal connectors which is useful for the earphone supplied, although I did swap to Souriau before I realised this, on the one in the picture. I use six pin connected as on the original unit and 3pin for the coil. The connectors are still sealed when unmated but it is better to use their protection covers. RS have them.

According to the VMH3CS spec. charging is possible through the 6 pin connector although at this stage I do not know if there is charging circuitry in the unit, or not. Firmware changes are also made via this connector but information is not available on either issue.

Eric.

[Skippy]

From your previous posts showing +6V and -6V levels, it's pretty apparent it will have a proper switching regulator inside, with most of the power drawn going through the regulator. So this variation in current draw is entirely consistent with such a circuit: the power input is almost constant.
Re. the charging circuit: it's probably not that useful to you, anyway, as your use of 4 cells (and a diode) would make it non-functional.

[Ferric Toes]

I lost the coil.

Latest venture with the Vallon was to make a probe with a ferrite core and see how it performs. Having a spare coil shaft, I cut off the swivel and cleaned up the end so that it was square. I have a few 4in x 0.5in rods and these nicely fit into the tube with room to spare for winding, shielding, etc. First thing was to wind a ferrite and see what inductance and resistance I get. I am aiming for 1.8 - 2.0mH and approximately 3 ohms resistance. Because this type of ferrite rod is conductive to a small degree (15K end to end) I always use a thin drain wire along its length and wind on top of that to hold it down. The drain wire in this case was one strand of 0.2 tinned copper wire. Leaving 1/4in at each end of the rod, I close wound as many turns as I could between two O rings, one at each end of the rod. The wire used was 0.2mm Teflon insulated, just because I have reels of it rather than it being necessary. The windings were then overwrapped with plumbers PTFE tape to hold the wire in place and form a base for further windings if needed. An inductance measurement showed that I had just over 1mH so more turns were needed. About 20 turns overwound in the centre of the rod gave a total of 1.8mH which is close enough. I anchored the ends of the overwind with adhesive tape and brought out all wires at the start end. I them put a heatshrink sleeve over the whole winding leaving the 1/4in clear at each end. A resistance measurement gave 2.3 ohms, which is an ohm short of what is needed.

Once the heatshrink had cooled, I then cut some copper fabric adhesive shielding tape and wrapped it round the heatshrink tube on the winding and core. I left a 1/16in gap so that you don't end up with a shorted turn, and then soldered a drain wire to join the three pieces of fabric tape used. The drain wire is brought out to the same end as the other wires. The whole core assembly is given a final wrap of adhesive pvc tape. This is necessary if finally potting the core in polyurethane resin, as something in the resin or catalyst will dissolve the adhesive on the copper fabric tape and it comes adrift.

Both drain wires and the ground side of the coil winding are connected together and in turn connected to the coax braid. Coil live obviously goes to the coax core.

On test , outside of the tube, the startup check failed, and after a bit of investigation I concluded that the major difference to the standard coils was the resistance. I added a 1 ohm wirewound resistor in series with the coax core at the connector end and startup was fine. The core, cable and connector were all mounted into the tube and a plastic end bung fitted. No potting was done at this stage as there is always the chance that the probe might have to come apart again. For potting I always use polyurethane resin as this is softer than the hard resin used for potting conventional coils. This then damps any mechanical resonance of the ferrite core caused by magnetostriction at higher drive currents. If this ringing isn't damped it will be seen as ringing on the RX output which changes in amplitude if you press a finger on the core at one end. Not to be confused with electrical ringing of the coil inductance and associated capacitance.

First tests look good and a nickel is detected at 8in and a beer can at about 2.5ft. No earth field effect which I have seen with some ferrite probes. Maybe the bipolar pulsing helps in this regard. I plan on winding another with slightly thinner wire so that I will hopefully get the winding all on one layer and also the correct resistance without having to add an external one.

There will be lots of uses for this probe as some of our local beaches have plenty of sand, but also lots of rocks where it is difficult to swing a normal coil. The military spec shaft is immensly strong so the next move is to attach a retractable carbon fibre mini shovel blade to the probe shaft end.

[Chet]

[QUOTE=Ferric Toes;218248]
"First tests look good and a nickel is detected at 8in and a beer can at about 2.5ft. No earth field effect which I have seen with some ferrite probes. Maybe the bipolar pulsing helps in this regard. I plan on winding another with slightly thinner wire so that I will hopefully get the winding all on one layer and also the correct resistance without having to add an external one."

Hi Eric

That is really good results; how much depth increase does the ferrite core provide over a coil without the core?

Have a good day,
Chet

[Ferric Toes]

[QUOTE=Chet;218253]

"First tests look good and a nickel is detected at 8in and a beer can at about 2.5ft. No earth field effect which I have seen with some ferrite probes. Maybe the bipolar pulsing helps in this regard. I plan on winding another with slightly thinner wire so that I will hopefully get the winding all on one layer and also the correct resistance without having to add an external one."

Hi Eric

That is really good results; how much depth increase does the ferrite core provide over a coil without the core?

Have a good day,
Chet

The core increases the magnetic field in proportion to its relative permiability (uR). Unfortunately I bought the cores many years ago and have mislaid the spec on the core material. The uR can be as much as 10,000 for ferrites, but the field can be limited by magnetic saturation before you reach that multiplying figure. The core also enhances the return signal from the metal target. Without the core I would expect maybe an inch or two on the nickel. The length of the core also has an effect. I used to have some 8in cores and I would expect another couple of inches as the field is projected further, the same as a long bar magnet would.

Not all ferrite cores work, as some give a huge viscosity signal. Avoid all cores used as radio antennas as this is a different type of ferrite.

Eric.

[Monolith]

[QUOTE=Ferric Toes;218262]

The core increases the magnetic field in proportion to its relative permiability (uR). Unfortunately I bought the cores many years ago and have mislaid the spec on the core material. The uR can be as much as 10,000 for ferrites, but the field can be limited by magnetic saturation before you reach that multiplying figure. The core also enhances the return signal from the metal target. Without the core I would expect maybe an inch or two on the nickel. The length of the core also has an effect. I used to have some 8in cores and I would expect another couple of inches as the field is projected further, the same as a long bar magnet would.

Not all ferrite cores work, as some give a huge viscosity signal. Avoid all cores used as radio antennas as this is a different type of ferrite.

Eric.

Your explanations about the ferrite cores are very interesting and enlightening.
The radio antenna cores used to be easily available some years ago, but in general, long ferrite cores are getting harder to find. What specifications do you recommend to be used for a pin-pointer?

[Ferric Toes]

[QUOTE=Monolith;218264]

Your explanations about the ferrite cores are very interesting and enlightening.
The radio antenna cores used to be easily available some years ago, but in general, long ferrite cores are getting harder to find. What specifications do you recommend to be used for a pin-pointer?

I will go through my old notebooks and see if I can find specs. I bought some on ebay a couple of years ago and that is what I am using now, and they work just fine. I have a plastic tube with a 200uH winding into which I insert a bare rod. Using a low power TX and just an RX amp, if the rod gives little or no signal when inserted, then it will work. I have had pinpointers running at just 1uS delay using ferrite cores although great care is needed in shielding and winding to achieve this short delay. Such sensitivity has little use in the field but great for some industrial uses. e.g. detecting broken off needle points in fabric manufacture, or finished goods such as shirt manufacture, baby clothes, etc.

[Ferric Toes]

MK2 Probe core for Vallon

I am in the process of making a second core and winding now that I have made a MK1 that fires up OK. This time I am using some Litz wire that has been kicking around for as long as I remember. 50/0.02 silk covered would you believe and it is thinner than the 0.2mm used in the MK1. I managed to achieve 2mH in just one layer on the core, which is good. Heat shrink tubing was slid on, just, and then by way of another change, I used Scotch 24 shielding tape for outer shielding as per the picture. I didn't realise that you could open the Scotch 24 into a tube into which the core and winding slide nicely. The Scotch is easier to solder to than the copper fabric used in MK1. I was hoping that the resistance would be higher so that I could do without the 1 ohm in series, but I still need it to get to 3 ohms required. A finishing layer of pvc tape holds the shielding neatly in place. The finished core fits into 20mm O.D. pvc or abs plumbing pipe.



When this is done and tested, I will make yet another with the more usual 300uH inductance, which seems to be standard for PI circuits. I may then go on to design the electronics for a 1 - 20uS delay probe. You would need this range of adjustment otherwise, unless the soil is absolutely non-conductive and non-magnetically mineralised you would be tearing out your hair. - we'll see.

Eric.

[Ferric Toes]

I am in the process of making a second core and winding now that I have made a MK1 that fires up OK. This time I am using some Litz wire that has been kicking around for as long as I remember. 50/0.02 silk covered would you believe and it is thinner than the 0.2mm used in the MK1. I managed to achieve 2mH in just one layer on the core, which is good. Heat shrink tubing was slid on, just, and then by way of another change, I used Scotch 24 shielding tape for outer shielding as per the picture. I didn't realise that you could open the Scotch 24 into a tube into which the core and winding slide nicely. The Scotch is easier to solder to than the copper fabric used in MK1. I was hoping that the resistance would be higher so that I could do without the 1 ohm in series, but I still need it to get to 3 ohms required. A finishing layer of pvc tape holds the shielding neatly in place. The finished core fits into 20mm O.D. pvc or abs plumbing pipe.

When this is done and tested, I will make yet another with the more usual 300uH inductance, which seems to be standard for PI circuits.

Eric.

Well it was done and tested - and it doesn't work
I undid the white pvc tape and removed the Scotch 24 screening and now - it does work. Obviously the strong magnetic field in close proximity to the screen generates sufficient late time eddy currents to cause a problem. So it's back to the woven fabric copper screen that worked on the first one. Interestingly the Vallon open coil is screened with the Scotch 24 and so is the 11in coil that I made, with no problem.

Yes, on my MVM it gives a reading of 100 at 10uS delay when held as a cylinder in the sample chamber, so it is acting as a shorted turn in the same plane as the coil winding. The Vallon switch on check identified it correctly as a coil short. Anything in the same plane as the winding should have a gap in it as does the Vallon coil and my 11in. To put a longitudinal gap in the Scotch 24 it would all fall apart.

Moral for me, and I should know better, - wait till it is fully tested before reporting!

[STAVROSNU]

Dear Eric,

I have just sent you a private message.
Can you please read it and send me your reply, at your convenience !


Best Regards

Stavros

[bbsailor]

Well it was done and tested - and it doesn't work
I undid the white pvc tape and removed the Scotch 24 screening and now - it does work. Obviously the strong magnetic field in close proximity to the screen generates sufficient late time eddy currents to cause a problem. So it's back to the woven fabric copper screen that worked on the first one. Interestingly the Vallon open coil is screened with the Scotch 24 and so is the 11in coil that I made, with no problem.

Yes, on my MVM it gives a reading of 100 at 10uS delay when held as a cylinder in the sample chamber, so it is acting as a shorted turn in the same plane as the coil winding. The Vallon switch on check identified it correctly as a coil short. Anything in the same plane as the winding should have a gap in it as does the Vallon coil and my 11in. To put a longitudinal gap in the Scotch 24 it would all fall apart.

Moral for me, and I should know better, - wait till it is fully tested before reporting!

Eric,
Try cutting the Scotch24 into a single layer and put a layer of tape over the long end to prevent the wire mesh from shorting out around the probe circumference. I have found that this method reduces the detection of the wire mesh while still providing a good shield.


Joseph J. Rogowski

[bbsailor]

[QUOTE=Ferric Toes;218265]

I will go through my old notebooks and see if I can find specs. I bought some on ebay a couple of years ago and that is what I am using now, and they work just fine. I have a plastic tube with a 200uH winding into which I insert a bare rod. Using a low power TX and just an RX amp, if the rod gives little or no signal when inserted, then it will work. I have had pinpointers running at just 1uS delay using ferrite cores although great care is needed in shielding and winding to achieve this short delay. Such sensitivity has little use in the field but great for some industrial uses. e.g. detecting broken off needle points in fabric manufacture, or finished goods such as shirt manufacture, baby clothes, etc.

Anyone wanting to experiment with ferrite cores should use what is called "soft ferrite material". Eric initially made this recommendation on his forum many years ago. This material does not hold the magnetism after the current turns off. I have used the soft ferrite rods found at surplussales.com, (ICH) ROD 7.5/50 made from soft ferrite 3C80 material on many PI machines. You can experiment with gluing two or more together to make the rod longer or bundle three rods side-by-side. Just make sure that all rods are grounded and shielded to keep the noise low.

I hope this helps?

Joseph J. Rogowski