Minipulse Plus Rev My time has come!

[Qiaozhi]

Question of the day,how does the EFE work?I imagine it takes a sample of the field,then invert it and substract it from original sample?To achieve that it should drive three op-amps one with sample another to produce the inverted sample and a third to substract them?

The main sample will contain a signal from the target plus a contribution from the Earth field. The EFE sample is usually taken a few hundred microseconds later, when the target signal has substantially died away, whereas the EF signal will remain the same. By subtracting the EFE sample from the main sample you are then left with the target signal.

No, preamp out from my target response tester. Don't have a MPP.

The reason your target response tester gives a signal for the magnet is because you don't take a second (EF) sample. Of course, since the target response tester doesn't actually move, there is no need to provide this.
This is an example of the white board magnets I'm talking about ->

https://www.amazon.co.uk/Scribble-Value-Whiteboard-Fridge-Magnets/dp/B004RYSUMU/ref=sr_1_3?keywords=whiteboard+magnet&qid=1557095021&s=gateway&sr=8-3

[godigit1]

https://www.rmcybernetics.com/science/diy-devices/diy-pwm-signal-generator

Here is a simple one works good. I have a version Im turning into a Tx as well.

[mushaba]

https://www.rmcybernetics.com/scienc...gnal-generator

Here is a simple one works good. I have a version Im turning into a Tx as well.

Allready design its PCB with Eagle,i will give it a try later this afternoon,thanks friend!!

[mushaba]

The main sample will contain a signal from the target plus a contribution from the Earth field. The EFE sample is usually taken a few hundred microseconds later, when the target signal has substantially died away, whereas the EF signal will remain the same. By subtracting the EFE sample from the main sample you are then left with the target signal.

So the Sec.Sample (TP10) shown in "Fig.10 Sample Timming" of building instructions is the EF..If i move this closer to main sample will the EFE work better but in danger of missing targets?Im looking for a way,without heavy mods to make it less sensitive to smaller items since i wont go on small gold or coin hunting.

[Qiaozhi]

So the Sec.Sample (TP10) shown in "Fig.10 Sample Timming" of building instructions is the EF..If i move this closer to main sample will the EFE work better but in danger of missing targets?Im looking for a way,without heavy mods to make it less sensitive to smaller items since i wont go on small gold or coin hunting.

If you want to make the detector less sensitive to small items, then just increase the sample delay.
Also, you don't need to spend time making a super fast coil, a simple jumble wound coil using enamelled copper wire will be perfectly ok.

[mushaba]

If you want to make the detector less sensitive to small items, then just increase the sample delay.
Also, you don't need to spend time making a super fast coil, a simple jumble wound coil using enamelled copper wire will be perfectly ok.

I thought "fast" had to do with fast response of coil so we can sample earlier to find non megnetic metals like gold,didnt know that could be that sensitive and detect all theese little quantity of metal..

Also when we place a can in front of the coil,we hear audio and seconds later the audio fades away,is that proof of EFE circuit working?

[mushaba]

I wonder if it's possible to use the MPP with a coil 1m x 1m square like Delta pulse,the damping procedure will be the same I believe Did anyone try it?

[waltr]

I thought "fast" had to do with fast response of coil so we can sample earlier to find non megnetic metals like gold,didnt know that could be that sensitive and detect all theese little quantity of metal..

Fast is to detect target with very short TC. A target's TC is dependent of composition (type and alloy of the metal which determine the conductivity) AND the size.
Lower conductivity => shorter TC
smaller size => shorter TC

Also when we place a can in front of the coil,we hear audio and seconds later the audio fades away,is that proof of EFE circuit working?

No that is not EFE circuit effect. It is the SAT circuit (U4a) effect of C9 & R6.

[mushaba]

Fast is to detect target with very short TC. A target's TC is dependent of composition (type and alloy of the metal which determine the conductivity) AND the size.
Lower conductivity => shorter TC
smaller size => shorter TC

At my air test I found that a fast coil has greater detection depth that a jumble wound(without shield) that's why I stayed with fast coil.Do not want to detect items smaller than a coin but also I don't want to compromise the detection range.George suggested to have a larger main sample delay...I wonder if a larger coil will ignore small items and the MPP will operate as it should(no change of original delays)..

[Qiaozhi]

At my air test I found that a fast coil has greater detection depth that a jumble wound(without shield) that's why I stayed with fast coil.Do not want to detect items smaller than a coin but also I don't want to compromise the detection range.George suggested to have a larger main sample delay...I wonder if a larger coil will ignore small items and the MPP will operate as it should(no change of original delays)..

There is no need for a fast coil if you're searching for items larger than a coin. To find larger items deeper, then go for a bigger coil. That will also help you to ignore small targets.

When the MOSFET is abruptly turned off, the magnetic field in the coil collapses. This generates eddy currents in any nearby metal targets. These eddy currents will die away more quickly for thin, small, and low conductivity items. Presumably these are the targets you want to ignore. Therefore, sampling with a delay that is short enough to pick up these items is just pointless.
In conclusion:
You should increase the sample delay, use a bigger diameter coil, and increase the TX pulse width. You may also need to lower the TX pulse rate. It should be obvious that Litz wire, basket weave coils, etc., do not provide any advantage in this situation.

[mushaba]

There is no need for a fast coil if you're searching for items larger than a coin. To find larger items deeper, then go for a bigger coil. That will also help you to ignore small targets.

When the MOSFET is abruptly turned off, the magnetic field in the coil collapses. This generates eddy currents in any nearby metal targets. These eddy currents will die away more quickly for thin, small, and low conductivity items. Presumably these are the targets you want to ignore. Therefore, sampling with a delay that is short enough to pick up these items is just pointless.
In conclusion:
You should increase the sample delay, use a bigger diameter coil, and increase the TX pulse width. You may also need to lower the TX pulse rate. It should be obvious that Litz wire, basket weave coils, etc., do not provide any advantage in this situation.

Ok George,i wasnt a good student after all

I didnt have the time to make a new coil but i increased the 1st sample delay to 60us (i think it as far as it can with R39).I lowered the total frequency at about 500Hz and left the Tx pulse at 120us,with a quick scan at my trashy backyard seems the detection became less sensitive to small items.Do you have any recomendation on timming values?

If i find the correct timmings i will have fixed values resistor to "replace" pots (R5-R14-R37) via a switch,i will think it through some more later..

As for the coil my plan is to try a 40cm OD coil with double enameld 0,4mm wire,the shielding is that frighttens me because i've never succesfully done one..I was reading Carl's notes on Hammerhead article about coil shielding,the Scotch24 is the optimal choice,unfortunatelly they dont have it in stock...I have 2 adhesive tapes,one with aluminum and one with coper but when i tried once to shield my coil its detection ragnge went down dramatically..How much space should be between coil windings and shielding material?

[waltr]

First delay time should be a user control. Many PI detectors have this as a control so the user can adjust as needed.
As needed can be due to the location, field verse salt beach, or which coil is connected, small fast verse large slow, and what object you are targeting, small gold jewelry verse coins.

I never found shielding to help on the coils I have built. So my three coils (30cm mono, 25cm DD & a 20cm Concentric) are all unshielded.
I had a thread to discuss this but it never went anywhere and everyone just quoted what has been written.

I also test coils with temporary mounting, foam board and wood, before doing all the work to encase them. This allows changes and re-testing.

[Qiaozhi]

I didnt have the time to make a new coil but i increased the 1st sample delay to 60us (i think it as far as it can with R39).I lowered the total frequency at about 500Hz and left the Tx pulse at 120us,with a quick scan at my trashy backyard seems the detection became less sensitive to small items.Do you have any recomendation on timming values?

As a rough guide, when searching for gold nuggets (8-10us delay) the coil must be shielded. This will provide electrostatic shielding between the coil and the ground, and also stops electrical noise.
Use > 15us for the beach, and 30-50us for coin shooting.
Longer delays (typically transmitting at 100-400pps) do not need to be shielded.

If i find the correct timmings i will have fixed values resistor to "replace" pots (R5-R14-R37) via a switch,i will think it through some more later..

If your coil has an inductance of 300uH and a series resistance of 5 ohms (including the series 3R3 resistor) then the tau during TX-on will be L/R = 300u/5 = 60us. To allow the coil current to build up to maximum requires 5*tau, which equals 300us. In this case I would recommend using a TX pulse rate of 500pps. However, in practice, there is little or no difference to using 1000pps with a TX pulse width of 185us. With 300us the energy in the coil will be 600uJ, and with 185us the energy is 536uJ.

As for the coil my plan is to try a 40cm OD coil with double enameld 0,4mm wire,the shielding is that frighttens me because i've never succesfully done one..I was reading Carl's notes on Hammerhead article about coil shielding,the Scotch24 is the optimal choice,unfortunatelly they dont have it in stock...I have 2 adhesive tapes,one with aluminum and one with coper but when i tried once to shield my coil its detection ragnge went down dramatically..How much space should be between coil windings and shielding material?

[ATTACH]46340[/ATTACH]

Wind spiral wrap on top of the coil before adding the electrostatic shield.

[mushaba]

As a rough guide, when searching for gold nuggets (8-10us delay) the coil must be shielded. This will provide electrostatic shielding between the coil and the ground, and also stops electrical noise.
Use > 15us for the beach, and 30-50us for coin shooting.
Longer delays (typically transmitting at 100-400pps) do not need to be shielded.

Why longer delays do not need to be shielded??There is "more time" to compare the 1st with 2nd sample ??

The embarrasing question of the day:the sampling interogator is configured as differential amp (U3a)??The inputs are dc voltages or ac waveforms...probaply dc because the 470nF acts as decoupling??

If your coil has an inductance of 300uH and a series resistance of 5 ohms (including the series 3R3 resistor) then the tau during TX-on will be L/R = 300u/5 = 60us. To allow the coil current to build up to maximum requires 5*tau, which equals 300us. In this case I would recommend using a TX pulse rate of 500pps. However, in practice, there is little or no difference to using 1000pps with a TX pulse width of 185us. With 300us the energy in the coil will be 600uJ, and with 185us the energy is 536uJ.

So R10 on the Tx circuit its kind of current regulator,it determines how much current the coil will draw??The difficult is to achive more current without spikes so the mosfet wont go avalance..More current=more strength on the magnetic field created??

I've just finished a simple 36cm OD and man thats hard to keep it rigid...I wound 18 turns of 0.4mm double enameld wire,stabilised with electrical tape,then plastic spiral wrap and then another layer of electric tape..Still did not figure out how to make it rigid,tommorow i plan go to a friends router and engrave a circular pattern on wood,stabilise it in there end then epoxy it...

Wind spiral wrap on top of the coil before adding the electrostatic shield.

I wont dare the shielding at the moment but i thing i will have to because the inductance read about 400uH,shielding will lower that right?

[mushaba]

Scratch that about DC...capacitor is a barrier for DC...but only DC makes sense because if the main sample and secondary was ac waveforms should compare at the same time and cancel each other...