Why like this?

[6666]

Back to the original question...

Yes, they are all PIs but their designs are different. The SurfPI and Barracuda are most similar. They have the exact same overall stage designs with only small differences in gains and taus. So it makes sense they perform about the same.

The MPP is quite different. First, it uses an N-drive transmitter instead of P-drive and has a 2-stage preamp instead of 1-stage. Neither of these should make much of a difference and arguably should allow for earlier sampling. The integrator stage is considerably different and appears to have lower gain than the SurfPI/Barracuda. The subsequent gain stage is also lower. Finally, the MPP audio is produced from a 555 that is part of the signal path, not simply a chopper applied to the signal. I don't know how this affects the overall signal gain.

I strongly suspect that a different coil is not the path to improving the MPP. Rather, focus on the integrator & gain stages. If you scope the preamp outputs you should be able to determine whether the 3 designs have roughly equivalent sensitivity at that point and I suspect they do. Then move to the integrator stage and repeat.

As to why the 3.2 coil is better on the MPP and a wash on the others, it may be that the MPP samples earlier. I think the MPP goes down to 10us whereas the others are 15us. Or because the MPP has a 400V FET instead of 200V on the others. There are various possibilities.

Its interesting that you mention the gain in the gain stages , how does the gain in your HH compare to SurfPI/Barracuda, would the HH be more or less or roughly equal ?

[green]

Maybe C12 R24 MPP, C3 R7 simulation is causing the problem.

No comments on simulation https://www.geotech1.com/forums/atta...&d=1602094352? Looks like MPP would be better if C12(MPP) C3(simulation) was shorted and an offset pot was added to zero amplifier. What am I missing?

[billr]

Thanks Carl
I checked and re checked the scope readings to the manual and all seems to be in spec. Adjusted a little here and there to no avail.
I wasn't too concerned about the MPP having less performance then the others, perhaps a little disappointment it was not better by a mile.
Have managed to increase the MPP to 13" on a new coil. It's OK at that. I'm sure others have far better performance than I do.
For now I am going to box up the other two and use them in battle.

[green]

Maybe C12 R24 MPP, C3 R7 simulation is causing the problem.

[Carl-NC]

Why the huge difference between both types of detectors and their preference for coil match? Aren't they all PI's?

If anyone has a simple mono coil that does have good range on the MPP can you tell me what it is, size of wire and how many turns etc.
Making the coils for the first two units was simple but the Beach is proving to be more demanding.

Back to the original question...

Yes, they are all PIs but their designs are different. The SurfPI and Barracuda are most similar. They have the exact same overall stage designs with only small differences in gains and taus. So it makes sense they perform about the same.

The MPP is quite different. First, it uses an N-drive transmitter instead of P-drive and has a 2-stage preamp instead of 1-stage. Neither of these should make much of a difference and arguably should allow for earlier sampling. The integrator stage is considerably different and appears to have lower gain than the SurfPI/Barracuda. The subsequent gain stage is also lower. Finally, the MPP audio is produced from a 555 that is part of the signal path, not simply a chopper applied to the signal. I don't know how this affects the overall signal gain.

I strongly suspect that a different coil is not the path to improving the MPP. Rather, focus on the integrator & gain stages. If you scope the preamp outputs you should be able to determine whether the 3 designs have roughly equivalent sensitivity at that point and I suspect they do. Then move to the integrator stage and repeat.

As to why the 3.2 coil is better on the MPP and a wash on the others, it may be that the MPP samples earlier. I think the MPP goes down to 10us whereas the others are 15us. Or because the MPP has a 400V FET instead of 200V on the others. There are various possibilities.

[surfdetector]

When I did my testing with/with out series resistance. I measured the DC Voltage out of the integrator.
With no capacitive coupling the Voltage out is DC. Then with various targets and different distances.
Results were very clear that some series resistance did not decrease detection distance and with some target the distance increased.

I'll go with Carl's explanation on why this is so.

Waltr regarding the addition of series resistance, is it generally 10 ohms that you insert?

[green]

Simulations do not always do what real circuits do. I have a feeling it does not properly simulate the target eddy currents. I have never gotten any of the target simulations that have been posted to match real circuit testing.

US nickel is easy, simulation and real target straight line decay linear log(TC=10us). US quarter harder. https://www.geotech1.com/forums/show...210#post258210 not exact but close. Does a nickel or quarter detect at a longer distance with added coil resistance with your detector. If not what target does?

[waltr]

Couple reasons higher coil resistance could cause higher signal. Circuit doesn't like higher Tx current for some reason(simulation doesn't care). Higher coil current takes longer to decay, not enough delay time(sampling when amplifier saturated). Any others?

Simulations do not always do what real circuits do. I have a feeling it does not properly simulate the target eddy currents. I have never gotten any of the target simulations that have been posted to match real circuit testing.

Would that be why at 14.5" I can still get a strong signal but have to wait a moment before passing the coin back in the opposite direction? Up to 14" when a coin is passed back and forth fairly rapidly the signal is made. Above that distance there is a lag, almost like something needs to empty before it sees the coin again. God. I'm so fluent at technical jargon. I sometimes amaze myself.

This 'lag' sound like to may be in the later circuit stages, like the integrators and/or SAT. These stages tend to have high gain plus largish value caps.

When I did my testing with/with out series resistance. I measured the DC Voltage out of the integrator.
With no capacitive coupling the Voltage out is DC. Then with various targets and different distances.
Results were very clear that some series resistance did not decrease detection distance and with some target the distance increased.

I'll go with Carl's explanation on why this is so.

[billr]

6666
They are just normal mono coils 9.5 to 10.5 inches with some slight variation in design to increase or decrease resistance. Cant remember any details. Have made so many since the start of this year. I grab one and see how it does. If it sucks I try another one. Mine are very average. Others make far better than I do.

[6666]

green . US Nickle
Surf . Yes stable
6666 As per the coil calc on this forum. Home made. None
I'm just curious as to why two units react similarly and yet the third is so much different to the same coils when it is of the same genre. Was looking for a proven coil build for the MPP with sizes dims etc to help ascertain what the difference between the three units where. This is the best way for me to learn.
I can develop a suitable coil to match the others in performance but the question of "why is the MPP so different in coil choice compared to its relatives " still remains. and why do other proven coils read less when connected up.
Call me curious. I like to learn by asking questions on even the simplest of subjects.

Ok so you used the forum coil calculator, can you at least tell us what radius you used for both coils, how many turns and the wire thickness thanks
and what construction they both are ? bundle wind etc? thanks

[billr]

Higher coil current takes longer to decay, not enough delay time(sampling when amplifier saturated)

Would that be why at 14.5" I can still get a strong signal but have to wait a moment before passing the coin back in the opposite direction? Up to 14" when a coin is passed back and forth fairly rapidly the signal is made. Above that distance there is a lag, almost like something needs to empty before it sees the coin again. God. I'm so fluent at technical jargon. I sometimes amaze myself.

[green]

Couple reasons higher coil resistance could cause higher signal. Circuit doesn't like higher Tx current for some reason(simulation doesn't care). Higher coil current takes longer to decay, not enough delay time(sampling when amplifier saturated). Any others?

[green]

Labeled R7 L6 quarter, not correct. Look at Rx for quarter decay, Quarter is combination of the three RL circuits.

Quarter simulation https://www.geotech1.com/forums/show...210#post258210

[green]

Increased PS to 12V. Tx on time 300us(current flat tops with higher coil resistance). Lower coil resistance still higher signal.

[Altra]

Green, Try 12 to 14v on your TX. The TDI, GS5 and Goldquest all used current limiting and a higher supply voltage.