New Project - MPS PI - Iron discriminating metal detector.

[old_gingerbeer]

I also think that the power supply for the operational amplifiers (analog part) should be separate from that of the digital part.

Concur with this view There are issues with noise whilst detection small targets However there are some design improvements that can be made to improve detecting performance in adverse conditions Eg looking for gold specks in reactive ground or near power lines

[GeoMax]

Hi,

I do not think that you need to separate the "analogue" and "digital" circuit part of this circuit with two power supplies. Most important is a good PCB design and a well designed front-end circuit + a fast coil. If you want to improve the circuit, you should concentrate on that parts.

[algan]

If Carl's bi-polar TX were added to the scheme, it would be even better.

[Detectorist#1]

Hi GeoMax,
I agree with you - Power Supply Reduction Ratio (PSRR) of now days OpAmps is very good and no need for separate power supplies for analog parts of the schematic. Delay of first sample in this project is set from 50us to 100us. This nulling the efforts of well designed front-end circuit + a fast coil. We need to make a real tests with 1.21.1 version. It is interesting what will happened with Fe-Not Fe discrimination if the delay for first sample is reduced to 5us and duration of the samples from Rx signal will be reduced to 5us. This will increase possibilities for detecting of small nuggets. If discrimination capabilities will be seriously reduced, this new idea for hybrid PI detector with discrimination will be seriously compromented. BTW, do you have information how are the discriminations possibilities for Voodoo project for really small nuggets?

[Detectorist#1]

Hi algan,
Using of bipolar TX pulses in this idea will need for very big changes in the project.

[GeoMax]

Hi Algan,

Carl's bi-polar TX is nice, but way too complex for this circuit. As Detectoris#1 already wrote, quite a lot of changes to the circuit are needed. Carl's bi-polar TX circuit is probably better used with a microcontroller board.

[algan]

All you need to do is run the TX without a 4053 with some kind of mosfet driver, you only need two input pulses. Carl will tell you what parameters they are, can it be programmed further with a simple microchip, why does the circuit need to be changed further?

[nejdet]

With a little modification it could work.

[Detectorist#1]

Hi nejdet,
The modification of DFBowers' "MPS-pi-iron-discriminating-metal-detector" schematic for use of bipolar TX pulses is not only concerned with generating of these pulses - this will need modification of the DISK channel.

[Detectorist#1]

Hi eduardo1979,

I know for your publication of a schematic for hybrid PI-VLF detector from 20 July 2022. But no further information for the practical results of this schematic. Do you realized it in practice or this is only theoretical design? The idea sееms interesting.

[eduardo1979]

Hello.
It's true, the idea seems interesting, but I didn't get around to putting it into practice. I've been working on other projects and haven't had time to thoroughly investigate hybrid detectors. Right now I'm focusing on PIs with bipolar TX since I have Carl's book and I find this way of working quite interesting.

[Detectorist#1]

Hi eduardo1979,
Thank you for the quick answer. Yes, bipolar schematic in ITMD3 is good but main problem in PI MD is discrimination according me. I'm also very impressed of the ITMD3 book. But the idea of designing of hybrid PI-VLF detectors with more reliable discriminating is on the agenda for now days according me. There is a wide field for experiments for amateur's designs and not only. Yes, "the only sure discriminator is the shovel" but hope dies last.

[eduardo1979]

In general, PI detectors with Ground Balance function can discriminate very effectively between high and low conductivity metals. Hybrid detectors, on the other hand, can discriminate between ferrous and non-ferrous metals, but with certain limitations. For a hybrid detector to work well in all types of ground, it needs a ground balance control. I do not yet have enough experience with hybrid detectors, but perhaps in the future I can investigate this technology further if time permits.

[old_gingerbeer]

Hi All brewers
Been off internet for a while Here some further Kicad changes Fixed duplicated labels tided up original cct and created auto PCB route. Unfortunately as Elektor book on Kicad acknowledges the auto routing is only for simple PCBs this is a bit complex so requires routing to be done by a Hooman 2 diodes have been added for battery protection reversal protection so new version is 1.22.2. BTW as the site doesn't like .net extension therefore rename Hybrid vlf ver1.22.2 net.txt > Hybrid vlf ver1.22.2.net

[Detectorist#1]

Hi old_gingerbeer,

The schematic 1.22.2 is full of errors in comparison with 1.22.1.
1. Common connection of D3 and R19 have to be connected to the anode of D4
2. Left end of R21 have to be connected to the cathode of D17, not to the anode
3. The lower ends of C45, U14 and C46 have to be connected to GND, not to -5VA
4. Capacitor C2 have to be moved on the place of R6. R6 have to be removed.
5. The value of R7 have to be reduced from 100K to 10K
6. Common connection of C28 and R30 have to be connected to GND
7. The lower end of R30 have to be connected to pin4 of U9B
8. The left end of C47 have to be connected to pin7 of of U10B
9. The resistor R36 have to be removed
10. On the pin2 of RV4 have to added capacitor 1uF to GND
11. To left unused parts of U1 with floating inputs is not good idea. These parts is possible to switch unpredictable from capacitive connections with near parts of the board and this will add additional noise on the board