New Project - MPS PI - Iron discriminating metal detector.

[Op04]

G'day all
Looking for proof readers Just drew up 1.22.1 in Kicad ver 6 using Pi4B running latest Bookworm OS
please find attached pdf of cct drawing Would verification the drawing is correct in detail before releasing the sch file
It's been a learning curve so far haven't reached the PCB generation bit

Wrong. There's no -5v supply

[Detectorist#1]

Hi GeoMax,
I will try to explain how works the discriminator from idea from Sascho from year 2008. First, this schematic haves only DISK channel. Also the sample from RX coil is taken during TX pulse like DFBowers suggests. Also DFbowers suggests for duration of RX sample to be in range 20us-30us as starting point ( see post #10). In this schematic both condition are satisficed. The duration of RX sample is appr. 25us and the RX sample is taken during TX pulse. Because as RX pulse is used the TX pulse - the duration of TX pulse is also 25us. Further, R13 and C10 works as first integrator. R18 and C13 works as second integrator. As attached picture from DFBowers, for ferrous targets the signal of the output of the second integrator will be more negative and for non-ferrous targets will be more positive. Here is very important to set working point (output voltage of CA3130) by P2 and P4. Without target the level of the volume of the sound (700Hz) have to be set at middle level. At ferrous target the level of the sound will decrease and for non-ferrous target this level will increase. This schematic is far from perfection - but this is very early attempt to use IB coil in amateur developed PI MD for discriminate of targets.

[old_gingerbeer]

Wrong. There's no -5v supply

Fixed the -5v but it appears to be 2 independent +5v and -5v supplies each supplying VLF and PI sections! ...... Need to check against PCB layout which devices are grouped as A or B. Suspect this was done to simplify the layout to minimized topside tracks! Tried to have PCBs manufactured as samples A problem is the solder mask covers all the pads Solution is to have all the Cu tinned

[GeoMax]

Fixed the -5v but it appears to be 2 independent +5v and -5v supplies each supplying VLF and PI sections! ...... Need to check against PCB layout which devices are grouped as A or B. Suspect this was done to simplify the layout to minimized topside tracks! Tried to have PCBs manufactured as samples A problem is the solder mask covers all the pads Solution is to have all the Cu tinned

Hi Old_Gingerbeer,

the "Coil Pulse Generator" circuit part has also an error.

[Op04]

Fixed the -5v but it appears to be 2 independent +5v and -5v supplies each supplying VLF and PI sections! ...... Need to check against PCB layout which devices are grouped as A or B. Suspect this was done to simplify the layout to minimized topside tracks! Tried to have PCBs manufactured as samples A problem is the solder mask covers all the pads Solution is to have all the Cu tinned

Hi Old_Gingerbeer,
In case you going to make alternative PCB. I hope you make through hole version, my guess you are trying to make SMD PCB. I don't know why you replace CD40106. There's nothing wrong with original version, except we need a high quality plated through hole manufacturing especially on this project, since making all track tinned will make a little more time to be carefully clean

[Detectorist#1]

The change of 40106B with 74HC14 is deeply wrong. 74HC14 have max voltage supply of 6V. It is not possible to supply this circuit from +5V and -5V and in this case is not possible to make right control voltage levels of 4066 circuit. In this schematic have additional mistakes. "Old_gingerbeer" is too young for quick correcting of all mistakes!

[GeoMax]

Hi GeoMax,
I will try to explain how works the discriminator from idea from Sascho from year 2008. First, this schematic haves only DISK channel. Also the sample from RX coil is taken during TX pulse like DFBowers suggests. Also DFbowers suggests for duration of RX sample to be in range 20us-30us as starting point ( see post #10). In this schematic both condition are satisficed. The duration of RX sample is appr. 25us and the RX sample is taken during TX pulse. Because as RX pulse is used the TX pulse - the duration of TX pulse is also 25us. Further, R13 and C10 works as first integrator. R18 and C13 works as second integrator. As attached picture from DFBowers, for ferrous targets the signal of the output of the second integrator will be more negative and for non-ferrous targets will be more positive. Here is very important to set working point (output voltage of CA3130) by P2 and P4. Without target the level of the volume of the sound (700Hz) have to be set at middle level. At ferrous target the level of the sound will decrease and for non-ferrous target this level will increase. This schematic is far from perfection - but this is very early attempt to use IB coil in amateur developed PI MD for discriminate of targets.

Thanks for the explanations... This helps to undestand how it works.

[GeoMax]

Hi Old_Gingerbeer,

C35 & C36 have also wrong value...

[old_gingerbeer]

Hi Old_Gingerbeer,
In case you going to make alternative PCB. I hope you make through hole version, my guess you are trying to make SMD PCB. I don't know why you replace CD40106. There's nothing wrong with original version, except we need a high quality plated through hole manufacturing especially on this project, since making all track tinned will make a little more time to be carefully clean

Initially started with ver 1.07 but then carried on with 1.22.1 so some of the differences came up in the Kicad ver1.22.1 of the diagram There was no real intention to replace 40166 just could locate it in the symbol library use decades ago 74C14 but found the 74HC14 symbol easy to find Thanks to Detectorist#1 for the education regarding 74HC14
Thanks to all the proof reading Many hands make work light the current schematic has many corrections included Error on cap C35 & C36 was due to poor printing of the original cct by my cheap printer

[old_gingerbeer]

Hi Old_Gingerbeer,
In case you going to make alternative PCB. I hope you make through hole version, my guess you are trying to make SMD PCB. I don't know why you replace CD40106. There's nothing wrong with original version, except we need a high quality plated through hole manufacturing especially on this project, since making all track tinned will make a little more time to be carefully clean

Yes hole thru version is on the cards but prefer to leave the PCB as Dip version as 99% of homebrewers cannot handle SMDs but will leave it to someone else to do a SMD version

[Op04]

Yes hole thru version is on the cards but prefer to leave the PCB as Dip version as 99% of homebrewers cannot handle SMDs but will leave it to someone else to do a SMD version

[GeoMax]

Initially started with ver 1.07 but then carried on with 1.22.1 so some of the differences came up in the Kicad ver1.22.1 of the diagram There was no real intention to replace 40166 just could locate it in the symbol library use decades ago 74C14 but found the 74HC14 symbol easy to find Thanks to Detectorist#1 for the education regarding 74HC14
Thanks to all the proof reading Many hands make work light the current schematic has many corrections included Error on cap C35 & C36 was due to poor printing of the original cct by my cheap printer

Hi old_gingerbeer,
some correction were made.... but the error in the oscillator is still there in your diagram and other vital informations are missing there as well. In case you do not see your error and what vital informations you are missing in your schematic, check-out the schematic which I re-drew already last year.

[old_gingerbeer]

Hi old_gingerbeer,
some correction were made.... but the error in the oscillator is still there in your diagram and other vital informations are missing there as well. In case you do not see your error and what vital informations you are missing in your schematic, check-out the schematic which I re-drew already last year.

Never seen ver 1.23 the cct is based on the diagram 1.07 version however 1.22.1 that mirrors the gerber plots post here on this thread Noticed in your drawing there is only one +5v and -5v supplies were as the published PCB has two

[nejdet]

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

[GeoMax]

Hi old_gingerbeer,

that is right, the second power supply is obsolet. With other words... no need. I assigned the version number 1.23 to distingush between all the varieties only. I could have used also the version # 1.22.2 but I have skipped the 2nd power supply and assigned some component parts differently. That's why I choosed the version number as 1.23.