Reverse Engineering a pulse induction metal detector...... Help required

[haseeb zaib]

Hey guys

I am reverse engineering this PI metal detector but I am stuck in this part of the circuit. I am calling it 4 stage amplifier but mainly this is the portion where all the sampling is being done. Before getting to a part where I am stuck, let me explain a circuit a bit, upper side of schematic we have reset circuit with lm393 as a comparator and 4063 is a multiplexer. now we get to main part of IC tl064, The first stage U1A is mainly responsible for signal coming from the front end section of the schematic which is signal from the pre-amp(oscilloscope output attached). according to my understanding the U1B is just inverting the signal back to its orginal shape, but here is where the problem occurs the signal coming out from switch (CD4066) U2A and U2D are complete gibrish and I dont understand it and frankly after that I dont get what this circuit is doing, I am completely lost. If anyone can give me some suggestion or any advice will be really help full thanks. And output of U1D is going straight into audio circuit.

All the respective oscilloscope outputs and schematic are attached

https://imgur.com/gallery/l9aGZnF

[kt315]

i say just my opinion... it's idiotic concept be cause every switch in high sensitive circuit has long wires.
long wires give a noise. this men did want to get low noise but get big noise.

[Qiaozhi]

Hey guys

I am reverse engineering this PI metal detector but I am stuck in this part of the circuit. I am calling it 4 stage amplifier but mainly this is the portion where all the sampling is being done. Before getting to a part where I am stuck, let me explain a circuit a bit, upper side of schematic we have reset circuit with lm393 as a comparator and 4063 is a multiplexer. now we get to main part of IC tl064, The first stage U1A is mainly responsible for signal coming from the front end section of the schematic which is signal from the pre-amp(oscilloscope output attached). according to my understanding the U1B is just inverting the signal back to its orginal shape, but here is where the problem occurs the signal coming out from switch (CD4066) U2A and U2D are complete gibrish and I dont understand it and frankly after that I dont get what this circuit is doing, I am completely lost. If anyone can give me some suggestion or any advice will be really help full thanks. And output of U1D is going straight into audio circuit.

All the respective oscilloscope outputs and schematic are attached

https://imgur.com/gallery/l9aGZnF

Please note that you only need to post your question in one location, not make multiple [identical] posts in different areas of the forum.
I have removed your other thread on the Schematics forum.

[haseeb zaib]

sorry

[Qiaozhi]

sorry

Since you have also made at least one post on the forums, you should now be able to attach images, schematics, etc., rather than hosting them on an external site. This will make the attachments permanent, unlike 3rd party sites where uploads tend to disappear after some time.
Click on the Go Advanced button and then Manage Attachments.

Also, it would be helpful if you stated which PI detector you are trying to back-engineer.

[haseeb zaib]

.

[haseeb zaib]

Since you have also made at least one post on the forums, you should now be able to attach images, schematics, etc., rather than hosting them on an external site. This will make the attachments permanent, unlike 3rd party sites where uploads tend to disappear after some time.
Click on the Go Advanced button and then Manage Attachments.

Also, it would be helpful if you stated which PI detector you are trying to back-engineer.

this is a blackdog metal detector xtreme

[waltr]

and please post the schematic here. I never like using external sources.

[Carl-NC]

Here is the schematic:



I understand the circuit except for Q4. Are you sure this part of the circuit is correct? Also confused about the timing, is that 200us/div? That would make the TX width = 100us which sounds right but the RX delay is 800us which does not.

I need to draw up a simplified schematic to explain what is going on, will try to do that later today.

[haseeb zaib]

Here is the schematic:

[ATTACH]55723[/ATTACH]

I understand the circuit except for Q4. Are you sure this part of the circuit is correct? Also confused about the timing, is that 200us/div? That would make the TX width = 100us which sounds right but the RX delay is 800us which does not.

I need to draw up a simplified schematic to explain what is going on, will try to do that later today.

yup thats the thing I am also confused about the part where transistor is used, I have checked it like 5 times already and that formation is used in metal detector. And Rx delay is not 800us.

[Carl-NC]

I've redrawn the schematic to simplify it; non-critical elements removed:



First, I assume there is an input resistor to U1a; I've labeled it Rin. U1a and inverting amp U1b drive the sampling switches for the main sample and ground sample. U1b typically has gain = -1; if there is no EFE compensation (might not be needed since this uses a 1m loop) then U1b might have some gain needed for GB.

The sampling feedback (R3,C2) is basically a DC restore. However, if it is sampling during the TX pulse then I don't know how it could work. You would typically sample during the main sample or just before the TX pulse.

The feedback on U1c is still a mystery to me. This should be the integrator so somewhere I expect to see a feedback cap. As drawn, the feedback appears as if it will rail out the opamp.

U10 is retune circuit. I tossed out all the switch debounce and now U11 is represented by a push-button (I assume S1 in the original schematic is a user switch). The divider R42,R43 provides a reference voltage and when you press the button the output of U1c is forced to equal the reference voltage. In the process C19 is charged to some voltage and when you release the button C19 holds the output offset of U1c at the reference level. This kind of circuit was widely used in TR-Disc detectors in the 1970s. In this case, it replaces the SAT circuit usually seen in PI designs.

I will poke some more at the U1c feedback but honestly it just looks wrong.

[haseeb zaib]

I've redrawn the schematic to simplify it; non-critical elements removed:



First, I assume there is an input resistor to U1a; I've labeled it Rin. U1a and inverting amp U1b drive the sampling switches for the main sample and ground sample. U1b typically has gain = -1; if there is no EFE compensation (might not be needed since this uses a 1m loop) then U1b might have some gain needed for GB.

The sampling feedback (R3,C2) is basically a DC restore. However, if it is sampling during the TX pulse then I don't know how it could work. You would typically sample during the main sample or just before the TX pulse.

The feedback on U1c is still a mystery to me. This should be the integrator so somewhere I expect to see a feedback cap. As drawn, the feedback appears as if it will rail out the opamp.

U10 is retune circuit. I tossed out all the switch debounce and now U11 is represented by a push-button (I assume S1 in the original schematic is a user switch). The divider R42,R43 provides a reference voltage and when you press the button the output of U1c is forced to equal the reference voltage. In the process C19 is charged to some voltage and when you release the button C19 holds the output offset of U1c at the reference level. This kind of circuit was widely used in TR-Disc detectors in the 1970s. In this case, it replaces the SAT circuit usually seen in PI designs.

I will poke some more at the U1c feedback but honestly it just looks wrong.

[ATTACH]55729[/ATTACH]

Thanks really helpful. I appreciate u took a time in your day to redraw this all really appreciate.