Geotech Baracuda - step 7 stucked

[mehdi394]

Hi everyone.
I'm building geotech baracuda, managed to get to step 7 but when measuring pin 1(u7), the voltage oscillates and never stops, the min and max has around 0.4 volt differences.
please tell me what I did wrong, I attached images of signal on pin4,6 of U4 and signal across R1 (damping resistor).
when I place target near the center of the coil nothing changes pint 1 of U7(except the oscillation I mentioned).

I adjusted minimum sample delay with oscilloscope but when I wanted to check it with multi meter , like step 7 its value its nothing close to -2.1 volts.
* coil : 440 uH
* pl2 pin 2 shorted to pin1.
* double checked the component values.
ps: not silverdog's board.I built it myself but double checked it in case I made a mistake. nothing found.

thanks

[waltr]

What is the scope's time base in your images??

I would like to see signal at output of pre-amp, U2 pin 6. Then at junction of R16, C15 & D7.

Also to see the timing relation of U4 pin 4 (TX pulse) to U4 pin 10 (first sample) then U4 pin 12 (second sample).

Next is to check that the two sample pulses actually turn on the FETs Q2 & Q3 (FET pinouts can vary so double check these).

Did you double check Voltage on each IC is correct?

[kt315]

I attached images of signal on pin4,6 of U4
--
probably you do not see that CMOS power is -5V and 0V/zero. so hight level of your o-scope signal must be zero, low level must be -5V.

[mehdi394]

thank you waltr ,
I found the problem based on test points you mentioned, one of the FETs was not working, I replaced it and now It works fine.
I completed the whole circuit and when I place a target near the center of the coil its voltage rises and speaker that I added starts to make sound,
however performance is very poor and also still I get voltage variations on U7 pin1, and thus I get noise from speaker sometimes even when not anything is around the coil.
changes on pin1 U7 does not have any pattern and its like a noise,
how can I get rid of this noise and have stable detector?
what about the performance ,it detects a pliers in 10 cm with a 15cm coil.
how can I improve the performance?
If any signal from scope is needed I will provide as soon as possible.
thanks

[mehdi394]

probably you do not see that CMOS power is -5V and 0V/zero. so hight level of your o-scope signal must be zero, low level must be -5V.

actually its just that the scope is on ac mode not dc,
no problem with biases.
thanks

[waltr]

thank you waltr ,
I found the problem based on test points you mentioned, one of the FETs was not working, I replaced it and now It works fine.
I completed the whole circuit and when I place a target near the center of the coil its voltage rises and speaker that I added starts to make sound,
however performance is very poor and also still I get voltage variations on U7 pin1, and thus I get noise from speaker sometimes even when not anything is around the coil.
changes on pin1 U7 does not have any pattern and its like a noise,
how can I get rid of this noise and have stable detector?
what about the performance ,it detects a pliers in 10 cm with a 15cm coil.
how can I improve the performance?
If any signal from scope is needed I will provide as soon as possible.
thanks

Ok on bad FET.

Are you testing this inside your house with LOTS of AC MAINS around?? I assume so based on your using an O'scope.
AC Mains and other electronic devices in your house can cause all sorts of interference and this is probably what you see & hear.
Give it a short test many 100's of meters away from ALL AC MAIN power.

The common solution is tuning the Pulse rate frequency to minimize the AC mains interference.

Sensitivity seems not great.
Have you tested with coins?
A 15cm coil is a bit small and distance is proportional to coil size.

Possible the AC mains noise is part of issue. Check away from AC.
Have you tried different sampling delay times?

[mehdi394]

Ok on bad FET.

Are you testing this inside your house with LOTS of AC MAINS around?? I assume so based on your using an O'scope.
AC Mains and other electronic devices in your house can cause all sorts of interference and this is probably what you see & hear.
Give it a short test many 100's of meters away from ALL AC MAIN power.

The common solution is tuning the Pulse rate frequency to minimize the AC mains interference.

Sensitivity seems not great.
Have you tested with coins?
A 15cm coil is a bit small and distance is proportional to coil size.

Possible the AC mains noise is part of issue. Check away from AC.
Have you tried different sampling delay times?

Thanks again for your tips.
Actually I was playing with sample delay, and by changing R14 the pulse width changed not delay. I EXPECTED that the start of the pulse move to right or lefy but what I see is pulse width changes.
MD works fine but not with very good performance so why is rhat happening. Did I miss something or confused pulse width with delay?
This is the clip I got from U4 pin 4 and 6. Notice how pulse width changes.
Any insight where I got wrong?
https://youtu.be/qsc0P6q-L4Y

[Qiaozhi]

Thanks again for your tips.
Actually I was playing with sample delay, and by changing R14 the pulse width changed not delay. I EXPECTED that the start of the pulse move to right or lefy but what I see is pulse width changes.
MD works fine but not with very good performance so why is rhat happening. Did I miss something or confused pulse width with delay?
This is the clip I got from U4 pin 4 and 6. Notice how pulse width changes.
Any insight where I got wrong?
https://youtu.be/qsc0P6q-L4Y

You've got yourself confused about the way this circuit works.

The TX pulse is delayed by C7, R14 and the external pot. Adjusting R14 and the external pot changes this [sample] delay. The sample delay pulse is then inverted by U4c, and further delayed by C12 and R15, before being inverted a second time (by U4e) to create the main sample pulse that drives Q3. The inverted sample delay pulse (from U4c pin6) also goes to C14 and R17 to create an additional delay for the EFE pulse. This is then inverted by U4d and fed to C13 and R21 to generate a delay, which is then inverted by U4f to create the EFE pulse that drives U2. This is all shown quite clearly in Fig.9 of the Build Document.

You can also see in your video that reducing R14 to too low a value will reduce the amplitude of the main sample delay so that the main and EFE pulses stop working. Hence the reason for the presence of R14, to set the minimum sample delay.

[mehdi394]

You've got yourself confused about the way this circuit works.

The TX pulse is delayed by C7, R14 and the external pot. Adjusting R14 and the external pot changes this [sample] delay. The sample delay pulse is then inverted by U4c, and further delayed by C12 and R15, before being inverted a second time (by U4e) to create the main sample pulse that drives Q3. The inverted sample delay pulse (from U4c pin6) also goes to C14 and R17 to create an additional delay for the EFE pulse. This is then inverted by U4d and fed to C13 and R21 to generate a delay, which is then inverted by U4f to create the EFE pulse that drives U2. This is all shown quite clearly in Fig.9 of the Build Document.

You can also see in your video that reducing R14 to too low a value will reduce the amplitude of the main sample delay so that the main and EFE pulses stop working. Hence the reason for the presence of R14, to set the minimum sample delay.

Thank you so much, now I have better understanding how the circuit works, my electronic knowledge is rusty, its been near 10 years I've been away from electronic projects, so sorry if my questions seems naive.
one other question, voltage on pin1 U7 changes, I eliminated any probable source of noise(ac mains) but still hopping in range of about 0.2 or 0.3 volts. Is it normal? if not what part of the circuit should I investigate? preamp or Sampling Integritor?
do you have any idea what may be the source of this voltage hopping?

[waltr]

one other question, voltage on pin1 U7 changes, I eliminated any probable source of noise(ac mains) but still hopping in range of about 0.2 or 0.3 volts. Is it normal? if not what part of the circuit should I investigate? preamp or Sampling Integritor?
do you have any idea what may be the source of this voltage hopping?

The output of the integrator will change when each of the sampling switches close. This will happen twice on each TX cycle.
Use one channel of scope on output of pre-amp and trigger this channel. Use the second scope channel to look at the integrator output.
Set the scope time base to see a full TX cycle.
Is the Voltage at integrator output changing with sampling switches? Or is there a random Voltage change?
Can you post a picture of this?

Are the R's & C'c on the integrator amp matched to the same values? R26 & R27 should be matched as well as C10 & C11, and R20 & R19.

If you are in a house then there will be AC mains noise and possibly other noise sources.
Did you check detection distance AWAY from the house?
If it is Noise then away for source should allow you to set the threshold much lower which increases detection distance.

[waltr]

The output of the integrator will change when each of the sampling switches close. This will happen twice on each TX cycle.
Use one channel of scope on output of pre-amp and trigger this channel. Use the second scope channel to look at the integrator output.
Set the scope time base to see a full TX cycle.
Is the Voltage at integrator output changing with sampling switches? Or is there a random Voltage change?
Can you post a picture of this?

Are the R's & C'c on the integrator amp matched to the same values? R26 & R27 should be matched as well as C10 & C11, and R20 & R19.

If you are in a house then there will be AC mains noise and possibly other noise sources.
Did you check detection distance AWAY from the house?
If it is Noise then away for source should allow you to set the threshold much lower which increases detection distance.

Another thing you can try is to add a few Ohms of resistance between the coil and TX pulse MOSFET (Q1). This can help stabilize the receive circuits by decreasing the current spikes on the board. I put 10 Ohms in series on my Hammer Head and it did not decrease detection distance.
The other thing the series R does is decrease the coil change time as per TC = L/R allowing the current to get to a steady state before switching off the coil.