TGSL continues beeping (problem with lm308)

[Victor1406]

Hi, a few weeks ago I finished my TGSL.
The problem is this: the first time I turned it on, it didn't make a sound. I noticed that when I removed the LM308 (U105), the detector was working but no more than 5 cm.
Then I changed the LM308N (these were also heated) for some NE5534.
Now the problem is that it continues to sound, when an object passes, it stops sounding. But it sounds again, that is, it works the other way around.




The Tx coil measures 6mH and Rx 6.5mH. I have measured the frequency of Tx with my multimeter and it gives me 14.5kHz perfectly. I changed Tx to Rx and I get about 13.9kHz according to the manual [When I measure Rx it gives me 200-300Hz below the Tx frequency (14.5kHz)].




The minimum I achieved on pin 7 of the LF353 was 200mV, it's high, but I don't know how to lower this. I have read that pin 7 of the LM393 (U102b) should be close to 0, but it gives me -2.5v.




I checked the voltages on all the integrated circuits (approximately -6v and 8v).
I checked all connections.
I have LM358P, I'm not sure if this last letter can be a problem, since I read that the LM308 works but LM308N does not.
I have the JFET BF245A, some say it works just like the BF245C, but in the Russian forums I read that it does not.
I have measured the current and I have about 80 mA while it sounds. If I remove the U105, the beep stops, the current drops 50 mA and only detects less than 5 cm.
Also, the L7808CV warms up, and I don't think I should, as it is 1.5A.




Can I measure Tx amplitude directly with the multimeter? I should be at 16v, but with the multimeter I get 15mV.




I don't know what to do anymore I guess it's an IC thing, but I don't know which one. Are there voltages I can compare?

[waltr]

Without an O'scope this will be very difficult to troubleshoot.
However, U104 and U105's output offset is critical for proper operation. Measure the output Voltage of these op-amps. They should be VERY close to Zero mV.

Since LM308's are hard to find and expensive I will guess the ones you have are Fakes. An NE5534 is not the correct op-amp since it has a large output offset.

Try some TL061, TL071 or TL081 for U104 & U105.

My TGSL draws about 30mA without beeping. Your 50mA may be due to the NE5534 but doesn't seem to high.

The biggest issue with the different JFETs is there is three different pin outs. You must insure what the pin out of the JFETs you are using then install them correctly.


200mV peak-peak on U101 pin 7 is good. With properly 'nulled' coils this can be over 1Vpp. You really need an O'scope to check this and to see the TX to RX (U101-7) phase shift.

Try to beg, borrow or steal an O'scope.

[Nandor]

What does approximately mean to you? You have to have at least -6.20 volt for proper opperation! Try low ESR capacitors an schottky diodes for the negative voltage converter so you can achive -6.50 volts.
If your regulator warms up there is a short circuit between the negative and the positive rails.

Regards

[Victor1406]

Without an O'scope this will be very difficult to troubleshoot.
However, U104 and U105's output offset is critical for proper operation. Measure the output Voltage of these op-amps. They should be VERY close to Zero mV.

Since LM308's are hard to find and expensive I will guess the ones you have are Fakes. An NE5534 is not the correct op-amp since it has a large output offset.

Try some TL061, TL071 or TL081 for U104 & U105.

My TGSL draws about 30mA without beeping. Your 50mA may be due to the NE5534 but doesn't seem to high.

The biggest issue with the different JFETs is there is three different pin outs. You must insure what the pin out of the JFETs you are using then install them correctly.


200mV peak-peak on U101 pin 7 is good. With properly 'nulled' coils this can be over 1Vpp. You really need an O'scope to check this and to see the TX to RX (U101-7) phase shift.

Try to beg, borrow or steal an O'scope.

The output of U104 gives me 230mV and the output of U105 gives 195mV, this voltage increases or decreases about 10mV if I change the NE5534 for another equal to this one, but they are always between 200mV. I understand that the detector sounds because both are positive at the same time.

When I touch the unconnected GB potentiometer pin with my hand, it stops sounding. Also when I touch pin 1 of the switch.

I have also checked the pins of the BF245A and they are correct, just like the Bf245C 1,2,3=G,S,D.

I measured the zero of the coils and they were 20mV, not 200mV (I connected an oscilloscope with the sound card of the cell phone, it seems to work well for this).

[Victor1406]

What does approximately mean to you? You have to have at least -6.20 volt for proper opperation! Try low ESR capacitors an schottky diodes for the negative voltage converter so you can achive -6.50 volts.
If your regulator warms up there is a short circuit between the negative and the positive rails.

Regards

I measured and got -6.4v at the negative voltage source, this with all ICs in place.
I have also measured ESR on the capacitors.
I have checked the whole circuit several times, and I don't find any short. The current without the NE5534 is 35mA, I will try to buy another LM7808.
I have also made another pcb, but it has the same problem, so I think it can be some integrated, LM358P, LM393 or LF353 perhaps. I changed the LM393(u102) and now I get -3.4v on pin 7.

Now I don't care about discrimination anymore, I just want it to work at a decent distance.

[waltr]

The threshold into the comparators (U106) is only about 0 to 36mV so this is why it beeps.
The NE5534 has way TOO HIGH of an offset which is why it is NOT recommended. It could work IF you add circuit to adjust offset of the NE5534 but this adds about 4 parts to each one so not easy to add mod and it may drift so more trouble than its worth.
Just get some TL0x1 op amps.

Everything else sounds OK.

Another thing that is important due to the high impedance and gains is to properly Clean All flux from the PCB.
Flux can conduct enough to make this circuit not work correctly.

[Victor1406]

The threshold into the comparators (U106) is only about 0 to 36mV so this is why it beeps.
The NE5534 has way TOO HIGH of an offset which is why it is NOT recommended. It could work IF you add circuit to adjust offset of the NE5534 but this adds about 4 parts to each one so not easy to add mod and it may drift so more trouble than its worth.
Just get some TL0x1 op amps.

Everything else sounds OK.

Another thing that is important due to the high impedance and gains is to properly Clean All flux from the PCB.
Flux can conduct enough to make this circuit not work correctly.

Thanks Waltr, I didn't think to change the reference. What I did was change R39 220k for a 15k, so I get a voltage of 0 to 500mV, so I regulate the potentiometer just before reaching 195mV and the detector stops beeping.
The distance improved a lot, but it could be better. How can I improve this? Does it depend on the Tx and Rx capacitors?

The outputs of U104 and U105 should have the same value (suppose 200mV)? maybe I can put two sensitivity potentiometers for each NE5534, ie one on pin 6 of the LM393 (U106b) and another on pin 2 of the LM393 (U106a). Although I don't know what this would do.


Now I have no discrimination, on what do the 20 degrees of phase depend? can I change something?


Anyway, I don't care too much about discrimination, since it's my first detector and I'll settle for this.

[waltr]

Ok, not the best solution but at least you are understanding how the circuit works.

The 20 degrees is between the TX waveform and the RX coils waveform (on C6).

I found this is not as important as the relationship between the GB sampling pulse at TR5 gate, and the RX output on U101 pin 7.
Put the scope on these two points and set the GB pot to center and switch to AM mode.
Adjust the coil 'null' so that U101 output's ZERO crossing is centered on the GB pulse.

Now use a ferrite and silver targets and adjust the GB pot (only slightly should be needed) so ferrite is barely detected and silver is strongly detected.
Disc should now work and should have best distance.

[Nandor]

Hello Victor

Can you post some high resolution pictures of your PCB? It would be easier to us to find the problem.

regards

[Victor1406]

Hello, I have separated pins 2 and 6 of the LM393(U106), to each one I assigned a potentiometer that regulates from 0 to 500mV. Since I have different voltages in the outputs of the NE5534 (230mV and 195mV).


As Waltr said the problem is in the NE5534. So I'll buy some TL0x1. Unfortunately I don't have an oscilloscope, so I could only put pin 7 of the U101 at 5-10mV with the multimeter. But this turned out well, in AM, I have adjusted GB until the ferrite is not visible. Silver detects it well.


Is it normal that when touching the GB cable it produces a beep?
The distance while he's still is good, but when I move him I have to adjust the sensitivity and it decreases a lot.
I have discrimination, what I don't understand is how to control the potentiometer... if I turn it to the left it doesn't detect iron, if I put it in the middle it doesn't detect copper.

[dbanner]

You are going down the proverbial "rabbit's hole".
You must assemble the PCB with the correct parts as indicated in the schematic of the build document.
Only then can you make a good assessment of what is right and what is wrong with your circuit. Otherwise you go around in endless circle. Like a dog chasing its tale.

[waltr]

Hello, I have separated pins 2 and 6 of the LM393(U106), to each one I assigned a potentiometer that regulates from 0 to 500mV. Since I have different voltages in the outputs of the NE5534 (230mV and 195mV).


As Waltr said the problem is in the NE5534. So I'll buy some TL0x1. Unfortunately I don't have an oscilloscope, so I could only put pin 7 of the U101 at 5-10mV with the multimeter. But this turned out well, in AM, I have adjusted GB until the ferrite is not visible. Silver detects it well.


Is it normal that when touching the GB cable it produces a beep?
The distance while he's still is good, but when I move him I have to adjust the sensitivity and it decreases a lot.
I have discrimination, what I don't understand is how to control the potentiometer... if I turn it to the left it doesn't detect iron, if I put it in the middle it doesn't detect copper.

Ok work around until you get the TLx01s.

Yes, ANY MOVEMENT in the coil and wires will produce a beep.
This circuit is very sensitive to be able to detect a small coin in the ground.

With switch in Disc mode, the disc pot should reject small iron at minimum. Then as it is increased first is rejected small aluminum foil (cut a piece about 1cm square), then small nickle coins, then larger foil, then gold ring, last is silver.
Very possible the Disc pot is wired backwards (I did that) so check with those target from each end of pot.

Also possible that coil null/tuning still not correct or RX coil connected backwards.
Do you a single or double beep with a coin about 10cm from coil? If yes RX swap RX coil connections and Re-tune coil null and GB.

[Victor1406]

Hello, I have bought some TL071CN, I have placed but the problem remains, as with the LM308N, these are heated and the current increases to 90mA. The output of U104 slowly drops from 1v to 0v when I measure. The output of U105 is kept at -5v.(as with LM308N) Obviously it does not produce any sound.


I don't know what's wrong, the NE5534 doesn't heat up and the current doesn't increase as much as with the TL071CN or LM308N. So I guess it's not a connection problem.


Can it be a problem with the TL071CN?

[kt315]

Can it be a problem with the TL071CN?
----
esspecial opamp tester was projected for geotech1ers. all problem with all kinds of opamps will be fixed in a second.
believe me now.

album https://www.geotech1.com/forums/album.php?albumid=109

[waltr]

Hello, I have bought some TL071CN, I have placed but the problem remains, as with the LM308N, these are heated and the current increases to 90mA. The output of U104 slowly drops from 1v to 0v when I measure. The output of U105 is kept at -5v.(as with LM308N) Obviously it does not produce any sound.


I don't know what's wrong, the NE5534 doesn't heat up and the current doesn't increase as much as with the TL071CN or LM308N. So I guess it's not a connection problem.


Can it be a problem with the TL071CN?

Possibly you installed them back-wards??
Double check since this is not hard to do. I did this on one opamp building my TGSL, current draw high and opamp hot with wrong output Voltage.