Got my TGSL 100x100 board working, but discrimination doesn't work

[yatahaze323]

Hello Everyone,

I've finally completed an almost fully working board of the TGSL 100x100 that I had posted awhile ago. Nandor had pointed out that my J107s might need to be reversed so after getting a cheap component tester to see what the pin outs were I tested the J107s and he was correct and they needed to be rotated to face the opposite direction.

I built my tx and rx coils and have them hooked up and I have them setup so I was able to test the ground balance to reject ferrite. Currently the area I'm in is electronically noisy but I'm able to detect a US quarter at about 8-9 inches however I'm not sure if I properly nulled the coil using the oscilloscope.

One issue I'm having is the discrimination doesn't seem to work at all, I can turn discrimination all the way up and it still is able to detect everything I wave over the coil. I checked my solder joints and they all look good I'm not quiet sure what could be causing this. Could this be because my coil isn't properly nulled?

Thanks everyone!

[kt315]

very soon i will get a first batch of PCBs of coil tester. everyone will have a chance to tune any coil. i promise.

[yatahaze323]

A coil tester sounds like a great idea! It would make building coils much easier, especially for people that are new to this hobby such as myself.

[ne555]

The problem is that building a pcb and coil as a forum plan does not give a good result

[dbanner]

There have been very many builds from scratch using plans posted here on forum. The TGS is tesoro circuit, we know it works very well. The most common problem is with construction of coils. The nulling of the coils and also the shield.
The coils must be made to correct inductances using correct wire gauge to get the appropriate resistances.
The shield must be done using preferably graphite. And correct hookup of shield must be adhered to.
The coils must be oriented correctly, observing star/ finish ends of wires.
And finally, the nulling performed with an oscilloscope.
If you pay careful attention to these details, you will have excellent working device.
Also using the recommended chips from the build document.
Also, If you use other smd layout may have small errors, use Silverado layout. TGSL EDU

[dbanner]

If you choose to use smd layout, or some other layout, I really think that you should check it against the schematic to make sure it is good layout without errors.

[ne555]

Built high precision concentric coil and DD and well shielded the result one was noticed that the transistor 2 n2907 I use a kind of plastic from the Wright transistor. Some of the member designs here were a metal type transistor that read the data****e for too transistor and found a lot of variation between the model, even though it is the same number

[yatahaze323]

The board I built was a slightly modified version of the one in the sticky post. The only difference is that I removed some extra copper so that the board would be 100x100mm.

For the coil I've been following the instructions that are in the sticky post as well. Currently both the RX and tx coils have been shielded but I haven't put them into a shell yet. I'm thinking that maybe after shielding my tx coil that I flipped it around so that the wire is now wrapped in the wrong direction. I'll have to try flipping it to see if that corrects the issue.

[dbanner]

After nulling, when coil is in wrong direction,(either the tx or Rx), you will observe the double beeping sound when a target is moved over coil. This is typical indication that you have reverse coil.
If after properly balanced coil you cannot get any discrim, this point to problem on pcb.
Assuming you have good balance coil, discrim should work good.
It is said by many that tesoro have one of the best discrimination separation.
Good luck with your project.
Another thing to check is oscillator frequency, it should be very close to 14.5khz. and also I put Rx on tx oscillator to check 14.9khz.
I do not have good inductance meter.

[Nandor]

Hi yatahaze.
If you have good ground balance it means that the coils are pretty ok, you should be able to balance the ground effects at almost middle position, this is important!
Connect the oscilloscope to your GEB channel and turn the GEB pot to maximum, it should shift it 45 degrees compared to the origo point! Now do this with the DISC channel and see if you turn the potentiometer how many degrees does it shift? It is a rounded square wave and it should shift it 90 degrees compared to the origo point, relative to the received RX signal! Maybe something is wrong with the components that build up the Disc circuitry, LM393 or other passive components. Check it with scope and post pictures with minimum pot and maximum pot positions so we can help you to find the problem.

[yatahaze323]

I left my tgal and coil at work since that's where my oscilloscope is, so I will check everything on it Monday.

Dbanner
I attached both coils to the TX to get the oscillating frequency and the tx was exactly at 14.5khz and the RX was exactly at 14.9khz.
I'll have to try flipping the tx could and see if that was the issue.

Nandor
I'll check the points on the oscilloscope on Monday and post the pictures.

Thanks again everyone!

[Nandor]

If Tx is 14.5kHz than Rx should be around 16.12kHz. So you have the proper 20 deg phase shift between Tx and Rx signals.

[kt315]

you have to look a reaction in channels. on ferrous target - in one channel you have to see negative response,
on non-ferrous - in both channel you have to see positive response. 5 and 3 pins of the comparators.
also check right connection of R40 and D13.

[yatahaze323]

I need to correct my last post. To test that the coils where correctly constructed I did what was suggested in the TGSL101 instructions where I connected the TX coil to the board and it measured 14.5khz with the oscilloscope, then to test the RX coil I connected it to the TX oscillator and it measured 13.9khz with the oscilloscope. If I understood the instructions the TX and RX coils should be the correct frequency's.

kt315 - Is the Comparator U106 (LM393)?

For when I get a chance I'll do the following to try and figure out the problem,
1. Flip the TX coil and test the discrimination and make sure I don't have the double beep problem.
2. I'll connect my Oscilloscope to the GEB channel to see if it shifts 45 degrees when turned to the max. I'll take pictures of the Oscilloscope with the potentiometer at min and max.
3. I'll connect my oscilloscope to the Discrimination Channel and see if it shifts 90 degrees. I'll take pictures of the Oscilloscope with the potentiometer at min and max.
4. I'll check pins 5 and 3 on the comparators, and post pictures.
5. I'll check the pins to the right of R40 and D13, and post pictures

Also I have extra IC's so I can switch them out if needed.

Anything else I should add to the list of things to check?

Thanks again everyone!

[kt315]

yes. U106. LM393.