'KROT2-XM' (mole), new project Xmega32

[Ropcsika]

This is not the correct results ...

Make such manipulations:
1. Disconnect the sensor.
2. Set "balance coil" A = 0 (synchronous press "-" and "+")
3. Short-circuit terminals RX
4. Press "0 ADC" (value should be about 50 mV., Then the buttons "-" and "+" to adjust around 0 mV.) And forget about this mode.
5. Connect the sensor.
6. Find resonance TX. (Current should be about 80-100 mA.)
7. Balance sensor. (50-150 mV.'s A good indicator.)
8. Go look for coins.


If the unit emits a lot of sounds wear:
1. Wash the PCB.
2. Check the connector (where to connect the sensor).


Good luck ...

Zero_adc 0 adjustment

[Tepco]

Geo, it is possible that Xmega analog ADC input is damaged during same event (ESD, overvoltage...) that damaged MCP601. Remove D2 and use DMM to measure voltage at pin 3 of empty socket, record this value. Then connect pin 6 and 3, so ADC input will be connected to this point. Measure voltage again. If it is different, D3 is probably damaged.

[Geo]

"Coil bl" A=0!
3. and 4. should show <50mV
Look at the oscilloscope pin_6 D2
What version of firmware you?

Hi Andrew.
It is not possible to make A=0 with MCP601 on board. With TL061 yes.....
I attach 3 photos.
First from input of Rx coil
Second of pin6 of D2 with MCP601 on board and
Third of pin6 of D2 with TL061.
With TL061 the device works very good but has a lot of erratic signals "lot of sounds wear".
Do you think that TL061 can be the problem??
I don't know the version of firmware because WM6 has programmed it. Maybe he can tell us.

Thanks

[Geo]

This is not the correct results ...

Make such manipulations:
1. Disconnect the sensor.
2. Set "balance coil" A = 0 (synchronous press "-" and "+")
3. Short-circuit terminals RX
4. Press "0 ADC" (value should be about 50 mV., Then the buttons "-" and "+" to adjust around 0 mV.) And forget about this mode.
5. Connect the sensor.
6. Find resonance TX. (Current should be about 80-100 mA.)
7. Balance sensor. (50-150 mV.'s A good indicator.)
8. Go look for coins.


If the unit emits a lot of sounds wear:
1. Wash the PCB.
2. Check the connector (where to connect the sensor).


Good luck ...

Thank you Smity for analytical info.
As i wrote to Andrew with TL061 in place of MCP601 all are ok, A=0 and ADC=3mv.
Problem is the erratic signals.
I washed many times the pcb without results. Coil is connected ok at pcb.

Regards

[Geo]

Geo, it is possible that Xmega analog ADC input is damaged during same event (ESD, overvoltage...) that damaged MCP601. Remove D2 and use DMM to measure voltage at pin 3 of empty socket, record this value. Then connect pin 6 and 3, so ADC input will be connected to this point. Measure voltage again. If it is different, D3 is probably damaged.

Hi Tepco.
My opinion is in agreement with yours.... , i think that i must order one MCP... and one ATXMEGA... before i make use of the hammer.
Later i"ll make the test that you say and i"ll inform you.

Regards

[Geo]

Geo, it is possible that Xmega analog ADC input is damaged during same event (ESD, overvoltage...) that damaged MCP601. Remove D2 and use DMM to measure voltage at pin 3 of empty socket, record this value. Then connect pin 6 and 3, so ADC input will be connected to this point. Measure voltage again. If it is different, D3 is probably damaged.

I just finished the measurements.
Voltage at pin3 without D2 =1.632.
When i connect together pin3 and pin6 the voltage is about the same = 1.5972, but i can't see any other problem.....
New D2 and D3 and i think that all will go right... otherway ... you know

Regards

[Tepco]

Not conclusive, Xmega may be good. 35mV difference is around 1uA leakage, not excessive. With damaged input pin will be much higher. I tested my unit, with similar results, 1.604\1.578V, it works, with perfect stability.

[Geo]

Not conclusive, Xmega may be good. 35mV difference is around 1uA leakage, not excessive. With damaged input pin will be much higher. I tested my unit, with similar results, 1.604\1.578V, it works, with perfect stability.

So.... what other can i do????

[Tepco]

Absolutely no idea why detector can be unstable without coil. OK, with D2 removed, ADC input is open and can pick up something, but... Some passive component failure (noisy resistor, leaky capacitor), or xmega damaged? When I first tested my unit, I used unshielded DD halves, 20cm, on TX resonant at some 11.5kHz, later added some capacitance, detector tuned at highest frequency but still not resonant, some 57mA TX current (coil was originally intended for PI detector RX coil, thin wire, high resistance). On RX side, only 120 turn, nonresonant, without tuning capacitor, all this connected with some 2m of ordinary wire, not even cable. With threshold set to nearly lowest sensitivity, this already detected 1E coin at 20+ cm. Bit unstable when I move around, otherwise maybe 1-2 beeps per minute (when neighbors turn on or off some appliance). I just finishing “original” coil, yesterday I made TX part, now i'm at 112. turn on RX coil. I will try one interesting method for custom built coil case, using XPS Styrofoam and epoxy, can be interesting to someone, will report more on this in Coils section.

[Geo]

Absolutely no idea why detector can be unstable without coil. OK, with D2 removed, ADC input is open and can pick up something, but... Some passive component failure (noisy resistor, leaky capacitor), or xmega damaged? When I first tested my unit, I used unshielded DD halves, 20cm, on TX resonant at some 11.5kHz, later added some capacitance, detector tuned at highest frequency but still not resonant, some 57mA TX current (coil was originally intended for PI detector RX coil, thin wire, high resistance). On RX side, only 120 turn, nonresonant, without tuning capacitor, all this connected with some 2m of ordinary wire, not even cable. With threshold set to nearly lowest sensitivity, this already detected 1E coin at 20+ cm. Bit unstable when I move around, otherwise maybe 1-2 beeps per minute (when neighbors turn on or off some appliance). I just finishing “original” coil, yesterday I made TX part, now i'm at 112. turn on RX coil. I will try one interesting method for custom built coil case, using XPS Styrofoam and epoxy, can be interesting to someone, will report more on this in Coils section.

I will be happy to read your method!!!!.
I made my coil as author says (45 turn Tx and 200 turns Rx and i wrap Scotch 24 tape for schield. Also i connected coils to pcb via 2 x 2pole shielding cables. All fine.... but the device don't like to work....
Never mind, all the years i constructed maybe 150 .... 200 detectors but did not work all so good. So another one without good results.

Regards

[WM6]

What version of firmware you?

It is KPOT_V06A.hex inside.

[Geo]

Thanks..

[WM6]

I washed many times the pcb without results.

No need for such washing. Using some thinner solvents you can destroy foil capacitors.

Board was properly cleaned, dried and protected again using high isolative (such as 25000V) spray lacquer.

[ANDREW K2]

Hi Andrew.
It is not possible to make A=0 with MCP601 on board. With TL061 yes.....
I attach 3 photos.
First from input of Rx coil
Second of pin6 of D2 with MCP601 on board and
Third of pin6 of D2 with TL061.
With TL061 the device works very good but has a lot of erratic signals "lot of sounds wear".
Do you think that TL061 can be the problem??
I don't know the version of firmware because WM6 has programmed it. Maybe he can tell us.

Thanks

Hi Geo.
If you believe your pictures then you gain Сoefficient MCP601 =1, TL061 =2. Must be> 50 .... Ostsilogram connected with the coil? If the coil was not where the signal? After all, you probably have A = 0 do? TL061 certainly not suitable look for at least 18V.. Frequency gain of 1 MHz...
Try this:
Remove R13 and concluded with IN string any point U/2. (R7/R

[Geo]

Hi Geo.
If you believe your pictures then you gain Сoefficient MCP601 =1, TL061 =2. Must be> 50 .... Ostsilogram connected with the coil? If the coil was not where the signal? After all, you probably have A = 0 do? TL061 certainly not suitable look for at least 18V.. Frequency gain of 1 MHz...
Try this:
Remove R13 and concluded with IN string any point U/2. (R7/R

Hi Andrew.
No,... gain at TL061 is about 75. At 3rd photo the selector is at 0.2V/ division and at 1st and 2nd at 10mv/ division...

TL061 is not a good replacement for MCP601 but i have not something better in my drawers.
Photos are with coil connected.

Regards