Carl's fluxgate Gradiometer ckt.

[Krzysztof]

Hi Damiafix.
Ah, the English.
Resolution FMX-1 is 256 units, because this is the data width 8byte (D0-D7).
Leave 2.2Kom is OK.
There is no need to use the switch to him.
The best is the logarithmic characteristics of the R6.
The worst characteristic is exponential because the too fast growing.
When you have a linear potentiometer R6 is just 2.5 Kom.
Best regards Chris.

[damiafix]

Thanks Chris,
I'll try again.

Have a good day, D.

[Geo]

Hi.
I have a problem with my gradiometer. I short pin 17 and 18 of SLC and connect one sensor. I put the sensor far from iron objects, I turn the unit on, let it for 10 sec for the calibration and measur the voltage on the pin13 of the SLC (with the DAC out of the pcb). Voltage is 0.00v, and if i will put an iron object close to the sensor the voltage goes to 4 or to 4.5 or 4.7V and then falls to 0V.
Is it right???

Regards

[Krzysztof]

Hi Geo.
Read whole plot, small tests are there.
To take out DAC first and to measure levels H / L on exits D0-D7 on SLC07.
Pin1 to Vcc.Low sensitivity mode.
Leave on table MAG, to does not touch him! Pin 17,18 together, one sensor.
To measure D0-D7. They should be all zero.
To to bring nearer the piece of metal, iron then and to measure.
Then again nearer and to measure.
Quite near all exit D0-D7 bedą level H.
You can use "Mini Logical Analyser for poor". (post # 12).
This the good simultaneous review of levels and nothing costs. You do from Digitals a lot of it - will add.
When you do not touch anything then exit levels have to be solid.
You will to bring nearer metal they - grow and they become solid.
To jump D0 maybe or even D1 - then fault EMI. Just problem.
Give to know how it went.
I looked for several days this the beaten in in the ground of engine Me109g with war. Dusk dropped already, I hurried and measurements are doubtful.
I walked for quickly and unequatly. Map went out ambiguous. To correction in year. Field is already sowed now.
Best regards Chris.

[Krzysztof]

Hi Geo.
Damiafix had similar symptoms (#11). Assembly and PCB were clean.
He spread, he measured, he folded and it is OK.
These spirits are gadfly? Maybe brownies?
Maybe this this Jin which transfers treasures a bit further and deeply when it sees from PD someone?
Do not place metal near sensora quite. Max 15 cm. You cross range measuring SLC07.
To muffle the trembling of generator FMG3 maybe even, though little real.
Indication MAG have to be stable. In practice this devout wishes.
Best regards Chris.

[damiafix]

Hi Geo, Hi Chris!
don't worry Geo, this result is normal; in a first moment I was in doubt that some component could to be damaged, but after few test (suggested from Chris) I've understood that the mag was ok. You must to test the components, out from the pcb, please read the post 11.
The real problem, for my own view, is that this tool is not really suitable for field use, a real use, but it can to be a interested and inexpensive way for to understand as work a flux induction magnetometer, and for to enter in the magnetometry world, trough the door of the electronic diy.
For me, it's been a good experience but now I'm building the epe mag, I should to buy the pic and the lcd in the next days, the sensors are the same.

Good lucky, D.

[Geo]

Hi Damiafix and Chris.
I am sure that there is a problem.
I made the test with the Logical Analyser. So.... when the sensor is far from iron objects i have a square signal on the sensor with a frequency of 110Khz and all the leds are off or maybe the led at pin13 flash. When i let an iron object near to the sensor the frequency is going at 40...45Khz and stay there. The leds are going at "on" for a moment but they return at Off condition. So i believe that the SLC is faulty.
Your opinion????

Regards

[Krzysztof]

Hi Geo.
Not necessarily corrupt SLC07. Maybe without showing a large load current outputs. May again be reset / configuration, and therefore sets a very high level as the primary signal. Try what the impact of PIN1, 3,4 - theoretically to Vcc. Is it because you FGM3H very large range of frequency changes. Or slide a little iron from the sensor. Normally the output signal level should be constant under specified conditions, not dynamic. Remove the stand ADC, may burden the SLC? Search for short-circuit or low power SLC (maybe a bad solder and is much resistance?).
Best regards Chris.

[Geo]

Hi Geo.
Not necessarily corrupt SLC07. Maybe without showing a large load current outputs. May again be reset / configuration, and therefore sets a very high level as the primary signal. Try what the impact of PIN1, 3,4 - theoretically to Vcc. Is it because you FGM3H very large range of frequency changes. Or slide a little iron from the sensor. Normally the output signal level should be constant under specified conditions, not dynamic. Remove the stand ADC, may burden the SLC? Search for short-circuit or low power SLC (maybe a bad solder and is much resistance?).
Best regards Chris.

Hi Chris.
As i remember first times who i made it, it worked good except the VCO who did not liked to me. I will check all around the SLC, but so little components or nets so to make problem!!!. Btw... do you have the hex file of the SLC???? or can you take it from your SLC07???

Regards

[damiafix]

Hi,
are you sure that is possible to dump the program from the SCL007?
I think of no, but for this I leave the word to Chris that is surely most expert of mine!
About the SCL007 integrity I still suggest you to check again all the connections, in my FXM-1 in a first time, I believed that also my SCL was damaged, but after few test and controls, I've found that all was ok. I remember that an evening I had the same problem that you have described, the problem in my case was a not solid contact, close the sensor; please check with attention the connection between the sensor and the pcb with a tester, if is ok, try with the other sensor. Yes, is possible that the SCL007 is damaged, but first of this conclusion you should control with much cure.
Check again please.

All the best, D.

[Geo]

Hi.
Really i don't know what to say!!!!, a lot of stranges phenomenons. Connections are very good. Voltage at pin 1,3,4 is to +5V. No short circuit... .
I put the sensor on a wooden chair and turn on and after self calibration i put a iron object near to sensor. What is happening????
Some times the leds turn on but don't stay on....
Sometimes they stay on (this is good). When i put a iron pincers near to the sensor the frequency is goes lower, but when i put a coffe metal box the frequency goes up (not all the times). What is happening???? The times who work ok i saw that it need a frequency shift of 1Khz so to have the blink of the first led, so i believe that it need more sensitivity.

Regards

[Geo]

Hi.
Today i made again some tests.
What i saw!!!!! the SCL007 works OK after 10minutes. I seem that it wants some minute to heating. At first 3..4 times it don't stay with leds "on" when i put an iron near to the sensor. But after about 10minutes it works OK.
Any idea......

Regards

[Krzysztof]

Hi Geo.
Dynamic damage is always the hardest to locate. I think that is a problem of frequent reset / configuration.
Check the power supply Vcc and / or voltage PIN3, 4 to ripple towards GND. Combine the two sensors may be like the original and see the frequency of each sensor. Only one of the metal (not close).

The sensitivity or full-scale range can be coded by programming the levels on two input pins on the IC, either by switches or software control through a computer output port. (S0 and S1). Each increment in this coded input from 00 to 11 increases the sensitivity by a factor of two.

This is the description of SCL6 but close to similar.
I think it is any IC PIC. The same technique reset to GND.
Hex protected.
See exactly what is etched on the top of the SLC, any subtitles?
If you have a stiff power to search for the instability in the Pin Control.
Best regards Chris.

[Geo]

Hi Geo.
Dynamic damage is always the hardest to locate. I think that is a problem of frequent reset / configuration.
Check the power supply Vcc and / or voltage PIN3, 4 to ripple towards GND. Combine the two sensors may be like the original and see the frequency of each sensor. Only one of the metal (not close).

This is the description of SCL6 but close to similar.
I think it is any IC PIC. The same technique reset to GND.
Hex protected.
See exactly what is etched on the top of the SLC, any subtitles?
If you have a stiff power to search for the instability in the Pin Control.
Best regards Chris.

Hi Chris.
I combined the 2 sensors and works the same. They have about the same operating frequency. Near to 100Khz there is 0.2Khz difference from sensor to sensor, so for me is OK. I will try to order another SCL007. SCL007 is a PIC16C54A. But i have not the hex file so to programme a new one. Maybe Carl can help.
Until to take the new one i will check the ADC and the other circuit so to be sure that all the other work OK.

Regards

[Geo]

Hi Chris,
I've removed the AD557 from the pcb and placed it on a breadboad for the test....



Those are the results for me:



Best regards, Dam.

Hi.
I tested the AD557 and here are the results
test_3a to 3F = 0.36V
test_3g =0.63V
test_3h =1.27V
test_3i =2.54V
test_3J to 3o =0.36V
test_3p =0.80V
test_3q =1.27V.

So and this IC (AD577) is faulty.
What happened????
Regards