VDI-RODERICO

[dbanner]

In addition to using a superior opamp, perhaps one could have a play with various band pass topologies to see what is optimum for this this application, in conjunction with the selected opamp. There is online software for this. I'm sure you all know about Filter Wizard. I just only discovered this. https://www.analog.com/designtools/en/filterwizard/

[waltr]

Modeled a simplified version of the White's Log amp meter drive in LTSpice. The White's circuit has many switches that I did not include.

I injected a 1Hz sine into one log amp and a 1/4Hz sine into the other. This is to 'see' what the log amps do and then what the output of the diff amp (sum) does. The input waveform is show on the plot as the current through the input R's, I(R7) & I(R. This puts the inputs on a different scale (i) to make them easier to see.
First to notice is the Gain compression, at V(x) & V(y), at higher input Voltage, the sine is 'distorted' into a rounded corner square.
Next is the output, V(vout). This would be the meter drive to indicate VDI.

Attached is screen capture and the LTSpice files zipped.

[SWL]

are there somewhere any tests, maybe in the Russian forum?

Can you put the schematic.

Here is the topic on the Russian site..... There is also a video test
http://md4u.ru/viewtopic.php?f=80&t=9955

[dbanner]

Here is the topic on the Russian site..... There is also a video test
http://md4u.ru/viewtopic.php?f=80&t=9955

Thanks, will have a look.

[dbanner]

Modeled a simplified version of the White's Log amp meter drive in LTSpice. The White's circuit has many switches that I did not include.

I injected a 1Hz sine into one log amp and a 1/4Hz sine into the other. This is to 'see' what the log amps do and then what the output of the diff amp (sum) does. The input waveform is show on the plot as the current through the input R's, I(R7) & I(R. This puts the inputs on a different scale (i) to make them easier to see.
First to notice is the Gain compression, at V(x) & V(y), at higher input Voltage, the sine is 'distorted' into a rounded corner square.
Next is the output, V(vout). This would be the meter drive to indicate VDI.

Attached is screen capture and the LTSpice files zipped.

Looks like a clown party there, Waltr. Well I can certainly see the compression. The sine waves now have flat tops and rounded corners indeed. So for a changing value of I(R7) and I(R8 ) , there respective outputs V(x) and V(Y) remain fairly constant over some range of input (flat top portion). V(out) is also compressed.
Are the inputs for X and Y in a real working circuit absolute value inputs?

[waltr]

Looks like a clown party there, Waltr. Well I can certainly see the compression. The sine waves now have flat tops and rounded corners indeed. So for a changing value of I(R7) and I(R8 ) , there respective outputs V(x) and V(Y) remain fairly constant over some range of input (flat top portion). V(out) is also compressed.
Are the inputs for X and Y in a real working circuit absolute value inputs?

I set a +-10mV sine which is a fairly weak but detectable signal.The picture looks very much the same, just larger Voltages, if inputs are increased.
The Log out signals are not flat top but have a slight curve.

I am leaving the analysis of VDI of output the readers here.

[dbanner]

Shallow curves. Alright.

[roderico]

I have redone waltr's ltspice circuit with values more similar to the original and it seems to work fine, the phase different is in the amplitude of X and Y (you can take them from the scope view) e.g. x = 2.75*sin(wt) and y = sin(wt), w=2*pi*5 (5Hz) then x/y = 2.75 and atan(2.75) = 70?, at Vout1 you should have something proportional to log(x/y)=0.44, i.e. a constant voltage and this value is then transformed into a different value but also constant (Vout2) I guess the resistors are then adjusted to have a value similar to 70

[roderico]

Here is the topic on the Russian site..... There is also a video test
http://md4u.ru/viewtopic.php?f=80&t=9955

ok thank you, it could be a very good improvement although it will saturate sooner, here it is the frequency response for a 5mV signal with ltspice in case someone wants to tune the values the old is Vout1 and the new Vout2, it would be very good to have a test in the fields where the real noise input is (maybe in the Russian forum says something about the field testing)

[roderico]

In addition to using a superior opamp, perhaps one could have a play with various band pass topologies to see what is optimum for this this application, in conjunction with the selected opamp. There is online software for this. I'm sure you all know about Filter Wizard. I just only discovered this. https://www.analog.com/designtools/en/filterwizard/

knowing the frequencies where the major contribution of noise comes from (e.g. from field testing) this filter wizard could make it easier to tune the parameters

[dbanner]

There appears to be some sort of DC offset adjustment (R230) for U206C.

[SWL]

Today I am disappointed. I waited a month when I received the package with the controller. And when I collected everything, I saw that PIC16F886 does not work with a 128x64 display. Now I will order 25K22///

[roderico]

because 128x64 firmware does not fit in pic16f886 program memory (only 14KB vs 32KB of pic18f25K22) but it does work with a 16x2 LCD display, you also need to make sure that your display generates a VEE otherwise you would need to provide a negative voltage check your display datasheet just in case... a blue display would also be difficult to read without backlight (white or green works better IMO)

[SWL]

Тhank you. I was doing DM-1 .. I didn’t like it .. Now I want to try your project. But only on a 128x64 display .. I will wait again for the order ...

[waltr]

Finally got the graphic display and op-amp circuits built and installed in my TGSL. Got out Sunday for field testing.
First restate design:
Display is a 126x64 OLED from Adafruit with SD1305 controller (code posted above).
Use the PIC18F25K22 but change pin defines for my universal display bd.
The op-amp circuit is on a separate board and mounted inside the TGSL enclosure. Four wires, gnd, +12, X & Y from TGSL EDU board.
12V passes through to PIC/display then linear 3.3V reg to run PIC, display and sent back the op-amps. This is a DIN connector on the TGSL enclosure using 6 pins, gnd, 12V, 3.3V, Bat divider, X op-amp, Y op-amp. The display is in a small box with a cable out to the mating DIN.
The display is 3.3V so easiest to simply run all from 3.3V.
Running everything from 3.3V works perfectly with no changes to the code.

When bench testing had a hard time trying to get consistent VDI values. Adjust the pot that is across the op-amps was not helping and was having a hard time understanding what the pot does. Then disconnected the pot so each op-amp (X & Y) went to the PIC's ADCs. VDIs then are very consistent and matched what is expected from the test targets.

Site I hunted is an old abandoned farm home with lots of iron.
The OLED display was easy to see and read. Only issue was glare from the tape I put over the display but moving the display to a slight angle fixed this.

The VDI become very useful. Due to the amount of iron junk the TGSL will give nice beep that seems to be a good target from the end of a large iron object. The X-Y Display however showed the TGSL Over loading, instead of a line within the graph the line would max at the graph edges. The TGSL audio sounded good but the graph showed a bad target.
I did check this with a few large iron object that I previously dug. These large iron object did give varying VDI from 45 to -80 so just the VDI number was not enough to say to dig or not.

Good target gave a VDI in the expected value range. 38-45 where all shotgun shell heads, cartridges or other brass items, one being a Victorian ladies pin (late 1800's).
VDIs of 80-87 were Indian head once cent coins of which I found three dating 1885 & 1887. Did not get and lower VDIs that could be nickles or rings.

I really like having this graphic display and feel it is very helpful for target ID.

I plan to post pictures of the boards etc but the charger for my camera died and the batteries are flat. Once I get the batteries charged I'll take and post pictures.