VDI-RODERICO

[roderico]

ok very good, very nice videos, your XY offset should match the values shown in the screen when there is no target near the coil and then you might decrease the level to 10mV for better sensitivity, something similar to the battery that should be calibrated with a measurement of the battery voltage. I see that the out-of-bound values were not fixed correctly in previous version, could you check test4? this version also sets the histogram range according to the selected vdi interval and slightly decrease the display latency...

[kroman01]

next videos from glcd_test3

https://www.youtube.com/watch?v=UT9x_So4TiY

https://www.youtube.com/watch?v=yxw73frkF-A

https://youtu.be/3fz5k3C7iqg

XY offset: 513
level: 20mV
scale: auto

[kroman01]

glcd_test3 works great, that's what I meant, good job
you can turn off the 180 range in the histogram when the inteval value is set to -90.90

[roderico]

I added the axis scaling in XY and scope views for testing: 1x (same as previous version), 2x, 4x and auto (this can scale the signal up to 16x). This version should also fix the values out of bounds in the histogram view, although it is still missing to reduce the display latency...

this version fixes a small bug with the axis auto-scaling related to the flatting of channel Y reported by kroman01, better to use this one for testing instead of previous one...

[roderico]

I added the axis scaling in XY and scope views for testing: 1x (same as previous version), 2x, 4x and auto (this can scale the signal up to 16x). This version should also fix the values out of bounds in the histogram view, although it is still missing to reduce the display latency...

[roderico]

and concerning the oscillator frequency a solution might be to reduce the latency of the GLCD somehow, e.g. minimizing the reads/writes to the GLCD...

[roderico]

ok I see what you meant, very nice videos, a possibility would be to auto-scale X and Y axis. I also saw a bug in the histogram view, sometimes it goes out of scale but it can be easily fixed. IMO the signal Y flattens in the scope view because the MPC602 is in saturation, so it cannot be done much, however it saturates quicker than channel X, it would be good to cross check this with a real oscilloscope...

[kroman01]

films from GLCD action, R4 and R6 resistors worth 1M
-

https://www.youtube.com/watch?v=NWTbO8B05eM

https://www.youtube.com/watch?v=TBfOhaZAVmQ

https://www.youtube.com/watch?v=q7bxj370gc4

-
on the scope you can see that something is cutting off negative indications

[roderico]

could you post a picture of XY and the histogram? are you using 1M or 3M?

[kroman01]

i programmed an existing PCB and it works, it seems to me that too low sensitivity to visualization, as shown by good VDI - weak signals, XY histogram and oscilloscope nothing can be seen R4 and R6 I have 1M maybe I should increase the U1 gain get and use resistors R4 and R6 for 3M
i will do field tests ... a bit slow, which will give a bigger 10 MHz - 20 MHz clock

[roderico]

this is a port of the vdi to a KS0108 based GLCD display for testing, it can be configured in three different modes:
- XY correlation,
- histogram view
- or like a small "oscilloscope"
different patterns in XY correlations might give you additional information about the target, e.g. some hints about its shape, the histogram view can be used for a better vdi "average" and the scope view is more just for debugging

[roderico]

PIC18 are better (and more expensive in general), however, this is mainly due to drive the LCD in 8-bit mode, in 4-bit mode you need double work and time to send just one character to the LCD

[kroman01]

I connected VDI to PIC16F690 with 1.4 software and I see that there is a significant difference compared to new systems that are faster, especially in the case of PIC18F25k22

[roderico]

my coil RX and TX signals are shifted by 90 degrees so I guess I would get the same vdi values as you if the same calibration is used for the vdi (e.g. with a small piece of foil)

[kroman01]

the only difference I noticed in connection with software 1.4 is the faster saving of settings in memory.
what phase shift is needed in X and Y