NEW !!! Full differential PI front end.

[Muntari]

Hi Moodz,
Yep, gottcha, just looked at PWM port line source / sink specs, I was having a seniors moment...

I was originally concerned with damage to port pins but given your resistor values, should be ok.


Cheers

Brian K

[WM6]

Excellent and very inventive moodz.
Someone will ask you for full schematic soon (and maybe for kit or preprogramed pic too).

[Aziz]

Industry standard analog switches

Hi Brian,

the CD4066 has not a logic level shifter built in (*shame*). There are DGxxx devices that do it. But there are also improved versions available:
MAX313
MAX392
MAX333A

These analog switches makes the driving with MCU logic levels easy. Some of them are precision type and cost more.

You can also try the CD4053 (only 3 switches).

I like your new design.
Aziz

[Carl-NC]

doh ... just realised that explanation is wrong sort of.. when the PIC port is high the 4066 is ON. When the PIC port is low the 4066 is OFF. The diodes prevent the PIC high level from driving the port ... this is done by the pull up resistor to get correct bias.

moodz (again )

I don't think this circuit will do what you want it to do. The pull-ups won't pull up to +15 when the PIC outputs are at +5. They will only go to 5.7V, which probably isn't enough to switch the 4066. You really need a bipolar (with base R) or a FET in place of the diodes.

- Carl

[Muntari]

Something like this perhaps...

Cheers

Brian K

[Muntari]

Hi Aziz,

Small point but .... credits to Moodz this is his design not mine....

cheers

Brian K

[Aziz]

Hi Aziz,

Small point but .... credits to Moodz this is his design not mine....

cheers

Brian K

Oh , I am sorry for this. I ment moodz of course.

I also would like to see your design Brian. It's long time since you didn't publish details of it.

Aziz

[gadgetmaker]

4066 just use a 74HCT4066

I have used 74HCT4066's for fast pixel overlays for video; they should work well.

[gadgetmaker]

Just a crazy idea

Moodz,

Why not try to drive both coils, one then the other by simply adding another fet and then use the two signals to cancel any dc offset during A to D conversion?

Karl

[Muntari]

Hi Aziz,

My design is undergoing some upgrades, new proccessor, new firmware.

Besides this, my time has been taken up with a sub-project related to the MPI and also I purchased a company late last year which has required most (read 99% !) of my time. So the MPI project has taken a back seat for a few months.

Will the MPI see light of day...yes it will...does it work....yes it does..when will it be ready.... when we have all the usual production issues ironed out and legal matters confirmed.

I will post on another thread when we are close.

In the meantime, Moodz is doing a great job with his beast...I like it.



Cheers

Brian K

[simonbaker]

Moodz,

Why not try to drive both coils, one then the other by simply adding another fet and then use the two signals to cancel any dc offset during A to D conversion?

Karl

I wonder if that's similar to crazy idea i had, using dummy coil just to cancel out TX signal... but I know next to nothing about PI circuits.

http://www.geotech1.com/forums/showthread.php?t=15465

-SB

[gadgetmaker]

More on Diff circuit

SB,

If you read Moodz work he is canceling noise out and his current circuit drives only half the coil. My idea is to drive both sides, not at the same time but one at a time and then use a +/- rail on the diff amp; the problem then becomes how to handle two opposite polarity signals? You could offset the output of this with a precision dc level before gating and input to the A/D or you could use an A/D that would handle the +/- input. Now you have a completely diff circuit. I will build a prototype up when I can and send some scope traces.

I have done a lot of PIC programming in C but have never used the DSP part from Microchip; I think it is time to learn about these parts!

One of the most intersting thing to me is to do active damping and do more analysis of the data using the DSP. It always seemed to me that damping the signal with a resistor limited the response of the system. Maybe there is some descrimating information here????

Well I may be talking to myself? I will actually do some work and post some data and see I can help.

Best to all,

Karl in AZ

[msysta]

To my knowledge it is the decay of the coil voltage that makes the eddy currents to the target and faster the decay, stronger the eddy current. So in that sense the coil flyback energy should just be drag down as fast as possible.

Maybe a mosfet could be used there, phasing it with the clocking device through schmitt-trigger operated switch. It would act like a drain, short circuiting the coil for a derivative short moment. Or maybe a decent capacitor could do the job.

[gadgetmaker]

Variable Damping

Msysta,

This active damping pulse could be switched in right after the flyback pulse reaches maximum level and then (out or the damping resistance changed) at a predetermined voltage level or after a certain duration. I wonder if this would allow one to sample the signal sooner thus increasing sensitivity?

One thing that is for sure; there is a lot of fun experiments with PI technology to try!

Best to you from Arizona.

Karl

[msysta]

Hmm.

Why not use spark gap? When current rises enough it would make an arc and null the voltage very fast

Theres at least spark gap tubes like this available.

http://commons.wikimedia.org/wiki/Fi...k-gap-tube.jpg