Alternative differential integrator.

[Teleno]

The differential integrator in MPP is based on a small signal model. It's is not a true differential integrator for larger signals.

I'm submitting this alternative based on a Sample/Hold auto-zeroing configuration. The LF398 chip is inexpensive.

The second sample (EF) is used to charge the S/H capacitor. The main sample is fed to one side of the S/H capacitor, thereby subtracting the "memorized" EF from the input signal (auto-zero). The output of the S/H is then fed on to a true integrator.

This is the concept (node references corresponding to the MPP Rev. D schematic):






J1 and U2 were chosen for convenience in LTSpice, but the original J113 and TL072 can be used just as well.

[eclipse]

Thanks Teleno, I guess this would help if I was to use the MPP with larger coil and 250-400uS TX pulse width?

I applied the changes on my schematic, here it is.

2092016.pdf

[Teleno]

Thanks Teleno, I guess this would help if I was to use the MPP with larger coil and 250-400uS TX pulse width?

I applied the changes on my schematic, here it is.

[ATTACH]37193[/ATTACH]

To maintain the polarity of the output we would need an inverting amplifier, like the original MPP:



Also, the capacitor C2 in my schematic should be 470nf (not 22nF).


You can also use the audio circuit I had proposed before: http://www.geotech1.com/forums/showt...172#post210172

[eclipse]

The audio section suggested I tested with the T3DX board and didn't work very well - sound was quite edgy.

[Teleno]

The audio section suggested I tested with the T3DX board and didn't work very well - sound was quite edgy.

Didn't test it myself. In any case using that audio circuit would allow you to remove the inverting amplifier U8b.

[eclipse]

Here's the updated version with the inverting op-amp. Attaching PDF file only because it's better than JPG, no preview sorry

20092016_2.pdf

[Teleno]

Here's the updated version with the inverting op-amp. Attaching PDF file only because it's better than JPG, no preview sorry

[ATTACH]37199[/ATTACH]

It looks good. Now we only need to replace the timing logic by an ATtiny uC for flexibility. Q3 can be replaced by a 900V MOSFET, too.

[eclipse]

I've already done that with 2 different controllers PIC12F683 and PIC16F684. Audio needs to be made through the controller though, right now relies on the 555 audio.

[6666]

I've already done that with 2 different controllers PIC12F683 and PIC16F684. Audio needs to be made through the controller though, right now relies on the 555 audio.

Been trying to make a pic emulate a 555 audio VCO, but it is very difficult , but still trying, I would like to hear from anybody who has got the right code working.

[Teleno]

Two more alternatives:


1. Switchable inverting-non-inverting voltage follower + integrator



2. inverting voltage follower + 2-path integrator.




LTSpice files attached as .zip

[eclipse]

What a variety Teleno!

Here's the full schematic with Option 1 with LF398 (not tested!).

20092016_3.pdf

[eclipse]

One more version (since there was one op amp channel available) - the extra parts there suppose to give non-motion behavior.

Probably won't work but there it goes (taken from SuperScan/6PI)

20092016_04.pdf

[eclipse]

Will the first submitted proposal be compatible with the GB mod by Devor

[Teleno]

Will the first submitted proposal be compatible with the GB mod by Devor

I wasn't aware of this mod, but if it is copatible with the original integrator then it should also be with my alternatives.

[eclipse]

Good stuff, I'm sure these mods will be of great help!