PI-1 Improvements? ( Carl?)

[Russ_NY]

Carl
In your response to Philippe on 6/6/03 you made reference to their being LOTS of possible improvements for the PI-1. Yes a micro of some sort to take care of timing would simplify the circuit, but as you say no better or worse performance if done properly. I can only think of a few things which would actually improve performance. One being a DSP with some good programming for noise reduction and target signal processing. Perhaps the use of a dual coil design. Or a combination of the DSP and a dual coil. Sure I can think of bells and whistles such as a signal level meter, but do those actually improve performance? I am not trying to pick nits here, I would really be interested in hearing about any potential improvements. I have read every scrap of information I can find on metal detector design, and looked closely at every schematic that I have come across. This project has been a great learning experience for me but I feel like I have run into a brick wall, or I have completely overlooked some aspect of detector design along the way. Could you please elaborate on these potential improvements. I would appreciate it very much.
Thanks

Russ_NY

[Carl]

Re: PI-1 Improvements? ( Carl?)

The most obvious improvement is to use a microcontroller. A uC is more stable & more consistent, making performance easier to maintain (no tweaking). I didn't use one because many people don't have the ability to program them, and the purpose of this design was to fully exercise all design parameters.

A while back, on Eric's forum, some folks explored energy recovery from the coil. Saves on power.

PI-1 does not have true ground balance, it has ground tracking. In really bad ground conditions, it probably won't work very well.

A differential front-end, using an IB coil, should offer better noise rejection.

There are a number of optimizations that can be explored. The coil design (I more or less chose 0.5uH as a target), the preamp (better opamp, lower noise configuration), the integrator (maybe move into the uC for more complex processing).

Also, a better front-end clamp, maybe active instead of passive, might help on the preamp overvoltage recovery.

The audio is still not where I want it to be.

Finally, use SMD & shrink the whole thing to a 1x2 inch board. With that, you could make a nice pinpoint probe, or a headphone detector, a la HeadHunter.

Some of these improve performance, some don't. The biggest performance limiter I see, is noise. It limits how much gain can be run. Optimizing the coil & clamp for better signal return, the front-end for lower noise, and the integrator for better signal processing, are Good Things To Do.

I do have plans to switch to a uC, and do the whole thing with SMD. Whether I do any other improvements, will depend on how much time I get to spend on it.

- Carl

[Russ_NY]

Re: PI-1 Improvements? ( Thanks Carl)

Carl
Thanks for the information. I had considered some sort of active clamp on the input to the amp. I think it would definitely improve on the recovery time. A uC is definitely a good move not only for timing, but it could supply the audio as well. I have a good friend who is an exceptional hardware and software engineer who is very interested in doing some DSP work with the design. If we get a chance to work on it I will certainly report on the results. Surface mount is fine if you want to make it tiny. I think though that my next revision will probably be double sided. I am still getting quite a bit of drift and need to resolve that before moving on. Thanks again Carl, Happy hunting.

Russ_NY