Low ohm connector problems Minelab P.I

[Woody.au]

Its nearly 2:00 am here and I am still up repairing detectors for customers..lol I have had this connector problem a few times with Minelab P.I detectors and it has a habit of catching me out. The coil impedance of these SD GP GPX detectors is around 0.5 ohm, the overall gain in these detectors is in the region of 600,000 so you can imagine that the slightest resistance in the coil connector can literally send the detector tropo!

This detector (a GP Extreme) would just sit on the bench and start screaming stopping and starting the audio and it sounded like something in the detector was breaking down or a dry solder joint. So after a few hours I could not find anything obvious so I turned it on and it started doing the same again. I wiggled the coil connector and it made no difference, only a slight signal when I pushed hard on it. But I have been caught out with this before, I took the coil connector off the detector and looked at the pins with a magnifier, all the Gold plating was gone and the Nickel plating was worn in parts down to the base metal.

I gave the connectors a good spray of contact cleaner and no more problem. I know this will play up in the future so lucky I gave a Gold plating setup and replated a new connector with hard 18k Gold.

Now its sitting on my bench and has not gone tropo for over an hour.
This just proves that with low ohms and reasonable currents the connectors have to be in A1 condition.

[nick_f]

Nice job, Woody. On my home built unit, I just solder the cable and the coil, there are fewer problems.
I think if you plate with gold, it will transfer 24k gold, not 18k as the electrode.
I want to ask you about the GPX4500, do you know what opamp are they using in the Rx preamp? I am curious if they moved to something better than the 5534 they used in SD2000.

Regards,
Nicolae

[Woody.au]

They all still use AD797 devices 1 for each channel.

[nick_f]

They all still use AD797 devices 1 for each channel.

Thanks for the info, Woody, much appreciated. I suspect this is the first stage, out of maybe three cascade opamps. How about the second and third stage of amplification?

Regards,
Nicolae

[Woody.au]

Well the first stage X 2 are AD797 those 2 go into a 4053 then go into 2 X TLE2142 A seperate Discriminator feed from the coil goes to 2 X TLE2072 . The 3 signals go into 4 X 4053 and come out as Ch1 CH2 GB and Disc. The integrators 4 X TLE2072 X 2 then go into the filters LT1056 X 4.

[nick_f]

Thanks again Woody. The TLE2142 is the first opamp where I see the parameter "Saturation recovery", which is 150ns. That seems to me a very good value.
Carl's Hammerhead diagram has problems detecting small gold because of the NE5534 is set to a high gain of 1000 and it has a long saturation recovery (about 4us if my measurements are correct). The saturation recovery time increases very much at higher gains, that's why the better diagrams use more stages of amplification, with lower gains for each stage.

There is a very interesting article here: http://www.analog.com/library/analog...omparator.html

Regards,
Nicolae

[Woody.au]

A common trait in Minelab designs is to use 4 small silicon diodes in 2 sets of 2 in series anti Parellel to limit the the voltage excusions of the opamp and thus not allow it to go into saturation.