Photo of GP Extreme Badness?

[bklein]

Well.... I think I have it working.
The flex ribbon interconnect cables had broken/intermittent connections on the outside conductors. These things are miserable. Painting them sure didn't help either. I should replace them but instead I just repaired them to see if the unit then worked. It seems to but I haven't taken it out of the garage yet (2:30am). Very possible something else could have been damaged by these open conductors but just have to see. I found a spool of 2N2907's that I used for the BC857C and just kept moving things, blowing transistors, until I pinned it down to the flex cable. Thanks to all of you for your help and suggestions. I hope this thread helps others if they are faced with what I was.

Barry

[bklein]

Actually the above was not the end of this saga...
Tried the unit and it worked for about 30 seconds. I took the coil off and attached it to my Surf PI clone board. The TX signal was good but the RX had an intermittant. (I wasn't able to catch this with an ohmmeter before). It threw me for a bit because the open signal was kinda looking like I would want - it showed the coil sensing metal (but had a ton of EMI noise and was too big). I pinned the issue down to the coil grommet area - sure enough the plastic insulation had cracked all over this area.
Cut that area off, resoldered things, touched up a tear in the shield and good to go. I feel much better thinking I've found the smoking gun although there really were others already (flakey cable connector, damaged PCB ribbons, bad parts).

[mario]

Actually the above was not the end of this saga...
Tried the unit and it worked for about 30 seconds. I took the coil off and attached it to my Surf PI clone board. The TX signal was good but the RX had an intermittant. (I wasn't able to catch this with an ohmmeter before). It threw me for a bit because the open signal was kinda looking like I would want - it showed the coil sensing metal (but had a ton of EMI noise and was too big). I pinned the issue down to the coil grommet area - sure enough the plastic insulation had cracked all over this area.
Cut that area off, resoldered things, touched up a tear in the shield and good to go. I feel much better thinking I've found the smoking gun although there really were others already (flakey cable connector, damaged PCB ribbons, bad parts).

As usual, when some is wrong, it's not easy narrowing down the issue to a particular problem.

I had an accident while my unit was on the bench, and now I'm working on fixing the issue.

I have a few components on order that will hopefully resolve my problem, but I don't expect things to be fully resolved so easily. More than likely I'll spend another couple of weeks down before it's finally "fixed."

Man it sure sucks to have down time on these units.

[bklein]

Yeah, from your contributions to the thread it looks like you know what you are doing and have some experience. For someone like me it was a high-risk situation but the money I spent on the detector was low enough I was willing to risk it. Luckily bad stuff got hot enough to show through the paint!
And you have the challenge that you don't know what the parts really are, and what replacements may be. Try replacing them and they all may fry again....
This is something that one obviously builds up experience with. Now I know to check the coil first, not last. This one was a real challenge and I was ready to give up a few times.
I pity the guys out there with these older detectors that die because Minelab is sort of giving up on them as they don't care to look for new replacement parts for those that are failing - so they say they basically are out of replacement parts.
I also have the same sick interest in repairing older music synthesizers. I once got a deal on a PPG Wave 2.2 synth that very well may have been the one used in that "Science" song of the 80's. There were less than 300 made. I fixed it but then the original owner called me every month wanting me to sell it back to him. I thought of the Motorola 6800 microcontroller system it uses and how old the parts are and eventually sold it back to him. A software synth package has the same sound, for $200, and never dies.
I am very happy to have my GP Extreme now though and am anxious to try it out in the Arizona desert.

[Woody.au]

I have been selling repair kits for the GP GPX detectors for the last 3 years, when the N channel Fets short it can damage the P (GP) or N (GPX) channel Fets on the other side of the board that feed into the AD797 Ic's.
The Fet gate driver also gets destroyed when the input Fets short Drain to Gate. Sometimes the Dual Gate Mosfets on the input of the AD797 ic's partially short causing noise and imbalance. On some occassions the AD797 devices fail or go
noisy. It is best to replace the complete front end as some faults cannot be detected on the work bench due to the extreme low level signals being dealt with. Make sure that the -15 volt rail is discharged before replacing the N Fets
as any soldering iron shorting the Fets will blow the driver transistor as the -15 volt supply has large storage capacitors.

[bklein]

I have been selling repair kits for the GP GPX detectors for the last 3 years, when the N channel Fets short it can damage the P (GP) or N (GPX) channel Fets on the other side of the board that feed into the AD797 Ic's.
The Fet gate driver also gets destroyed when the input Fets short Drain to Gate. Sometimes the Dual Gate Mosfets on the input of the AD797 ic's partially short causing noise and imbalance. On some occassions the AD797 devices fail or go
noisy. It is best to replace the complete front end as some faults cannot be detected on the work bench due to the extreme low level signals being dealt with. Make sure that the -15 volt rail is discharged before replacing the N Fets
as any soldering iron shorting the Fets will blow the driver transistor as the -15 volt supply has large storage capacitors.

Hello Woody,
My detector has been working after my repair but a couple days ago it failed again - while I was testing my various used coils. I had tested the coils first with my LCR meter but perhaps missed seeing an intermittant. The detector had a shorted BC547C transistor and right fet. I replaced them but haven't powered it back up yet. I'm wondering if just the metal all near my in my garage lab could have overloaded the front end and killed it? What can cause this susceptability - are there no protection circuits for possible shorted or open coils (which is the culprit?). What do you guys do to fully rash out a coil and cable and be sure it has no issues? I think in this case it may be due to a simple spread open pin in the cable plug(?) I used the same coil just a few weeks ago with no problems. I'd hate to take this on a search trip and have it happen in the field...

[detectormods]

The failure is caused by the small driver pnp trasistor that that drives the gates of the signal input N fets. It can be damaged if the Fets are shorted source to gate, you install new fets and it will fail again with the -15 volt series 100 ohm resistors getting hot (brown paint from heat) The trick is to relace the N Fets the P secondary blocking Fets and before replacing the PNP driver short out ever supply rail electrolytic capacitor as and mistake will pop the PNP driver again.