Tweet
X-rays are already posted in the thread (above). You could be right about a capacitor being at fault as the failure was gradual to the point it only senses metal right on the coil. Semiconductors would usually fail hard not gradually. But finding the bad cap with all this epoxy is not likely. I could try I suppose? Hmm though, the heat. Wonder if this shorted part is a tantalum?
-
-
How I see it you'll probably need to sell the rest for parts and cut on the losses.
Otherwise you may end up spending 4K on a new module on a device which doesn't worth much more on the second hand market soon after new model is released.Comment
-
Yeah the Axiom might be devaluing it already.Comment
-
I returned the pod to Minelab AU and am waiting for a hopefully better repair estimate.Comment
-
You give up too easily! So close, I would have persisted with a small screwdriver and hot air and a lot of patience.
One thing that is nice about that module is, once it is reverse engineered a new improved module(or even just a working one) can be made and installed. No need to get the gunk off.
The diode you found that was shorted could likely have been the problem. If the cause of the fault was a badly wired coil adapter lead or a badly made coil, allowing high voltage to the Rx, then it would be likely to cause an input protection diode to go short.
One thing for sure, Minelab really don't want people poking around in that module.... Challenge accepted
Cheers Mick
Ps, anyone know of an Xray machine going cheap?
Just looking at the xray images again, at the top left of the module there is an S08 chip and it does not appear to connect to anything in the module. This is probably a chip, like what is in the coil connector, so you can't just install another one unless it is programmed to the main board. Nice Job!Comment
-
Yeah, normally I would go all the way with it. But the epoxy was not softening. It would go brittle if soaked in the MC and then heated. The stuff would just pop off but would most likely take parts with it. I tried putting another diode in the location and it did not work any differently. Under a microscope it didn?t look like a diode - no markings on it. I suppose the problem could have been a nearby sot23. It also may be that ML may want still more than I care to pay for the repair - I?ll know next week. I don?t think they normally will replace the module - they would toss both boards and install new ones plus it then requires a calibration.Comment
-
Yeah, most likely you'll get the same quote as you got from Minelab USA.
I already have an idea to defeat the "protection" chip in the event of replacing the Rx module
No gunk removal required.
Comment
-
The crypto chip is working. It detects (poorly) with a coil and says the coil is missing if it is (and shuts down).Comment
-
That module likely has it's own crypto chip(top left corner) so you can't just swap it with another.
That is what I'm referring to as having a possible solution for. But one would need to come up with a working replacement RX board before needing to worry about that.Comment
-
So do you know of others that may have bad modules? What would have caused it? He didn’t try a xcoil on this one.Comment
-
I have only heard rumors, not personally seen them.
By chance, did you try a different coil on the detector? If the coil nulling of the loops was not correct/moved over time, it is possible that the preamp and input see's higher voltages than it should, heating up and losing sensitivity due to lower forward voltage of the protection diodes when hot. There could be nothing actually wrong with the detector!Comment
-
The original owner had sent the whole unit in originally to ML AU. If it were just a calibration issue they should have recognized it and it should have been a cheap repair - instead the $4600 quote. ML trashes all boards and replaces them with new ones. Rarely does component level repair. It’s a shame. Component level obsolescence may really be a problem for them (and us) in coming years.Comment
-
Ahhh, that makes sense then. Yes I know Minelab doesn't seem to do component level repairs anymore. A bloke was quoted good money to have an sdc fixed, once again replacing boards. The actual fault, bad connection in the header pins that connect the 2 boards together.
I guess it will keep me busy for many years to come, as long as I can get parts, which is becoming an ever growing battle.
Potting that module and then offering no way to replace it apart from replacing both boards is just silly.Comment
-
That company with their user unrepairable boards are fueling the Chinese copy market, the Cheebo manufactured GDX8000 can be fixed up to out perform the GPX4500, (In my opinion) I had a few that do not ground balance and are noisey but a bit of work made for a fully useable detector.Comment
-
Minelab AU took very special care of me this time and I have a working GPZ. I’m hoping to come across a deal on a coil.Comment
Comment