Tweet
For the last two weeks I've been faffing around at home playing with metal detector stuff. What bliss... 
Unfortunately back to work next week. A necessary evil, I'm afraid.
Anyway, one drawback of having some much free time is that eBay is not far away. So recently I won a faulty Fisher 1266-XB that apparently (so the owner told me) was "as dead as a door nail". Ahh... I though. It's probably only a simple fault. Might even just be a broken wire.
Think again!
The detector arrives in reasonable time, and looks ok, although it's obviously been well used. The coil cable has been previously repaired with some electrical tape, and immediately I'm thinking that could be a potential problem. No batteries in the detector - but then it was advertised as faulty, so not surprised. Put in some new batteries, and turn on ... faint click in the speaker, and the battery test does the same. Hmmm ... not completely dead then.
Next - quick check of the coil with the LCR meter. Checks out ok.
Now it's time to open her up.
First stop - the power supply. There's about +/-2.5V on the power supply rails, and now I'm getting confident that it's a simple problem after all. The trouble is that the 1266 schematic that someone posted on the Geotech site some time ago doesn't exactly match this detector. Luckily I have an official schematic for a 1265-X which is virtually identical, except for a few minor differences that are covered by the 1266 schematic. In fact, both the 1265-X and the 1266-XB are almost identical on paper, apart from the PCB layout, which is completely different.
After some careful eyeballing of the PCB - no obvious problems there, nothing burnt or broken - I tried to identify the 7 transistors that make up the power supply, and to take some voltage measurements to try and isolate the fault. There was something strange with the voltages around the first two transistors that didn't make sense, and then I suddenly noticed that one of these transistors was inserted back-to-front!
A bit more eyeballing of the PCB, and I realised that this detector had already undergone some repairs. 
The transistor should have been an MPSA13 but an MPSA14 had been fitted. Actually this is ok as it's an equivalent. Surely this couldn't be the problem after all? The desoldering tool went into action and the offending transistor stuffed into the transistor tester. Good - it seems to have survived. Back in the right way round and switch on ... still dead. 
But this time the voltages make more sense, and it's clear that something is not connected right. Damn! There it is - a broken track on the top layer leading to the transistor base connection. Thoughts of "what a botched repair someone's done here" come to mind. Anyway, one jump lead later ... switch on ... guess what? ... yep, still dead. However, the good news is the power supply is now +/-4V, which is correct!
What next? TX is driven by a crystal oscillator - this is OK. Voltage across TX coil also looking good. Output of RX preamp also OK. Phase reference circuits all working. Phase discriminators also functioning. Starting to get even more confident.
I then jumped forward a few stages and measured the outputs of the static, resistive, and disc channels. Wow! All OK!
Now at this point I know most of the design is working, but there's still no output. I also know that the speaker is connected because it clicks when I turn the detector on. What if the audio oscillator is not functional? Sure enough, dead as a dodo. After a lot of probing around with the scope and some eyeballing again, decided that the LM339 comparator is dead. Trying to get this out of the board without destroying any of the tracks was a challenge, but achieved without mishap. On a whim I then decide to fit an IC socket, which is just as well because (believe it or not) the first LM339 I fitted was faulty.
Now there's a faint sound from the speakers and/or headphones. Damn - still another fault. This time I traced the fault to the final transistor in the audio amp. Another bl**dy MPSA13. Replaced with a BC517 (which is also a darlington pair) - and ..... a really loud audio signal!
Finally I adjusted all the internal presets according the manufacturers instructions, as I couldn't trust they were correctly set up by the previous "repairer", and resoldered a broken wire to the charging socket. One wire going to the DISC1 pot was hanging on by a thread, and also needed attention, and I replaced the bolts holding the front panel to the PCB with some that had a smaller head, since the existing ones had caused a shorting problem at some previous time.
So now I have a fully working Fisher 1266-XB.
Tomorrow I plan to remove the electrical tape from the coil cable and fit some heatshrink sleeving instead. Might even extend the cable, as I'm sure it's shorter than it should be.
Anyway, just thought I'd share my experience with you about buying broken detectors on eBay. In my feedback I've mentioned that it's all repaired and working. Now I'm just waiting for the email asking me how I did it!!
Unless of course he reads this ...
Unfortunately back to work next week. A necessary evil, I'm afraid.
Anyway, one drawback of having some much free time is that eBay is not far away. So recently I won a faulty Fisher 1266-XB that apparently (so the owner told me) was "as dead as a door nail". Ahh... I though. It's probably only a simple fault. Might even just be a broken wire.
Think again!
The detector arrives in reasonable time, and looks ok, although it's obviously been well used. The coil cable has been previously repaired with some electrical tape, and immediately I'm thinking that could be a potential problem. No batteries in the detector - but then it was advertised as faulty, so not surprised. Put in some new batteries, and turn on ... faint click in the speaker, and the battery test does the same. Hmmm ... not completely dead then.Next - quick check of the coil with the LCR meter. Checks out ok.
Now it's time to open her up.
First stop - the power supply. There's about +/-2.5V on the power supply rails, and now I'm getting confident that it's a simple problem after all. The trouble is that the 1266 schematic that someone posted on the Geotech site some time ago doesn't exactly match this detector. Luckily I have an official schematic for a 1265-X which is virtually identical, except for a few minor differences that are covered by the 1266 schematic. In fact, both the 1265-X and the 1266-XB are almost identical on paper, apart from the PCB layout, which is completely different.
After some careful eyeballing of the PCB - no obvious problems there, nothing burnt or broken - I tried to identify the 7 transistors that make up the power supply, and to take some voltage measurements to try and isolate the fault. There was something strange with the voltages around the first two transistors that didn't make sense, and then I suddenly noticed that one of these transistors was inserted back-to-front!
A bit more eyeballing of the PCB, and I realised that this detector had already undergone some repairs. The transistor should have been an MPSA13 but an MPSA14 had been fitted. Actually this is ok as it's an equivalent. Surely this couldn't be the problem after all? The desoldering tool went into action and the offending transistor stuffed into the transistor tester. Good - it seems to have survived. Back in the right way round and switch on ... still dead. But this time the voltages make more sense, and it's clear that something is not connected right. Damn! There it is - a broken track on the top layer leading to the transistor base connection. Thoughts of "what a botched repair someone's done here" come to mind. Anyway, one jump lead later ... switch on ... guess what? ... yep, still dead. However, the good news is the power supply is now +/-4V, which is correct! What next? TX is driven by a crystal oscillator - this is OK. Voltage across TX coil also looking good. Output of RX preamp also OK. Phase reference circuits all working. Phase discriminators also functioning. Starting to get even more confident.
I then jumped forward a few stages and measured the outputs of the static, resistive, and disc channels. Wow! All OK!
Now at this point I know most of the design is working, but there's still no output. I also know that the speaker is connected because it clicks when I turn the detector on. What if the audio oscillator is not functional? Sure enough, dead as a dodo. After a lot of probing around with the scope and some eyeballing again, decided that the LM339 comparator is dead. Trying to get this out of the board without destroying any of the tracks was a challenge, but achieved without mishap. On a whim I then decide to fit an IC socket, which is just as well because (believe it or not) the first LM339 I fitted was faulty.
Now there's a faint sound from the speakers and/or headphones. Damn - still another fault. This time I traced the fault to the final transistor in the audio amp. Another bl**dy MPSA13. Replaced with a BC517 (which is also a darlington pair) - and ..... a really loud audio signal!
Finally I adjusted all the internal presets according the manufacturers instructions, as I couldn't trust they were correctly set up by the previous "repairer", and resoldered a broken wire to the charging socket. One wire going to the DISC1 pot was hanging on by a thread, and also needed attention, and I replaced the bolts holding the front panel to the PCB with some that had a smaller head, since the existing ones had caused a shorting problem at some previous time.
So now I have a fully working Fisher 1266-XB.
Tomorrow I plan to remove the electrical tape from the coil cable and fit some heatshrink sleeving instead. Might even extend the cable, as I'm sure it's shorter than it should be.
Anyway, just thought I'd share my experience with you about buying broken detectors on eBay. In my feedback I've mentioned that it's all repaired and working. Now I'm just waiting for the email asking me how I did it!!
Unless of course he reads this ...
Comment