Noise-Reduction for Fisher 1266-X (evtl. 1270-X & 1236-X2)

[Funfinder]

Yesterday I modified my old Fisher 1266-X because of too much noise. The "snap, crackle and pop" was even with minimum sensitivity too extreme because of nowadays electromagnetic pollution. The blue marks on the platine represents the 150kOhm resistor connection.

I'm also added 25 Ohm resistor to the speaker because of too much loudness and for power saving.

Following 2 digipix:
I haven't located oder added this modification on the schematic by now but if someone is really intererrested I could look where this new resistor has to be added.

Now the Fisher is still as sensitive like before but almost without any maddriving sounds / metal-signal overloading interferences! Still tunable to more "background-noise" by nearby variable resistor.

[Qiaozhi]

Yesterday I modified my old Fisher 1266-X because of too much noise. The "snap, crackle and pop" was even with minimum sensitivity too extreme because of nowadays electromagnetic pollution. The blue marks on the platine represents the 150kOhm resistor connection.

I'm also added 25 Ohm resistor to the speaker because of too much loudness and for power saving.

Following 2 digipix:
I haven't located oder added this modification on the schematic by now but if someone is really intererrested I could look where this new resistor has to be added.

Now the Fisher is still as sensitive like before but almost without any maddriving sounds / metal-signal overloading interferences! Still tunable to more "background-noise" by nearby variable resistor.

I have a Fisher 1266-XB, and would be interested to see where this resistor is located on the schematic.

[Funfinder]

Hi Qiaozhi, I will check it out.
If platine layout is the same, you can solder the resistor directly at the smooth side.

[Funfinder]

@ Qiaozhi
It's between Op Amp TLC27 (U 301) Pin7 (Output 2) and the middle of R327-R328.
or in other words: between [STATIC 1] and [DISC 1]

Replacing variable resistor R329 100k with 150k-200k (if such exist it all...) should also work.

If needed I'll add the mod to the whole schematic found here and upload it again:

http://www.sendspace.com/file/poce13
http://www.geotech1.com/forums/showt...t=14422&page=6

[Funfinder]

Update:
If someone don't want to mod the unit he can use a very small screwdriver and turn the little screw of the blue box (Variable Resistor 305) a few 360° turns clockwise.

This affects the sensitivity push-pull knob for motion and pinpointing which has a relatively large bandwidth.

Turn the knob completly clockwise (to highest sensitivity), push the pinpoint trigger and adjust the blue V.R. 305 (on the above picture it's directly below the other variable resistor) so you still can hear the threshold sound, but only 1/3 loud or even below that.

That way you have still enough sensitivity at maximum turn, but much less background or almost no noise at low sensitivity.

I also heard that sometimes the Fisher coil-cables have breaking up or soldering problems. Shake the cable with caution and listen if the rattling noise starts and stops. If so you have to open it up, solder and isolate until the connection is ok again.

[Qiaozhi]

Update:
If someone don't want to mod the unit he can use a very small screwdriver and turn the little screw of the blue box (Variable Resistor 305) a few 360° turns clockwise.

This affects the sensitivity push-pull knob for motion and pinpointing which has a relatively large bandwidth.

Turn the knob completly clockwise (to highest sensitivity), push the pinpoint trigger and adjust the blue V.R. 305 (on the above picture it's directly below the other variable resistor) so you still can hear the threshold sound, but only 1/3 loud or even below that.

That way you have still enough sensitivity at maximum turn, but much less background or almost no noise at low sensitivity.

I also heard that sometimes the Fisher coil-cables have breaking up or soldering problems. Shake the cable with caution and listen if the rattling noise starts and stops. If so you have to open it up, solder and isolate until the connection is ok again.

As far as I can see then, this mode simply changes the discrimination bias, which leads to some loss of depth. This is probably not a bad thing though, as the 1266 is almost impossible to use in the field on maximum sensitivity, and in most cases needs to be set to minimum.

[Funfinder]

I've heard alot Fisher Models have big problems with electricity-lines and heavy mineralized soil.

My Fisher 1266-X "hears" a High-Voltage-Line still from around a half kilometer or half mile and near railroads the same prob.

Reducing the bias is good as long as the noise destroys good signals even at minimum sensitivity or highest discrimination level (but in this case it drives discrimination useless...)


Better would be eleminating or filtering out those terrible "wave peaks" at all. Because those also start on heavy mineralized soil and hot bricks the problem is somewhere within in the circuit.


A very simple solution to the whole problem:
Raising the bias again but sending the final audio signal before the speaker through an IC who only let passes through signals with a certain lenght, perhaps above 0,5 seconds! This circuit should be bypassed in pinpoint mode.

There is a special amount of time usable or good signals have, below it's "snap crackle and pop". Good targets create almost a 1sec signal.


A second chance could be eliminating the short peaks already before they arrive at the amplifier circuit. Because they're having 0,3 to 0,0x Hz even a large condenser won't work for shortcutting those.

If we just knew what creates those high peaks at all. Overloads the coil itself at many occasions and why do other MDs don't have this problem?

I will continue the mod tests.

[Qiaozhi]

I've heard alot Fisher Models have big problems with electricity-lines and heavy mineralized soil.

My Fisher 1266-X "hears" a High-Voltage-Line still from around a half kilometer or half mile and near railroads the same prob.

Reducing the bias is good as long as the noise destroys good signals even at minimum sensitivity or highest discrimination level (but in this case it drives discrimination useless...)


Better would be eleminating or filtering out those terrible "wave peaks" at all. Because those also start on heavy mineralized soil and hot bricks the problem is somewhere within in the circuit.


A very simple solution to the whole problem:
Raising the bias again but sending the final audio signal before the speaker through an IC who only let passes through signals with a certain lenght, perhaps above 0,5 seconds! This circuit should be bypassed in pinpoint mode.

There is a special amount of time usable or good signals have, below it's "snap crackle and pop". Good targets create almost a 1sec signal.


A second chance could be eliminating the short peaks already before they arrive at the amplifier circuit. Because they're having 0,3 to 0,0x Hz even a large condenser won't work for shortcutting those.

If we just knew what creates those high peaks at all. Overloads the coil itself at many occasions and why do other MDs don't have this problem?

I will continue the mod tests.

Have you tried calibrating your detector according to the service manual?

[Funfinder]

I have no service manual, just user guide.

But it wouldn't make much difference, 'cause I know the few pots already well and what they're good for.

Those e.m.-interference-noise is a generally problem of old (and afaik even new!) Fisher models, but perhaps we can solve it somehow.

[Funfinder]

Anti-Noise Modification works great!

I've done it!

Back to almost full bias without noise!

Depth gain: 50% !!!
Ring: 10cm before & 15cm after mod.

Disc Bias variable R. 325 still tunable.

Only 1 small 10 pF condenser needed.
Soldering very easily on smooth side.

Conclusion: The already in use C311
with 22 pF should have 10 pF more.

Schematic with modification is here:

http://www.multiupload.com/24F4VVEKZQ

[WM6]

I've done it!

Back to almost full bias without noise!

Depth gain: 50% !!!
Ring: 10cm before & 15cm after mod.

Congratulation!
Almost unbelievable without afect other caracteristic?

[Funfinder]

5pF is even better!

http://www.multiupload.com/JOZ2JAJ2ZI

With 5pF a discriminated object remains as good as completly silent while with 10pF a very short pulse can still be heard. It's possible to distingiush a good signal from that one with ease, but 5pF is almost the same in bias-level and makes discrimination much easier.

On the very above picture with the electronic parts you can see at the center right border the white variable resistor for bias.

Tuning it ca. 1/3 clockwise from the very left start gives a very good sensitivity and almost no noise.

Good luck!

[Funfinder]

Thanx for your interrest, WM6, perhaps even you can learn something here.

The effect was intended and works fantastic!

[Qiaozhi]

I've done it!

Back to almost full bias without noise!

Depth gain: 50% !!!
Ring: 10cm before & 15cm after mod.

Disc Bias variable R. 325 still tunable.

Only 1 small 10 pF condenser needed.
Soldering very easily on smooth side.

Conclusion: The already in use C311
with 22 pF should have 10 pF more.

Schematic with modification is here:

http://www.multiupload.com/24F4VVEKZQ

Hi Funfinder,

By changing the capacitor value in the feedback loop you are modifying the cutoff frequency of the second-derivative filter. However, you stated in the text that you have added only 10pF more capacitance, and that the original capacitor value was 22pF. In fact, the original capacitor value was 22nF (0.022uF), and according to your photo of the PCB, you have actually fitted a 100nF (104) cap.

This means the original cutoff frequency was 7.2Hz, and you have lowered this to 1.3Hz.

[Funfinder]

Wow, that was really a good explanation! Thank you!

Of course: nano!

Seems around 40-60 nF in total would work best.

But anyway, we can use already the cutoff-frequency for eliminating more or less disturbing interference with best sensibility.

The Fisher is no longer an Electro-Smog Detector!