LM394 on my Minelab PCB

[Mechanic]

Hi Mario,

LOL! You can try.... Some guys have replaced the ne5534's with 797's, I'm sure it will work but how well? I tried looking for a better alternative for the 5534 but could not find one and gave up! The critical stage in the front end is the preamplifier. The next critical stage is the integrators.

When I go looking for noise, I have the coil connected, connect the receive wire to coil gnd. Any noise seen on the testpoints on the audio board is then internal noise(20mv/div .2s sweep speed). I then short the output of the preamp to gnd to determine if any noise is coming from the preamp. I then do the same for each stage after that up to the low pass filters. You will most likely find that some noise is from the front end the rest from the integrators.

Cheers Mick

[Stefan]

Hi Mick,

I have been poking around the front end amps and the two NE5534 downstream of the AD797 form a 4 pole LP filter with a 3 db cutoff around 150Khz. That is probably why Mario is seeing a slightly different waveform.

Regards,

Stefan

[mario]

Is that where that annoying chirping originates? I notice it worsens when i play with the earthfield trimmers. Maybe I should skip replacing those three 5534 ICs and change the 4066 chips... ?

Hi Mario,

LOL! You can try.... Some guys have replaced the ne5534's with 797's, I'm sure it will work but how well? I tried looking for a better alternative for the 5534 but could not find one and gave up! The critical stage in the front end is the preamplifier. The next critical stage is the integrators.

When I go looking for noise, I have the coil connected, connect the receive wire to coil gnd. Any noise seen on the testpoints on the audio board is then internal noise(20mv/div .2s sweep speed). I then short the output of the preamp to gnd to determine if any noise is coming from the preamp. I then do the same for each stage after that up to the low pass filters. You will most likely find that some noise is from the front end the rest from the integrators.

Cheers Mick

[Mechanic]

You been fiddling with the trimmers? Naughty boy! After work I will explain what they do and how to adjust them properly, otherwise you will wrap the detector around the nearest tree!

Cheers Mick

[mario]

[gef12]

Thank you Mick,

so it sounds like the LT1028 will be a slightly better replacement, but only for the AD797.

Now that I got my machine stable I will be replacing the AD797 and search for a higher impendance chip to replace the 5534 chips later.

What impendance and/or other characteristics should I match?

I had started playing and swapping out my front end opams... if i was you ..leave the last 5534 there .. I tried many different combos..
currently my 2000 has lt1028 x 2 and the 5534 and thus far think its the best for my unit.

[mario]

thanks Gef!

I had started playing and swapping out my front end opams... if i was you ..leave the last 5534 there .. I tried many different combos..
currently my 2000 has lt1028 x 2 and the 5534 and thus far think its the best for my unit.

I greatly appreciate the advice Gef. I will probably begin by replacing the two 5534 IC's between the AD797 and the last 5534 with LT1028s. I have 4 LT1028 chips, but I really can't justify replacing the 797 chip or the last 5534 since I do need it for the higher impedance issue mentioned by Mick.


Anyways, I'm done with my 3 hr testing and I can confirm a slight improvement on stability of the threshold and very noticeable improvements on start up times of the detector and EMI. Also, the detector recovers a tad faster after a large trigger.





After researching alternatives for other IC's found on the PCB, I found the following:
LT1057S8 with OP249 (a direct replacement!). I believe the SD2000 uses LT1057ACJ8, which is not pin compatible with the LT1057S8. Nevertheless, I also found a better direct replacement with the OPA2134 for the LT1057ACJ8.

Some specs for the OP249:

Fast Slew Rate: 22 V/us Typ
Settling Time (0.01%): 1.2 us Max
Offset Voltage: 300 uV Max
High Open-Loop Gain: 1000 V/mV Min
Low Total Harmonic Distortion: 0.002% Typ
Improved Replacement for AD712, LT1057, OP215,
TL072, and MC34082

Anyways, I notice that we seem to be improving the threshold characteristics. However, I'm really interested in improving first and for most, sensitivity to smaller objects, secondly, depth.

Do you guys think we need to work on the transmit side to gain greater sensitivity?

An easy way to increase depth and possibly sensitivity is by increasing the voltage. However, I really don't want to increase the voltage, since this will be akin to following a brute force approach.

Any suggestions?

[Mechanic]

Hi Mario,

To start off with we should try speeding up the oscillator for better sensitivity to smaller targets and slow it down for a bit more depth on larger targets. I think the section I have pointed out in the pic is the oscillator that drives the rest of the cct. There should be a fast square wave from that cct going to the part marked !?X?!. The pin on that chip will be the clock pin, which puts another piece in the jigsaw puzzle as to identifying what it is! I have a feeling that if you change the values of some of the capacitors in the oscillator section that you will be able to change the frequency of the oscillator. But be aware in doing so you will change the range of targets that the detector will hear. If you make it sensitive to smaller targets it may not hear the bigger ones as well. Also there could be an increase in ground noise. But why not fiddle and learn! Don't blow it up

I still must go over the earth field cancellation procedure.... next time!

Cheers Mick

[Mechanic]

Hi Mario,

Just thought I'd add a bit to the last post.

When deciding which caps do effect the clock frequency start by just adding say a 1nf cap to the existing cap without removing it. When you find an area of interest carefully remove the cap but be careful that you do not remove the ring around the bottom of the via with the cap! Then when you have the cap removed solder some pins into the holes and then solder your experimental caps to the pins. This way you will preserve your board a bit longer

Cheers Mick

Ps Also the part that I told you originally was a micro, it is actually an eeprom

[gef12]

Tks Mario
I had replaced my LT1057 with the OP249 .. and did not really notice a great improvement.. .. all thou there could have been a minor improvement .. so I have just left them in ... still works great

[mario]

Thanks Gef,

I think we've done a pretty good job at stabilizing the threshold.

Most of those other ICs are Op amps: the 797, 5534s, even the LT1057. Unfortunately for us, those ICs had some respectable specs, so there is not much we can improve by replacing them with other chips with better specs.

My next step will be identifying that "X" IC. I want to see how it works in relation to the surrounding ICs. It will be awesome if we can get the sensitivity up to par with the gpx. I think understanding that IC is the key...

Unfortunately for me, I blew a IN4148 diode, the one that is connected to the "Source" of the P fet. I must have weakend it last time I incorrectly soldered those fets and blew the 4093. Now it's shorted and I'm getting an erratic signal. Anyways, my unit is going to be down until my replacements arrive.

I have already started fiddling with that "X" chip. I removed the white gunk from its surface and was some what amused about the minelab guys: they went out of their way to scratch off all identifiable numbers. They scratched deep, very deep!

[mario]

Does this start looking familiar?

I'll try to finish connecting the dots soon

[Mechanic]

Hi Mario,

Looks like the green trace is one of the power rails either +5v or -5v referenced to coil ground. What voltage is present? The yellow trace is the other power rail. Those jumpers do something, put your scope on the coil and swap the jumpers around and see what happens.

The fact that the trimmer for the frequency is in that box I have hit the nail on the head for the oscillator section. Perhaps take some time and draw up a diagram of the parts there and make yourself a schematic. When you tune the trimmer do you see the frequency at those capacitors(they are not xtals) change a little? It won't change a lot. What is the frequency?

Cheers Mick

[mario]

K, I must have been there a few hours tracing those pins....anyways, hopefully google refreshed the picture down at your end. I think I've traced enough of that IC to at least get a few ideas as to what it is. Any suggestions?

I'll eventually post a schematic, that way some one could one day improve on the design (and hopefully tell us about it).

I'll play with the scope/jumpers tomorrow.

[gef12]

Mario .. once your done with fixing it all up and running nice.. try to up the resisistor 470 ohm across the 797 ... maybe 500 or 510 ..should increase gain slightly on smaller stuff
gef