Threshold on the SD2000

[Ismael]

To start this thread I have always been looking at ways to stabilise the threshold but couldn't see how to make any changes. Basically the threshold circuit is a pot attached to +5 and -5V rails with the wiper going to the -input (via a 3.3mohm resistor) to a TLC074 (pin 9 op-amp 3) and also to pin 12 on a 3083 (discussed in the technical forum in post https://www.geotech1.com/thuntings/s...ad.php?t=13385 ) the 074 output of opamp 3 then goes to pin 10 on the 3083. The wiper is also connected via a 10k resistor to pin 6 on a TLC27M7 (-2) which is also shorted to 2out (same chip) through a cap and resistor to pin 14 on a 4053. This is all the threshold pot connects to. I will try to post the circuit later on just to visualise things a little better. The + input on the 074 goes to pin 12 on the 074 (+input #4 opamp) the rest I can't remember. Now I don't know whether by changing the 074 op-amp makes a difference, I don't think changing the set resistor does anything. Also as an after thought sometimes ML puts a diode between pins 10 & 12 on the 3083 with the anode going to pin 12. I have tried this on detectors which don't have it and could hear any difference.

I hope I am making sense so far. It seems that all the threshold does is change the differential between the 2 voltages to the op-amp (+5 one way and -5 the other).

If there was someway of stabilising this part of the circuit or maybe I am looking at it from the wrong way, maybe I should try to stabilise the signal before the threshold. Any thoughts on this?

I will over the next few days post some drawings I have on the 2000 but they are not complete but do show most. I had started to do the layout but never seem to keep up with it but now with all this activity I might continue. Does anyone know of any Electronics program that can create a circuit diagram from a PCB layout with nets? I am using DXP and Altium explorer but had designed the layout with old Protel 99 and the problem now is that when I started most of the ic's were unknown when I try now to use the proper IC the footprints are bigger or smaller than what I originally designed. I suppose I will just have to bite the bullet and start all over with the proper IC & footprints and re-lay the tracks.

Regards Ismael
www.minelabmods.com

[Ismael]

Here are some drawings and boards:

Both PCBs from my first 2000 still gutted!!


These are a couple of layouts with some tracks that I was given: Board 1



Board 2:



Regards Ismael
www.minelabmods.com

[Woody.au]

Well Ismael, I totally stuffed up on what I thought was the coil driver fet, it is not the IRF710 but the pair of BUK455 that are connected in parallel. Now I suspect that the IRF710 is for energy recovery as the detector works with this fet disconnected from the circuit! I think that there is meant to be a pulse that goes to the gate of this fet from the output of the TLC271 but I only have 2 volts of DC at this point. No wonder the fet was getting hot...

The 2 BUK455 fets get a little warmer with the irf710 out of circuit, The out put from the TLC271 also goes to the input of a LM393 as well as the fet gate resistor, I replaced the 393 but no change, will have some TLC271 chips in the morning to if it is the culprit. Now I wish I had looked at the gate of the irf710 before going the hog and carrying out so many mods What I think happened is that I grabbed the board while it was turned on and the 200 volt flyback on the drain went into the gate compliments of my finger.

Keep up the good work.

[ZED]

Yea Woody, with the irf7 series fets i have looked at the gate signal on both the 2000 and the 2200 and have never been able to see any pulse signals.
I dont think there is anything wrong with your machine,if the fet gets warm to hot then every thing should be ok.

ZED

[Woody.au]

I have a feeling that the IRF710 is only a voltage limiter, I traced the output from the Fet and it goes through one diode and 2 resistors to the input of the TLC297, no pulses there? After checking all the associated parts I think that the job of the IRF710 is to limit the voltage across the coil, the detector works fine without the IRF710 in circuit. To me it looks like the IRF710 is operated in its resistive region, I think it maybe has something to do with the way M/L snap off the coil current before saturation. I will do some more experiments just to make sure the detector is working the way it is supposed to operate.

[ZED]

Woody you might like to consider some IRF640N as alternatives to the buk455.
If you do try them have a look at the voltage transients in the power supply when the coil switchs.

ZED

[Woody.au]

I put in a pair of W20NK50Z in a To247 package. That IRF710 just clamps the voltage so that the flyback pulse does not get over 190 Volts,When I disconnected it the flyback voltage shot up to 382 volts!!! I have just installed the new front end fets that I had roll up today, that P channel fet in the input when changed makes a huge difference. Look at the pointer in the pic pointing at the new fet, can you see the little black spot? The solder pads are bigger than the fet.