Replies to Minipulse Plus

[green]

Already tried the 100R trimpot - you could be right about it not making much difference. Would a higher value make any difference and, if so, what would you recommend?

Look at integrator out, TP4 with a scope. Hold ceramic or fridge magnet next to coil and pull away quickly. Try with pot adjusted each end, if pot is high enough resistance polarity at TP4 should change. Don't have a MPP but works with my detector.

[Goaty]

I don't think that will make any difference, as it's only used for fine tuning if you're not using 1% tolerance resistors.

What are the details of your coil?
Also, I remember a long time ago that I once accidentally connected the coax screen to the the TX output, and the central core to 0V by mistake. It worked fine on the bench, but was falsing everywhere in the field. So, I'm just wondering if it might be a shield problem.

Already tried the 100R trimpot - you could be right about it not making much difference. Would a higher value make any difference and, if so, what would you recommend?

The coil at the moment is bundle wound stranded Cu wire with Cu foil shield. 431 uH & 1.6R resistance 250mm diam. Damping R 700R - set with variable resistor network to just remove any ringing.
Meter check confirms connections correct.

I did not follow up shorting C12 and removing R19 as it seemed pointless when increasing C12 to 22uF showed no improvement.

The component values in the Sampling Integrator were supposedly 1% (Chinese) but I checked them all with a meter as I was building it.

Starting to get splinters in my fingers with all of the head scratching over this - may just have to slow down and live with it for now. The warmer weather is starting to arrive here so I will give it a good test on the beach soon and see how much of a nuisance the falsing is.

Thanks once again for your suggestions - really appreciated!

[Qiaozhi]

I think I will try a 100R balance trimpot between R24 & R25 with wiper to C12. Will leave the 22uF across C12 in place

I don't think that will make any difference, as it's only used for fine tuning if you're not using 1% tolerance resistors.

What are the details of your coil?
Also, I remember a long time ago that I once accidentally connected the coax screen to the the TX output, and the central core to 0V by mistake. It worked fine on the bench, but was falsing everywhere in the field. So, I'm just wondering if it might be a shield problem.

[Goaty]

Recalibrated damping resistor on bundle-wound coil. Checked all test point waveforms - all good and to spec.

Still falsing!!

Swapped out U5 as TP3 showed slight distortion - still falsing!

I think I will try a 100R balance trimpot between R24 & R25 with wiper to C12. Will leave the 22uF across C12 in place

[Goaty]

It would be easier to test the Earth field elimination by waving a white board magnet a few inches away from the coil.

Agreed but I was curious as to why it only sounded when the swing changed direction and it would indicate the the falsing is not due to any mechanical deflection with direction change as the forces are all constant with circular motion other than the way the coil is interacting with the lines of force of the Earth's magnetic field.
(Also, I don't have any suitable ceramic magnets handy - only rare earth types with metal coatings ex hdd)

This would still indicate that the problem is in the EFE circuit somewhere - It's just a matter of finding it!

[Qiaozhi]

Strange test – Strange results!



A search through my ‘Junk Box’ turned up a 22uF electro so I tacked it across C12 and took the detector outside for a quick test. It still seemed to be falsing on anything but a slow swing, even at head height.
Then I had an idea – what would happen if I swung it in a complete circle? (other than getting dizzy and falling over )

With the shaft horizontal at waist height and the coil vertical I rotated clockwise in complete circles and it only falsed on the Eastern side

When I rotated anti-clockwise the falsing occurred on the Western side of the circle.
Not too sure what to make of the results – seems to be magnetic field related but too dizzy to figure it out at the moment
.

Has anyone any ideas on how to interpret these results?
PS I am in the Southern Hemisphere

It would be easier to test the Earth field elimination by waving a white board magnet a few inches away from the coil.

[Goaty]

Strange test – Strange results!



A search through my ‘Junk Box’ turned up a 22uF electro so I tacked it across C12 and took the detector outside for a quick test. It still seemed to be falsing on anything but a slow swing, even at head height.
Then I had an idea – what would happen if I swung it in a complete circle? (other than getting dizzy and falling over )

With the shaft horizontal at waist height and the coil vertical I rotated clockwise in complete circles and it only falsed on the Eastern side

When I rotated anti-clockwise the falsing occurred on the Western side of the circle.
Not too sure what to make of the results – seems to be magnetic field related but too dizzy to figure it out at the moment
.

Has anyone any ideas on how to interpret these results?
PS I am in the Southern Hemisphere

[Goaty]

Thanks for that info George. I will play around with it next week.

[Qiaozhi]

(5) Play with the value of C12 (coupling cap.) as suggested in a post by Eric. He suggested an electro up to 22uF or even DC coupling and that scoping across R10 should show a flat DC baseline. From memory, I think one of the pulses in this area showed a slight overshoot – will have to go back and check that again.

When I back-engineered the original Minipulse design, the value of C12 was found to be 470nF, and that's why the Minipulse Plus uses the same value.
With 470nF, the corner frequency of this high-pass filter is ~34Hz, and with the 22uF (suggested by Eric) it's 0.7Hz.

While experimenting with a different design to the MPP, I encountered the same problem with less than perfect Earth field elimination (EFE), so decided to investigate further. This design was also using a high-pass filter between the preamp and the sampling integrator, with a capacitor value of 470nF, and there was a definite improvement in EFE when this capacitor was increased in value. However, there was a much greater improvement when the capacitor was shorted out, and the resistor to ground (R19 [10k] in the MPP) was removed. I'm not sure whether this is possible with the MPP, as it does not possess the ability to set the null-offset of the preamp, so maybe you can try this and report back. If increasing C12 to 22uF, or shorting C12 and removing R19 does not do the trick, then you will need to look elsewhere for the problem.

[Goaty]

Jfet swap (Q3 & Q4) made no difference to falsing - starting to think sampling integrator section is OK.

Tried tapping/bumping both coils - no problems.

Next!!!!????

[Goaty]

@ Qiaozhi

Yes, I'm sure I'm missing something obvious but at least I'm learning as I investigate.

I contacted Teleno as he was having similar problems with his but he had not found a solution either.

I think I will try a more closely matched pair of jfets next before giving the sample pulse generator section another going over.

A bundle wound coil I tried earlier gave the same effect as the 3DSS I have on it now so I assumed that the coil was not part of the problem.

BTW, the resistances quoted above for R20 & R23 were 'in-circuit' values - when I pulled them and re-measured they were both 6k8.

The puzzling thing is that the rest of the circuit works quite well - the detection depth is as good, if not better that, the Surf PI 1.2.- or I could just learn to ignore the falsing and learn to swing more slowly!

Anyway, thanks again for your suggestions.

Cheers
Goaty

[Qiaozhi]

I've finally found the time to start removing and checking some of the components in the sampling integrator section and would appreciate some advice on a couple of my findings -

(1) The jfets Q3 and Q4 give the following results on my component tester - Q3 IDSS 1.5Ma Vgs 1.1v, Q4 IDSS 1.5Ma 1.1V. Are these closely matched enough?

(2) R20 & R23 measure 6k25 ohms each instead of 6k8. Should I change them to 6K8 even though they match.

I have measured C13 and c14 they match OK at 450nF instead of 470nF.

Any advice would be appreciated.

I don't understand why you're having problems with Earth Field Elimination on the MPP. It should be working ok. As you can see from the Completed Projects list, several others have not experienced this problem.

I'm wondering whether you're chasing a red herring, and you either have a fault on your board or there's something up with the coil.
What are the details of your coil?

[Goaty]

I've finally found the time to start removing and checking some of the components in the sampling integrator section and would appreciate some advice on a couple of my findings -

(1) The jfets Q3 and Q4 give the following results on my component tester - Q3 IDSS 1.5Ma Vgs 1.1v, Q4 IDSS 1.5Ma 1.1V. Are these closely matched enough?

(2) R20 & R23 measure 6k25 ohms each instead of 6k8. Should I change them to 6K8 even though they match.

I have measured C13 and c14 they match OK at 450nF instead of 470nF.

Any advice would be appreciated.

[Goaty]

Just a (lack of ) progress report :-


After much reading forum posts and studying the MPP circuit diagram I have listed my options for dealing with this problem -


(1) Fit a low value balance trimpot between C12 (wiper) and R24, R25.


(2) Measure and match C13 & C14 (470nF)


(3) Match the parameters of Q3 and Q4 JFETs.
I’m not sure just which parameters – possibly IDSS and Vgs, and how close in value they should be.


(4) Match the ‘main’ and ‘sample’ pulse widths more accurately.


(5) Play with the value of C12 (coupling cap.) as suggested in a post by Eric. He suggested an electro up to 22uF or even DC coupling and that scoping across R10 should show a flat DC baseline. From memory, I think one of the pulses in this area showed a slight overshoot – will have to go back and check that again.


Can anyone suggest any others?

[Goaty]

@ kt315

Thanks for the words of wisdom kt.
Yes, this metal detector caper has opened up a whole new aspect of electronics for me.
I have been a hobbyist/part-time instructor/ amateur radio operator/ alarm tech in electronics for close on 50yrs and still learning. Metal detector technology has turned out to be a whole new ball game!


My biggest problem nowadays is that since retiring I don’t have time to do everything I want .

I’ve built 3 or 4 different detectors but seem to spend more time building them than actually using.


BTW, my intro to computing came mid way through my teaching career when it involved marking ‘punch cards’ and then sending them off to a computing centre to be processed. The success rate with a class of 30 kids was not very high!


@Wahyu


Thanks for that link – more good info there to add to my collection.
I’m starting to formulate a plan of attack now – almost time to fire up the soldering iron.