XP deus coil modification

[Skippy]

18mV pk-pk is not so bad, but it's the kind of null that entry-level machines (Garrett ACE250, Fisher F2,etc) tend to have. The machine obviously seems to work with this level, but you should aim to better it. As you are now powering it from the 18650 cell, I guess you could put the foil-pouch LiPo cell inside the coil, where it originally was ,and see what happens. Just for curiosity, really.

[Henrikras]

Good idea Skippy - to do a test with the Stock Battery in place. Even could find a small board with som few electronic components and place that allso, just to see the Null Maybee changes.
A Shame you couldn't watch the video, It shouldn't Be so difficult to do. I need your or others help to understand the phase change. If you or some had time to explain that Maybee I got it allready, just hard see so small differents at scope !!!
Henrik

[Skippy]

I will try and watch the video tomorrow, on a better computer.
I was wondering if the two SOT23-5 packaged devices were part of the amplifying or demodulating circuitry - there is two identical looking circuits, caps, resistors, very analogue , on the opposite side of the PCB to the pre-amp. But... the pinout that opamps normally use doesn't seem to match the way they are connected. It might be worth a poke around with a meter / scope, see what's going on there.
I'll try and come up with some explanation of the phase business tomorrow.

[Henrikras]

I will try to do a test, the Stock Battery vs the new 18650 cell... Be cause the Stock cell got higher voltage than I thought. 3.9-4.0 volt.
I need to know the excact discharge curve both cells.
And one more thing I could do ! Scope prope 1 at RX and prope 2 at pin 1 / preamp.. To see any phase / small differences.

[Skippy]

It's normal to charge both types of Li cell (LiPo and LiIon) up to 4.2 V, I don't know what low-voltage cut-off is used, I guess 3.6V?

[Skippy]

OK, I just watched the video. I guess your null is too large to see much change on the scope when you bring targets near. Because the machine seems to work OK with this high null level, I reckon you should be able to make your coil good enough without doing any more complicated measurements. Maybe you did not notice, but pin 1 of the pre-amp connects to a square test pad on the PCB. You could probably solder a short piece of flexible wire onto this pad if needed.

[eagle52]

My Chinese 3.7 blue Li Ion always charge up to 4.2 V

[Henrikras]

Here is some test and reading... 4.20 volt should Be the max voltage (the cell can Be damage If it overcharge )
link 1. http://lygte-info.dk/review/batterie...mary%20UK.html
link 2. http://www.rchelicopterfun.com/rc-lipo-batteries.html

Concerning the Deus project, I just want to be sure that the Deus cell isn't different from others cells. Remember the charging unit is at the board (XP construction). I Need to know the excact charging cut off voltage (properly 4.2 V) and the excact low Battery level voltage.. (Led blinks).
In theory could XP got a special develop cell up to 4.35 volt ! The Battery issue at the Deus is Very sensitive, If the indikator shall work.
Not to forget the boards stabilized working voltage 3.3 volt - a different cell with 0.05-0.1 volt lower voltage could in my theory make a different, and I dont like that. Just want to Be sure...
Henrik.

[Skippy]

The charging circuit looks a bit mysterious. There's a diode in series with the negative connection, so that you can't short out the battery at the two charging pins. And there's a strange module of some kind, could be all sorts of things. Does the plug-in charger regulate the charging current?
The Deus cell will be a standard LiPo type, I am sure.

[Henrikras]

No - the plug in charger dosn't regulate. Ist a USB standard power supply. (5 V).
And you nailed it, the Stock Battery was a standard know cell. Ist started with 4.25 volt and after 7 hours with 45 ma discharge it was down to 3.77 Volt and the Battery controller said 14% remaining.. Same discharge time and current at the 18650 cell I got said 3.986 Volt and 80% remaining. The 18650 cell started with 4.18 volt.
Result; the 18650 will Be a good choice.
Henrik.

[golfnut]

Just sharing my view on the null etc.

If the ferrite is rejected, and the respective targets are ringing out correctly - then the 18 mV null is just what you get when coil is phased up correctly.


Any movement/null adjust will get lower null, But the phasing will change and then you may Not reject ferrite/play iron tone for Iron, etc.


I would leave well alone - sounds perfect as it is to me. only a view, I could be wrong !!

[Skippy]

Quote:"But the phasing will change and then you may Not reject ferrite/play iron tone for Iron"
I disagree. The detector looks at the changes in signal coming from the coil, and static levels, such as the 18mV pk-pk here, make no difference. .# This is good, of course, because 18mV is 2000 x bigger than a small target signal [rough estimate]. The potential problem with the null is this: it may not be the total amplitude of the null that is important, but rather the in-phase and 90 degrees phase components of the null signal. So 18mV might be fine if it was in-phase with the ground signal when demodulated, but if it were 18mV at 60 or 100 degrees, it may be too 'strong' for the circuit. Two (of several) difficult problems here: 1: we don't know what the demodulated signals look like - it's perfectly possible that demodulation is done in 'software' inside the PIC, and not on the analogue bit of the circuitry. 2: We don't know where to measure the signal, to say something is zero degrees, something is 60 degrees. Everything will be at different phases - the pre-amp likely causes a phase shift, the TX signal won't be in-phase with the logic signal that switches it etc, etc. So you need to aim for a good null, regardless of phase, and maybe be prepared to fine-tune it afterwards as well.

[golfnut]

Hmm maybe ?

Assuming the preamp has 25dB gain, then the coil has 80dB null currently, whites ~65dB

[Skippy]

To be honest, we know almost nothing about the circuit here, and thankfully, I don't think we need to know to get this project working well. My gut feeling is that the ADC is a 24-bit converter, and the pre-amp etc gain is not that high, the sensitivity is achieved simply because of the fine resolution. This in turn makes it less sensitive to the size of the null signal. This is handy, because it likely varies a bit from one operating frequency to another.
So, as an example: Instead of having a low-null coil, a pre-amp gain of 200, and a 18-bit ADC, it has a modest-null coil, pre-amp gain of 20, and 24-bit ADC. This is illustrative, I haven't any evidence, but I hope you get the idea.
I do know that in another modern machine, fairly conventional analog techniques are used at the front-end, with roughly 10 times more gain on the 90 degrees channel than the zero degrees. Hence why tolerable null levels are phase dependant.
Also, the circuit for the X Terra machine (705?) is available (those Russians, eh) and that uses little analogue, and 24 bit ADC.

[Henrikras]

Can't quit follow you, my knowledge is not good enough. Never mind me, I will learn.
Just tryed to put the Stock Battery back in place and the Null did go lower, not much but a little. (16-17 mv PP)
Then I tryed with the small piece of ferrite, and the Null can Be finetuned down to > 5mv, but I can't see ANY different in performance.
Everything Tested; ID at 5 different targets from iron to high conducted, same depht, same chattering with high sens., and same weak ID from deep targets.. Its seems to me that the Deus isn't Very difficult or sensitive to a good Null.
Just finish with the charge plug.. ! Mini USB just like the Deus control unit and Headphones. (See picture)
From now - coil production. Got all the AWG. Shell from DB is allso at its way, to make a test with a 7".
Thank you so fare..
Henrik.