IGSL

[liekwang]

?Davor I do not understand what is meant by AB--C

[Davor]

There are amplifier classes. Class A conducts current through an active element all the time, and it is mostly linear of all, but it sucks lots of current. Class B is a class where active element is conducting either positive or negative going signal, and switchoff happens at precisely zero volts. Class AB is in between, and most oscillators use it because it is linear for small signal, so oscillations may start. Once oscillation is established the linearity around zero volts is not crucial, so class C can be employed to reduce power consumption. Therefore AB->C

[liekwang]

?Thanks davor for the information

[ivconic]

........
Some have, more power can give you more depth, but use the car headlamp analogy.
...........
S

I like the analogy!

[sini]

Hi all,

I made IGSL musketeer coil. Everyone on this talked about deep coil nulling, so I deeply nulled my coil. I nulled my coil ( on the output of preamplifier ) to 14 mV. After this nulling was sensitivity to 1 EUR about 30 cm. I was satisfied.
Next day I switched on my IGSL. Epoxy in the coil dried and coil disballanced to 126 mV. Damn. But...I tried sensitivity and sensitivity is 40 cm to 1 EUR !!! Really. Discrimination and GB is perfect. I thought, that deep nulling is the main goal ( the holly grail ) of nulling. Can someone explain me, why is disballanced coil more sensitive than deeply ballanced coil ?

[ivconic]

Hi all,

I made IGSL musketeer coil. Everyone on this talked about deep coil nulling, so I deeply nulled my coil. I nulled my coil ( on the output of preamplifier ) to 14 mV. After this nulling was sensitivity to 1 EUR about 30 cm. I was satisfied.
Next day I switched on my IGSL. Epoxy in the coil dried and coil disballanced to 126 mV. Damn. But...I tried sensitivity and sensitivity is 40 cm to 1 EUR !!! Really. Discrimination and GB is perfect. I thought, that deep nulling is the main goal ( the holly grail ) of nulling. Can someone explain me, why is disballanced coil more sensitive than deeply ballanced coil ?

"on the output of preamplifier" ! ha!

Maybe your preamp involves some hums on existing signal!? Dc offset?
If you followed TGSL thread you may noticed my constant insisting on monitoring residual voltage at the coil pins and not after the preamplifier.
You can now observe residual voltage at very coil pins and see the difference, the "real" residual voltage.
Besides; good nulled coil doesn't mean lowest residual voltage on coil. Not the rule.
At such setups as IGSL is (Tesoro's) usually TX operates at one frequency and RX at slightly different (once i founded rough "rule" as 1.9-2.2kHz difference) , so is obvious that sightly "unbalanced" coil will perform better than "stiff" nulled one.
So if coil is nulled with lowest possible residual voltage at TX=8kHz; be sure it will not perform good at significantly different frequency than.
Yet, IGSL (TGSL too) setup allows slight residual voltage offset at coil balance and most probably, by funny coincidence, your coil balance entered in to most desired balance range, when dried!
With frequency change - residual voltage will change too. If your coil with present 126 mV performs nice than i can only suspect dc offset at your preamp.
But hey; if your setup works nice: than don't bother to analyze it much, leave it that way and be happy!

[Qiaozhi]

Hi all,

I made IGSL musketeer coil. Everyone on this talked about deep coil nulling, so I deeply nulled my coil. I nulled my coil ( on the output of preamplifier ) to 14 mV. After this nulling was sensitivity to 1 EUR about 30 cm. I was satisfied.
Next day I switched on my IGSL. Epoxy in the coil dried and coil disballanced to 126 mV. Damn. But...I tried sensitivity and sensitivity is 40 cm to 1 EUR !!! Really. Discrimination and GB is perfect. I thought, that deep nulling is the main goal ( the holly grail ) of nulling. Can someone explain me, why is disballanced coil more sensitive than deeply ballanced coil ?

The answer is simple [or maybe not that simple] - deep nulling is not the Holy Grail. It is much more important to have the correct phase-offset between the TX and RX signal.
From your scope shot, the phase-offset appears to approximately 7.5us, and the TX frequency is 8.4kHz. Therefore, 7.5u/(1/8.4k)*360 = 22.7 degrees. which just happens to be [more or less] the correct offset for this design.

[ivconic]

The answer is simple [or maybe not that simple] - deep nulling is not the Holy Grail. It is much more important to have the correct phase-offset between the TX and RX signal.
From your scope shot, the phase-offset appears to approximately 7.5us, and the TX frequency is 8.4kHz. Therefore, 7.5u/(1/8.4k)*360 = 22.7 degrees. which just happens to be [more or less] the correct offset for this design.

Yes, 20 degrees was the golden rule at TGSL.

[ivconic]

Just for curiosity , Sini can you measure residual voltage at the very coil pins?

[WM6]

What of relationship exists between TX-RX correct phase-offset and setting of maximum sensitivity? Is there any contrast between those two?

[Davor]

I suspect resin to act as a kind of "Q-dope" and improve Tx performance. Tx noise is very capable of reducing Rx sensitivity, and I lately focus on Tx improvement.
Reality check - when batteries go low, usually a whole rig becomes quirky and noisy, and the only component being directly affected by voltage fluctuation is Tx, so there you have it.
20° is a designed phase offset for this design, being a successor of TGSL. However, I learned that going for 30° may be useful in extending the Fe discrimination.

[sini]

Davor, I am looking forward to your new Tx oscillator design. I hope, that your new Tx design will not need center tapped Tx coil ( old musketeer coil will be useful ).

Where to measure phase offset ? Directly on the Rx coil pins or at the output of preamplifier ( C3 100pF feedback capacity increases phase offset )? I measure everything at the output of preamplifier ( residual voltage nulling, phase, noise ). I measure residual nulling voltage by osciloscope ( voltage is measured Peak to Peak - Dc offset is not counted ) . Is it correct method ?

[Davor]

IMHO yes, because a coil is put into it's working conditions within a tank, and oscilloscope probe is hooked to a low impedance preamp output. Otherwise there is little difference between a preamp input and a oscilloscope, and the difference comes as additional shift by feedback and coupling capacitors of a preamp.
My vision of a good oscillator is a solution that uses the same coil as the original. only with minor changes to improve noise and purity.
So in case of IGSL it must be a single ended solution with conduction angle just below 180° to promote signal purity.

[sini]

Hi,
I made VDI which I connected to IGSL ( X = U6 PIN14, Y = U6 PIN 1 ). I indicate few problems:
After first VDI switching ON - PIC controller is not right inicialized ( AD converters ). For right inicialization I have to do: swich OFF - wait 3 seconds - switch ON.( as is on the video )
VDI is not stable - if detector does not indicate any metal - VDI number varies from -80 to + 80 ( as is on the video ) . I tried to set threshold number from 1 to 10 but no success.
Is X,Y gain of MCP602 too high ?? What is the threshold voltage of threshold setting 1,2,..10 ?
Thanks, Sini

http://www.youtube.com/watch?v=L0dfbFoOCVA
http://www.youtube.com/watch?v=p_wCnwZ0sto

[Davor]

IGSL has a bit different approach to discrimination than the Fisher descents. Those devices have a fixed X and Y channels that get manipulated only after the points where VDI signal is taken, but not IGSL. With IGSL phases are manipulated at the switchers. This is beneficial for compression, but makes VDI a bit troublesome.
VDI can work with IGSL, provided you keep the discrimination setup for the channel you are sampling for VDI constant. That way you are reducing your discriminating abilities in audio, but your VDI becomes meaningful.