Stability of the VDI scale

[pito]

to detect metal you need to swing the coil all the time because synchronous detectors are AC coupled, in static mode (pinpoint ) they are DC coupled.
AC coupled amplifiers are not sensitive to DC drift.

[JoyJo]

Everything is right. We're talking about the same things. I wanted to clarify this. For a fully analog detector, what you have indicated is undoubtedly true: the timing of the demodulators depends on the TX signal. Therefore, the system automatically adjusts to the "changes" in the temperature caused by the temperature and something else. I am interested in the question in relation to detectors with fully digital signal processing.

[pito]

https://tehnodium.ru/thread-967.html

[Altra]

You can sample at 4x the tx frequency. Four equal samples alternating R,X,-R,-X this will be less sensitive to synchronization.

[Carl-NC]

no matter what parts and coils are used, there will still be phase drift.

Why is that, what are the sources of drift?

[JoyJo]

I mean, it will not be possible to ideally select the components depending on the change in parameters due to temperature. It may be possible to select resonant capacitors for the TX and RХ circuits so that the capacitance is compensated. But the influence of the soil on the inductance of the TX and RХ coils cannot be canceled. The inductance will change - the phase will shift away.

[moodz]

[JoyJo]

I had this detector and many more homemade ones. I know how ferrite calibration works. The Quasar ARM is a very good detector. But, unfortunately, it also has a drawback, which I started the conversation about: over time, the VDI scale shifts slightly. Subsequent ferrite calibration eliminates it. But this effect will manifest itself again. That's why I'm interested in how to make sure that one particular detector with one particular coil has a stable VDI scale that does not have a bias, as in branded detectors.

[Carl-NC]

I mean, it will not be possible to ideally select the components depending on the change in parameters due to temperature. It may be possible to select resonant capacitors for the TX and RХ circuits so that the capacitance is compensated. But the influence of the soil on the inductance of the TX and RХ coils cannot be canceled. The inductance will change - the phase will shift away.

Ground effect also causes the TX circuit loading to change, and you can detect this with careful monitoring of the power supply. Some commercial detectors do this, and adjust to compensate. But ground effect is instantaneous, as soon as you lower the coil to the ground you get loading. You have been alluding to a more gradual shift, so I doubt this is the cause.

I had this detector and many more homemade ones. I know how ferrite calibration works. The Quasar ARM is a very good detector. But, unfortunately, it also has a drawback, which I started the conversation about: over time, the VDI scale shifts slightly. Subsequent ferrite calibration eliminates it. But this effect will manifest itself again. That's why I'm interested in how to make sure that one particular detector with one particular coil has a stable VDI scale that does not have a bias, as in branded detectors.

What does ferrite calibration do? Does it rotate the software demod timing until the R value is nulled?
After calibration, the only way the VDI can shift is if the RX phase relative to the TX phase is changing. This is something you can monitor and measure on the bench and probably determine what is causing it. I suspect caps, either the series TX cap or RX caps if the RX coil is resonated. It could also be TX wire resistance. What type of caps are you using? What is the TX coil wire gauge? Something in the circuitry is causing this problem and there is no way to solve it until you figure out the cause. There is no magical "commercial detectors do it this way" solution.

[Gwil]

Something I don't understand - What does "calibration" mean in this context? If I measure a voltage with different voltmeters they should (ideally) read the same. The process of calibration should ensure this, within limits.
If presented with one particular kind of target, should all brands/models of detector give the same VDI reading? Or does it vary depending on the design?

Thanks

[Gwil]

Nobody feels like answering my question?

[Carl-NC]

Different companies have their own different VDI scales, and even different models within a company can have different scales. But 2 units of the same detector model should reasonably give the same VDI result, which is what they are calibrated to do.

[Gwil]

Thanks Carl - That's exactly what I wanted to know.