newbie with PI questions

[mikebg]

WARNING! VHF ringing
If you notice excessive heating of the MOSFET, it may be due to the VHF ringing because parasitic capacitances and inductances form Colpitts oscillator circuit . The VHF ringing is difficult to monitor with oscilloscope, but you can hear it with a FM radio. This effect is known to Ham radio designers using class C MOSFET amplifiers. To avoid ringing at turn off process, they connect one 10-20ohm resistor directly in gate lead of MOSFET. When two or more MOSFET are connected in paralel, damp resistor 10ohm must be connected directly to each gate.

At the time interwal when MOSFET or VDR or ZD is conducting because of breakdown , the energy accumulated in L and in both capacitances is dissipated as heat with great power, because in the formula for power P = Vz.I,
voltage Vz is great. When the zener current stops, the rest of energy dissipates at a maximum rate if the ratio between the resistors corresponds to the published above formula.
CONCLUSIONS:
1. The oscilloscope connected to resistor Rm can not measure this part of coil current that flows through coil's stray capacitance C.
2. The designer should take measures to reduce not only coil capacitance, but also TX networks capacitance.
3. When you measure resonance frequency of coil, dont disconnect it from TX network.
4. If the measured resonance frequency is too low, reduce coil inductance, but this reduces ampere-turns of TX coil.
　
　
　

[johnandles]

tryin to scare new members away ?

Thanks mikebg for your techno babble ..a new member only asked a simple question ..that required a considerate answer..
Please be kind.
Johnno

[Qiaozhi]

Quote: "The damping coefficient of a parallel RLC circuit is 1/(2RC), so low R results in longer damping. If you remove the damping resistor (R=∞) the PI coil will oscillate, which represents the fastest drop time". Carl
Carl, you reinveted a special case of Pulse induction named "Sine induction":-). However your network "LC without damping" represents infinitely long drop time; it oscillates continuously :-)

This is an incorrect statement.
The dc resistance of the coil (which you have termed "r") will cause the sine wave amplitude to reduce over time, as energy is dissipated. Not oscillate continuously, as you said in your reply.

[msysta]

Thank you again for good explanation Carl and Mikebg. This field just turns more interesting day by day. I should have understood this RLC thing from Mikes formula in the other thread but repetition is sometimes needed...

And how did you know that I was just wondering why paraller mosfet needs separate gate resistor? Well now know that too. Thanks.

And sorry for stealing the thread

Markus

[msysta]

A good reference to RLC damping and circuit in general also here:

http://en.wikipedia.org/wiki/RLC_circuit

[mikebg]

Re: Wikipedia damping

Msysta, IMNHO this is an useless article.
It considered an ideal case where no coil resistance. Can not produce such a coil for metal detector. If you do SPICE simulation of such a network, the result will be very different from reality.
I have posted in this forum an useful information and formulas for damping of TX current, reinvented by REMI group. In the future, REMI group will publish formulas for damping of RX voltage.
Mike BG - spokesman of REMI group.
P.S. The abbreviation IMNHO means "In my not humble opinion" or "In my opinion immodestly".

[Carl-NC]

I intentionally kept the equivalent circuit simple, to illustrate why a lower damping R gives a longer settling time. Series coil R does not invalidate the illustration.

[msysta]

Hi,

Thanks for reply.

I was thinking the coil internal resistance as a paraller one instead of series. In this way it could be counted in with the ideal formula.

How should a real world coil be modelled? As having serial or paraller resistance? Same question goes for internal capasitance also.

br,
Markus

[mikebg]

To ring or not to ring?

In fact, need we a critical damping?
I rediscover the SelfDamping Coil which will be named in my honor "Mike's coil ":-). The SelfDamping Coil does not need Damping resistor for critical damping. It is wound with thin wire so that its resistance "r" corresponds to the above formula calculated at R = infinity. The advantages of "Mike's coil" :-) are:
1. Saving is a powerful hot resistor. In the famous structures it is placed on PCB in control box to hot all items located there. Instead, at SelfDamping "Mike's coil ":-) the coil energy is dissipated only in the sensing head. It can therefore be used as an IceSmelter :-)to make alien circles in the snow.
2. It can be shown that the "Mike's coil" :-) is maximal lightweight because is winded with the tolerable most thin wire.
3. The "Mike's coil" :-) is the cheapest because it contains a minimum of metal.
4. The "Mike's coil" :-) will be wound with a slightly thicker wire of calculations to obtain high resistanceresistor R for critical damping. This allows you to shunt the coil with a small variable resistor for fine adjustment of critical damping. Because the coil is almost self damping, the resistor is too low-powered with high resistance.
Why we need a critical damping?
What is achieved with critical damping?
Can metal detector work in the presence of ringing?
In fact, you need a precise setting of critical damping?
I can answer these questions but not in time domain. As HAM radio designer, I prefer formulas for analysis in frequency domain. In this domain,the term "critical damping" means quality factor Q = 0,707 or damping factor d = 0,707. [ Q and d are different quantities; the relation of both is 2.Q.d = 1 ]
What happens with TX current when we increase the Q of TX tank above 0.707 or do the damping d <0,707? In time domain, the TX current gets remittent ringing. In frequency domain, the TX current gets more uniform frequency spectrum. This is like the targets are illumined with white light to distinguish them by color.
The more the form of ringing TX current resembles a sinc function pulse, the more uniform is irradiated frequency spectrum. Attached figures show the form of two sinc pulses in time domain and their corresponding frequency spectrum. Note that the more narrow sinc pulse has more wide frequency spectrum and vice versa.
What will happen if you illumine targets with a steady spectrum containing all frequencies from audible range of 20Hz to 20KHz? You will hear the voice timbre of each target.You will recognize what is the target of his speech as you can guess who talk without seeing it.

[mikebg]

The ringing family


Thе method of ringing or insufficient damping is used in one ancient family metal detectors. The attached circuit diagram represents principle of oldest family member - Bridge Balance Metal Detector TIMBRE 1898. It iluminates environment with whole decade ELF radio frequency spectrum 300Hz - 3KHz and even a decade below it. Simply put, this is wide band Audio Spectrum mashine. The operator recognizes the targets in timbre of its sound, as we can recognize by the voice who speaks without seeing it. This is very similar to targets illuminated with white light to identify them by color, but in this case we see with our ears :-).
Details of TIMBRE family are posted in section "Commercial Design Group".
Let we examine the design of resistors R because L, C and other resistances are given. The RL is coil resistance, the Rfg is output resistance of function generator FG. In the attached figure is derived formula for the calculation of R. In the end of 19th century, capacitors were quite expensive, so the original circuit diagram use only one capacitor instead two capacitors shown.
Keep in mind the following design rule for damping a network in time domain: All resistances in the network affect the damping. In the ancient bridge network, all resistors in series with TX coil act as if they increased the value of coil resistance "r".
In frequency domain the above design rule is: All resistances in a network affect the bandwidth BW.
The formulas for calculation of bandwidth are:
BW = f / Q where Q is quality factor,
BW = 2.d.f where d is damping coefficient and
BW = 1 / (2.p.Ts) where Tc is coil timeconstant L / r.
Note that Q = 2.p.f.L / r = 2.p.f.Tc.
From the Q formula follows that each coil is characterized by an important design parameter that can be called Tc or Q. The smaller the design parameter, so a smaller, cheaper and lighter is the designed coil. The "Mike's coil" :-) to meet this requirement, but is to analyze its shortcomings. They are very unpleasant.

[Nc-Dirty-digger]

Hi Digger,

I think the "Goldscan 4" or "GS4" has real ground balance and some limited discrimination. There is much info on this site if you do a search.

Mark

if im not mistaken these are where the Whites TDI came from.

[Carl-NC]

The SelfDamping Coil does not need Damping resistor for critical damping. It is wound with thin wire so that its resistance "r" corresponds to the above formula calculated at R = infinity. The advantages of "Mike's coil" :-) are: ....

The disadvantage of "Mike's coil" is that the resulting series r will limit the coil current to a few mA.

Why we need a critical damping?

The point is to get the coil response out of the way ASAP, so we can look at the target response. The only criterion is to get the coil response low enough so that the preamp is in the linear region when you sample. So it really doesn't matter if you're underdamped or overdamped, as long as you have headroom to look at the signal.

- Carl