New PI technology proposals.

[robby_h]

Oh and I am looking for tiny gold nuggets,

so I want a fast turn off. Now in and undamped state

the flyback voltage is really going to jump up, especially in

your ideal spice simulations, so say we put a 1 MegOhm damping

resistor and then switch in a 600 Ohm damping resistor in parallel.

Hi JC,

I have tried a lot of these ideas but the switch capacitance wrecks the idea and even if we could find a zero capacitance switch we would still be countering the extra energy which results by not having the damping resistor switched in during the pulse. It seems everything we do has it's own counter effect.

[robby_h]

unsubstantiated, speculation!

To unregistered Doug,
The implications are very significant as should be obvious in the designs I mentioned and the picture I posted. Rather than being your usual serial pest, how about telling me why I am wrong? Otherwise this conversation is much like me saying a stone will fall if I drop it and you saying the statement is "unsubstantiated, speculation!"

[Carl-NC]

So the fastest turn off (of the current in the coil) would be to turn off without a damping resistor or avalanche MosFet and let the current fall as in an undamped state. Then as the voltage falls to say 10 volts switch in the damping resistor to make it a critically damped coil and will slide into zero.

Does this work?????? I never tried it.

I'm using a switched-in damping R in a security wand design, but only for power reduction, and I'm switching it in right at turn-off, not later.

I have done some sims with switching in Rd during flyback and there is (potentially) a benefit, but timing is ridiculously touchy. Too soon and no benefit, too late and it rings. Probably not usable.

- Carl

[Unregistered]

I think it's only a factor with some of the high pulse frequency / low coil current designs Eric does.

Thank you Carl,
I would rather believe the comments of one of the most respected and clever electronic engineers in the USA and those of a legend in the Pi world, Eric Foster rather than somebody totally unknown and with no credentials or evidence whatsoever to support his assertions!

[Carl-NC]

Thank you Carl,
I would rather believe the comments of one of the most respected and clever electronic engineers in the USA and those of a legend in the Pi world, Eric Foster rather than somebody totally unknown and with no credentials or evidence whatsoever to support his assertions!

Sorry, but I'm still learning! And that means being willing to say, "I could be wrong." I am leaving tonight for 3 weeks of travel, so my work on PI is suspended until I return. One area I will research more is the effect of turn-on time. But to do that, I want to use my USB-PIC board, so I need to get it running first.

- Carl

[gwzd]

Different way of thinking and of course by doing many mistakes. This is a normal engineering process. Isn't it?


Aziz

No, it aint. Only if you want to get fired...
Normal engineering process as opposed to researching process involves a solid knowledge of the technology being used in order to apply it to the solution of an specific problem in a traditional or novel way.
"doing many mistakes" is a rather costly way of working, for your employer that can be devastating.

Do you know how much your employer would pay in social fees if you would actually work in Germany?
Do you think a real engineering office could afford that way of thinking?
Do you think your eventual customer would patiently wait until you are done with Spice and finally start producing things that actually work?

Get real, as a hobby yes, do as many mistakes as you wish, as a part of "engineering process", dream about it!

Regards,

[Aziz]

Hi gwzd,

No, it aint. Only if you want to get fired...
Normal engineering process as opposed to researching process involves a solid knowledge of the technology being used in order to apply it to the solution of an specific problem in a traditional or novel way.
"doing many mistakes" is a rather costly way of working, for your employer that can be devastating.

Do you know how much your employer would pay in social fees if you would actually work in Germany?
Do you think a real engineering office could afford that way of thinking?
Do you think your eventual customer would patiently wait until you are done with Spice and finally start producing things that actually work?

Get real, as a hobby yes, do as many mistakes as you wish, as a part of "engineering process", dream about it!

Regards,

Have you something about PI related so we all can profit from your excellent knowledge? I know, you never do any mistakes. Don't you?
Please, stop your way of discrediting people. This is not efficient for our purpose.

Aziz

[Aziz]

Hi all,

Eric described exactly the flyback process of what is happening during the avalanche breakdown voltage of the MOSFET. Thanks Eric.

You can see the SPICE simulation results, which exactly confirms this. The second picture shows the target induction voltage (target eddy current has same behaviour). So clipping the flyback voltage will decrease the target stimulation. The target voltage induction is not instantly high. You can not see the rise of this in the picture below.

Aziz

[Aziz]

"So clipping the flyback voltage will decrease the target stimulation."

Sorry for my bad english.
It should be:
"So clipping the flyback voltage will limit the target stimulation."

[gwzd]

Hi gwzd,



Have you something about PI related so we all can profit from your excellent knowledge?

Yes

I know, you never do any mistakes. Don't you?

I do make mistakes, I just don't call them "engineering process", but plain stupidity and ignorance from my side.

Please, stop your way of discrediting people.

I'm not discrediting anyone, certainly not you, you manage pretty well yourself. I was answering to your question, remember?

This is not efficient for our purpose.

Aziz

On the contrary, it prevents other people following your "many mistakes" and thinking that not working efficiently is the "correct way".

As always, stay calm and please note that I don't call you names or anything like that, but do not expect me to agree with whatever rubbish comes out of your keyboard.
My best regards,

[Aziz]

Hi gwzd,

if you have something PI related, then let us know now. We are very interested on your contribution.

Aziz

[JC1]

Nice !!!

Hi Aziz,

Nice simulation, which shows everything that is

been talked about.

[Ferric Toes]

"So clipping the flyback voltage will decrease the target stimulation."

Sorry for my bad english.
It should be:
"So clipping the flyback voltage will limit the target stimulation."

Not necessarily. As I stated before, provided the downward current (field) ramp takes place in 1/10 of the target time constant, then the object is fully stimulated. Designing for a higher flyback voltage and faster switch off will not get you any more signal. Very high flyback voltages have their own problems. i.e very high peak currents in the RX input clipping diodes, and input resistors working outside their voltage ranges. I had an input resistor fail on an early Goldscan for this reason. It didn't fail completely, it just went horribly noisy.

Basic PI theory can be considered by just examining one step in a magnetic field. If a field has been set up for an infinitely long time and then reduced to zero in a infinitely short time, we can see what happens in a conductive object within that field. All the time the field is static, then nothing happens within the object. As soon as the field starts to change, then eddy currents will be generated which try to maintain the field within the object at its pre-switch off value. These eddy currents first flow on the objects surface, and then gradually diffuse inwards, until for times after one time constant, they become exponential. This diffusion is caused by the object's resistance that cause the eddy currents to lose energy (heat loss). Eventually, there is no internal emf and hence no current and the field and object become static again. In this theoretical consideration the initial field could be generated by means other than a coil of wire carrying a current. Something to ponder on during the upcoming holiday period - if a long standing magnetic field (not coil generated) is instantaneously switched off in free space with no conductive objects nearby, is there a flyback voltage?

Eric.

[Aziz]

Hi JC1,

Hi Aziz,

Nice simulation, which shows everything that is

been talked about.

Thanks. I have still many many interesting "rubbish" thoughts, that might surprise you all at the end. It will take several weeks to present all them. So you all will be well entertained. Step by step.

Now an interesting question to be proofed next:
Has the coil a different time constant during transmit on and off?
(My rubbish statement was: NO).

You mentioned this very interesting related question. Why the flyback damping is faster than charging the coil during transmit on?

But I want let the people to think for oneself this question. (I have the "rubbish" proof already.)

Aziz

[Aziz]

Hi Eric,

Not necessarily. As I stated before, provided the downward current (field) ramp takes place in 1/10 of the target time constant, then the object is fully stimulated. Designing for a higher flyback voltage and faster switch off will not get you any more signal. Very high flyback voltages have their own problems. i.e very high peak currents in the RX input clipping diodes, and input resistors working outside their voltage ranges. I had an input resistor fail on an early Goldscan for this reason. It didn't fail completely, it just went horribly noisy.

Basic PI theory can be considered by just examining one step in a magnetic field. If a field has been set up for an infinitely long time and then reduced to zero in a infinitely short time, we can see what happens in a conductive object within that field. All the time the field is static, then nothing happens within the object. As soon as the field starts to change, then eddy currents will be generated which try to maintain the field within the object at its pre-switch off value. These eddy currents first flow on the objects surface, and then gradually diffuse inwards, until for times after one time constant, they become exponential. This diffusion is caused by the object's resistance that cause the eddy currents to lose energy (heat loss). Eventually, there is no internal emf and hence no current and the field and object become static again. In this theoretical consideration the initial field could be generated by means other than a coil of wire carrying a current. Something to ponder on during the upcoming holiday period - if a long standing magnetic field (not coil generated) is instantaneously switched off in free space with no conductive objects nearby, is there a flyback voltage?

Eric.

a higher target stimulation with higher flyback voltage or fast damping will not necessarily cause more RX signals. My recent simulation models showing really interesting results, which will be presented soon.
They might very likely be simple, but they contain very useful informations. Your practical experiences will help to understand the process. I appreciate that. Pity, that I haven't the means for some practical experiments at the moment.


"if a long standing magnetic field (not coil generated) is instantaneously switched off in free space with no conductive objects nearby, is there a flyback voltage?"

Of course not. Why?
Free moving electrons in the conductur will see a force (EMF) when there is magnetic field change. So electrons in the conductor will start moving and building an electrical potential (voltage induction).
Non conducting objects haven't free moving electrons. So no flyback voltage can be induced.

Regards,
Aziz