PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-03-29, 10:24:27
News: If you have a suggestion or need for a new board title, please PM the Admins.
Please remember to keep topics and posts of the FE or casual nature. :)

Pages: 1 [2] 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 ... 39
Author Topic: Itsu's workbench / placeholder.  (Read 105227 times)

Group: Professor
Hero Member
*****

Posts: 3354
Dear Itsu

Could you explain why you were searching for NMR in a core in the manner described? What were the original thoughts or direction to proceed using the method shown in the video?
I put him up to it in relation to the these patents and this thread since using a Spectrum Analyzer to see if there is any RF energy absorption at 45.55Mhz in iron powder is simple and relevant.

The internal 33T field of iron ferromagnetic domains effectively marginalizes the effects of any small ambient fields, removing the need for creating extremely homogeneous external fields, which are typically necessary for NMR experiments.

Also, it is known, that an RF field which is experienced by nuclear spin is enhanced due to the motion of the magnetization of ELECTRON SPINS by the applied RF field thus the NMR signal is enhanced by the same factor. Furthermore, the enhancement factor in domain walls is generally much larger than in domains themselves.

All of the above leads me to believe that some 45.5MHz signal should be observable.

What are the conclusions from yourself and verpies on the negative outcome?
That power levels and impedance matching (RF current coupling to core) need to be better or the typical perpendicular biasing magnetic field will be necessary to see anything.  I was hoping the remanence will not leave domains so perfectly randomized, that their signals cancel out.

According to the attached paper, the rotation of domains in iron powder begins at 300mT and ends around 750mT. 
« Last Edit: 2019-01-18, 17:05:12 by verpies »
   
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
Dear Verpies

Thanks for taking the time in explaining the background information regarding Itsu's tests. I will attempt to assimilate the supplied information.

Regards


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   

Group: Moderator
Hero Member
*****

Posts: 4056
Ok guys, thanks for the new .asc files, i also agree with F6FLT and ION to use the standard layout, so i adopted
the Poynt99/F6FLT one for now.

I have setup my real life circuit using the MOSFET and after some adjustments (taking care the signals are equal
in amplitude on the vertical division scales like both 500mA/Div. etc.) it seems that the mean (avg) current through
csr1 and csr2 are the same (also the same as the current DMM in the 12V supply lead).

» i still keep some distance here as i have seen these equal currents before (also with the transistor) but
the next day it had again some  offset somehow «.

So it seems to me that there is some "feedthrough" in the transistor during its off time which causes the earlier
seen higher csr2 current.

 
The final test now would be for Tinman to replace his transistor for a MOSFET and see how his currents values are now.


Video of the testing here: https://www.youtube.com/watch?v=FUexkxmYJU4

Screenshot:  yellow csr1,  blue csr2


Regards Itsu
« Last Edit: 2019-01-18, 20:01:05 by Itsu »
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 901
Hi guys,

Have a look at my test. I don't get the same results as Itsu

https://youtu.be/F3C5opU3koY

Regards
Luc
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4500


Buy me some coffee
Hi guys,

Have a look at my test. I don't get the same results as Itsu

https://youtu.be/F3C5opU3koY

Regards
Luc

It is good to hear from you Luc.

As we know from past experience,things always stray from the original design,and then more time spent searching for answers as to how the !non original! device is showing what it is showing.

Then the !Eureka! moment come-we have found the source of the extra current ;D.

It would seem that very few listen to,or looked at the pages of data i supplied.

Seems all the extra current comes from the FG C.C
So lets do the math.
In my case,on 12v with a duty cycle of 5%,my circuit draws 36mA average through CSR1. Through CSR2,i have 70- 72mA average of current flowing.
My base resistor is 100 ohms.
What dose the voltage across a 100 ohm resistor have to be to have an average of 70mA flowing through it ?. Now,as we have only a 5% duty cycle,what would the voltage across the 100 ohm resistor have to be during that 5% on time to have an average of 30mA of curtent flowing through it.

As i said in the other thread,i have a feeling that everything i have shown will fall on deaf ears.
So many things overlooked ,as things must conform to known science.


Brad.


---------------------------
Never let your schooling get in the way of your education.
   

Group: Moderator
Hero Member
*****

Posts: 4056
Hi guys,

Have a look at my test. I don't get the same results as Itsu

https://youtu.be/F3C5opU3koY

Regards
Luc


Hi Luc,

thanks for the tests and results.

But the circuit used and tests carried out are not the same as i (we) are doing in the real circuit or the simulations
so results can not be readily compared.

Please allow me some comments:

First the diagram shows the red plus (+) sign at C2 in the wrong position.
Secondly, the yellow input probe was never at that position in our tests, it was always across the source (battery).
Then, your Caps are different (much higher in value), same with the load (we use leds).
Not sure about the switching device, we used a transistor at first, now a MOSFET.

Concerning the measurements, please use "mean" for the currents (so not rms or cyc rms) and many cycles so we can
compare with our results, see my screenshot above.

You can use cyc mean, but that will take only 1 cycle no matter how much cycles you put on the screen and we have seen
that 1 cycle could mislead you here.

Try to use the same vertical / div. setting when comparing currents through csr1 and csr2, i notice it can differ
somewhat when using different settings.

Finally, we use a floating FG which is directly connected across the base / emitter (transistor) or gate / source
(MOSFET) to prevent from any FG currents to being added to csr2.
It did add "SOME" (so not ALL) extra current through csr2.

Regards Itsu
   

Group: Moderator
Hero Member
*****

Posts: 4056
It is good to hear from you Luc.

As we know from past experience,things always stray from the original design,and then more time spent searching for answers as to how the !non original! device is showing what it is showing.

Then the !Eureka! moment come-we have found the source of the extra current ;D.

It would seem that very few listen to,or looked at the pages of data i supplied.

Seems all the extra current comes from the FG C.C
So lets do the math.
In my case,on 12v with a duty cycle of 5%,my circuit draws 36mA average through CSR1. Through CSR2,i have 70- 72mA average of current flowing.
My base resistor is 100 ohms.
What dose the voltage across a 100 ohm resistor have to be to have an average of 70mA flowing through it ?. Now,as we have only a 5% duty cycle,what would the voltage across the 100 ohm resistor have to be during that 5% on time to have an average of 30mA of curtent flowing through it.

As i said in the other thread,i have a feeling that everything i have shown will fall on deaf ears.
So many things overlooked ,as things must conform to known science.


Brad.


Hi Brad,

i don't think one can say: "Seems all the extra current comes from the FG C.C"

I does add SOME current through csr2, but certainly not ALL.

After using a floating FG which is attached across the base / emitter (transistor) or gate / source
(MOSFET) it was eliminated from the csr2 current, but still i had more current through csr2 then csr1.

Finally now after using a MOSFET instead of the transistor, (plus the floating FG) i see equal currents
through csr1 and csr2.
That issue (why more current through csr2 when using a 2N3055 transistor) is still open!


So i invite you to do the same, find a MOSFET, preferrable a IRF530 and use that instead of your transistor.

Your csr1 and csr2 currents should also be equal almost (because of the extra FG current if not floating,
but that would be only a minor contribution).

Set up your scope to measure the both currents using the same vertical/div. setting and avg current over many cycles.



Thanks,   Itsu
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 901
Thanks for your detailed reply Itsu. Below are my answers to your questions

First the diagram shows the red plus (+) sign at C2 in the wrong position.

Thanks for pointing out the schematic error. I have fixed it. I'm sure you know it wasn't that way in my test circuit but good to point out.

Secondly, the yellow input probe was never at that position in our tests, it was always across the source (battery).

I know the Yellow voltage probe was originally across C1. However, since I also included CSR 3 position in my test to measure both the in and out (flyback) power of the coil it had to be in that position. I'm sure you know that it won't affect the current probe measurement when I moved it from CSR 3 to CSR 2 and then to CSR 1 position. However, I would be please to provide new scope shots with Yellow voltage probe only across C1 for further confirmation.

Then, your Caps are different (much higher in value), same with the load (we use leds).

Sorry but I was under the impression that this investigation was to understand why two different position of current measurement being CSR 1 (pre C1) and CSR 2 (post C1) are giving different current measurements when an Inductor is pulse?  Please correct me if I have misunderstood.
If I haven't misunderstood, then why would cap value or inductive discharge cap load have any influence on why CSR 1 or CSR 2 have different readings.
What I'm trying to demonstrate is, if you find a way to correctly measure both (csr1 & csr2) values then that method should hold up or work with different Cap, Inductor or flyback load values, and not just with specific values, no?

Not sure about the switching device, we used a transistor at first, now a MOSFET.

My switch is an Opto Isolated MOSFET

Concerning the measurements, please use "mean" for the currents (so not rms or cyc rms) and many cycles so we can
compare with our results, see my screenshot above.

I will also switch the Current Probe to "mean" in the redo test with Voltage probe across C1

You can use cyc mean, but that will take only 1 cycle no matter how much cycles you put on the screen and we have seen
that 1 cycle could mislead you here.

My scope does not have "cyc" in its selection menu. I know 1 sample gives false reading. Poynt has tough me well on that.

Try to use the same vertical / div. setting when comparing currents through csr1 and csr2, i notice it can differ
somewhat when using different settings.

Yes, I know that and why all my multi sample scope had the same vertical / div. settings.

Finally, we use a floating FG which is directly connected across the base / emitter (transistor) or gate / source
(MOSFET) to prevent from any FG currents to being added to csr2.
It did add "SOME" (so not ALL) extra current through csr2.

My MOSFET switch is Opto Isolated from the FG and used an Arduino as FG since it give me more accurate microsecond pulse width control.

Looking forward in your reply and or corrections.

Regards
Luc
   

Group: Moderator
Hero Member
*****

Posts: 4056

Hi Luc,

thanks for the reply, it clears up some things, sorry if i sounded to picky, not my intention.

Understood about the yellow probe, indeed, it does not matter concerning the currents through the csr's, but as you
also showed the Math traces presenting the power across those csr's, then it matters where you take the input voltage from.

About the caps, in my tests it did matter how big they are, and especially how they relate to each other.
I used both 330uF caps at first, and the csr current traces looked different then when i went over to what
Brad used (10000uF and 470uF). 
You might be right that it should not matter, but seeing the different traces i wanted to replicate as close as
Brads situation as possible.

Ok on the switch, so no FG current should creep through into the csr2 current measurements, thats good.

Looking forward to your csr current measurements in mean.

Thanks  Itsu
   

Group: Administrator
Hero Member
*****

Posts: 3198
It's not as complicated as it may seem...

As we know from past experience,things always stray from the original design,and then more time spent searching for answers as to how the !non original! device is showing what it is showing.

Then the !Eureka! moment come-we have found the source of the extra current ;D.

It would seem that very few listen to,or looked at the pages of data i supplied.

As i said in the other thread,i have a feeling that everything i have shown will fall on deaf ears.

Brad.

You mean like every time I am asked to come here and look at something, I find that everything I have been instructing through the years; basic, fundamental, subtle, and even some intricacies of power measurement, has either fallen on deaf ears, been misunderstood or misinterpreted, been bastardized, been overlooked or forgotten, and/or old bad habits and erroneous assumptions still abound, and hence the 360 degree process you describe above?

Yep, I get it.
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 901
Hi Luc,

thanks for the reply, it clears up some things, sorry if i sounded to picky, not my intention.

Understood about the yellow probe, indeed, it does not matter concerning the currents through the csr's, but as you
also showed the Math traces presenting the power across those csr's, then it matters where you take the input voltage from.

About the caps, in my tests it did matter how big they are, and especially how they relate to each other.
I used both 330uF caps at first, and the csr current traces looked different then when i went over to what
Brad used (10000uF and 470uF). 
You might be right that it should not matter, but seeing the different traces i wanted to replicate as close as
Brads situation as possible.

Ok on the switch, so no FG current should creep through into the csr2 current measurements, thats good.

Looking forward to your csr current measurements in mean.

Thanks  Itsu

Thanks Itsu for confirming my understanding is correct and the test I put forward can also be considered.
Please find the below test circuit and scope shots with the changes you requested.
Looking forward to better understanding the huge measurement differences.

Kind regards
Luc
   

Group: Moderator
Hero Member
*****

Posts: 4056

Hi luc,

well, looking at your screenshots, i cannot make much sense out of them.

The signals shown (except for CVR2 Single sample.jpg) do not match the signals in my circuit nor the Sims.

"CVR1 Single" shows an AC like signal which is not present on mine or others, there should be an elevated (positive)
sawtooth like signal there see my post #2 screenshot or Poynt99 his post #3 middle picture.

Also the negative mean value's are strange (they are there, no doubt as the CVR1 multi sample show them)

Could you mention the used capacitors, load resistor and inductor value's (plus coil resistance) so i can simulate?
I see you use 40V as a source.

To be clear you are using NO csr's, you just use your current probe on the CSR positions?

Thanks,  Itsu
   

Group: Moderator
Hero Member
*****

Posts: 4056
I put him up to it in relation to the these patents and this thread since using a Spectrum Analyzer to see if there is any RF energy absorption at 45.55Mhz in iron powder is simple and relevant.

The internal 33T field of iron ferromagnetic domains effectively marginalizes the effects of any small ambient fields, removing the need for creating extremely homogeneous external fields, which are typically necessary for NMR experiments.

Also, it is known, that an RF field which is experienced by nuclear spin is enhanced due to the motion of the magnetization of ELECTRON SPINS by the applied RF field thus the NMR signal is enhanced by the same factor. Furthermore, the enhancement factor in domain walls is generally much larger than in domains themselves.

All of the above leads me to believe that some 45.5MHz signal should be observable.
That power levels and impedance matching (RF current coupling to core) need to be better or the typical perpendicular biasing magnetic field will be necessary to see anything.  I was hoping the remanence will not leave domains so perfectly randomized, that their signals cancel out.

According to the attached paper, the rotation of domains in iron powder begins at 300mT and ends around 750mT.


Concerning this 45.5Mhz NMR item i found a website for calculating Amidon toroid (iron powdered) responses here:
https://www.changpuak.ch/electronics/amidon_toroid_calculator.php

So did some tests with the calculated 4 turns @ 45Mhz.

Dumping some pictures here for verpies

Calculator  input/output data
picture of the 4 turn setup (parallel mode)
response of the TG/SA

We see a nice dip at 45.5Mhz which for me points to a good impedance match in and out of the TG/SA using these 4 turns


Itsu 
   

Group: Professor
Hero Member
*****

Posts: 3354
How does it respond to an axial (perpendicular) DC magnetic field from 300mT to 750mT when zoomed at 45.3MHz-45.7Mhz with a low VBW ?

P.S.
Please sketch the "parallel mode" before someone asks....
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 901
Thanks for your reply Itsu.
Please find the replies to your questions below

well, looking at your screenshots, i cannot make much sense out of them.

This scope model has never been very good at displaying good data when it's sharp peaks. Works fine with sine waves.
That's why I included the 500us / div. single shots to get a good visual which obviously show tremendous differences between CVR 1 & CVR 2.

The signals shown (except for CVR2 Single sample.jpg) do not match the signals in my circuit nor the Sims.

Yes, I agree and why I took the time to demonstrate that in certain conditions one can produce even greater differences then Brad showed.

Also the negative mean value's are strange (they are there, no doubt as the CVR1 multi sample show them)

I think that's related to the scopes inability to obtain good data from the narrow fast rising current pulses

Could you mention the used capacitors, load resistor and inductor value's (plus coil resistance) so i can simulate?
I see you use 40V as a source.

Yes, that was all mentioned in the video but here it is re-tested in writing. 40vdc in 100kuf C1, 27kuf C2, 478 Ohms load, 0.250 Ohm Toroid coil @ 100mH

To be clear you are using NO csr's, you just use your current probe on the CSR positions?

Yes, that is correct. I use the same current probe and just move it to each CSR position.
If you wish I can test each position using a 0.1 or 0.01 Ohm 1% metal film resistors. Let me know.

Regards
Luc
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4500


Buy me some coffee
You mean like every time I am asked to come here and look at something, I find that everything I have been instructing through the years; basic, fundamental, subtle, and even some intricacies of power measurement, has either fallen on deaf ears, been misunderstood or misinterpreted, been bastardized, been overlooked or forgotten, and/or old bad habits and erroneous assumptions still abound, and hence the 360 degree process you describe above?

Yep, I get it.

Well i dont know how you have come about that conclusion,as all my measurements have been done exactly as you have taught.

You have taught that a DMM is very good at averaging out voltage and current,and that is one method i have used.

You have taught that the RMS value is to be used to calculate dissipated power of a resistor of known value,and that is what i have done.

You have taught that the mean or average voltage across a known resistance is used to calculate current flowing through that resistance,and that is what i have done.

But i have also gone above and beyond that,in that these values have been confirmed in many more ways.
All these methods are in the other thread,along with all being on my video's.

Then there is the concern about the base current being the extra value seen through CVR2,when in fact ,that value dose very little to contribute to the extra current value flowing through CVR2.

So no,your methods have not been misunderstood or misinterpreted, been bastardized,or anything else--unless you can point out where i have done so.



Brad


---------------------------
Never let your schooling get in the way of your education.
   
Group: Elite Experimentalist
Hero Member
*****

Posts: 901
Hi Itsu,

I decided to make a new test using two 1% 0.1 Ohm metal film resistors as CVR 1 & 2
Please find attached schematic and scope shots.
Note the multiple samples still don't give any useful data.
However, it's visually clear on the 500us single scope shot that CVR 1 and 2 are nowhere close.
CVR 2 has about 12 times the height of CVR 1 and a voltage rise on a resistor is squared, so it's more like 4 x 12 = 48 times more power.
I know this doesn't mean OU but why can't we easily explain this large different?
Looking forward to Poynt's or verpies explanation

Regards
Luc

PS. I also included (last pic) a shot with probe 1 @ 50mv voltage per div. (instead of 200mv per div.) to better see the wave of CVR 1
   

Group: Tinkerer
Hero Member
*****

Posts: 3055
Perhaps the anomalous current in the Transistor Emitter
is due to Transistor Leakage at the moment of turn-off?

At turn-off the Collector to Emitter Voltage should
increase dramatically due to Inductive Kick-back.
Transistors are not capable of instantaneous turn
off on account of charge storage.

Could it be?  The Capacitor plays a role as well.

Didn't Verpies suggest this as possibility some time ago?


---------------------------
For there is nothing hidden that will not be disclosed, and nothing concealed that will not be known or brought out into the open.
   

Group: Moderator
Hero Member
*****

Posts: 4056
How does it respond to an axial (perpendicular) DC magnetic field from 300mT to 750mT when zoomed at 45.3MHz-45.7Mhz with a low VBW ?

P.S.
Please sketch the "parallel mode" before someone asks....

I did some manipulating with magnets and you see the dip move, but will try that again later today in that specific range and low Bandwidth (was at 30Khz).

With parallel mode i mean like shown below.


Itsu
   

Group: Moderator
Hero Member
*****

Posts: 4056
Hi Itsu,

I decided to make a new test using two 1% 0.1 Ohm metal film resistors as CVR 1 & 2
Please find attached schematic and scope shots.
Note the multiple samples still don't give any useful data.
However, it's visually clear on the 500us single scope shot that CVR 1 and 2 are nowhere close.
CVR 2 has about 12 times the height of CVR 1 and a voltage rise on a resistor is squared, so it's more like 4 x 12 = 48 times more power.
I know this doesn't mean OU but why can't we easily explain this large different?
Looking forward to Poynt's or verpies explanation

Regards
Luc

PS. I also included (last pic) a shot with probe 1 @ 50mv voltage per div. (instead of 200mv per div.) to better see the wave of CVR 1



Hi Luc,

Thanks for the data on your circuit.

good idea to use real csr's as i found that current probes generate problems on their own.

I agree however that this data still don't give any usefull data.
The signals still look strange, like the tapering off of the csr2 amplitude in the middle screenshot (DSO2102.jpg)
In 5 cycles (still low for a good average calculation, try 20 or 40 or so) it decreases 40%.

Seems there is a sinewave like signal superimposed on it, could you increase the time base to see if it lateron
increases again or even goes negative?
I saw the same thing (even going negative) on your earlier posted (post #35) screenshot on CSR1 (CVR1 Multi samples.jpg)

 
Seems you are running at 10Hz (110ms between cycles) so this superimposed (sine?)wave is still much lower, like 1 or 0.5Hz

Anyway, this data is not useable, so i also invite others to give their opinions and meanwhile i will simulate
your circuit to see how that behaves there.

Thanks,   Itsu
   

Group: Moderator
Hero Member
*****

Posts: 4056
Perhaps the anomalous current in the Transistor Emitter
is due to Transistor Leakage at the moment of turn-off?

At turn-off the Collector to Emitter Voltage should
increase dramatically due to Inductive Kick-back.
Transistors are not capable of instantaneous turn
off on account of charge storage.

Could it be?  The Capacitor plays a role as well.

Didn't Verpies suggest this as possibility some time ago?

muDped,

i think its worth looking into this transistor leakage to get to the bottom of it.
Not sure how to setup things to make it visible if real.
Could it be in the datasheet?

I did not see any info from verpies on Brads thread, but F6FLT (and yourself) mentioned something about it there:

Post #215  from F6FLT:
Quote
What is the collector voltage of Q1?
If it is less than the base voltage+0.6v, then the pulses are rectified by the base-collector junction and the mean DC current goes through R7-L1-R6-R5 and can add up in R2.

But i don't think the collector voltage (±12.5V) was ever less then base voltage+0.6v

Anyway,  thanks for thinking along.

Itsu
   

Group: Renaissance Man
Hero Member
*****

Posts: 2711


Buy me a cigar
Hi Guys.

Luc, are you pulsing the coil, Steel core and PM as per Brad's motor drive?

Could the tapering be the PM demagnetising?

Cheers Graham.


---------------------------
Nanny state ? Left at the gate !! :)
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 4500


Buy me some coffee
muDped,


But i don't think the collector voltage (±12.5V) was ever less then base voltage+0.6v

Anyway,  thanks for thinking along.

Itsu

Collector voltage is 0v + V/drop across CVR2 when transistor is on.
Base voltage is only 3v,as it is 6VPP,so 3v is a negative value.


Brad


---------------------------
Never let your schooling get in the way of your education.
   

Group: Moderator
Hero Member
*****

Posts: 4056

Thanks Brad,   you could be right, i will take a look at the collector voltage more carefull.

Perhaps F6FLT has something there then.

Itsu
   

Group: Administrator
Hero Member
*****

Posts: 3198
It's not as complicated as it may seem...

good idea to use real csr's as i found that current probes generate problems on their own.
Part of the problem with the scope readings (esp. on csr1) may be because the csr's are 0.1R as opposed to 1R values.

Quote
I agree however that this data still don't give any usefull data.
The signals still look strange, like the tapering off of the csr2 amplitude in the middle screenshot (DSO2102.jpg)
In 5 cycles (still low for a good average calculation, try 20 or 40 or so) it decreases 40%.

Seems there is a sinewave like signal superimposed on it, could you increase the time base to see if it lateron
increases again or even goes negative?
When sine waves appear superimposed on the measured wave form, it is often due to under-sampling by the scope, so it is an artifact generated by the scope.

I think for this particular investigation of avg currents in the csr's, the zoomed-in shots are sufficient to see the disparity between them.

Looking forward to your sim of Luc's version.

   
Pages: 1 [2] 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 ... 39
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-03-29, 10:24:27