Storing Cookies (See : http://ec.europa.eu/ipg/basics/legal/cookies/index_en.htm ) help us to bring you our services at overunity.com . If you use this website and our services you declare yourself okay with using cookies .More Infos here:
https://overunity.com/5553/privacy-policy/
If you do not agree with storing cookies, please LEAVE this website now. From the 25th of May 2018, every existing user has to accept the GDPR agreement at first login. If a user is unwilling to accept the GDPR, he should email us and request to erase his account. Many thanks for your understanding

User Menu

Custom Search

Author Topic: 3v OU Flashlight  (Read 563435 times)

Vortex1

  • Hero Member
  • *****
  • Posts: 518
Re: 3v OU Flashlight
« Reply #480 on: May 05, 2014, 10:31:10 PM »
Itsu:

Your waveforms look quite normal for TKLABS v1.4 schematic.

The modulation and cutoff of the switcher resulting in the sine wave seem reasonable for a secondary loaded with a 100uF capacitor. This is like the switcher looking into a near dead short at high frequencies (not considering leakage inductance and esr of the capacitor.)

If you want more "action" try the Vadik Guk or verpies schematic version posted above. The switcher chip will operate in a more normal mode without such a difficult capacitive load on the secondary.

Regards, ION
« Last Edit: May 06, 2014, 12:50:20 AM by Vortex1 »

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Re: 3v OU Flashlight
« Reply #481 on: May 06, 2014, 04:18:34 AM »
This looks just spot on.
As a matter of interest, have you checked the native resonance frequency for the core?

~A
Spot on -- nearly --, but at 1/20 the frequency of Akula's scopeshot. Also the embedded sinus is showing adequate supply voltage to the 34063: it is switching cleanly rather than following the sinusoid during the bursts.

verpies

  • Hero Member
  • *****
  • Posts: 3473
Re: 3v OU Flashlight
« Reply #482 on: May 06, 2014, 10:59:29 AM »
Guys,


Let's think about this more analytically.

The two core halves apparently are vibrating mechanically, creating a variable air-gap between them.
The variable air-gap means that any coil wound on this core  possesses variable inductance (dL/dt).

This variation is in all probability sinusoidal because elastic bouncing and mass-spring vibrations are inherently sinusoidal.

The variation in this inductance can be sensed on-the-fly with high frequency LC oscillator formed by one of the windings resulting in frequency modulation (FM) or by measuring the rise-time of the current ramp (di/dt) when this winding is charged by a constant voltage source.

When a constant frequency PWM modulator chip is supplied with smooth voltage source and subjected to a constant load while its output is connected to such variable inductance winding and its feedback loop is correctly configured - its driving output will experience pulse width modulation (PWM) proportional to the inductance of the winding.

If the PWM modulator chip is operating at relatively high frequency, the integration of these variable pulse widths will yield a low frequency signal that is proportional to the air-gap between core halves. 
It can be argued that this LF signal also represents an instantaneous width of that air-gap ...or relative position of core halves.

All very straightforward and conventional like in a linear inductive encoder.
« Last Edit: May 06, 2014, 02:23:19 PM by verpies »

itsu

  • Hero Member
  • *****
  • Posts: 1845
Re: 3v OU Flashlight
« Reply #483 on: May 06, 2014, 10:24:59 PM »

Inductance of the transformer coils reduced to 56mH and 27mH.
As always, the core halfs firmly forced together by electrical tape with a piece of paper inbetween the 2 outer legs of the e-core.

Highest inductance to W2 as the other way around gave no stable signals no matter how i varied the voltage and/or the pot meter.
No sign of any prolonged functioning after removing the supply voltage (here again 2.3V @50mA)
The transistor (MJE13009) still in use.
Note the pin 13 sign wave now increased to 53Hz.

Next will be further reducing of the inductance to 5.6 and 2.7mH

Regards Itsu


magpwr

  • Hero Member
  • *****
  • Posts: 1168
    • Youtube Channel
Re: 3v OU Flashlight
« Reply #484 on: May 07, 2014, 12:22:56 AM »
Inductance of the transformer coils reduced to 56mH and 27mH.
As always, the core halfs firmly forced together by electrical tape with a piece of paper inbetween the 2 outer legs of the e-core.

Highest inductance to W2 as the other way around gave no stable signals no matter how i varied the voltage and/or the pot meter.
No sign of any prolonged functioning after removing the supply voltage (here again 2.3V @50mA)
The transistor (MJE13009) still in use.
Note the pin 13 sign wave now increased to 53Hz.

Next will be further reducing of the inductance to 5.6 and 2.7mH

Regards Itsu

hi Tinselkoala and itsu,

There is one small noticable difference between Akula's version and Tinselkoala along with itsu.
I have attached the edited scope shot to show where i spotted something like "resonance within resonance" at these area.Upon observing carefully i can't help noticing for Akula there is much more of resonance within resonance base on scope shot provided from him.
I only spotted the difference only after itsu provided his latest scope shot which i compared against Tinselkoala's base on my memory then compared against Akula version.
Got to rush off bye for now.....









 



verpies

  • Hero Member
  • *****
  • Posts: 3473
Re: 3v OU Flashlight
« Reply #485 on: May 07, 2014, 01:18:29 AM »
8) I think one of Akula's traces might be inverted.
(My shot: Both channels at 2v/div, 5 ms/div horizontal. Bottom channel is inverted in scope.)
So I surmise that Itsu's scopeshots are different because none of his channels are inverted

verpies

  • Hero Member
  • *****
  • Posts: 3473
Re: 3v OU Flashlight
« Reply #486 on: May 07, 2014, 01:22:27 AM »
As always, the core halves firmly forced together by electrical tape with a piece of paper in between the 2 outer legs of the e-core.
How does the LF sinus signal frequency respond to various tensions squeezing the core ?

d3x0r

  • Hero Member
  • *****
  • Posts: 1433
Re: 3v OU Flashlight
« Reply #487 on: May 07, 2014, 01:45:46 AM »
How does the LF sinus signal frequency respond to various tensions squeezing the core ?
I think with the screw through a pot core, that it is very tightly compressed... maybe 20-50 inch-pounds




don't go chasing waterfalls?
Kinda wish there was just a thumbs up button (for magpwr's post)
might be fun to strip the black backgrounds to alpha and overlay things
(not for me just sayin)

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Re: 3v OU Flashlight
« Reply #488 on: May 07, 2014, 02:29:15 AM »
hi Tinselkoala and itsu,

There is one small noticable difference between Akula's version and Tinselkoala along with itsu.
I have attached the edited scope shot to show where i spotted something like "resonance within resonance" at these area.Upon observing carefully i can't help noticing for Akula there is much more of resonance within resonance base on scope shot provided from him.
I only spotted the difference only after itsu provided his latest scope shot which i compared against Tinselkoala's base on my memory then compared against Akula version.
Got to rush off bye for now.....

I think that happens when the supply voltage isn't adequate for the 34063. Instead of dropping all the way to the baseline, the bottom of the chip's pulse oscillations only follows that sinus shape. If you increase the supply voltage a tiny bit, or even change the setpoint pot, the chip will switch fully and the embedded sinus within the pulse "boxes" or bursts goes away. The parts between the pulses will still be there.

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Re: 3v OU Flashlight
« Reply #489 on: May 07, 2014, 02:32:41 AM »
How does the LF sinus signal frequency respond to various tensions squeezing the core ?
The tighter you have the core squeezed, the lower the frequency of the sinus. If you separate the core the frequency goes up and up until it drops out.
This of course is because this changing gap changes the inductance of the coils. It also must change the dimensional resonant characteristics of the core.
I think I illustrated this in one of the videos.
https://www.youtube.com/watch?v=kHAc7kDY4-w
around 4 minute mark. I think I got the directions right.

TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Re: 3v OU Flashlight
« Reply #490 on: May 07, 2014, 03:34:05 AM »
hi Tinselkoala and itsu,

There is one small noticable difference between Akula's version and Tinselkoala along with itsu.
I have attached the edited scope shot to show where i spotted something like "resonance within resonance" at these area.Upon observing carefully i can't help noticing for Akula there is much more of resonance within resonance base on scope shot provided from him.
I only spotted the difference only after itsu provided his latest scope shot which i compared against Tinselkoala's base on my memory then compared against Akula version.
Got to rush off bye for now.....

Sigh....

I apologize for having a better screen resolution on my scope.....    :-\

(Note that the frequency of the sinus and bursts is about 300-310 Hz, not the  higher figure given in the "measurement" box.)
 

d3x0r

  • Hero Member
  • *****
  • Posts: 1433
Re: 3v OU Flashlight
« Reply #491 on: May 07, 2014, 03:52:52 AM »
I took a little time to grok the skematic.


If the LED is on (which is a result of switching on the self triggered primary)
then the chip is on... (the second half the k561(unspellable)h2 )


which I guess then becomes another signal geneartor? not sure of the internals of the  k561? ya okay they're just gates looped to be a pll of sorts that allows the secondary to back pulse...


I seee..... it's one of those phase-matching things too... so it's using a transformer bidirectionally...


I dunno it looks like a bleed valve ....


... so the VCC from the chip is gates to the gate ..


-------
Okay so right; the right side is a frequency delay probably with self feedback in lack of other stimulus... driven by the primary from the mc403063 thing;
so it's a bidirectional communication ... so side a drive results on side b and side b drive (but how is it you can drive and receive? well.. that's the magic of the right side, it allows a bleed in the opposite direction ..


and I guess on the source side, the noisier/sharaper the square waves, every wave is part of a perfect square pulse but we know that


hmmmm


so that's what happens when the second side doesn't work...


so okay schematic done; and the right side turns out to be intended to work but doesn't... so that terminates in a capacitor.


Now if the drawing of the foil windings connected to the others was correct....






d3x0r

  • Hero Member
  • *****
  • Posts: 1433
Re: 3v OU Flashlight
« Reply #492 on: May 07, 2014, 03:56:55 AM »
Mangled schematic.


Edit2 : Added dots to indicate start of winding.. both windings in same direction




TinselKoala

  • Hero Member
  • *****
  • Posts: 13958
Re: 3v OU Flashlight
« Reply #493 on: May 07, 2014, 11:00:26 AM »
Mangled schematic.


Edit2 : Added dots to indicate start of winding.. both windings in same direction
You are "assuming" that the designer of the circuit intended for the burst oscillations to drive the lower-frequency switching of the secondary, thus achieving something like phase-synchrony or locking, at different frequencies. But what reason do you have for your assumption? Maybe the 100R resistor is supposed to add energy by converting some of its mass, under the influence of the spiky oscillation bursts, and feed this energy back into the system via radiant energy. That could be why the CMOS - type inverter is specified. Couldn't it?

My point is that there is no evidence that the various circuit tricks to synchronize two sets of oscillations, fast and slow, have anything to do with the LEDs remaining lit. You are chasing Red Herrings. 

But if you must, you can try the original Guk version or one of the more recent suggestions up above, where the right-hand coil part is actually switched by the transistor/mosfet in one way or another.  However I think you will find that your resulting waveforms may be different _than what Akula actually showed_.

d3x0r

  • Hero Member
  • *****
  • Posts: 1433
Re: 3v OU Flashlight
« Reply #494 on: May 07, 2014, 11:15:11 AM »
You are "assuming" that the designer of the circuit intended for the burst oscillations to drive the lower-frequency switching of the secondary, thus achieving something like phase-synchrony or locking, at different frequencies. But what reason do you have for your assumption? Maybe the 100R resistor is supposed to add energy by converting some of its mass, under the influence of the spiky oscillation bursts, and feed this energy back into the system via radiant energy. That could be why the CMOS - type inverter is specified. Couldn't it?

My point is that there is no evidence that the various circuit tricks to synchronize two sets of oscillations, fast and slow, have anything to do with the LEDs remaining lit. You are chasing Red Herrings. 

But if you must, you can try the original Guk version or one of the more recent suggestions up above, where the right-hand coil part is actually switched by the transistor/mosfet in one way or another.  However I think you will find that your resulting waveforms may be different _than what Akula actually showed_.
no I'm assuming that akula is correct when he states, he started probing the circuit,  and found out the other half was actually defunct. (so beyond your lots of words to hide the fact that the coil is wrong what's your point?  you're not even commenting about my anaylysis of the circuit in d3x0rese terms....

and as then no schematic matches the coil akula uses.

but besides that; could put it in a sim and give it sufficient power and see what it does given a sine wave with proper power... or you could jump proper power to it and pulse it inline and see....


Edit: and I'm not assuming the designer intneded for the busts... I *think* he was trying to head in a different direction and found that worked out better anyway.