PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2022-01-20, 02:21:48
News: Check out the Benches; a place for people to moderate their own thread and document their builds and data.
If you would like your own Bench, please PM an Admin.
Most Benches are visible only to members.

Pages: 1 ... 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 [32] 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
Author Topic: Akula0083 30 watt self running generator.  (Read 827594 times)
Sr. Member
****

Posts: 281
Hi,

There is new video from akula (obviously it was attempted to delete very quickly, so no original link) - https://www.youtube.com/watch?v=ogMcs6iBJwY




Good day All:

OK, the SMPS in the video says '100-240@.35a  to 5vdc @ 5amps' and when he reads the LED voltage he shows 12.44vdc.  So, the SMPS is ONLY powering the ICs? This is the equivalent of the 7812 reg in the 30watt design..... Where is the 12.44vdc generated??

LEDs are two parallel strings of 4 in series.....


That type of SMPS is a 50/60hz AC input..... So the output Frequency from the 2nd coil side is 50/60hz????   What is the input frequency?

If you notice also the Yellow color tampering seal on the side of the SMPS has been half way removed..... Maybe the SMPS has been modified?

ON the primary side of the coil,  two of the wires on the return side  are joined together at the ferrite and then return to the circuit board as one cable only.
So the Primary side of the coil has THREE wires that connect to the Circuit board.   Is that wire connected back unto itself or does it connect to a copper foil????

Two ICs and a minimum of components....   maybe an integrated drive IC?
Both ICs have the numbers scraped off :'(

It is a shame the quality of the video is so blurry....

take care, peace
lost_bro

« Last Edit: 2014-04-18, 17:32:08 by lost_bro »
   
Group: Guest


That type of SMPS is a 50/60hz AC input..... So the output Frequency from the 2nd coil side is 50/60hz????   What is the input frequency?

If you notice also the Yellow color tampering seal on the side of the SMPS has been half way removed..... Maybe the SMPS has been modified?


Very likely zero frequency with the device powered from a battery or super cap inside the SMPS.  :(
   

Group: Moderator
Hero Member
*****

Posts: 2448


Buy me a cigar
Dear Verpies and All.

Having carried out the Capacitor mod earlier today. Please see YouTube video below.

https://www.youtube.com/watch?v=yfRb80SYAI8

Awaiting further instructions !!  :)

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
Group: Guest
I finish soldering...

There are 5 little problems:

1 in the elfa order list is missing C7 100uF/35
2 the holeof  big resistor are a little to close..
3 The Foil trimmer is to close to the Capacitor 2200uF 50V
4 2 more capacitor 100 nF 50V
5 Distrelec didn't ship me the toroid .. I temporary fix one similar..
« Last Edit: 2014-04-19, 09:19:41 by TutorialFE »
   

Group: Professor
Hero Member
*****

Posts: 2683
Copper foils in that orientation will create a radial electric field perpendicular to the coil windings.
Indeed

One of the forces has to rotate (i.e. change position) and the other force is static. Magnetic fields can also be used.  
Like in the MAGVID ?
   

Group: Professor
Hero Member
*****

Posts: 2683
Having carried out the Capacitor mod earlier today. Please see YouTube video below.
It shouldn't do that.
What is on Ch2?
   
Sr. Member
****

Posts: 336
I finish soldering...

There are 5 little problems:

1 in the elfa order list is missing C7 100uF/35
2 the holeof  big resistor are a little to close..
3 The Foil trimmer is to close to the Capacitor 2200uF 50V
4 2 more capacitor 100 nF 50V
5 Distrelec didn't ship me the toroid .. I temporary fix one similar..


TutorialFE,

1. The missing 100uF is added on the corrected board design.
2. Yes but is it not a problem to have the resistors a little up in the air, better cooling.
    I also made the board holes a little larger. More easy to fit the resistors.
3. I know, it depends on the type used. Your solution looks OK to me.
    Made more room around the foil capacitor on the new design.
4. There should be 5 pc. of 100nF, 2 pc. of 22nF, 1 pc. of 3nF and 1 pc. of 68pF.
5. Your replacement toroid is OK.

Remember to lift the plus wire of the C4 electrolytic capacitor and solder a wire from the plus of C4 to pin 14 on the TL494!
Also, did you lift the heat sink for the BD transistor a little off the PCB so that the heat sink do not short the pads underneath?
(This is fixed on the new design.)
I also recommend NOT to solder the L1/L2 bobbin into the board. Use some wires from the board to the L1/L2 coils. This will
make it easier to do experiments.

GL.
« Last Edit: 2014-04-18, 20:27:43 by Groundloop »
   
Group: Guest
@GL I would know also how to turn the trimmers....

Thank you for your patience
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3340
Quote
What about increasing R5 (Akula diagram) to say 10 Ohm? Will it not increase the feedback signal to a level where it does affect the duty cycle?

It could.
Also, changing R5 to an inductive wirewound resistor will dramatically change the signal at the "feedback takeoff point".

My R5 (1 Ohm) is a wirewound resistor, so must have some inductance.

I changed R5 to a 10 Ohm resistor, and the signal across R5 increases as expected, but no change in the pin 3 signal or duty cycle, see screenshots
first one with normal 1 Ohm R5,  2th with 10 Ohm R5.

Signals are:

gate (yellow),
voltage across R5 (blue)
signal on pin3 (purple)
current through L2 (green)

Regards Itsu 
   
Hero Member
*****

Posts: 520
@Itsu

Thanks for the vid. Noticed that your L1 choke has few turns while the others have many turns. Is that irrelevant?

@TutorialFE

Thanks for your photos of your circuit build. I am building up my confidence to start the circuit on my own and those photos will help. Better not to solder in the E-core, plus maybe put long wires at D5 and R3 in case you need to change it out (I will have my Diode Carousel at D5).

@verpies

If the L1 choke is there for when C11 can feed it, why not add a diode across it pointing to R1. This way the starting energy can enter without overtaxing the choke.

I still say R3 is the triple wrong way. A diode there will let the L2 move to the L1 impulse only. This way you will consume much less energy plus L1 will now be able to pulse and get rid of the flux via L2 thus run more smoothly because right now it is creating a mess. Consumption would drop dramatically and then leave the field open to do the fine tuning of the system because this will allow for new resonant frequencies.

Only at that point would remain a question of wind method, turns, D5 diode choice, etc. So the wind method is still a variable as well. My money is on standard mode.

You asked about why the L1/L2 current is the same timing. I guess it will depend if the E-core was wound to T-1000 method or standard method. If T-1000 method with two layers of L2 under the L1, then L1 is hitting the L2 right away without any out of phase core flux bouncing first. The core becomes irrelevant.

Actually in the present set-up, I see the E-core as being useless. May as well send the juice directly to load. When L1 gets its pulse, it has a north and south polarity that is clear, but meanwhile the L2 is being force fed positive from R3 and positive from R2. So where do the north-south flux polarities in L2 go to create positive and negative potential difference in L2? There is no where for the L2 negative to go so it has to cancel out somewhere. If R3 was a diode, you would no longer force feed L2 with positive leaving L2 to set-up its potential difference all while lowering total power consumption.

Thanks for the diagram. Now we're takin the same language (or almost).

wattsup


---------------------------
   

Group: Professor
Hero Member
*****

Posts: 2683
gate (yellow),
No comment ;)

voltage across R5 (blue)
The current through the LED bank increases slightly each time the MOSFET turns off and charges up C3 through D5 due to the flyback pulse generated by L1 as the MOSFET opens.
This is normal and expected behavior of a boost converter.  So far so good.

The narrow spike, at the moment the MOSFET turns OFF is EMI or is due to R5's inductance.  It does not last long enough to charge up the differentiating capacitor at pin 2 and trigger the associated comparator.  To trigger it, the pulse would have to be high for a longer time and then abruptly fall. You have the abrupt fall but not the former.  Also, a feedback pulse going below ground would trigger the second comparator at pin 15.  
You have neither waveform feature in your feedback signal so your feedback loop is not doing anything :(

signal on pin3 (purple)
That one is starting to annoy me.  Apparently it is idle but the scope shows positive and negative spikes on it.  The TL494 cannot produce negative spikes so I must assume that those spikes are an artifact due to EMI.  Time for EMI countermeasures!

current through L2 (green)
Is this channel AC-coupled?  
How can the current in L2 be bidirectional if there is D6 in series with L2 ?

I can see that you did not reverse the polarity of L2 and the current in L1 flows at the same time as the current in L2 - like in a plain transformer  >:(
If L2 was reversed, its current would start to increase the moment the MOSFET turned OFF.
« Last Edit: 2014-04-18, 23:09:16 by verpies »
   

Group: Professor
Hero Member
*****

Posts: 2683
If the L1 choke is there for when C11 can feed it, why not add a diode across it pointing to R1. This way the starting energy can enter without overtaxing the choke.
Because such diode would provide a low impedance path to spikes generated by L2 even if this diode was pointing that direction.
L3 is not overtaxed.  It is not a part of the major power loop - only the auxiliary power flows through it.  Together with C13 and C6, its job is to filter out spikes from this auxiliary power rail because the sensitive TL494 is powered from it.  If this choke is not doing its job, then EMI shows up at the TL494.  Just look at Itsu's impossible negative spikes on pin 3.

I still say R3 is the triple wrong way. A diode there will let the L2 move to the L1 impulse only.
No. There is already a diode in series with L2.  Why can't you see it?
I redrew the diagram especially for you and swapped the position of R3 with the D6 block.  Electronically that swap makes no difference because series connections are commutative... but maybe it will open your eyes graphically.
« Last Edit: 2014-04-19, 00:18:33 by verpies »
   
Group: Guest
In practical terms, the closing of the 'STOP' switch causes an increase in supply current and decrease in LED lumen level, irrespective of coil phasing, so I can see no advantage in having L2 in circuit.

Hoppy
   
Group: Guest
I noticed something strange in the Pcb board...

Groundloop tracks so thin can cause problems in electronics???  
« Last Edit: 2014-04-19, 15:11:54 by TutorialFE »
   
Sr. Member
****

Posts: 331
Hi,

Here is higher quality video from akula - https://www.youtube.com/watch?v=7V8vC0UL-XY
There he showed PCB connnections and some measurements as well.

Cheers!
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3340
@Itsu

Thanks for the vid. Noticed that your L1 choke has few turns while the others have many turns. Is that irrelevant?


Well,  i copied it from the original picture which had 7 turns on a similar sized core.
As i have it connected via a PCB connector i can swap it out easely which i did with severall other (more turns) chokes, but no real difference is noted.

Regards Itsu
   
Group: Guest
@ T-1000 I think we are close to the replica .....
« Last Edit: 2014-04-19, 11:17:53 by TutorialFE »
   
Group: Guest
Here is higher quality video from akula
There he showed PCB connnections and some measurements as well.

He clearly wants to be a teacher and is pretty good at it, especially the one handed camera work.

What do we need to do to form a classroom and let Akula be the instructor until we all understand and can do it ourselves?

http://www.youtube.com/watch?v=pUW-3yXQIWQ

Seems he tries pretty hard to teach us what is needed to make this work.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3340

Quote
voltage across R5 (blue)
Quote
You have neither waveform feature in your feedback signal so your feedback loop is not doing anything :(

How is that with other builders?   Is the feedback loop also not doing anything?

Quote
signal on pin3 (purple)
Quote
That one is starting to annoy me.  Apparently it is idle but the scope shows positive and negative spikes on it.  The TL494 cannot produce negative spikes so I must assume that those spikes are an artifact due to EMI.  Time for EMI countermeasures!

Ok,  where is my stack of 0.1uF caps......

Quote
current through L2 (green)
Quote
Is this channel AC-coupled?  
How can the current in L2 be bidirectional if there is D6 in series with L2 ?

AC coupled would display a little sine wave symbol after the BW (bandwidth reduced to 150MHz) symbol,  so no, not AC coupled
I measured D6 and it seems to function correctly  :D
EDITTED:  its an AC current probe,  guess this would ALWAYS show AC only,   Right?

Quote
I can see that you did not reverse the polarity of L2 and the current in L1 flows at the same time as the current in L2 - like in a plain transformer  >:(
If L2 was reversed, its current would start to increase the moment the MOSFET turned OFF.

Well that could be because the way i insert the current probe into the circuit (when i turn it 180° it will show the opposite)
Notice that i sometimes invert this current channel to what i think would be good (current flowing when gate signal is high), but thinking about it that would not be the correct thinking in ALL situations, so i have to come up with a solid solution on how to insert the current probe into the circuit.


Regards Itsu
   

Group: Professor
Hero Member
*****

Posts: 2683
Groundloop's tracks are so thin, that they can cause problems in electronics???  
The thin darker lines are the gaps between tracks...not the tracks themselves.
   
Group: Guest
Hi,

Here is higher quality video from akula - https://www.youtube.com/watch?v=7V8vC0UL-XY
There he showed PCB connnections and some measurements as well.

Cheers!


It looks to me like the Kapanadze 9V battery latching circuit trick. The short battery pulse produces HV from the coil into the SMPS which latches a switch to supply battery (in SMPS) current back to the device - simple! He is relying on the Kapanadze 9V battery ruse to convince us that he has achieved self-running.  :(

Hoppy
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3340

I will from now on use the below current probe setup when measuring the L1 / L2 currents.
I adjust the L1 current screenshot to show positive current when the gate is positive and use the same setting for the L2 current.
That way both should show the actual direction of the current in relation to each other.

Regards itsu

   

Group: Professor
Hero Member
*****

Posts: 2683
Here is higher quality video from akula - https://www.youtube.com/watch?v=7V8vC0UL-XY
There he showed PCB connnections and some measurements as well.
I will let Hoppy have a go at it.
I predict that he will take it apart because Akula did not take the AC/DC-DC converter apart.
Maybe even the phrase "hidey hole" will be resurrected.
   

Group: Professor
Hero Member
*****

Posts: 2683
In practical terms, the closing of the 'STOP' switch causes an increase in supply current and decrease in LED lumen level, irrespective of coil phasing, so I can see no advantage in having L2 in circuit.
The conventional circuit analysis supports this conclusion.

However if there is any unconventional magic* to be had in this system then it will manifest itself exactly across this L2 as a high power spike that will recharge C11 through D6 ...and most importantly, such L2 pulse will cause an abrupt decrease the voltage at the 'feedback takeoff point".

If you were following my conventional circuit analysis, TL494 feedback loop is configured to react exactly to such abrupt voltage decrease at this point. AFAIR: -dV/dt > -6V/ms


* e.g. magnetoacoustic pulse or beta current pulse due to NAR or NMR, etc...
   
Group: Guest
The conventional circuit analysis supports this conclusion.

Howeve,r if there is any unconventional magic* to be had in this system then it will manifest itself exactly across this L2 as a high power spike that will recharge C11 through D6 ...and most importantly, such L2 pulse will cause an abrupt decrease the voltage at the 'feedback takeoff point".

If you were following my conventional circuit analysis, TL494 feedback loop is configured to react exactly to such abrupt voltage decrease at this point. AFAIR: -dV/dt > -6V/ms


* e.g. magnetoacoustic pulse or beta current pulse due to NAR or NMR, etc...

I think its high time we all wise-up to this Akula chap!!
   
Pages: 1 ... 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 [32] 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2022-01-20, 02:21:48