PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2020-07-03, 11:52:43
News: Registration with the OUR forum is now by invitation only.

Pages: 1 2 [3] 4
Author Topic: Akula0083 Lantern v3  (Read 82322 times)
Newbie
*

Posts: 23
So how found frenquency of ferite? Akula say, wound one turns and then LC circuit frenquency be much higher and we see frenquency of ferite.
   
Group: Elite
Hero Member
******

Posts: 3639
It's turtles all the way down
There is a serious difference between the hand drawn schematic schem2.jpg and later cleaned up schematics of the Dual TL494 Akula lamp driver.

The hand drawn schematic has an separate external generated power supply (19 turn winding, 3 diodes and 2200 uF cap) that feeds just the 7 and 13 turn windings, it is not connected to a main bus  powering the chips and LED's bus as in the other cleaned up schematics. This has a big advantage.


Upon startup current is fed from the battery through the 13 turn winding and through SF 56 diode to allow the chips to power up.

After power up, the internal chip power bus is allowed to be pumped up higher than the battery applied voltage by the flyback action of the 13 turn winding and SF56. The LED bank is attached to the internal power supply bus.

It is the disconnection of the two separate supplies that allows the internal chips and LED  power supply to be pumped higher than the applied power when first touched by the battery.

This also regulates the burst mode (shutdown of H4 oscillator) via the 10k voltage preset pot on pin 4, as well as regulating the internal supply.

All of the cleaned up schematics (except the Chris Sykes version) have both the internal and external (secondary) supplies connected together. This causes the problem with the flyback diodes.

I called attention to the flyback diodes VD3 and VD6 in a prior post. The way they are drawn in the cleaned up schematics, they will allow little to no power in the 19 turn secondary as they completely snub and recirculate the current in the 7 and 13 turn windings

I am told the hand drawn schematic was reverse engineered from captures of the printed circuit board.  The hand drawn schematic makes much more sense to me.

Edit: Added the corrected schematic AkyDoubleTL494.gif
« Last Edit: 2014-05-07, 23:26:46 by ION »


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

Group: Elite Experimentalist
Hero Member
*****

Posts: 2177

Hi Ion,

to add to the confusion, there seems to be an "updated" hand drawn schematic, see the one posted by Groundloop on page 1 of
this thread:  http://www.overunityresearch.com/index.php?topic=2420.msg38125#msg38125

Spot the differences: 

removed SF56 on the bottom MOSFET
removed text "12mA 9v" above the pin 12 lines
Added + - symbols plus text right of the lower MOSFET
added text "FNR 109" below the varistor symbol

Regards Itsu
   
Group: Elite
Hero Member
******

Posts: 3639
It's turtles all the way down
Regarding the two hand drawn schematics and their differences:

In the uncorrected schematic (schem2.jpg, it is the wire drawn just to the left of the switch to the upper power rail that is one concern. This line effectively ties the chips and led supply voltage to the battery / 19 turn winding supply. It also causes the upper SF56 to shunt the primary 13 turn winding.

It is well known that shunting the drive winding with a diode in a flyback converter kills any transfer of energy to the secondary. The energy is all wasted as heat in the diode and the resistance of the winding. Snubber networks are used instead to tame the overshoot and ringing due to leakage inductance to a safe value, while allowing as much energy to be transferred to the secondary as possible.

SF56 on the lower FET to the rail also circulates all flyback energy in the 7 turn winding, wasting it. Same for the upper FET and it's SF56 if that wire is in place. The wire was removed on the schematic you posted, a derivitive of AkyDoubleTL494.gif, as was the lower SF56, now allowing the transfer of energy to the 19 turn and allowing for a possible "bootstrapped" operation with no waste of energy. (except for the LED's and chip quiescent current).

I believe the schematic you posted is the more workable corrected version.

Unfortunately, the early cleaned schematics do not reflect all of these changes.

When I get back to the bench I will incorporate them into my build.

Note: I just added the corrected schematic to my earlier post



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

Posts: 337
Regarding the two hand drawn schematics and their differences:

In the uncorrected schematic (schem2.jpg, it is the wire drawn just to the left of the switch to the upper power rail that is one concern. This line effectively ties the chips and led supply voltage to the battery / 19 turn winding supply. It also causes the upper SF56 to shunt the primary 13 turn winding.

It is well known that shunting the drive winding with a diode in a flyback converter kills any transfer of energy to the secondary. The energy is all wasted as heat in the diode and the resistance of the winding. Snubber networks are used instead to tame the overshoot and ringing due to leakage inductance to a safe value, while allowing as much energy to be transferred to the secondary as possible.

SF56 on the lower FET to the rail also circulates all flyback energy in the 7 turn winding, wasting it. Same for the upper FET and it's SF56 if that wire is in place. The wire was removed on the schematic you posted, a derivitive of AkyDoubleTL494.gif, as was the lower SF56, now allowing the transfer of energy to the 19 turn and allowing for a possible "bootstrapped" operation with no waste of energy. (except for the LED's and chip quiescent current).

I believe the schematic you posted is the more workable corrected version.

Unfortunately, the early cleaned schematics do not reflect all of these changes.

When I get back to the bench I will incorporate them into my build.

Note: I just added the corrected schematic to my earlier post



ION,

I have checked the two hand drawings. The later one does not have the wire up to plus. So the version you posted
above is a newer version that was not available to me when I did the cleaned up version.

GL.
   
Group: Elite
Hero Member
******

Posts: 3639
It's turtles all the way down
ION,

I have checked the two hand drawings. The later one does not have the wire up to plus. So the version you posted
above is a newer version that was not available to me when I did the cleaned up version.

GL.
GL
I understand that and hope it is a help that I have analyzed the two schematics. Your work is greatly appreciated.


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

Group: Elite Experimentalist
Hero Member
*****

Posts: 2177
ION,

I have checked the two hand drawings. The later one does not have the wire up to plus. So the version you posted
above is a newer version that was not available to me when I did the cleaned up version.

GL.

Hmmm,  i do not agree here, as the version you posted Groundloop surely was already updated (newer) looking at the erased parts (lower SF56,  "12mA 9v" text above the pin 12 lines, wire up to plus, etc.).


Regards itsu
   
Group: Elite
Hero Member
******

Posts: 3639
It's turtles all the way down
Hmmm,  i do not agree here, as the version you posted Groundloop surely was already updated (newer) looking at the erased parts (lower SF56,  "12mA 9v" text above the pin 12 lines, wire up to plus, etc.).


Regards itsu


Itsu:

Can you name the schematic GL posted, as far as I know, the two power supplies were accidentally tied together with most of the later schematics and the lower SF56 is an error.

The correct schematic should not have the two supplies tied together (line from junction of 13 turn, 7turn, switch, 1000uF and battery terminal should not connect to upper TL494 and LEDs), and should delete the lower SF56 diode. Also the line from pin 12 of lower chip to the 1000uF capacitor should be removed.

The attached schematic is the correct one:


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

Posts: 337
Hmmm,  i do not agree here, as the version you posted Groundloop surely was already updated (newer) looking at the erased parts (lower SF56,  "12mA 9v" text above the pin 12 lines, wire up to plus, etc.).


Regards itsu


Itsu,

There is so many versions floating around on the net. I do not think it matter much anyway.

GL.
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 2177
Itsu:

Can you name the schematic GL posted, as far as I know, the two power supplies were accidentally tied together with most of the later schematics and the lower SF56 is an error.

The correct schematic should not have the two supplies tied together (line from junction of 13 turn, 7turn, switch, 1000uF and battery terminal should not connect to upper TL494 and LEDs), and should delete the lower SF56 diode. Also the line from pin 12 of lower chip to the 1000uF capacitor should be removed.

The attached schematic is the correct one:


Correct, thats the one Groundloop posted in his post on page 1:   http://www.overunityresearch.com/index.php?topic=2420.msg38125#msg38125

Regards Itsu
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 2177
Itsu,

There is so many versions floating around on the net. I do not think it matter much anyway.

GL.


well, if its that version that gives ION a good feeling, i would go for that one first.

Regards Itsu
   
Group: Elite
Hero Member
******

Posts: 3639
It's turtles all the way down
well, if its that version that gives ION a good feeling, i would go for that one first.

Regards Itsu

Regarding reply #12, the upper schematic (AkyDoubleTL494.gif) was a rev2 reverse engineer of the board from the Russian site, all other schematics are derivatives of that schematic or the rev1 (shem2.jpg) version and include additional errors because they conform to  rev1.

Not to say that this schematic is 100% correct, but, the derivative errors would be now compounding.

To clarify the first hand drawn schematic (shem2.jpg) had the errors, these were erased out in the second hand drawn schematic (AkyDoubleTL494.gif.(you can see the erase marks on the paper). The first hand drawn schematic was the basis for most all of the the cleaned up derivatives, errors and all.

If you are good at switch mode circuit analysis, you will see why it is my preferred schematic when you compare it to the derivatives.

I have explained my reasoning in earlier posts, especially post #53 above. I will be modifying my build and report in soon.


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

Group: Renaissance Man
Hero Member
*****

Posts: 2205


Buy me a cigar
Regarding reply #12, the upper schematic (AkyDoubleTL494.gif) was a rev2 reverse engineer of the board from the Russian site, all other schematics are derivatives of that schematic or the rev1 (shem2.jpg) version and include additional errors because they conform to  rev1.

Not to say that this schematic is 100% correct, but, the derivative errors would be now compounding.

To clarify the first hand drawn schematic (shem2.jpg) had the errors, these were erased out in the second hand drawn schematic (AkyDoubleTL494.gif.(you can see the erase marks on the paper). The first hand drawn schematic was the basis for most all of the the cleaned up derivatives, errors and all.

If you are good at switch mode circuit analysis, you will see why it is my preferred schematic when you compare it to the derivatives.

I have explained my reasoning in earlier posts, especially post #53 above. I will be modifying my build and report in soon.

Dear ION.

I have been a little busy with the 30 W device but I am following the other threads as well.

Would you kindly re post the schematic and might I humbly request it be given a new suffix.  Perhaps on the lines of OUR lantern replication.  I myself am finding it difficult to follow all these versions !!  New broom ??   O0

Just my thoughts.

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
Group: Elite
Hero Member
******

Posts: 3639
It's turtles all the way down
Dear ION.

I have been a little busy with the 30 W device but I am following the other threads as well.

Would you kindly re post the schematic and might I humbly request it be given a new suffix.  Perhaps on the lines of OUR lantern replication.  I myself am finding it difficult to follow all these versions !!  New broom ??   O0

Just my thoughts.

Cheers Grum.

The "AkyDoubleTL494" schematic is the correct one. I hate to rename it as it will just add to the confusion. The link you sent to me from the Russian site also had the correct schematic, but a strange name. It is a much lower resolution version, I wouldn't use it unless you have no problem reading the fine stuff.

John Blade is also reverse engineering the board and started a new thread for it at OU.com.

On another note I was able to make the changes to my breadboard and it does indeed now produce the packets, but requires very careful adjustment( I'm using 10 turn trimmers for good resolution of setting). There are interesting very crisp changes in light intensity as the frequency is adjusted. This occurs in bands dim / bright /dim /bright for very small changes in frequency. I suspect it is the oscillators locking in then losing sync.

No stand alone operation though, but I believe the circuit is very close to working correctly, it might be just a matter of finding the right settings.

More to come.


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

Posts: 337
The "AkyDoubleTL494" schematic is the correct one. I hate to rename it as it will just add to the confusion. The link you sent to me from the Russian site also had the correct schematic, but a strange name. It is a much lower resolution version, I wouldn't use it unless you have no problem reading the fine stuff.

John Blade is also reverse engineering the board and started a new thread for it at OU.com.

On another note I was able to make the changes to my breadboard and it does indeed now produce the packets, but requires very careful adjustment( I'm using 10 turn trimmers for good resolution of setting). There are interesting very crisp changes in light intensity as the frequency is adjusted. This occurs in bands dim / bright /dim /bright for very small changes in frequency. I suspect it is the oscillators locking in then losing sync.

No stand alone operation though, but I believe the circuit is very close to working correctly, it might be just a matter of finding the right settings.

More to come.

ION,

Attached is the circuit that John Blade found when he did a reverse engineer drawing of the actual PCB that Akula did use.
On his circuit there is a direct connection between the two plus rails. I have attached his drawing here.

One other issue, I think I have tracked down the "FNR" component. It is most probably a ZINC OXIDE LEADED VARISTOR.
See attached data sheet.

GL.
   
Group: Tech
Sr. Member
****

Posts: 291

No stand alone operation though, but I believe the circuit is very close to working correctly, it might be just a matter of finding the right settings.


....and failing that, further pondering on that green cap strapped across the supply rails! It appears to have a 5V rating, which would indicate a high capacity.

Hoppy
   
Group: Elite
Hero Member
******

Posts: 3639
It's turtles all the way down
....and failing that, further pondering on that green cap strapped across the supply rails! It appears to have a 5V rating, which would indicate a high capacity.

Hoppy

Yes, a 5V rating would indicate high capacity for that size, like 0.1 F or more. Are we wasting our time here?

Regarding the schematic discrepancies, I have built it both ways and find the schematic I posted "AkyDoubleTL494.gif" works the best, but is not exhibiting self run.

I set up each oscillator separately, then ran them both together.

The H4 oscillator I set up for as low a duty cycle as possible and maximum high frequency. Then I set the voltage adjustment so that the H4 oscillator triggers on and off "in a binary manner" by triggering on the ripple voltage on the internal bus. This then gives the HF packets (picket fence) with about a 50% duty cycle of "packets on to packets off".

Then I introduce the B4 oscillator with a much lower frequency and low duty cycle. There are many frequency adjustments of this oscillator that cause very discrete bands of dimming then getting very bright, a lot more action to the controls than with the earlier schematic.

Since I don't have a clue where the sweet spot is, one could spend an eternity hunting for the best setting position of the 5 controls.

GL: Yes, those varistors also come in aqua color from other manufacturers. Regarding John Blade's reverse engineering: I am trying to verify this but need a higher quality scanned version of the print. There are some circled sections of the initial version that need developing, as well as other areas. I will try compare it to the other reverse engineered print of the board.

All for now, will experiment more as time permits.

For Peter: embellishments (emoticons, quotes, bold, italics etc.) to the post are not working for me, is this my computer problem? I'll try a different computer.
« Last Edit: 2014-05-09, 14:31:43 by ION »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Tech
Sr. Member
****

Posts: 291
Yes, a 5V rating would indicate high capacity for that size, like 0.1 F or more. Are we wasting our time here?

Not necessarily because its possible that the first digit is missing off the sleeve, which as Peterae commented some while ago, is not uncommon. However, I have since viewed many angle variations and cannot see any clear indication that a digit is missing. it is fairly clearly not a 50V rated cap from close observation. Another factor that indicates to me that this capacitor is powering the device (when the 9V battery is removed), is that there is no other reason to place an electrolytic cap as close as possible to the wire connections to a 9V battery that is just connected for short time to start the device. It is not smoothing anything, or in a feedback loop as far as I tell from the schematics that have been so far produced from back engineering the circuit board. Added to this IMO Akula has clearly faked his big table top garden demo and given that he is capable of this where does faking stop with him.

I do feel that there is an interesting capacitive effect with his copper tape wound ferrite coils which I'm currently investigating but think he is simply trying to enhance the effect by adding a bit of added energy in order to light a few extra LED's to make the demo look more convincing as self-runner.

Hoppy
   
Sr. Member
****

Posts: 277
Hello All:
This is a post  I made earlier today @ O.U.com:

My potcores arrived (4.7cm / Al 9500) so i went for the 5.4 / 2.7mH setup using a pot core.
......
Regards Itsu

Good day Itsu

Have you tried to 'characterize' the 'ferro-resonance' frequency of this new POT core?

Do you have access to a spectrum analyzer?

I agree with Vortex1 (ION) that one could spend an eternity trying to guess the tuning for this animal without  some kind of starting point for reference.

I would love to know if anyone has used a spectrum analyzer, like the Rigol seen in the Akula videos to 'characterize'  their cores?

To carry out a thorough analysis of the problem one must consider *all* of the datum .... maybe we should consider to look at it from the frequency domain perspective.

take care, peace.
lost_bro
   

Group: Renaissance Man
Hero Member
*****

Posts: 2205


Buy me a cigar
Hello All:
This is a post  I made earlier today @ O.U.com:

Good day Itsu

Have you tried to 'characterize' the 'ferro-resonance' frequency of this new POT core?

Do you have access to a spectrum analyzer?

I agree with Vortex1 (ION) that one could spend an eternity trying to guess the tuning for this animal without  some kind of starting point for reference.

I would love to know if anyone has used a spectrum analyzer, like the Rigol seen in the Akula videos to 'characterize'  their cores?

To carry out a thorough analysis of the problem one must consider *all* of the datum .... maybe we should consider to look at it from the frequency domain perspective.

take care, peace.
lost_bro

Dear lost_bro.

I have set up a thread here..... http://www.overunityresearch.com/index.php?topic=2430.msg38352;topicseen#msg38352

This is for exactly what you have just commented upon.  If you or any members have any ideas how to pursue this  line of investigation, I for one would be most grateful !!

Cheers Grum.


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

Group: Renaissance Man
Hero Member
*****

Posts: 2205


Buy me a cigar


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

Group: Elite Experimentalist
Hero Member
*****

Posts: 2177
Hello All:
This is a post  I made earlier today @ O.U.com:

Good day Itsu

Have you tried to 'characterize' the 'ferro-resonance' frequency of this new POT core?

Do you have access to a spectrum analyzer?

I agree with Vortex1 (ION) that one could spend an eternity trying to guess the tuning for this animal without  some kind of starting point for reference.

I would love to know if anyone has used a spectrum analyzer, like the Rigol seen in the Akula videos to 'characterize'  their cores?

To carry out a thorough analysis of the problem one must consider *all* of the datum .... maybe we should consider to look at it from the frequency domain perspective.

take care, peace.
lost_bro

Hi Lost_bro,

no, i did not try to 'characterize' the 'ferro-resonance' frequency of this new POT core, nor do i have a spectrum analyzer, but i did not had my birthday yet this year  ;)

Regards Itsu
« Last Edit: 2014-05-09, 21:52:19 by Itsu »
   
Sr. Member
****

Posts: 277
Hi Lost_bro,

no, i did not try to 'characterize' the 'ferro-resonance' frequency of this new POT core, nor do i have a spectrum analyzer, but i did not had my birthday yet this year  ;)

Regards Itsu

Hi Itsu:

Best bang for the buck is the Rigol DSA815-TG;

 http://www.tequipment.net/spectrum-analyzers/spectrum-analyzers-1/

Akula has the Rigol DSA815 basic model without the tracking generator.  The Rigol DSA815-TG has the tracking generator which in my opinion is worth the extra $200.00 USD, and is only available as a factory installed option.

The link has the latest price... not to bad, thinking about getting one myself. O0

take care, peace
lost_bro
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 2177

Yes,  i have my eye also on the Rigol DSA815-TG for a while, but i have my doubts about the 1.5GHz.

I think it will be to low, but then again, its better then having no spectrum analyzer and the difference between one which can reach say 3GHz is very big.

Regards Itsu
   

Jr. Member
**

Posts: 90
Transipnotica state? Leaves the cave!!!
@All

I've lucky found another ferrite from tv and I've discover the thikness of two (piece of paper that is shown in akula disassembled video but they are not paper) piece of plastic.

They are two piece of plastic 0.5 mm thikness

Also I think that the flyback that I've found is the easiest to take ferrite without any effort.
« Last Edit: 2014-05-11, 19:32:00 by TutorialFE »
   
Pages: 1 2 [3] 4
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2020-07-03, 11:52:43