PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
OverUnity Research > Benches > Grumage (Moderator: Grumage) > Akula0083 30 watt self running generator.
2024-11-26, 22:12:58
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 ... 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 974992 times)
Topic Options img
img  ion
Re: Akula0083 30 watt self running generator. « Reply #975 on: 2014-04-30, 14:56:46 »
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
Quote from: wattsup on 2014-04-30, 00:58:21
@All

Just received my ecores and I am very disappointed at the size of these little critters. The ETD 29 fits the board but I am so afraid it is way to small to push 30 watts or more of transfer to recharge the system and supply a load. How did we wind up with this core size? The former only has 7 terminals per side wheres the AK30 former has 10. That is a big difference in size. Dammit, I also received some slightly larger ETD 34's but they did not ship the formers. But I am sure those are too small as well. If the core is not the right size, then pumping all this juice into them is useless because you will never get it back.

I think I am going to order real ecores that have a 10 terminal former otherwise we cannot compare this to Akula's device.

wattsup

Size of the core is usually inversely proportional to the operating frequency. You can push 30 watts through a very tiny core if the operating frequency is high enough. This is the trend in modern switchmode converters.  Whereas early designs used 20 kHz and fairly large cores, modern designs approaching a megahertz use very tiny cores, all other factors being equal e.g. amount of power thruput.

The problem with HF operation is core loss and switching loss goes up unless one is an experienced designer and pays careful attention to core material type and reducing switching loss to a minimum.

One can actually do away with the core altogether and use just air or vacuum for coupling in the RF range, as is the practice in large and small radio transmitters.

For FE research, I prefer operating my switchers at lower frequencies, between 20 kHz and 100 kHz where switching loss and core loss are easily minimized, such that possible OU  operation in a looped back converter does not have to overcome these sometimes formidable losses.

Naturally when replicating a design it is best to stick with the size core specified by the inventor.


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
img  Matt Watts
Re: Akula0083 30 watt self running generator. « Reply #976 on: 2014-04-30, 15:33:36 »
Group: Guest
ION, how did you become so practical?   I like that in an engineer.  We could use more of them.   :)
   
img  forest
Re: Akula0083 30 watt self running generator. « Reply #977 on: 2014-04-30, 21:11:24 »
Sr. Member
****

Posts: 472
ION

Do you have practical way to determine the max power the choosen core can pass depending on frequency ? Something which could help with unknown core ?
   
img  T-1000
Re: Akula0083 30 watt self running generator. « Reply #978 on: 2014-04-30, 22:10:00 »
Sr. Member
****

Posts: 375
After spending some time with Grumage I did find out stuff about TL494 which is not working fully as per supposed function. The error amplifier inputs stay always on DC level which is not good at least for Grumage.

For easy fix:
The pot R11 which is supposed to adjust generator ON and OFF modes has to be fed back from feedback on C5/R5. So you can arrange switch off level as soon as LEDs get lit and wait until C3 gets discharged then generator will kick in for another cycle.
Note: if there will be not enough voltage to make generator shut down when LEDs are on you might get voltage reference right from capacitor C3.

Please see modified circuit diagram attached:
   
img  lost_bro
Re: Akula0083 30 watt self running generator. « Reply #979 on: 2014-04-30, 23:14:04 »
Sr. Member
****

Posts: 281
Quote from: T-1000 on 2014-04-30, 22:10:00
After spending some time with Grumage I did find out stuff about TL494 which is not working fully as per supposed function. The error amplifier inputs stay always on DC level which is not good at least for Grumage.

For easy fix:
The pot R11 which is supposed to adjust generator ON and OFF modes has to be fed back from feedback on C5/R5. So you can arrange switch off level as soon as LEDs get lit and wait until C3 gets discharged then generator will kick in for another cycle.
Note: if there will be not enough voltage to make generator shut down when LEDs are on you might get voltage reference right from capacitor C3.

Please see modified circuit diagram attached:

Thanks for that info!!!

Yes, it makes perfect sense in as much that the Vc of the Circuit is a constant voltage and that was used as the pos. reference for the old position of the voltage divider.... now it's pos. reference is  the cycle by cycle variable charge on C5......

So,  do you think this error was an oversight on Akula's part ???  or something intentional to throw a wrench in the works......


take care, peace
lost_bro
   
img  NerzhDishual
Re: Akula0083 30 watt self running generator. « Reply #980 on: 2014-04-30, 23:48:03 »
Group: Guest

Hi Itsu,

Thanks+++ to having taken your time to make this experiment and this vid.

I have to apologize. My circuit was not accurate enough.
A more accurate version is:


I gave some precisions in: New_index2.html
I have figured out that adding this f' "Optional Cap" is kinda cheating!

The right circuit to be tested is:


You can put your scope probes where you want! It could merely change the resonant freq.

I have noticed only KHZ ranges (30-120) and 20 Volts PK-PK input versus about, saying,  120/150 volts
output, IE, enough to lighting this bulb. This, with, also, only few mH (single) coil.

Now, if you PM me your snail mail address, I would be very pleased to dispatch you one of my
(old) not 'unpackaged'  1 watts/ 220vols (round shaped) leds bulb!

OK! I now guess that I will have to make a vid! My first (NRG) one. In English, preferably. No?

Cheers,
Jean
   
img  NerzhDishual
Re: Akula0083 30 watt self running generator. « Reply #981 on: 2014-05-01, 00:23:05 »
Group: Guest

Hi Itsu,

Sorry, I did not fully read your post! Shame on me. :-[

Quote from: Itsu on 2014-04-30, 11:34:07
.............
 so perhaps you can mention this website in a reply in case other wants to join in.
..............

Of course I can: http://freenrg.info/4I/New_index-3.html

I have, right now, figured out that  the best way to be understood(?) - at least  listened (to(?))?
but  also  copiously mocked is to design a vid, preferably commented in English.
Why not?

En Français, cependant: Meilleurs Sentiments,
Jean
   
img  ion
Re: Akula0083 30 watt self running generator. « Reply #982 on: 2014-05-01, 02:50:09 »
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
Quote from: Matt Watts on 2014-04-30, 15:33:36
ION, how did you become so practical?   I like that in an engineer.  We could use more of them.   :)

Too many (50+) years on the bench. ;)

forest asked:

Quote
Do you have practical way to determine the max power the choosen core can pass depending on frequency ? Something which could help with unknown core ?

You have to determine how much power you are willing to waste in heating the core, then the answer is easy.

To clarify, it's all about your energy budget and your starting point e.g if you are told to design a power supply that must be 90% efficient, then how you spend your 10% must be practically divided among core loss, winding loss, switching loss, rectifier loss, etc.

You must also take into account many other factors too numerous to mention e.g. expected operating temperature range of the complete unit, size, weight, cooling method and so on.

But for a practical answer, if you can keep your finger on the core, you are probably not too close to the limit.


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
img  forest
Re: Akula0083 30 watt self running generator. « Reply #983 on: 2014-05-01, 08:29:41 »
Sr. Member
****

Posts: 472
ION

Thank you.
What I asked was the common situation for me. I get ferrite cores from scrap yard.In case of EI cores dismantled from computer PSU or DC-DC converters I can find TDK pdf and by measuing dimentions I can find what type of core it is and then by carefully underestimate magnetic flux density for saturation this core become usable and I have build successfully some converters using it. The magic is described wonderfully in this article : http://tahmidmc.blogspot.com/2012/12/ferrite-transformer-turns-calculation.html

However , in case of flyback ferrite or yoke ferrite I can only find some rough documents about material types and flux density, but how to compute effective cross-sectional area ? In other words how to know what  current I can pass through primary to avoid core saturation .... and so on.... We need a practical methods to use such cores without burning a lot of mosfets. I'm sure it is perfectly possible , yet tricky maybe!
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #984 on: 2014-05-01, 10:05:15 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: NerzhDishual on 2014-05-01, 00:23:05
Hi Itsu,

Sorry, I did not fully read your post! Shame on me. :-[

Of course I can: http://freenrg.info/4I/New_index-3.html

I have, right now, figured out that  the best way to be understood(?) - at least  listened (to(?))?
but  also  copiously mocked is to design a vid, preferably commented in English.
Why not?

En Français, cependant: Meilleurs Sentiments,
Jean


Hi Jean,   ok,  i got the picture, its basically how i was using the circuit.

I got me 2 led bulbs see picture (6.5W and 3W), but only the 3W bulb keeps the resonating frequency of about 40Vpp at 810KHz.
When connecting the bigger 6.5W, the resonance drops.
Neither led will turn on   :(

I will send you a PM with my snailmail address

Yes,  a video will be great, just show me the circuit with its connections to the FG, scope etc, you then can make some written comments here.

Thanks,  regards Itsu
   
img  ion
Re: Akula0083 30 watt self running generator. « Reply #985 on: 2014-05-01, 12:01:12 »
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
forest said:

Quote
However , in case of flyback ferrite or yoke ferrite I can only find some rough documents about material types and flux density, but how to compute effective cross-sectional area ? In other words how to know what  current I can pass through primary to avoid core saturation .... and so on.... We need a practical methods to use such cores without burning a lot of mosfets. I'm sure it is perfectly possible , yet tricky maybe!

You will never burn up your mosfets if you use a few tricks of the trade, like always observe current through the inductor using a small 0.1 ohm current shunt resistor. For a given core and inductance, if there is a sudden upturn at the end of the ramp, you are entering saturation and will need to either decrease the duty cycle or increase the frequency. The importance of a gap in a flyback converter is important in flushing out remnant magnetism, which is also a cause of saturation. It is always a balance between required inductance, gap size and # of turns.

Best also to current limit your power supply while testing until you gain confidence with the design. I also sometimes add a small incandescent lamp in series with the power supply and avoid large onboard caps which can take out a fet from instantaneous current during a hiccup.


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
img  wattsup
Re: Akula0083 30 watt self running generator. « Reply #986 on: 2014-05-01, 12:26:36 »
Hero Member
*****

Posts: 520
@ION

Thanks for your reply but let me articulate my point with this photo below.

Here you see the actual size of ETD-29 that fits both circuit boards compared to an ETD-34, which is still very small and also fits the circuit boards. Consider that the ETD-29 ecore is being asked to dish out a good 30 something watts plus looping energy and compare this to the ccore used in the v3 device that's only putting out a few watts. There is no logic possible to explain any more clearly.

@All

I have ordered some ETD-54 - Type 87 non-gapped to do my tests. If you are using the ETD-29, maybe consider changing your load to not more then 2-3 watts to start and then there may be some small changes required in the circuit to lower the overall throughput. Maybe put just enough load so when the circuit is energized (large caps full) and when the power is removed, the leds would take at least 2-3 seconds to turn off. That would be one way to choose maximum load with the ETD-29.  

Come to think of it, this one core size comparison, that I did not realize earlier until I had an ETD29 in my hands, now confirms to me that the x-named41 video was faked 2000%. That little ecore that he held between his fingers, like I have in my hand right now, could never do the job he is purporting. Wrong core size. So he just spiked his circuit board from under the panel and faked it, like I had already shown.

It takes time to figure things out. On the way, we discover new aspects that then build on others. Call it a process. If you are running the AK30 board, you will require more core if yo want to produce the design output.

wattsup



---------------------------
http://etherimpress.com
   
img  ion
Re: Akula0083 30 watt self running generator. « Reply #987 on: 2014-05-01, 14:55:05 »
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
Maybe I can clarify by using an analogy.

The size of the inductor and it's inductance represents the size of the shovel moving coal from the coal bin to the steam boiler / engine. For a certain horsepower output, you need to shovel coal at a certain rate (frequency).

If you use a very large shovel, (large inductance) you can go at a slow rate (low frequency) to satisfy required coal input.

If you are using something the size of a spoon (tiny inductance), you need to shovel very quickly (high frequency) to satisfy the input coal requirement.

The coal represents the magnetic energy being transferred through the inductor on each cycle.

Each bite of coal is the amount of charge put into the inductor during the current ramp up cycle.

Dumping it into the steam boiler represents the discharge (flyback) portion of the inductor cycle.

The work performed by the operator and spillage in moving the coal represents the switching loss, core loss, winding loss.

You can imagine that trying to move the coal very rapidly will result in increased spillage (above mentioned losses)

Inductance is increased with the use of a core, but a core is not necessarily required, it just makes for a more compact design by concentrating flux into a preferred path.

The beauty of using a core is that it reduces the amount of windings required for a certain inductance, thereby reducing resistive winding losses.

In my opinion, if you are trying for an OU device, relatively low frequency, large cores and minimal winding ohmic losses would make sense to increase the efficiency of conversion.

If there is a "magic frequency"  and expected power thruput, this should be the starting point for the inductor design.





---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #988 on: 2014-05-01, 15:14:03 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Grumage on 2014-04-28, 18:53:13
Dear Verpies.
As requested.
I'm back.

Is sth wrong with your components?
Where did you connect the scope probes and their groundclips?
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #989 on: 2014-05-01, 17:22:18 »

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
Quote from: verpies on 2014-05-01, 15:14:03
I'm back.

Is sth wrong with your components?
Where did you connect the scope probes and their groundclips?


Dear Verpies.

Yes drastically wrong !!  I had lost both 1 ohm 2 W resistors !!

I posted an updated reply here.   http://www.overunityresearch.com/index.php?topic=2358.msg38168#msg38168

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #990 on: 2014-05-01, 18:51:30 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Grumage on 2014-05-01, 17:22:18
Yes drastically wrong !!  I had lost both 1 ohm 2 W resistors !!
Oh! that explains it.

So if the scopeshot below shows currents flowing in L1 and L2 then it appears that you have the connection to L2 reversed because the absolute values of both currents increase at the same time (between the red and yellow line) like in a normal transformer.

So please reverse the connections of L2, set Ch2 to DC-coupling and invert the display of Ch1 on your scope so you don't have negative going peaks ...and show scopeshots of what's happening at the same duty cycle as well as a higher duty cycle.
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #991 on: 2014-05-01, 19:48:22 »

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
Dear Verpies.

How are these ??

Please let me know if I have the connections to the transformer correct ?

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
img  MenofFather
Re: Akula0083 30 watt self running generator. « Reply #992 on: 2014-05-01, 20:30:46 »
Group: Guest
Quote from: Grumage on 2014-05-01, 19:48:22
Dear Verpies.

How are these ??

Please let me know if I have the connections to the transformer correct ?

Cheers Grum.
You must get something like boxes of pulses going from TL494, I think.
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #993 on: 2014-05-02, 11:18:37 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Grumage on 2014-05-01, 19:48:22
How are these ??
Still indicative of a normal transformer mode.
Did you reverse the connection of only L2 ?

Are your diodes polarized correctly and do they work?

P.S.
You can use AC coupling on gate voltage signal (Ch3) because it is riding on a large DC component.
   
img  wattsup
Re: Akula0083 30 watt self running generator. « Reply #994 on: 2014-05-02, 13:31:47 »
Hero Member
*****

Posts: 520
Quote from: verpies on 2014-05-02, 11:18:37
Still indicative of a normal transformer mode.
Did you reverse the connection of only L2 ?
Are your diodes polarized correctly and do they work?
P.S. You can use AC coupling on gate voltage signal (Ch3) because it is riding on a large DC component.

@verpies and @grumage

I was going to ask the same thing because both waveforms flipped by only reversing the L2. This is the crazy stuff I was afraid of because don't forget that @grumage is using the T1000 wind method. So his L1 is two layers away from the core (weaker impress), and the L2 is creating potential cancellation hence very lazy peaks plus if the core is an ETD-29, it is very small to start with.

To confirm this, I would prefer to see 2 screen shots without using any ground reference, so only the probe at....

Ch1 - L1 - Feed Side
Ch2 - L1 - Pulsed Side
and
Ch1 - L2 - Return Side
Ch2 - L2 - Diode Side

You will notice in some of Akulas videos he is using his probe only. I only use my probe as well. I never use the ground unless I want to see a differential output, but in all other points you are better off working without a reference. Reference becomes universal so it is applicable to all points.

@all

Ordered yesterday bigger cores from Digi-Key and they arrived today. Got extra formers for all of these to simplify different wind tests without having to unwind them afterwards. I finally received .22 ohm resistors (had no luck finding them here) so should be ready to set-up some tests.

Photo below shows all four ecore sizes relative to a standard flyback core.  I will start with the ETD-54. I will save the bigger one for when I know a little more about this system.

I will not use the circuits. Way too much hassle and it will not answer my questions but just create more distraction for me.

I am only interested in what's going on from the mosfet and from the battery feed through the coils and looping and load. If I can remove the battery and keep the looping in run mode with my FG controlling the mosfet base to better find the sweet spots, that is what I want to know for now. Once that is accomplished, then I will have found the true pulse parameters I need to maintain with a very simple pulse circuit. This would also tell me the feedback that is available for looping back to the mosfet gate.

wattsup



---------------------------
http://etherimpress.com
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #995 on: 2014-05-02, 14:24:04 »

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
Dear all.

Your thoughts would be very welcome ?? !!

https://www.youtube.com/watch?v=7qUXZB8tom0

Dear Wattsup.

I will endeavour to try your request later. Unfortunately the weekly shop has to come first !!  :)

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
img  NerzhDishual
Re: Akula0083 30 watt self running generator. « Reply #996 on: 2014-05-02, 14:28:03 »
Group: Guest

Hi Itsu,

I have sent you a PM.

I have noticed that these big round shaped LEDS and the small square shaped ones
did not work with Joule Thief or Super Joule Ringer (SJR) circuits.
On the contray, with a personal SJR I can light small round shaped 220V leds with
very few current (less than 10 mA) and a mere 1.5 volt AA or AAA bat.
Not in full brightness with only about 10 mA but enough to read a book during the night...

Regards,
Jean
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #997 on: 2014-05-02, 15:09:10 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: wattsup on 2014-05-02, 13:31:47
I was going to ask the same thing because both waveforms flipped by only reversing the L2.
L2's current flipped because its connections were reversed.
L1's current flipped because Grum inverted the display of Ch1 on his scope (right Grum?)

Absv of both of these currents are still increasing at the same time which is indicative of a normal transformer mode.  :(
See the scopeshots below for comparison ( Ch1: Gate voltage, Ch2: L1's current, Ch4: L2's current. The display of Ch2 and Ch4 is inverted )

Quote from: wattsup on 2014-05-02, 13:31:47
...his L1 is two layers away from the core (weaker impress), and the L2 is creating potential cancellation hence very lazy peaks plus if the core is an ETD-29, it is very small to start with.
Why would a winding method affect the L1 & L2 currents like this ?

Quote from: wattsup on 2014-05-02, 13:31:47
To confirm this, I would prefer to see 2 screen shots without using any ground reference,
That will not show us currents flowing in L1 and L2.
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #998 on: 2014-05-02, 16:03:05 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4159
Quote from: Grumage on 2014-05-02, 14:24:04
Dear all.

Your thoughts would be very welcome ?? !!

https://www.youtube.com/watch?v=7qUXZB8tom0

Dear Wattsup.

I will endeavour to try your request later. Unfortunately the weekly shop has to come first !!  :)

Cheers Grum.


Hi grum,

that noisy signal from the transformer sounds familiar to me as i also did had it for a while.
It turned out i had a dry solder joint at the 1 Ohm resistor (R3) coming from L2.
It was unsoldered, but still making irregular contact producing this noise.

Regards Itsu
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #999 on: 2014-05-02, 16:28:48 »

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
Quote from: verpies on 2014-05-02, 15:09:10
L2's current flipped because its connections were reversed.
L1's current flipped because Grum inverted the display of Ch1 on his scope (right Grum?)

Absv of both of these currents are still increasing at the same time which is indicative of a normal transformer mode.  :(
See the scopeshots below for comparison ( Ch1: Gate voltage, Ch2: L1's current, Ch4: L2's current. The display of Ch2 and Ch4 is inverted )
Why would a winding method affect the L1 & L2 currents like this ?
That will not show us currents flowing in L1 and L2.


Dear Itsu.

No. My scope tells me with a little arrow if I have deliberately inverted the signal. And it would appear that I can only do it on Ch 2 !!

I have also looked for dry joints, non found so far !!

Cheers Grum.
« Last Edit: 2014-05-02, 16:54:46 by Grumage »


---------------------------
Nanny state ? Left at the gate !! :)
   
Pages: 1 ... 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 2024-11-26, 22:12:58