PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-11-26, 22:15:34
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 ... 13 14 15 16 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 ... 48
Author Topic: Akula0083 30 watt self running generator.  (Read 975026 times)

Group: Professor
Hero Member
*****

Posts: 3499
Anything else noticed and relevant?
The copper shields are coated with some adhesive.  The inner shield is gapped.  The outer one might be also.

Perhaps more details can be gleaned from this video, by people with better eyes. Such as:
1) Turn count
2) Layer count
3) Layer advancement
4) The top winding direction in relation to the bottom winding direction ( a.k.a. dot convention )
5) Soldered connections on those coppers shields.

   

Group: Professor
Hero Member
*****

Posts: 3499
For the hell of it, I'll initially try the ICL7660 inverting driver as a comparison to the two versions of the bipolar transistor drivers. I'm currently right out of non-inverting drivers.
Yes, you can do that as long as you feed this inverting driver with the signal from pin 11.  Don't forget a pullup resistor around 200Ω from pin 11 to +Vcc.

At a risk of being boring, I will reiterate for others what modifications should be made to the original Akula circuit for using an inverting integrated MOSFET driver to drive the MOSFET without inverting the total signal path (because of double inversion):


IN REFERENCE TO Akula's DIAGRAM:
1) Remove R18, R19, C12, VT2, VD1.
2) Pull out one side of R16 and one side of R17 (the sides at their junction) out of the PCB and connect them together in the air (not touching the PCB)
3) Optionally, you can change R17 to a 0Ω resistor (e.g. wire).
4) Bypass the power supply pins of the integrated MOSFET driver chip with a 470nF (or greater) ceramic capacitor (directly soldered across the supply pins at the underside of the driver chip)
5) Superglue the integrated MOSFET driver chip on its back to the power MOSFET.
6) Connect the positive power supply pin of the integrated MOSFET driver chip to the output pin of the three-terminal 12V linear voltage regulator (or to the positive terminal of C6) with at least 0.5mm dia. wire (short & stranded  wire is better)
7) Connect the negative power supply pin (a.k.a. GND) of the integrated MOSFET driver to the ground pin of the three-terminal 12V linear voltage regulator (or to the negative terminal of C6) with at least 0.5mm dia. wire (short & stranded wire is better).
8) Connect the output of the integrated MOSFET driver chip to the gate of the MOSFET through a 4.7Ω carbon resistor.
    a) use the non-inverting output, or
    b) ...or use the inverting output
9) Connect the input of the integrated MOSFET driver chip to the junction between R16 and R17 that is hanging in the air.
    a) if pt.8a was done, then use the inverting input
    b) if pt.8b was done, then use the non-inverting input

For example the TC4428 or the UCC27511 driver can be used in this configuration like this:
« Last Edit: 2014-04-16, 11:32:57 by verpies »
   

Group: Professor
Hero Member
*****

Posts: 3499
At a risk of being boring, I will reiterate for others what modifications should be made to the original Akula circuit for using an inverting integrated MOSFET driver to drive the MOSFET without inverting the total signal path (because of double inversion):

IN REFERENCE TO Groundloop's PCB:
1) Remove R2, R4, C1, Q1, D2.
2) Pull out one side of R1 and one side of R3 (the sides at their junction) out of the PCB and connect them together in the air (not touching the PCB)
3) Optionally, you can change R1 to a 0Ω resistor (e.g. wire).
4) Bypass the power supply pins of the integrated MOSFET driver chip with a 470nF (or greater) ceramic capacitor (directly soldered across the supply pins at the underside of the driver chip)
5) Superglue the integrated MOSFET driver chip on its back to the power MOSFET.
6) Connect the positive power supply pin of the integrated MOSFET driver chip to the output pin of the three-terminal 12V linear voltage regulator (or to the positive terminal of C6) with at least 0.5mm dia. wire (short & stranded  wire is better)
7) Connect the negative power supply pin (a.k.a. GND) of the integrated MOSFET driver to the ground pin of the three-terminal 12V linear voltage regulator (or to the negative terminal of C6) with at least 0.5mm dia. wire (short & stranded wire is better).
8) Connect the output of the integrated MOSFET driver chip to the gate of the MOSFET through a 4.7Ω carbon resistor.
    a) use the non-inverting output, or
    b) ...or use the inverting output
9) Connect the input of the integrated MOSFET driver chip to the junction between R1 and R3 that is hanging in the air.
    a) if pt.8a was done, then use the inverting input
    b) if pt.8b was done, then use the non-inverting input


For example the TC4428 or the UCC27511 driver can be used in this configuration like this:
« Last Edit: 2014-04-16, 11:32:07 by verpies »
   
Group: Guest
The copper shields are coated with some adhesive.  The inner shield is gapped.  The outer one might be also.

Perhaps more details can be gleaned from this video, by people with better eyes. Such as:
1) Turn count
2) Layer count
3) Layer advancement
4) The top winding direction in relation to the bottom winding direction ( a.k.a. dot convention )
5) Soldered connections on those coppers shields.

Cannot discern turn count, but outer winding is one complete layer; inner winding is almost two complete layers with no tape between.  It appears winding direction is opposite between primary and secondary.  Inner winding has two parts; one of which is only a small number of turns and is positioned on one side of the bobbin, not centered over the gap.
   

Group: Professor
Hero Member
*****

Posts: 3499
It appears winding direction is opposite between primary and secondary.
Didn't he drop the spool and picked it up turned around, making it appear that the winding direction is opposite ?
« Last Edit: 2014-04-16, 07:36:43 by verpies »
   
Group: Guest
Yes, between being clumsy and the camera focus changing, it is hard to tell.

I do now see the TWO windings on the inner portion.  He unspools to a point where he pulls the bobbin out of his hand.  When he picks it back up, he breaks the entire winding which looks to be about a dozen turns.
   

Group: Tinkerer
Hero Member
*****

Posts: 3948
tExB=qr
When he disassembles the transformer, is there any indication that the core is magnetized, rather than ferromagnetic?
   

Group: Administrator
Hero Member
*****

Posts: 3961


Buy me some coffee
Grumpy I did an analysis of the core sometime ago, one of his cores had the manufacturers part code on, unfortunately they were no longer in business but i found the data which suggested it was MnZn ferrite material, it appears to be a black almost crumbly looking core material, it has a very low resistance something like 6Ohms per meter.

Guys if he is using a double sticky sided tape on the copper foil then that will also insulate it when a layer is placed on top of itself, i dont think there would be any shorted turns in that copper.
   
Sr. Member
****

Posts: 336
@All,

First test run of my Akyla board. I have connected a 12 Volt 3,5 Watt LED lamp to the output.
The Input is 8,2 Volt @ 0,26 Amp. The voltage over the LED lamp is 13,7 Volt. LED lamp
seems to be full brightness. Nothing is heating up in my circuit and the MOSFET is cold.
I'm running with the "STOP" switch open. I did try to close the "STOP" switch but the
input current just went up to over 1 Ampere so I guess I did get the windings on the L2
coil the wrong way around. Only the OSC adjustment had any effect. I just adjusted
the OSC pot-meter until I got the highest voltage output. I'm running a N27 core right
now. My coil data: L1 = 76 turn 0,5mm, L2 = 186 turn 0,5mm, L1/L2 Ratio = 1:2,447.
Coil is standard with L1 first then L2 on top of that. The LED lamp is slightly warm.

Added: Test 2
IN = 08,00 Volt @ 0,22 A = 1,76 Watt
OUT= 12,63 Volt @ 0,10 A = 1,26 Watt
COP = 0,72

GL.
« Last Edit: 2014-04-16, 12:16:08 by Groundloop »
   
Group: Guest
@ all,

I have now modified the GL PCB to drive the mosfet from an ICL7600 chip. Current consumption is much better at 2.4A @ 12V (as read from my bench PSU) on max frequency of 131.6KHz. Lowest frequency is 18.9KHz. See attached scope shot 16 taken at mosfet gate and shot 17 across R17.

The 'duty' pot can be adjusted to allow a sharp transition from LED lit to LED extinguished by operation of the 'STOP' switch. Duty cycle is 42% at top of frequency range and 43.9% at bottom of range.

Must now get some work done!
   
Group: Guest


Nice neat build GL.  :)

I note that you have a lot more turns on your L1 & L2 coils than Akula.

Hoppy
   
Sr. Member
****

Posts: 336

Nice neat build GL.  :)

I note that you have a lot more turns on your L1 & L2 coils than Akula.

Hoppy

Hoppy,

Yes, that is correct. I want lower current usage and I wanted to see if more turns makes
any different result for the feedback from L2. But I think I got my L2 backwards.

GL.
   
Sr. Member
****

Posts: 375
Hi guys :)

Akula showed more details as per my request - https://www.youtube.com/watch?v=_DkjdsQ9vug

In first part he showed effect what happens when Tesla coil effect is applied perpendicularly to normal windings on 90 degrees. Obviously this big toroid was not good success for first intended OU device. The ground wire is used there as well.
In second part of video the circuit board is from LEDs 1W schematics using toroid from 1st part. This is where it was OK on self running it seems...

Cheers!
   

Group: Professor
Hero Member
*****

Posts: 3499
scopeshot 17 across R17.
With C12 removed on GL's PCB ?
   

Group: Professor
Hero Member
*****

Posts: 3499
Obviously this big toroid was not good success for first intended OU device.
What's under all these windings?
A ferrite core?
   
Group: Guest
With C12 removed on GL's PCB ?

No, C12 is in.
   
Group: Guest
Hi guys :)

This is where it was OK on self running it seems...

Cheers!


Do we take it then, that the two 30W versions we are working on never self-ran?

   
Sr. Member
****

Posts: 375
What's under all these windings?
A ferrite core?

I can assume, - yes, most likely just still question what made of ferrite - hard or soft. All OU effects I observerd over last few years was with hard ferrite http://en.wikipedia.org/wiki/Ferrite_%28magnet%29#Hard_ferrites

Do we take it then, that the two 30W versions we are working on never self-ran?
Interesting question which can be answered only after 1:1 replication.
I can assume it was self running if there was NMR effect involved.
Also I had conversation with Delamorto (where this saga began from with LEDs self runners in Youtube) and asked him to check something on ferrite because he did not know what type of ferrite it was and the compass needle was reacting to that ferrite even without pre-magnetization so this ferrite is presumably hard type which was same as in Lithuanian Yoke experiment material...

Cheers!
   

Group: Professor
Hero Member
*****

Posts: 3499
the compass needle was reacting to that ferrite
A ferrite can concentrate Earth's magnetic field and make the compass needle react to this concentration.

The inductance with known turn count, core size and gap width, as well as resistivity between two drops of salt water (5mm apart - not touching!) and the sound the ferrite makes after being struck while being suspended on a thread, can aid in ferrite identification better than the compass test.
   
Sr. Member
****

Posts: 375
A ferrite can concentrate Earth's magnetic field and make the compass needle react to this concentration.
If you try that with soft ferrite - won't work ;)
Also the hard ferrite has distinctive black color and soft ones from China - greyish...
   

Group: Professor
Hero Member
*****

Posts: 3499
Color is a much better indicator than a compass test.
   

Group: Professor
Hero Member
*****

Posts: 3499
No, C12 is in.
So R17 bypassed by a cap is not a good Current Sensing Resistor.
   
Group: Guest
So R17 bypassed by a cap is not a good Current Sensing Resistor.

Yes, I do know that and just wanted to record the waveform for comparative purposes for other builders.
   
Group: Guest
Color is a much better indicator than a compass test.

Do you maybe have a close up picture of a ferrite type u think will work?
And thank you very much for ur proposed integrated MOSFET driving mod, i think i will try directly with that mod when i assemble my akula 30w the next days. I will use the MIC4422BM, its working pretty good in my other builds at the moment, as long as switching frequency is below 300kHz, more than 300-400kHz destroys it.
kind regards and peace,
From other Planet
   

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
Dear All.

I carried out the mod's to the circuit today. I have used a TC4428 A.

Unfortunately I have no gate pulse at all !!  Do I need to isolate the Mosfet gate from the board ?? As I see it R4 ( GL's board ) is also in circuit !!

HELP please !!

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
Pages: 1 ... 13 14 15 16 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 ... 48
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 22:15:34