PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2022-01-20, 02:49:49
News: Forum TIP:
The SHOUT BOX deletes messages after 3 hours. It is NOT meant to have lengthy conversations in. Use the Chat feature instead.

Pages: 1 ... 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 44 ... 48
Author Topic: Akula0083 30 watt self running generator.  (Read 827598 times)

Group: Administrator
Hero Member
*****

Posts: 3848


Buy me some coffee
   

Group: Moderator
Hero Member
*****

Posts: 2448


Buy me a cigar
Dear All.

Ok. I have now got a pulse on Pin 10, but only at around 8.1 V dc. If I raise the voltage higher the pulse vanishes !!  I started with a process of, install TL 494. Then I attached the TC 4428 A this increased the current to 1.6 A !! NOT GOOD methinks !!

So although things are starting to work I still need some advice.

Cheers Grum.


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

Group: Elite Experimentalist
Hero Member
*****

Posts: 3340
And for those already having a non-inverting integrated MOSFET driver such as the IXDD609 or UCC27511 or TC4428 and wanting to go to ersion 3, the list of modifications is below:

IN REFERENCE TO Akula's DIAGRAM:
1) Remove R16, R17, R18, C12, VT2, VD1.
2) Connect pin 11 of the TL494 to pin 12.
3) Bend out pin 10 out of the socket and connected it to where the junction between R19 and C12 used to be.

   NOTE: at this point you can connect the junction of R19&C12 to the gate of the MOSFET even without the integrated driver.  The downside of this is that the MOSET will not switch as fast as with the integrated driver.

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 non-inverting output of the integrated MOSFET driver chip to the gate of the MOSFET through a 4.7Ω carbon resistor.
9) Connect the non-inverting input of the integrated MOSFET driver chip to where the junction between R19 and C12 used to be

I followed this mod using an ucc27511.

screenshot shows:

gate voltage (yellow)
drain voltage (blue)
pin3 signal (purple
drain current  (green)

lowest frequency setting (16KHz)   input 14V @ 10A


Regards Itsu
   
Sr. Member
****

Posts: 331
Guys,

When playing with this circuit do not forget to attach 10-20W bulb from car in series from battery so if it lights up in full, there i something wrong in circuit. We are talking about 12V very low power battery to power up final circuit here.. ;)
   
Sr. Member
****

Posts: 281
I followed this mod using an ucc27511.

screenshot shows:

gate voltage (yellow)


Regards Itsu

Good Day Itsu

Now, that's the type of Gate drive pulse I like to see!  O0

take care, peace
lost_bro
   

Group: Professor
Hero Member
*****

Posts: 2683
Sorry for the late reply but the O.U.R. forum was down.
If it wasn't down, you'd have had this reply at 9pm GMT..

I carried out the mod's to the circuit today. I have used a TC4428 A.
Did you use the TC4428 in the inverting mode or non-inverting mode ?

Unfortunately I have no gate pulse at all !! 
What do you have at the input of the TC3328 ?

Do I need to isolate the MOSFET gate from the board ??
No.
The removal of C1, D2 and R3 (even one sidedly) isolates the gate of the MOSFET from everything except the Zeners.

As I see it R4 ( GL's board ) is also in circuit !!
R4 does not touch the gate because it is either removed or separated by removed C1 and D2 on Groundloop's PCB.
R3 does not touch the gate because it is pulled out of the PCB on one side or completely.
   

Group: Professor
Hero Member
*****

Posts: 2683
Ok. I have now got a pulse on Pin 10,
The fact that you are mentioning pulses on pin 10, means that pin 11 must be connected to +Vcc and you are using the TC4428 in the non-inverting mode.

This also means that you'd done the mods to Groundloop's PCB according to this list and pin 10 is isolated from the PCB and connected to one side of R4 (the side that used to connect to C1).
The latter is connected to pin 4 of the TC4428.

Is the above correct?

...but only at around 8.1 V DC.
That 8.1VDC is the power supply voltage applied across C8, isn't it?

If I raise the voltage higher the pulse vanishes !!
What happens at the pin 3 of the TL494 as the pulse at pin 10 vanishes ?
   

Group: Professor
Hero Member
*****

Posts: 2683
gate voltage (yellow)
drain voltage (blue)
pin3 signal (purple
drain current  (green)
What was the state of the STOP switch and the polarity of L2 when this scopeshot was taken?

Gate voltage ref GND - too good (less than 50% duty cycle as expected).  Pulses at pin 3 can decrease this duty cycle even down to 0%.

Drain voltage ref GND - Enough to break down a 50V rated diode (D5).  @Lost_bro, help me to figure out what causes that little falling edge in the second half of the OFF period (? voltage across L2 relative to voltage across C11 ?).

Pin3 voltage ref GND - Hmm, are those ±2V spikes an EMI artifact, Dirty ground (e.g. ground loop) or real steering pulses ?   What is going on at pin 2 ?

Drain current - Drain current does not mean much.  L1 or L2 current is significant.

input 14V @ 10A
That's too much. We need to get the feedback loop (pins 2, 15, 3) to work properly so it can decrease the MOSFET's duty cycle and decrease the current draw with it.
The feedback loop is now negative so it is capable of accomplishing that.

P.S.
Please do not push this circuit to 14V 10A because sth will break.  We should be striving for minimum current draw.
« Last Edit: 2014-04-17, 11:36:07 by verpies »
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3340
What was the state of the STOP switch and the polarity of L2 when this scopeshot was taken?

Stop switch is in its normally closed position.  When in open position i will specifically mention this.

Quote
Gate voltage ref GND - too good (less than 50% duty cycle as expected).  Pulses at pin 3 can decrease this duty cycle even down to 0%.

too good      Hmmm is that good?   :D

Quote
Drain voltage ref GND - Enough to break down a 50V rated diode (D5).  @Lost_bro, help me to figure out what causes that little falling edge in the second half of the OFF period (? voltage across L2 relative to voltage across C11 ?).

Not sure which D5 is, but my IDH12SG60C schottky diodes are rated 600V/12A
  
Quote
Pin3 voltage ref GND - Hmm, are those ±2V spikes an EMI artifact, Dirty ground (e.g. ground loop) or real steering pulses ?   What is going on at pin 2 ?

Let me check tonight

Quote
Drain current - Drain current does not mean much.  L1 or L2 current is significant.

OK, Let me check tonight

Quote
That's too much. We need to get the feedback loop (pins 2, 15, 3) to work properly so it can decrease the MOSFET's duty cycle and decrease the current draw with it.
The feedback loop is now negative so it is capable of accomplishing that.

OK, when increasing the frequency, the current input goes down.  At 75KHz its below 1A and i can keep it on continuously, so i can experiment better.


Quote
P.S.
Please do not push this circuit to 14V 10A because sth will break.  We should be striving for minimum current draw.


Right, i will operate in the higher frequency range.

Thanks,   regards Itsu
   
Group: Guest

That's too much. We need to get the feedback loop (pins 2, 15, 3) to work properly so it can decrease the MOSFET's duty cycle and decrease the current draw with it.
The feedback loop is now negative so it is capable of accomplishing that.


I have found a point where the 'Duty' pot does progressively reduce the duty cycle but only when the supply voltage rail is reduced to around 8.5V. Of course at the duty cycle shown in the attached scope shot (taken at the mosfet gate), LED brightness is much reduced but I see no signs of self-running.

Hoppy
« Last Edit: 2014-04-17, 11:11:15 by Hoppy »
   
Group: Guest
Another reply I found these videos on the project 30w akula!!!


http://www.youtube.com/watch?v=plORzm-4i18

http://www.youtube.com/watch?v=Qu2lN1w-zBA
   

Group: Professor
Hero Member
*****

Posts: 2683
In another reply I found these videos on the project 30w akula!!!
http://www.youtube.com/watch?v=Qu2lN1w-zBA
...but that video does not show if the power supply was eventually disconnected
   

Group: Professor
Hero Member
*****

Posts: 2683
Stop switch is in its normally closed position.  When in open position i will specifically mention this.
but what about the L2 polarity. Which way is it connected?

too good      Hmmm is that good?   :D
It's good if it does not hurt the desired operational principle.  I don't know what the author of this circuit intended.
A crisp gate waveform like you have now is close to perfection in digital systems. It creates a lot of EMI, for sure.
Fortunately your UCC MOSFET driver is very universal and you can degrade its driving speed by increasing resistors at its outputs by e.g. 100x.
At this point, I do not know why degrading the performance if the UCC would be necessary (except for decreasing the EMI) but it is possible nonetheless.

Not sure which D5 is, but my IDH12SG60C schottky diodes are rated 600V/12A
It is the diode connected to the drain of the MOSFET.  On Groundloop's PCB it is called D5.  It does not have a designation on Akula's diagram  :/
Anyway, with the 600V rating you do not have to worry about it breaking down in reverse at these voltage levels.
  

OK, when increasing the frequency, the current input goes down.  
That is an indication that the L1 is limiting the current, because during a shorter MOSFET's ON time the current in L1 does not have the time to build up much.
   

Group: Professor
Hero Member
*****

Posts: 2683
I have found a point where the 'Duty' pot does progressively reduce the duty cycle but only when the supply voltage rail is reduced to around 8.5V.
Interesting.  How does the feedback signal look at pin 15?
   
Group: Guest
Interesting.  How does the feedback signal look at pin 15?

See attached scope shot taken at pin 15. The duty pot adjustment is quite critical and only a small adjustment in the pot results in the progressive reduction of duty cycle. At anything above about 8.6V, the duty cycle locks into around 43%.

Hoppy
   
Group: Guest
Maybe could be usefull for further information .....
   

Group: Elite Experimentalist
Hero Member
*****

Posts: 3340
but what about the L2 polarity. Which way is it connected?

well let me try to explain  :-)

i have followed the V2 diagram from grumage, see below
Then i have followed the below picture, meaning 1 coil is CW, the other CCW.
Finally to comply to the dot notation, i have connected the both outer leads of the coils as represented by the dots on the exploded view of the transformer

 
Regards Itsu
   
Hero Member
*****

Posts: 520
@verpies

I many of your posts when you refer to circuit components I cannot even see those on the diagrams.

@all

I've been following the progress while working on FTPU stuff that I cannot stop. hehehe.

Just got two circuit boards, GL2014 and Shark-01. Thanks for the sends. I think I will stand by and see you guys move forward a little more. I  am too novice in the circuit end to do major mods, etc, but.......

That R3 really bugs me. Seems to me it should be a diode pointing to source instead. That would change everything. Maybe another one of Akulas diagram tricks.

Can any one please try these;

1) Open SB1, run circuit and scope across L1 until L1 is in resonance, then close sb1. What frequency and what happens?

Then try...

2) Open SB1, run circuit but scope across L2 until L2 is in resonance, then close sb1. What frequency and what happens?

Why is this consuming what seems like 140 watts when it's supposed to only produce 30 watts? It should be running at 20 watts max if the OU is double less loses for looping.

Choked DC on R3 to L2 continuous then to R2 then feeding the load while filling C3 and pushing against the diode after L1. This means L2 is holding the core hostage at all times. The core is never left to resonate. L1 is simply banging its head on a wall. Then add the fact that some have overlapped the center gap with the secondary and that side is held hostage as well. 

Seems to me the frequency of such a scheme has to be much higher then where you guys are to be able to counter the constant effect of L2 on the core. Otherwise that R3 is a big mistake. How can you pass energy looped back to source via R3?

140 watts when run at brute force may sound OK, but for 30 watts, this should be running at resonance with very little feed to get OU and looping. I know you will see waveforms rise anywhere between 0 and 300k range but those are usually token areas where the best resonances can only occur higher up. Plus no one has wound the coils in tuned mode so this whole thing is just a crap shot hoping for the right mix, but the gold could be buried right from the start under non-tuned coil winds. I had posted how (#517) but removed it because there was no interest. Guys should consider putting their circuit aside and use their FG on the core and see where it really comes to life, then work around that. I have ordered a few cores so eventually I will do it myself since for me this is a prerequisite in order to know where to run the coils.

wattsup


---------------------------
   
Sr. Member
****

Posts: 281
 @Lost_bro, help me to figure out what causes that little falling edge in the second half of the OFF period (? voltage across L2 relative to voltage across C11 ?).



Hello Verpies,

If I told you I knew for sure what the artifact was, I would be telling you a lie....  >:-)

But, I'll throw in my 2 cents worth anyway....

Ok,  After the initial *discharge spike* of the L1, we see a gradual linear discharging of C11 thru  L1 From 20 > 16vdc (possible feeding the L2 and/or LED array?).....  then about 20us into the affair we see a very fast discharge, abruptly pulling the signal towards Zero potential which just as fast reaches an equilibrium @ what appears to be Vcc potential (from the scope shot appears about 12 - 13vdc).

Question:  What is the operational Vcc of this Circuit when the scope shot was taken?

Question:  What is the Voltage Value of C11 in this circuit when the scope shot was taken?

Question:  What is happening simultaneously @ LED array at this moment?

Question:  What are the inductances of L1 & L2?

Question:  What about the copper tape to ground shown in the Akula diagram,  Is it being used here?


I played around with LTspice last night just to see if I could duplicate this effect, and with standard models I could *not* duplicate it.   I did notice some interesting signal results when the system was about to enter resonance. RLC....  R1; C11; L1;....

take care, peace
lost_bro



   
Sr. Member
****

Posts: 281


P.S.
Please do not push this circuit to 14V 10A because sth will break.  We should be striving for minimum current draw.

Good day all:

Yes, has bothered me now for a while:  the 7812 is a 1 amp voltage regulator..... the LM317 =1.5amps  so,  how can this circuit consume 10 or more amps and be expected to self run from a 1 amp regulator?

And...... 1amp X 12vdc = 12watts power..... so, how can this be a 30 watt device using just a single 1amp voltage regulator @ 12Vcc?

Please let me know what I have missed here...

take care, peace
lost_bro
   
Group: Guest
Good day all:

Yes, has bothered me now for a while:  the 7812 is a 1 amp voltage regulator..... the LM317 =1.5amps  so,  how can this circuit consume 10 or more amps and be expected to self run from a 1 amp regulator?

And...... 1amp X 12vdc = 12watts power..... so, how can this be a 30 watt device using just a single 1amp voltage regulator @ 12Vcc?

Please let me know what I have missed here...

take care, peace
lost_bro

Hi lost_bro,

The regulator only supplies the PWM. The regulator is drawn back-to-front from convention.

Hoppy
   
Sr. Member
****

Posts: 281
Hi lost_bro,

The regulator only supplies the PWM. The regulator is drawn back-to-front from convention.

Hoppy

Good day Hoppy.

Ok, so timing between L2 & L1 is extremely important: ie L2 must discharge into L1 when MOSFET is ON so L1* discharge pulse* can recharge L2 during MOSFET OFF cycle.......

Has anyone checked the timing between the L1 & L2?

take care, peace
lost_bro
   

Group: Professor
Hero Member
*****

Posts: 2683
See attached scope shot taken at pin 15. The duty pot adjustment is quite critical and only a small adjustment in the pot results in the progressive reduction of duty cycle. At anything above about 8.6V, the duty cycle locks into around 43%.
That waveform bothers me.  
I wonder if it would still exist if you disconnected the feedback point form the junction of R6 & R7 (R5 & R7 on Akula's diagram) and connected it to ground near where the R7 is grounded (R5 on Akula's diagram).

If it still exists then it means that it is just reacting to EMI.
   

Group: Professor
Hero Member
*****

Posts: 2683
@verpies
I many of your posts when you refer to circuit components I cannot even see those on the diagrams.
Because when I reply to Itsu I use the component designations from Akula's diagram1.

...and when I reply to Grumage I use the component designations from Groundloop's diagram, because Grum he is using Groundloop's PCB.

...with Hoppy I forgot what board he is using so am am stating both ;)

I know it's a mess ...but what can I do besides linking to a diagram each time I write a component designation ! ;(
   

Group: Professor
Hero Member
*****

Posts: 2683
The regulator only supplies the PWM. The regulator is drawn back-to-front from convention.
Yes, and its terminals are not marked on Akula's diagram.
L1 and L2 are supplied from the positive terminal of C11
   
Pages: 1 ... 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 44 ... 48
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2022-01-20, 02:49:49