PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
OverUnity Research > Benches > Grumage (Moderator: Grumage) > Akula0083 30 watt self running generator.
2024-05-06, 14:01:20
News: Registration with the OUR forum is by admin approval.

Pages: 1 ... 10 11 12 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 ... 48
Author Topic: Akula0083 30 watt self running generator.  (Read 932745 times)
Topic Options img
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #600 on: 2014-04-13, 11:38:29 »
Group: Guest
Quote from: Groundloop on 2014-04-13, 10:38:45
Hoppy,

Thank you for providing the o-scope shots.

>>the LED cluster shows no sign of illumination

This indicates a problem in you circuit. Also the capacitor going warm indicate a high current draw to ground somewhere.
Have you used a voltmeter and checked what voltage you get in C10? Is the LEDs connected the right way? Do the LEDs
have a 1 Ohm resistor current path to ground? Any shorts in your soldering? Etc.

GL.

GL,

I have checked my LED array as functional on a DC supply and have checked all output tracks and connections as good on the PCB and not grounding. The overall current draw varies considerably but can run high at certain pot settings. Its possible that as Verpies commented, the problem is that reverse blocking voltage of the diode is being exceeded. The LED array has the 1R resistor path to ground and there is no shorts that I can find.

Hoppy
   
img  Groundloop
Re: Akula0083 30 watt self running generator. « Reply #601 on: 2014-04-13, 11:50:01 »
Sr. Member
****

Posts: 336
Quote from: Hoppy on 2014-04-13, 11:38:29
GL,

I have checked my LED array as functional on a DC supply and have checked all output tracks and connections as good on the PCB and not grounding. The overall current draw varies considerably but can run high at certain pot settings. Its possible that as Verpies commented, the problem is that reverse blocking voltage of the diode is being exceeded. The LED array has the 1R resistor path to ground and there is no shorts that I can find.

Hoppy

Hoppy,

Did you check the C10 voltage? I also agree that your Schottky diodes may be "busted" by high voltage coil spikes.

GL.
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #602 on: 2014-04-13, 11:57:41 »
Group: Guest
Quote from: Groundloop on 2014-04-13, 11:50:01
Hoppy,

Did you check the C10 voltage? I also agree that your Schottky diodes may be "busted" by high voltage coil spikes.

GL.

Yes, I have seen voltages well in excess of its rating. What are your experiences with your replication?

Hoppy
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #603 on: 2014-04-13, 12:34:33 »

Group: Professor
Hero Member
*****

Posts: 3366
Quote from: Hoppy on 2014-04-13, 10:04:55
Shot 06 shows the waveform well before it breaks into audible ferrite core oscillation by adjustment of the 'Duty' pot. Shot 07 shows the waveform just before it breaks into audible oscillation.
That means that the acoustic oscillation starts as the conduction time of the MOSFET decreases.
The gate driver seems to have problems with rising slew rate (BJT's job). The TL494 is responsible only for discharging the gate (falling slew rate).

Quote from: Hoppy on 2014-04-13, 10:04:55
Likewise, shot 08 shows the waveform as scoped at cathode of D6 well before it breaks into audible oscillation. Shot 09 shows the waveform just before it breaks into audible oscillation.
Too bad these waveforms are not measured in reference to the positive terminal of C11, because D6 is oriented in such a ways as to conduct only when the voltage generated by L2 exceeds the voltage across C10.
This can be seen on the scopeshot as the negative level of the waveform, which is established by the conduction of D6 above the voltage level of C10.

Quote from: Hoppy on 2014-04-13, 10:04:55
I have not found it possible to get a stable shot of the waveform whilst in oscillation (tried all trigger modes).
The core must be ratting randomly.  That is not necessarily bad.

Quote from: Hoppy on 2014-04-13, 10:04:55
At all points of pot adjustments, the LED cluster shows no sign of illumination. However, during periods of ferrite core oscillation (sounds loud, like a squealing pig!) cap C10 gets hot.
Maybe the diodes are conducting in reverse or C10 is overvoltaged by strong pulses from the flyback action in L1 ...or magnetoacoustics from L2.
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #604 on: 2014-04-13, 14:25:50 »

Group: Moderator
Hero Member
*****

Posts: 2740


Buy me a cigar
Dear All.

Please find attached scope shot of volt drop across R19 (1 ohm )  Ch1 and R17  Ch2  ( 0.2 ohm )  Gnd ref on  +  C11.

Cheers Grum.

PS. I may have got these positions wrong as I have forgotten what Verpies  told me yesterday !! Should have written it down !!  ;D


---------------------------
Nanny state ? Left at the gate !! :)
   
img  T-1000
Re: Akula0083 30 watt self running generator. « Reply #605 on: 2014-04-13, 15:52:40 »
Sr. Member
****

Posts: 375
http://www.overunity.com/14378/akula0083-30-watt-self-running-generator/msg397514/#msg397514

Ok, to save lots of time I am offering voice communications over Teamspeak 3.

Teamspeak 3 - client download:
http://www.teamspeak.com/?page=downloads
My private Teamspeak 3 server address is: sirius.proxydns.com
Please use your nickname and no password when connecting.
howto - http://youtu.be/S32AGZdY5yA?t=1m56s

Looking forward to see everyone speaking there :)

P.S> All voice communications are encrypted in this server.
   
img  Groundloop
Re: Akula0083 30 watt self running generator. « Reply #606 on: 2014-04-13, 15:53:15 »
Sr. Member
****

Posts: 336
Quote from: Hoppy on 2014-04-13, 11:57:41
Yes, I have seen voltages well in excess of its rating. What are your experiences with your replication?

Hoppy

Hoppy,

My replica is waiting for parts. I hope to get the parts on Monday or Tuesday.

GL.
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #607 on: 2014-04-13, 16:01:36 »

Group: Professor
Hero Member
*****

Posts: 3366
Quote from: Grumage on 2014-04-13, 14:25:50
Please find attached scope shot of volt drop across R19 (1 ohm )  Ch1 and R17  Ch2  ( 0.2 ohm )  Gnd ref on  +  C11.
C12 removed ?
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #608 on: 2014-04-13, 16:54:01 »
Group: Guest
@ Verpies,

I've scoped around 80V DC across C11, with no discernible waveform. Its just enough to puff-up the top of my 63V rated cap C3.

I need to clear my bench next week to start some work - work but will look-in whenever I can.

Hoppy
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #609 on: 2014-04-13, 17:35:03 »

Group: Moderator
Hero Member
*****

Posts: 2740


Buy me a cigar
Quote from: verpies on 2014-04-13, 16:01:36
C12 removed ?

Dear Verpies.

No C12 was not removed.  These double sided boards are really tough to work with so I left it there !! I am beginning to think that I should have done a breadboard first !!

Did I pick the correct resistors to scope on ??  And was the ref position correct ??

Thinking out loud I feel that the pulse width is FAR to wide, I was seeing much more interesting results with the very narrow pulses provided by my old PWM.

Cheers Grum.



---------------------------
Nanny state ? Left at the gate !! :)
   
img  lost_bro
Re: Akula0083 30 watt self running generator. « Reply #610 on: 2014-04-13, 18:00:09 »
Sr. Member
****

Posts: 281
Quote from: Grumage on 2014-04-13, 17:35:03


Thinking out loud I feel that the pulse width is FAR to wide, I was seeing much more interesting results with the very narrow pulses provided by my old PWM.

Cheers Grum.



Hi Grum

Funny that you mention this about the pulse width.....    I remember seeing another schematic when the thread started that indicated an Emitter Follower configuration for the MOSFET driver part of the circuit.

Well the current schematic is Common Emitter , sinking current (with pull-down resistor) and this inverts the TL494 signal (with respect to MOSFET gate), and the Emitter Follower (sourcing current, without pull-up resistor)  configuration does NOT invert the signal coming from the TL494 as does the Common Emitter config.

So, as of now when the pulse is Narrow @ Pin 11 >> WIDER pulse @ MOSFET GATE, and as stated by Verpies:
" because of this inversion, the functions of error amplifiers (pins 2 &15) are also inverted and their activation (pin 3 high) causes the lengthening of MOSFET's ON-state.  
Longer MOSFET conduction time translates to more average current in the transformer/inductor."

OF course more average current in the transformer translates to more average  power consumed and apparently more heat generated.

So,  the Emitter-Follower dirver config. that I saw on the other Akula 30watt schematic would pulse the MOSFET gate 'in time' (non-inverted with respect to gate) with the pulse width of the TL494  logic.

This would give a shorter pulse 'in sycn' with the pulse width generated by the TL494 in response to it feedback loop.....

I guess the point of this commentary is that it strikes me odd to believe that both so very different circuit driver topologies can be correct, when apparently their action is so very different!

Let me know your ideas on this.

take care, peace.
lost_bro
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #611 on: 2014-04-13, 18:18:10 »
Group: Guest
Quote from: lost_bro on 2014-04-13, 18:00:09

I guess the point of this commentary is that it strikes me odd to believe that both so very different circuit driver topologies can be correct, when apparently their action is so very different!

take care, peace.
lost_bro


Good point, which begs the question, exactly what is the two versions of this circuit correct for??  :-\

Hoppy
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #612 on: 2014-04-13, 19:38:22 »

Group: Professor
Hero Member
*****

Posts: 3366
Quote from: lost_bro on 2014-04-13, 18:00:09
...it strikes me odd to believe that both so very different circuit driver topologies can be correct, when apparently their action is so very different!
Good point.
What driver topology is shown by the pertinent Akula video?
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #613 on: 2014-04-13, 19:43:26 »

Group: Professor
Hero Member
*****

Posts: 3366
Quote from: Grumage on 2014-04-13, 17:35:03
No C12 was not removed.  
Unfortunately this needs to be done because a CSR cannot be bypassed by a capacitor.  Cutting one leg of the cap is enough too...

Quote from: Grumage on 2014-04-13, 17:35:03
Did I pick the correct resistors to scope on ??  And was the ref position correct ??
Yes x2 if you were referring to Groundloop's schematic diagram component designations.

Quote from: Grumage on 2014-04-13, 17:35:03
Thinking out loud I feel that the pulse width is FAR to wide, I was seeing much more interesting results with the very narrow pulses provided by my old PWM.
Lost_bro might have a good take on the inverted MOSFET conduction.
   
img  lost_bro
Re: Akula0083 30 watt self running generator. « Reply #614 on: 2014-04-13, 20:07:05 »
Sr. Member
****

Posts: 281
Quote from: verpies on 2014-04-13, 19:38:22
Good point.
What driver topology is shown by the pertinent Akula video?

Hello Verpies

OK, I remember that I posted over at Over unity.com many weeks ago something pertaining to a discrepancy with the video in as much, if I remember correctly that the Schematic shown on said video did NOT match the schematic shown in the link of the same. (different driving circuits)

Don't remember now which was which, but when I have time will check my posts at O.U.com.

take care, peace.
lost_bro
   
img  Peterae
Re: Akula0083 30 watt self running generator. « Reply #615 on: 2014-04-13, 20:18:23 »

Group: Administrator
Hero Member
*****

Posts: 3867


Buy me some coffee
If you want to play around with the pins on the chip to swap the internal driver over, i would bend the 4 legs so they don't plug into the chip socket and then use 4 pins from another chip socket plugged into the 4 sticking out pins and run wires to where you need to connect them, it saves cutting tracks  O0
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #616 on: 2014-04-13, 22:07:10 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102

Hi all,

i am back home, so i picked up where i left off and as my leds arrived, i hooked them up and made the same measurements as before.

I have my R1 (0,22 Ohm) starting to smoke after a few seconds which is similar as Hoppy has if i remember well.

I still have to read up on what is published up till now.

Anyway,  here a short video of what i did, and a screenshot of the severall signals measured when momentarely powering on the circuit.

Yellow trace is the gate voltage,
blue the drain voltage,
green the drain current and
purple the pin 11 signal

The cores "clack" together when the voltage is applied.
When i remove the current limiter from my PS, the current running into the circuit is 10A at 14V, the video and screenshot where taken when the current was limited to 7A causing the voltage to drop to 12V.

https://www.youtube.com/watch?v=ag_ry97wktA&feature=youtu.be

Regards Itsu
   
img  lost_bro
Re: Akula0083 30 watt self running generator. « Reply #617 on: 2014-04-13, 22:23:32 »
Sr. Member
****

Posts: 281
Quote from: Itsu on 2014-04-13, 22:07:10

Anyway,  here a short video of what i did, and a screenshot of the severall signals measured when momentarely powering on the circuit.

Yellow trace is the gate voltage,


Good day Itsu

I will comment on the Yellow trace:    8.4 volts is NOT enough voltage to drive the MOSFET into conduction fast enough, usually one would want a signal in amplitude of about 10-11 volts just to make sure the MOSFET is completely ON.

take care, peace.
lost_bro
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #618 on: 2014-04-14, 00:05:48 »

Group: Professor
Hero Member
*****

Posts: 3366
Quote from: Itsu on 2014-04-13, 22:07:10
Yellow trace is the gate voltage,
blue the drain voltage,
green the drain current and
purple the pin 11 signal
A scopeshot of current flowing in the windings and voltage signal on pin.3 (with different pot settings) when the scope is triggered on MOSFET's gate voltage signal, would reveal much more.
« Last Edit: 2014-04-14, 00:41:52 by verpies »
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #619 on: 2014-04-14, 00:15:34 »

Group: Professor
Hero Member
*****

Posts: 3366
Quote from: Hoppy on 2014-04-13, 22:19:48
Waveform 11 is scoped at the cathode of D6 with the LED brightly lit.
That's 80V above ground!  ...+ spikes.
What was the voltage across C10 and C11 during this measurement?
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #620 on: 2014-04-14, 00:35:41 »

Group: Professor
Hero Member
*****

Posts: 3366
Quote from: lost_bro on 2014-04-13, 22:23:32
8.4 volts is NOT enough voltage to drive the MOSFET into conduction fast enough, usually one would want a signal in amplitude of about 10-11 volts just to make sure the MOSFET is completely ON.
8V significantly exceeds VGS_ON thus this voltage is sufficient to fully turn ON most MOSFETs (see attachment).

You are correct, that 8V might not be sufficient to turn on some MOSFETs fast enough but they will turn ON eventually with that VGS.  
Whether that happens in 100µs or in 100ns depends on the total gate charge of a particular MOSFET.
   
img  lost_bro
Re: Akula0083 30 watt self running generator. « Reply #621 on: 2014-04-14, 02:52:52 »
Sr. Member
****

Posts: 281
Quote from: verpies on 2014-04-14, 00:35:41
8V significantly exceeds VGS_ON thus this voltage is sufficient to fully turn ON most MOSFETs (see attachment).

You are correct, that 8V might not be sufficient to turn on some MOSFETs fast enough but they will turn ON eventually with that VGS.  
Whether that happens in 100µs or in 100ns depends on the total gate charge of a particular MOSFET.

Hi Verpies,
Yes, it is above Vgs, but looking at the waveform in yellow and speaking of QGD, It looks to me like the MOSFET will spend too much time passing thru ( or should I say; sleeping) in the Miller Plateau....
When the MOSFET switches, the gate voltage becomes clamped to the plateau voltage (Miller Plateau) and stays there until sufficient charge has been added/ removed for the device to switch.

The wave form, Leading edge *should* be much more abrupt.  This is Definitely NOT nano-pulsing.

Take care, peace.
lost_bro
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #622 on: 2014-04-14, 08:49:57 »
Group: Guest
Quote from: verpies on 2014-04-14, 00:15:34
That's 80V above ground!  ...+ spikes.
What was the voltage across C10 and C11 during this measurement?


I've deleted that post (619) as my probe was incorrectly set to x1  C.C The correct scoped voltages are as follows: -

Across C10 with 'STOP' button normally closed:  11.4V
Across C10 with 'STOP' button open:                 15.4V
Across C11 with 'STOP' button normally closed:  7.4V
Across C11 with 'STOP' button open:                 7.6V

Hoppy
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #623 on: 2014-04-14, 09:51:51 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102
Quote from: lost_bro on 2014-04-13, 22:23:32
Good day Itsu

I will comment on the Yellow trace:    8.4 volts is NOT enough voltage to drive the MOSFET into conduction fast enough, usually one would want a signal in amplitude of about 10-11 volts just to make sure the MOSFET is completely ON.

take care, peace.
lost_bro

Thanks Lost,  i guess my current limiter on my bench PS (set at 7A) probably limited the voltage to this 8.4V as it should have put 12V on the gate via the driving transistor.
I will try again with the current unlimited to see if this changes the gate signal.

Regards Itsu
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #624 on: 2014-04-14, 09:54:20 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102
Quote from: verpies on 2014-04-14, 00:05:48
A scopeshot of current flowing in the windings and voltage signal on pin.3 (with different pot settings) when the scope is triggered on MOSFET's gate voltage signal, would reveal much more.

verpies, 

i can do that, any specific winding (L1 / L2) or pot or both?


Regards Itsu
   
Pages: 1 ... 10 11 12 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 ... 48
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-05-06, 14:01:20