PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
OverUnity Research > Benches > Grumage (Moderator: Grumage) > Akula0083 30 watt self running generator.
2024-05-17, 16:51:31
News: A feature is available which provides a place all members can chat, either publicly or privately.
There is also a "Shout" feature on each page. Only available to members.

Pages: 1 ... 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 41 ... 48
Author Topic: Akula0083 30 watt self running generator.  (Read 934115 times)
Topic Options img
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #625 on: 2014-04-14, 13:13:05 »

Group: Professor
Hero Member
*****

Posts: 3377
You don't have two current probes to measure current in two windings simultaneously.  You could use the R3 as a CSR if it is non-inductive but then your probe's ground clip would gave to be at the positive terminal of C11, which would create a ground loop if you wanted to scope pin.3 and the gate at the same time in reference to pin 7 (GND) :(

Pot R11 affects the signal at pin.3 the most.
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #626 on: 2014-04-14, 13:33:43 »

Group: Professor
Hero Member
*****

Posts: 3377
Quote from: lost_bro on 2014-04-14, 02:52:52
It looks to me like the MOSFET will spend too much time passing thru ( or should I say; sleeping) in the Miller Plateau....
Yes, the gate spends a long time at the Miller Plateau but is faster switching called for by the designer of this circuit?
Who said that nanosecond pulsing is desirable in this application at all?
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #627 on: 2014-04-14, 18:52:24 »

Group: Moderator
Hero Member
*****

Posts: 2743


Buy me a cigar
Dear All.

Please find attached scope shot of...

Ch 1.  Signal at gate of Mosfet.

Ch 2. Voltage at L1.

Ch 3. Signal on pin 3.

All with ref to Gnd.

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #628 on: 2014-04-14, 19:43:23 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102
Quote from: verpies on 2014-04-14, 13:13:05
You don't have two current probes to measure current in two windings simultaneously.  You could use the R3 as a CSR if it is non-inductive but then your probe's ground clip would gave to be at the positive terminal of C11, which would create a ground loop if you wanted to scope pin.3 and the gate at the same time in reference to pin 7 (GND) :(

Pot R11 affects the signal at pin.3 the most.

Ok, 2 screenshots, first one with:
gate voltage (yellow)
pin 3 voltage (blue)
L1  current  (green)

second one with:
gate voltage (yellow)
pin 3 voltage (blue)
L2  current  (green)

picture is from bench PS no current limit at full load (same as the screenshots)

R11 has no influence on any of the traces.

Next i will swap one of the L1 or L2 leads and remeasure.

EDITTED:  i had 2 the same screenshots presented,  now i have corrected that.
Be aware that the current needs to be taken times 2 due to my current terminator settings.


Regards Itsu
 

« Last Edit: 2014-04-14, 20:56:15 by Itsu »
   
img  Peterae
Re: Akula0083 30 watt self running generator. « Reply #629 on: 2014-04-14, 20:54:34 »

Group: Administrator
Hero Member
*****

Posts: 3867


Buy me some coffee
You all seem to have a duty of 50% or more, is that right ? Hence high current draw.
   
img  4Tesla
Re: Akula0083 30 watt self running generator. « Reply #630 on: 2014-04-14, 20:54:45 »
Group: Guest
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #631 on: 2014-04-14, 20:59:24 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102
Quote from: Peterae on 2014-04-14, 20:54:34
You all seem to have a duty of 50% or more, is that right ? Hence high current draw.


Correct, 56% (or 44% depending how/where you look   :) )

But there is no way to change this, the one pot (R15) changes the frequency, the other pot (R11) the voltage on pin1 which does not change the duty cycle.

Regards Itsu
   
img  4Tesla
Re: Akula0083 30 watt self running generator. « Reply #632 on: 2014-04-14, 21:03:38 »
Group: Guest
Quote from: Itsu on 2014-04-14, 20:59:24
Correct, 56% (or 44% depending how/where you look   :) )

But there is no way to change this, the one pot (R15) changes the frequency, the other pot (R11) the voltage on pin1 which does not change the duty cycle.

Regards Itsu

Does the chip allow the duty cycle to be changed?  Maybe another pot is needed?  Pins 3 or 4?

Datasheet:
http://www.ti.com/lit/ds/symlink/tl494.pdf

This might be helpful:
http://pisarski.info/?p=19
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #633 on: 2014-04-14, 21:15:58 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102
Quote from: 4Tesla on 2014-04-14, 21:03:38
Does the chip allow the duty cycle to be changed?  Maybe another pot is needed?
It does,  but we have discussed this before, and making changes to this would change the circuit we are trying to replicate which should be avoided.

Regards Itsu
   
img  4Tesla
Re: Akula0083 30 watt self running generator. « Reply #634 on: 2014-04-14, 21:19:28 »
Group: Guest
Quote from: Itsu on 2014-04-14, 21:15:58
It does,  but we have discussed this before, and making changes to this would change the circuit we are trying to replicate which should be avoided.

Regards Itsu

But if it doesn't work as is, why not change?
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #635 on: 2014-04-14, 21:20:27 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102
Here i swapped the L1 leads.

The screenshot with L2 now looks almost identical of that of Grumage above.
Noticeable is also that the input current went back to 8A.

Same probe points as before.

R11 again did not make any difference.

Regards itsu
   
img  From Other Planet
Re: Akula0083 30 watt self running generator. « Reply #636 on: 2014-04-14, 21:23:39 »
Group: Guest
Im also thinking the duty cycle described by you guys is way too high, wasting lots of energy here IMHO. Btw did you notice, that akula uses different types of wire for L1 and L2, though in schematic both are specified as same diameter?
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #637 on: 2014-04-14, 21:36:28 »

Group: Moderator
Hero Member
*****

Posts: 2743


Buy me a cigar
Dear All.

With a great deal of Skyping over the weekend and today I have found out a bit more. For both Hoppy and myself with L2 connected the LED's are extinguished and current increases drastically !! Just running L1 I have a very efficient LED driver !! So efficient I took out the string (due to overheating !!) I am now unable to continue as some of the LED's were damaged when the "meltdown" occurred !! China to the rescue again !!  :)  

I will have to take a back seat for a couple of weeks while the new ones are on the way.

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
img  Peterae
Re: Akula0083 30 watt self running generator. « Reply #638 on: 2014-04-14, 21:44:52 »

Group: Administrator
Hero Member
*****

Posts: 3867


Buy me some coffee
Fair enough about the duty, as you say if it's set using component values then it must be as intended, although i understand the duty changes depending on the chips output pins, which are different in the dual transistor version, maybe that version has 0-50%.

Other Planet
He is using fiberglass high temp sheathing that's a sort of woven fabric, in one winding he has stretched it to make it's diameter smaller due to the required length i think, so is the wore thicker on one winding? or is it just the sheathing has been pulled.
   
img  From Other Planet
Re: Akula0083 30 watt self running generator. « Reply #639 on: 2014-04-14, 22:05:44 »
Group: Guest
Quote from: Peterae on 2014-04-14, 21:44:52
Other Planet
He is using fiberglass high temp sheathing that's a sort of woven fabric, in one winding he has stretched it to make it's diameter smaller due to the required length i think, so is the wore thicker on one winding? or is it just the sheathing has been pulled.

What are the differences of this type of insulation other than thickness and max temperature compared to PVC/teflon/... ? Maybe different dielectric behaviour?
Is akula also using different conducting medium in L2?
   
img  Itsu
Re: Akula0083 30 watt self running generator. « Reply #640 on: 2014-04-14, 22:07:22 »

Group: Elite Experimentalist
Hero Member
*****

Posts: 4102

What i also tried was opening the stop switch, then activating the start switch (power on).
This causes the LEDs to almost explode, so needing to switch off again the start switch asap.
The leds will keep on glowing for some minutes after this start switch is opened (power removed) gradually decreasing their lights.
Closing the stop switch will cause this glowing to stop immediately.

Video here:  https://www.youtube.com/watch?v=c6ldTllVXLY&feature=youtu.be

This was with the L1 leads swapped.

Regards Itsu
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #641 on: 2014-04-14, 22:37:00 »
Group: Guest
My original strip-board was built as the push-pull version and this also pulled a heavy current at supply voltages much above 9V. However, it does appear to be a fairly efficient low voltage LED driver if the load is matched to the supply voltage. IMO this is what Akula designed it to be, thus his conversion of a mains operated LED lamp to battery operation. My 2 parallel strings of 5 x 3W LED chips is not a 30W loading, so given that Akula presumably used a full 30W load, then a battery voltage of 12V may turn out to be optimum for this loading in terms of best efficiency. I will when I have the time, increase loading using 4 x 12V / 9W LED lamps in parallel and report on my results. I will also revert back to BC546 / 556 push-pull driver for comparison.

Hoppy
   
img  ion
Re: Akula0083 30 watt self running generator. « Reply #642 on: 2014-04-14, 22:48:42 »
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
Thank you so much Peter and GL, I received the two " Shark-01" boards in todays morning mail.

I will get to building them promptly  right after "tax day" tomorrow.

On another note, there seem to be two boards, one by Matt Watts, one by GL. Are there more?

Would it be a good idea to put all the latest info to build the a given board into a pdf appropriately named e.g Shark-01.pdf?

Notes scattered through the forum could be added and the rev # updated as errors / improvements are found.

I have been very busy with some crisis on the home front I'm afraid I may missed some important stuff.

I'll try to go back over all 20+ pages as time permits to see if I did.

Thanks again guys.


---------------------------
"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 #643 on: 2014-04-15, 00:16:21 »

Group: Professor
Hero Member
*****

Posts: 3377
Quote from: Peterae on 2014-04-14, 20:54:34
You all seem to have a duty of 50% or more, is that right ? Hence high current draw.
With pins 13 & 14 connected together, the TL494 can output duty cycle from 0% to 48%, but because its output is inverted, the MOSFET can be driven only from 100% to 52% duty cycle.  
The only way to change these boundaries is to ground pin 13 or to not invert the TL494's output to the MOSFET's gate.
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #644 on: 2014-04-15, 00:18:04 »

Group: Professor
Hero Member
*****

Posts: 3377
Quote from: Grumage on 2014-04-14, 18:52:24
Please find attached scope shot of...
Ch 1.  Signal at gate of Mosfet.
Ch 2. Voltage at L1.
Ch 3. Signal on pin 3.
All with ref to Gnd.
Negative gate pulses, WTF?
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #645 on: 2014-04-15, 00:32:30 »

Group: Professor
Hero Member
*****

Posts: 3377
Quote from: Itsu on 2014-04-14, 19:43:23
Ok, 2 screenshots, first one with:
gate voltage (yellow)
pin 3 voltage (blue)
L1  current  (green)
Slow gate rise time.  Miller plateau is clearly visible.  VT2 is not doing its job well :(

Pin 3 voltage not showing any activity, unless these 1V spikes are not artifacts caused by EMI.  The waveform at pin 3 should have a 5V amplitude as shown in this video.
Is the junction of R5 & R7 showing any negative pulses or quickly falling edges?

Quote from: Itsu on 2014-04-14, 19:43:23
R11 has no influence on any of the traces.
That would be expected if there is no negative or quick activity on the junction of R5 & R7

Quote from: Itsu on 2014-04-14, 19:43:23
Next i will swap one of the L1 or L2 leads and remeasure.
So, I will not analyze transformer currents at this time.
« Last Edit: 2014-04-15, 01:09:37 by verpies »
   
img  PhysicsProf
Re: Akula0083 30 watt self running generator. « Reply #646 on: 2014-04-15, 00:33:30 »
Group: Professor
Hero Member
*****

Posts: 2995
Today, I received 14 of the "Shark-01" boards - finally arrived!    O0

Peter has some left for those on the east side of the Atlantic pond, I understand.
So I am very happy to send to anyone in the US/west of the pond one of these boards, that is, to anyone interested in doing the build.

Quote from: ION on 2014-04-14, 22:48:42
Thank you so much Peter and GL, I received the two " Shark-01" boards in todays morning mail.


Would it be a good idea to put all the latest info to build the a given board into a pdf appropriately named e.g Shark-01.pdf?

Notes scattered through the forum could be added and the rev # updated as errors / improvements are found.

Thanks again guys.

Totally agree with ION -- it would be helpful to have the latest info on the components etc. for the build.
 O0
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #647 on: 2014-04-15, 08:28:14 »
Group: Guest
@ verpies,

Here are two scope shots at junction of R5 / R7, referenced to ground. Shot 11 is with 'STOP' switch open and LED array lit. Shot 12 is with 'STOP' switch closed and LED array extinguished.

Hoppy
   
img  T-1000
Re: Akula0083 30 watt self running generator. « Reply #648 on: 2014-04-15, 09:09:20 »
Sr. Member
****

Posts: 375
Hi,

My intuition tells the second step should be finding right materials for transformer.
If you manage to find sources selling ferrite cores made fro hard ferrite(this is essential for effect, these black ferrite keep magnetic field residue!) please let me know. In USSR most pulse transformers was made from this type of ferrite.
Also on top of that you should try using iron wire instead of copper as they have opposite magnetic properties...

Cheers!
« Last Edit: 2014-04-15, 09:51:36 by T-1000 »
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #649 on: 2014-04-15, 10:08:51 »

Group: Professor
Hero Member
*****

Posts: 3377
Quote from: Hoppy on 2014-04-15, 08:28:14
Here are two scope shots at junction of R5 / R7, referenced to ground. Shot 11 is with 'STOP' switch open and LED array lit. Shot 12 is with 'STOP' switch closed and LED array extinguished.
Both of these waveforms appear to have sufficiently steep falling edges (>18V/ms) to trigger the comparator in TL494 at pins 1&2.
The 2nd waveform (with 'STOP' switch closed) also has -500mV negative excursions, that would trigger the second comparator in TL494 at pins 15&16.
Do these comparators take notice of these pulses?  Can you see the evidence of this on pin 3 ?

I wish I could see scopeshots like this from Itsu and Grumage.

P.S.
Are you using an inverting MOSFET driving circuit topology like this or non-inverting like this?
« Last Edit: 2014-04-15, 14:30:39 by verpies »
   
Pages: 1 ... 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 41 ... 48
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-05-17, 16:51:31