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, 22:27:41
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 ... 6 7 8 9 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 ... 48
Author Topic: Akula0083 30 watt self running generator.  (Read 975202 times)
Topic Options img
img  4Tesla
Re: Akula0083 30 watt self running generator. « Reply #500 on: 2014-04-05, 21:22:54 »
Group: Guest
Quote from: Grumage on 2014-04-05, 18:45:05
Dear 4Tesla.

These are the LED's I am using for a load. I have 5 in series times two. I have then put them in parallel to provide a 10W load.

Cheers Grum.

Hi Grum,

I found those on US ebay:
http://www.ebay.com/itm/New-50pcs-High-Power-1W-Warm-White-LED-Lamp-Bead-Chip-Light-90-100LM-3-2-3-4V-/251435918634?pt=LH_DefaultDomain_0&hash=item3a8abfa52a

Nice to have a lot of 50.. just in case some fail.  :)
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #501 on: 2014-04-05, 21:45:49 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Hoppy on 2014-04-05, 17:23:38
Shot taken across R5 with probe on top of resistor. Supply voltage 6.8V with LED's very bright and the mosfet getting very hot!!
Quote from: Grumage on 2014-04-05, 18:24:26
I don't think there is any doubt about that negative going pulse !!  :)
Indeed, that seems to indicate, that the TL494 feedback loop makes sense after all.  I had doubts about that ...and others had too.

Could you scope pin 3 while triggering on pin 5 to check if the TL494 reacts to those pulses at R5/R7.
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #502 on: 2014-04-05, 21:47:31 »

Group: Professor
Hero Member
*****

Posts: 3499
@Grum

Did you try scoping pin 3 and 11 while triggering the scope on pin 5 ?
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #503 on: 2014-04-05, 22:00:25 »
Group: Guest
Quote from: Grumage on 2014-04-05, 18:24:26
Good evening Hoppy.

I don't think there is any doubt about that negative going pulse !!  :)

I have a couple of questions.

1.  Is your transformer wound 15 one end and 45 the other, split between ?

2.  What is your duty cycle ?

I am still testing with a simple Mosfet PWM with less than 10% duty and current measured with DVM from 12VDC 7 AH LAB.  With this set up I rarely reach even 1 Amp !!

Cheers Grum.

Evening Grum,

I think the spike is due to my resistor R5 being slightly inductive at 0.4uH.

My winds are as proposed by Wattsup, as I don't see how Akula's L2 could possibly have 45 turns from looking at the photo inset on his circuit schematic. L1 is 3 layers of 5 turns. L2 is 4 layers of 7 turns, with a full depth split between L1 and L2.

Duty cycle is 35% at 26KHz, falling to 16.2% at 138KHz, as measured from mosfet drain to source.

Regards
Hoppy
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #504 on: 2014-04-05, 22:35:22 »

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
Quote from: verpies on 2014-04-05, 21:47:31
@Grum

Did you try scoping pin 3 and 11 while triggering the scope on pin 5 ?

Good evening Verpies.

I am still waiting on parts to arrive for the "Proper" drive board. In the mean time I have been "Transformer bashing" with a simple Mosfet PWM. In the hope of finding that elusive anomaly !!  :)

I had a little accident with my cores, be warned !! Make sure they are firmly clamped together whilst testing as when loose they tend to "clack together" and break !!  Luckily "THE BOSS" allowed me to buy another set so I now have two transformers ( one super glued) !! and wound conventionally. 15 T one end and 45 turn the other. My original wound the T-1000 way.

Cheers Grum.
 


---------------------------
Nanny state ? Left at the gate !! :)
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #505 on: 2014-04-05, 23:02:44 »
Group: Guest
Further investigation into the 'spike' reveals that it disappears at around 31KHz and increases fairly linearly up to max frequency.
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #506 on: 2014-04-05, 23:05:54 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Grumage on 2014-04-05, 22:35:22
I had a little accident with my cores, be warned !! Make sure they are firmly clamped together whilst testing as when loose they tend to "clack together" and break !!
I think Peterae has arrived at the conclusion, that "the anomaly" in this transformer depends on these cores "clacking" together, ...just not so hard that they break.  
His observations were quite deft and I don't see any logical errors in his reasoning.
   
img  Matt Watts
Re: Akula0083 30 watt self running generator. « Reply #507 on: 2014-04-05, 23:38:47 »
Group: Guest
So Verpies, would this be what Akula means be "resonance in resonance" ?

You have a carrier frequency tuned for the transformer and then a feedback frequency that targets the NMR/NAR effect?  Would that be the correct way to look at this circuit?
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #508 on: 2014-04-06, 00:06:50 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Matt Watts on 2014-04-05, 23:38:47
So Verpies, would this be what Akula means be "resonance in resonance" ?
Maybe.  If that is the case then it would mean "acoustic resonance in LC resonance".

Quote from: Matt Watts on 2014-04-05, 23:38:47
You have a carrier frequency tuned for the transformer and then a feedback frequency that targets the NMR/NAR effect?  Would that be the correct way to look at this circuit?
The NMR frequency will depend on the material being resonated and on LF polarizing magnetic flux density (B) during the T1 time.  A perpendicular orientation of the HF resonating magnetic field is also a prerequisite.
With NAR, the acoustic wavelength in relation to the core's size also become a factor but this additional complication is offset by the lack of the detrimental skin effect.
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #509 on: 2014-04-06, 10:19:16 »
Group: Guest
Further investigation into the feedback loop reveals the attached four scope shots all taken with DC coupled (x10 probe) on mosfet gate with signal referenced to ground. The first shot 07 shows the waveform with feedback connected at 47KHz. The second shot 08 shows a reduced amplitude waveform with feedback connected at 68KHz. The third shot 09 shows the waveform with feedback connected at 90KHz. The fourth shot 10 shows no waveform when feedback removed. The transition from waveform to no waveform is sharp and has the effect of switching the output LED's on and off. Disconnecting the feedback below 90KHz does have the effect of reducing the signal amplitude and LED light level slightly. This reduction increases as the frequency rises towards the 90KHz switching point. Supply voltage 6.8V.

Hoppy
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #510 on: 2014-04-06, 10:44:19 »

Group: Professor
Hero Member
*****

Posts: 3499
The N-Ch MOSFET gets stuck in the ON state?  (100% DUC)
Even without the scopeshot on pin 3, it is evident that the TL494 reacts to the feedback signal and the whole feedback loop is not bogus.
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #511 on: 2014-04-06, 11:39:36 »
Group: Guest
Quote from: verpies on 2014-04-06, 10:44:19
The N-Ch MOSFET gets stuck in the ON state?  (100% DUC)
Even without the scopeshot on pin 3, it is evident that the TL494 reacts to the feedback signal and the whole feedback loop is not bogus.

Yes, it would appear to be intended as a form of current limiting which makes sense given an LED load. However, I still cannot see how this device operates on 12V because I connected a 12V battery with a clamp meter to measure current and fleetingly recorded at least a 12A draw without feedback and a steady7.5A (at lowest frequency setting) with feedback and LED's bright. That was with a 0.68R big power resistor in place of the 0.24R (5 x paralleled 2W 1.2R), which all very quickly smoked with and without feedback! Also, as Grum has reported, I heard a loud 'clack' from the ferrite core on connection to the battery.
   
img  Peterae
Re: Akula0083 30 watt self running generator. « Reply #512 on: 2014-04-06, 11:51:24 »

Group: Administrator
Hero Member
*****

Posts: 3961


Buy me some coffee
If you start gluing the core, only glue to the 2 outside parts of the E Core, leave the middle faces just touching with no glue, the outside glue will hold it together OK hopefully.

The other way is to use elastic bands around the outside of the core to force the 2 halves together, at least this way there's space for inserting spacers etc.
   
img  Hoppy
Re: Akula0083 30 watt self running generator. « Reply #513 on: 2014-04-06, 12:02:57 »
Group: Guest
Quote from: Peterae on 2014-04-06, 11:51:24
If you start gluing the core, only glue to the 2 outside parts of the E Core, leave the middle faces just touching with no glue, the outside glue will hold it together OK hopefully.

The other way is to use elastic bands around the outside of the core to force the 2 halves together, at least this way there's space for inserting spacers etc.

It might also be an idea to vary the DC bias current by adjustment of R1 with and without spacers to see if we can reach a point where the core can 'chatter'. I have a feeling that Akula could be testing us and having a laugh when we create smoke and crack ferrites!!!
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #514 on: 2014-04-06, 14:58:46 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Hoppy on 2014-04-06, 12:02:57
I have a feeling that Akula could be testing us and having a laugh when we create smoke and crack ferrites!!!
There is always a risk of that if he monitors this forum.  Is he able to?
IMO clacking/chattering core should be free to do so.  Dampening materials will attenuate that ...especially in the gap.
   
img  wattsup
Re: Akula0083 30 watt self running generator. « Reply #515 on: 2014-04-06, 15:12:01 »
Hero Member
*****

Posts: 520
@all

I decided to remove this post due to lack of interest. Just do what you want to do. I will make the circuit when it arrives and do this on my own. It will be much simpler then trying to explains things here.

wattsup

« Last Edit: 2014-04-08, 14:15:05 by wattsup »


---------------------------
http://etherimpress.com
   
img  Peterae
Re: Akula0083 30 watt self running generator. « Reply #516 on: 2014-04-06, 15:33:02 »

Group: Administrator
Hero Member
*****

Posts: 3961


Buy me some coffee
Grumage
where are you buying your core & Bobbins from in the UK?

Thanks
Peter
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #517 on: 2014-04-06, 17:05:09 »

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
Quote from: Peterae on 2014-04-06, 15:33:02
Grumage
where are you buying your core & Bobbins from in the UK?

Thanks
Peter

Dear Peter and all.

Both Itsu and I got our bobbins and cores from here.

http://powermagnetics.co.uk/calculator.     From this handy inductance calculator you can go to the shop.  If you order in the AM it will usually be with you the following day (UK)  for Itsu 2 days !!

We bought these. http://powermagnetics.co.uk/pace-components/ferrite-cores/etd-cores/etd49-cf139   And the bobbin. http://powermagnetics.co.uk/pace-components/coil-formers-bobbins/bobbins-for-etd-cores/etd49-90794-20p

Hope this is of help ??

Cheers Grum.


---------------------------
Nanny state ? Left at the gate !! :)
   
img  Groundloop
Re: Akula0083 30 watt self running generator. « Reply #518 on: 2014-04-06, 17:41:47 »
Sr. Member
****

Posts: 336
Quote from: Grumage on 2014-04-06, 17:05:09
Dear Peter and all.

Both Itsu and I got our bobbins and cores from here.

http://powermagnetics.co.uk/calculator.     From this handy inductance calculator you can go to the shop.  If you order in the AM it will usually be with you the following day (UK)  for Itsu 2 days !!

We bought these. http://powermagnetics.co.uk/pace-components/ferrite-cores/etd-cores/etd49-cf139   And the bobbin. http://powermagnetics.co.uk/pace-components/coil-formers-bobbins/bobbins-for-etd-cores/etd49-90794-20p

Hope this is of help ??

Cheers Grum.

Grum,

Only EPCOS ETD29/16/10 bobbin and cores will fit the boards that I have designed.

GL.
   
img  Grumage
Re: Akula0083 30 watt self running generator. « Reply #519 on: 2014-04-06, 19:42:51 »

Group: Moderator
Hero Member
*****

Posts: 2765


Buy me a cigar
To all in chat !!


---------------------------
Nanny state ? Left at the gate !! :)
   
img  Peterae
Re: Akula0083 30 watt self running generator. « Reply #520 on: 2014-04-06, 21:23:45 »

Group: Administrator
Hero Member
*****

Posts: 3961


Buy me some coffee
OK so why is aKula not using standard preferred resistor values, the only time I've seen that is when either the resistor value is really critical for correct operation, or you get a theoretical non practical person designing the thing.

I think we know he is indeed a practical guy so then why R8 & R9 are 4.02K, i did notice that this value is orderable, also
R10 6.98K again totally weird value to use.

and then we have a 4K7 pot, that's an old world pot value, normally these days they make 5K
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #521 on: 2014-04-06, 23:15:13 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Peterae on 2014-04-06, 21:23:45
OK so why is aKula not using standard preferred resistor values, the only time I've seen that is when
either the resistor value is really critical for correct operation, or
you get a theoretical non practical person designing the thing.
Yes, nonstandard resistor values usually mean critical or theoretical reasons.
...but are those values really so non-standard for 1% resistors ?
« Last Edit: 2014-04-06, 23:42:40 by verpies »
   
img  Groundloop
Re: Akula0083 30 watt self running generator. « Reply #522 on: 2014-04-06, 23:22:40 »
Sr. Member
****

Posts: 336
Quote from: Peterae on 2014-04-06, 21:23:45
OK so why is aKula not using standard preferred resistor values, the only time I've seen that is when either the resistor value is really critical for correct operation, or you get a theoretical non practical person designing the thing.

I think we know he is indeed a practical guy so then why R8 & R9 are 4.02K, i did notice that this value is orderable, also
R10 6.98K again totally weird value to use.

and then we have a 4K7 pot, that's an old world pot value, normally these days they make 5K

Peter,

Akyla did use 4K, 5K, 7K and 7R in the drawing. Looking at his board it seems that he used 5% or 10% resistors.
In my part design I did use 1% 4,02K, 5,1K and 6,98K and 6,8R. I think the values I have chosen is well inside the
5% and 10% tolerances of the resistors Akyla did use.

So if we use %5 resistors for the Akyla values we get:

4K can be from 3,80K to 4,20K
5K can be from 4,75K to 5,25K
7K can be from 6,65K to 7,35K
7R can be from 6,65R to 7,35R

For 10% resistors we get even more spread.

GL.
   
img  verpies
Re: Akula0083 30 watt self running generator. « Reply #523 on: 2014-04-06, 23:38:21 »

Group: Professor
Hero Member
*****

Posts: 3499
Quote from: Grumage on 2014-04-06, 19:42:51
To all in chat !!

What are the scope probes positions for this scopeshot?
Is Ch1 the voltage on pin 3 ?

   
img  T-1000
Re: Akula0083 30 watt self running generator. « Reply #524 on: 2014-04-07, 00:29:50 »
Sr. Member
****

Posts: 375
Quote from: Grumage on 2014-04-06, 17:05:09
Dear Peter and all.

Both Itsu and I got our bobbins and cores from here.

http://powermagnetics.co.uk/calculator.     From this handy inductance calculator you can go to the shop.  If you order in the AM it will usually be with you the following day (UK)  for Itsu 2 days !!

We bought these. http://powermagnetics.co.uk/pace-components/ferrite-cores/etd-cores/etd49-cf139   And the bobbin. http://powermagnetics.co.uk/pace-components/coil-formers-bobbins/bobbins-for-etd-cores/etd49-90794-20p

Hope this is of help ??

Cheers Grum.

I just got http://powermagnetics.co.uk/pace-components/ferrite-cores/etd-cores/etd49-cf129 x2 + http://powermagnetics.co.uk/pace-components/coil-formers-bobbins/bobbins-for-etd-cores/etd49-90794-20p there. Hopefully can find time to assemble 555 circuit (will see if I can make feedback as per original circuit for duty cycle) on next weekend when order will arrive. Very curious if I can find anything unusual within transformer. As in way I suggested windings there it should be interesting to see things going inside of transformer with main frequency and feedback blocking pulses... :)

Quote from: verpies on 2014-04-06, 23:38:21
What are the scope probes positions for this scopeshot?
Is Ch1 the voltage on pin 3 ?
Channel 1 - VT1-C3 (isolated ground)
Channel 2 - across C4 (isolated ground on C3)


Cheers!
   
Pages: 1 ... 6 7 8 9 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 ... 48
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 22:27:41