Akula0083 30 watt self running generator.

Home Help Search Login Register

Welcome,Guest. Please login or register.

OverUnity Research > Benches > Grumage (Moderator: Grumage) > Akula0083 30 watt self running generator.

2024-05-03, 06:26:53
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 ... 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 37 ... 48

« previous next »

Author Topic: Akula0083 30 watt self running generator. (Read 932422 times)

Peterae	Re: Akula0083 30 watt self running generator. « Reply #525 on: 2014-04-07, 08:41:58 »
Group: Administrator Hero Member Posts: 3867	Thanks GL Actually i see he fixed those resistor values on the later circuit with the push pull pre driver transistors. R6 = 220R was 210R R5 = 4K7 was 4.02K R9 = 4K7 was 4.02K R10 = 6K8 was 6.98K C5 = 3n3 was 3nF but as you say if he was using 5% or 10% probably wont matter.

Hoppy	Re: Akula0083 30 watt self running generator. « Reply #526 on: 2014-04-07, 10:41:54 »
Group: Guest	Further investigation on the feedback loop using a 0.68R power resistor has enabled me to increase the supply voltage to 8.9V without my bench supply current clamping. The feedback loop now has far less effect and I now think that this is because the mosfet gate is now being driven with a much better waveform. With R1 at the previous 0.24R, the gate voltage was around its threshold on a 6.8V supply rail, the mosfet was getting "stuck in the ON state" as Verpies suggested and which was clear from the scope shots . I tried varying the value of R1 up to 4R and this does of course reduce running current and did result in the ferrite core 'buzzing' (no gap and ferrite faces touching) at around 3R but I could not hold this condition as the 'sweet spot' was very sharp indeed. This condition resulted at the onset of ringing in the waveform as scoped from mosfet drain to source brought about by careful adjustment of rail voltage. Its just possible that the 0.22R value for R1 was meant to be 2.2R?? This would make a lot more sense insofar as a sensible current consumption is concerned. My final build on the PCB will be with R1 at 2.2R.

Groundloop	Re: Akula0083 30 watt self running generator. « Reply #527 on: 2014-04-07, 10:47:45 »
Sr. Member Posts: 336	Quote from: Peterae on 2014-04-07, 08:41:58 Thanks GL Actually i see he fixed those resistor values on the later circuit with the push pull pre driver transistors. R6 = 220R was 210R R5 = 4K7 was 4.02K R9 = 4K7 was 4.02K R10 = 6K8 was 6.98K C5 = 3n3 was 3nF but as you say if he was using 5% or 10% probably wont matter. Peter, I did use the first circuit drawing I found, I think, attached here. The changes I did was to use a smaller core. This because the core I did choose is well capable of shuffling 30 Watt. And the core from EPCOS, ETD29 can be delivered with different core material. I have ordered N27 and N87 Ferrite cores to my bobbin, so I can try both. The other change I did was to make the R2/C4 combo tunable. But if people do not want that then it is possible to just solder in a fixed value resistor and capacitor. BTW, the "Bread board" type board is "air snail mailed" to members. The boards look like in the attachment under. GL. ------------------------ Akula0083 30 watt self running generator. aky30w.jpg (246.29 kB, 1600x900 - viewed 1939 times.) Akula0083 30 watt self running generator. aky30wpcbB.gif (46.65 kB, 529x797 - viewed 1685 times.) « Last Edit: 2014-04-08, 10:45:12 by Groundloop »

Grumage	Re: Akula0083 30 watt self running generator. « Reply #528 on: 2014-04-07, 14:12:00 »
Group: Moderator Hero Member Posts: 2740	Dear All. A question for our more erudite members !! How does the Al value of a Ferrite change the characteristics ?? For example. Say we had a core of Al 4000 set against one of Al 2000. What would be the practical outcome ?? A little bird has just sent a frequency clue !! Try around 17 KHz. With a core of Al 2000. Cheers Grum. --------------------------- Nanny state ? Left at the gate !!

Peterae	Re: Akula0083 30 watt self running generator. « Reply #529 on: 2014-04-07, 15:32:35 »
Group: Administrator Hero Member Posts: 3867	Nice board, next time you order from those guys, they will charge you per hole LOL

Black_Bird	Re: Akula0083 30 watt self running generator. « Reply #530 on: 2014-04-07, 16:38:08 »
Group: Guest	Quote from: Grumage on 2014-04-07, 14:12:00 Dear All. A question for our more erudite members !! How does the Al value of a Ferrite change the characteristics ?? For example. Say we had a core of Al 4000 set against one of Al 2000. What would be the practical outcome ?? A little bird has just sent a frequency clue !! Try around 17 KHz. With a core of Al 2000. Cheers Grum. Hi Grum, Basically the inductance equals AL times the number of turns squared ( L= ALNN). So if you have one core with AL 4000and the other with AL 2000, for the same number of turns, the first one will have double of the inductance of the second. Cheers, Black Bird

Grumage	Re: Akula0083 30 watt self running generator. « Reply #531 on: 2014-04-07, 17:33:12 »
Group: Moderator Hero Member Posts: 2740	Quote from: Black_Bird on 2014-04-07, 16:38:08 Hi Grum, Basically the inductance equals AL times the number of turns squared ( L= ALNN). So if you have one core with AL 4000and the other with AL 2000, for the same number of turns, the first one will have double of the inductance of the second. Cheers, Black Bird Dear Black Bird. Many thanks for that info. Cheers Grum. --------------------------- Nanny state ? Left at the gate !!

verpies	Re: Akula0083 30 watt self running generator. « Reply #532 on: 2014-04-07, 21:00:57 »
Group: Professor Hero Member Posts: 3363	Quote from: Black_Bird on 2014-04-07, 16:38:08 Basically the inductance equals AL times the number of turns squared ( L= ALNN). Yes, the A_L value of a core tells you how many Henries you will get for every turn ...squared. ( L=A_LN² ) NOTE: The A_L value will decrease if you gap or cut the core ...or saturate it or bring a magnet near it, or cool it. Heating, squeezing it with a vise (or with sound) can increase the A_L value. P.S. The A_L value of some cores changes even in the presence of external electric fields. (see: IEEE Transactions on Magnetics, Vol. 41, No. 10 from October 2005 by Konrad & Brudny). I think Osakosa wrote, that old Russian TV yoke ferrites were of this type. « Last Edit: 2014-04-07, 21:29:01 by verpies* »

Grumage	Re: Akula0083 30 watt self running generator. « Reply #533 on: 2014-04-07, 21:43:25 »
Group: Moderator Hero Member Posts: 2740	Quote from: verpies on 2014-04-07, 21:00:57 Yes, the A_L value of a core tells you how many Henries you will get for every turn ...squared. ( L=A_L*N² ) NOTE: The A_L value will decrease if you gap or cut the core ...or saturate it or bring a magnet near it, or cool it. Heating, squeezing it with a vise (or with sound) can increase the A_L value. P.S. The A_L value of some cores changes even in the presence of external electric fields. (see: IEEE Transactions on Magnetics, Vol. 41, No. 10 from October 2005 by Konrad & Brudny). I think Osakosa wrote, that old Russian TV yoke ferrites were of this type. Dear Verpies. As always, a thorough explanation. Please read this sentence from Hoppy's post earlier today. Quote. I tried varying the value of R1 up to 4R and this does of course reduce running current and did result in the ferrite core 'buzzing' (no gap and ferrite faces touching) at around 3R but I could not hold this condition as the 'sweet spot' was very sharp indeed. This condition resulted at the onset of ringing in the waveform as scoped from mosfet drain to source brought about by careful adjustment of rail voltage. Un quote. I am finding that by rapidly altering (back and forth) my frequency pot on my PWM I seem to be able to lengthen my ringdown on switch off ?? This may of course be my imagination !! But to me something in Hoppy's statement above might just be key ?? At certain frequencies my signal sort of hiccups down !! As if it is sort of rebooting !! I don't really know quite how to describe it !! Your thoughts ?? Cheers Grum. --------------------------- Nanny state ? Left at the gate !!

PhysicsProf	Re: Akula0083 30 watt self running generator. « Reply #534 on: 2014-04-08, 04:13:49 »
Group: Professor Hero Member Posts: 2993	Interesting observations, Grum: Quote I tried varying the value of R1 up to 4R and this does of course reduce running current and did result in the ferrite core 'buzzing' (no gap and ferrite faces touching) at around 3R but I could not hold this condition as the 'sweet spot' was very sharp indeed. This condition resulted at the onset of ringing in the waveform as scoped from mosfet drain to source brought about by careful adjustment of rail voltage. Un quote. I am finding that by rapidly altering (back and forth) my frequency pot on my PWM I seem to be able to lengthen my ringdown on switch off ?? This may of course be my imagination !! Smiley But to me something in Hoppy's statement above might just be key ?? At certain frequencies my signal sort of hiccups down !! As if it is sort of rebooting !! Do you have any output/input power data (to go along with these intriguing observations)?

Peterae	Re: Akula0083 30 watt self running generator. « Reply #535 on: 2014-04-08, 11:03:54 »
Group: Administrator Hero Member Posts: 3867	I will be interested to hear from anyone when they build Groundloops PCB. I have suspicions that Akula quickly moved to a new version with push pull driving of the fet base. If anyone is seeing drive problems regarding the fet, i would start looking at modding GL's PCB to use a Fet driver which we know Akula used in his latest version. If someone need help doing this i will work out a mod, a little bit of vero with the driver chip soldered to the fet gate should do the job. I may yet be wrong about this but the circuit that the PCB is built on has component value corrections in the later versions that then also use Push/Pull indicating to me that he never bothered updating the earlier version circuit with these value changes and instead moved straight onto a new version, and then we have the version in the video using a fet driver.

Hoppy	Re: Akula0083 30 watt self running generator. « Reply #536 on: 2014-04-08, 11:41:20 »
Group: Guest	Quote from: Peterae on 2014-04-08, 11:03:54 I will be interested to hear from anyone when they build Groundloops PCB. I have suspicions that Akula quickly moved to a new version with push pull driving of the fet base. If anyone is seeing drive problems regarding the fet, i would start looking at modding GL's PCB to use a Fet driver which we know Akula used in his latest version. If someone need help doing this i will work out a mod, a little bit of vero with the driver chip soldered to the fet gate should do the job. I may yet be wrong about this but the circuit that the PCB is built on has component value corrections in the later versions that then also use Push/Pull indicating to me that he never bothered updating the earlier version circuit with these value changes and instead moved straight onto a new version, and then we have the version in the video using a fet driver. Most likely the case Peterae! I would go so far as to advise builders not to power up the original circuit as laid on the PCB, unless they have a current limited bench supply.

Peterae	Re: Akula0083 30 watt self running generator. « Reply #537 on: 2014-04-08, 13:06:55 »
Group: Administrator Hero Member Posts: 3867	Hi Hoppy So it is looking like your excessive current is because the Fet is not being driven very well, thanks for that

Peterae	Re: Akula0083 30 watt self running generator. « Reply #538 on: 2014-04-08, 14:06:29 »
Group: Administrator Hero Member Posts: 3867	Dam i have been trying to find the video someone posted, which showed the Akula circuit using a fet driver, anyone got a link again for it.? EDIT I found it, the post was by Itsu http://www.overunityresearch.com/index.php?topic=2358.msg36954#msg36954 https://www.youtube.com/watch?v=EwCS15pRtH0&list=PLC7684829E98CAD74 ------------------------ Akula0083 30 watt self running generator. AkulaFetDriver.jpg (70.97 kB, 1280x584 - viewed 1829 times.)

Hoppy	Re: Akula0083 30 watt self running generator. « Reply #539 on: 2014-04-08, 18:16:48 »
Group: Guest	I modified my circuit to the updated circuit that shows a transistor push-pull mosfet driver (BC546 / 556 comp pair) and it now works fine without excessive heating. The feedback loop when broken will cause a cessation of output at all frequency settings. The 'stop' switch also works to cease output when the switch is closed. I think the feedback loop was likely intended as a safety switch to turn-off the PWM, should the output become open circuit. I have yet to observe anything approaching a self-running condition. However, this setup does appear to be quite efficient and a very good light output can be had from well under a 12V supply rail. It is possible that I cleared a track fault when I implemented the push-push mosfet driver, which may have caused an excessive running current, so try the original circuit before implementing the modification to enable a comparison to be made. « Last Edit: 2014-04-08, 19:16:07 by Hoppy »

Peterae	Re: Akula0083 30 watt self running generator. « Reply #540 on: 2014-04-08, 18:35:41 »
Group: Administrator Hero Member Posts: 3867	Nice work Hoppy Thanks Is this your vero build or GL's PCB?

Hoppy	Re: Akula0083 30 watt self running generator. « Reply #541 on: 2014-04-08, 18:37:56 »
Group: Guest	Quote from: Peterae on 2014-04-08, 18:35:41 Nice work Hoppy Thanks Is this your vero build or GL's PCB? Vero build.

verpies	Re: Akula0083 30 watt self running generator. « Reply #542 on: 2014-04-08, 18:45:21 »
Group: Professor Hero Member Posts: 3363	Quote from: Hoppy on 2014-04-08, 18:16:48 The feedback loop when broken will cause a cessation of output at all frequency settings. Why? The TL494 outputs pulses on pins 10/11 even when the error amplifiers/comparators are not activated (pin 3 low). Activation results in shortening of output pulses.

Peterae	Re: Akula0083 30 watt self running generator. « Reply #543 on: 2014-04-08, 19:17:19 »
Group: Administrator Hero Member Posts: 3867	Quote Vero build. I've just changed my mind about the fet driver being the lastest version of the Akula device. It has a few potential problems, he has not included the needed decoupling caps for the fet driver and these devices really do need that, this fet driver can drive with 9Amps and he is only using a 100uF on the output of the 7812, i think that the drive waveform in this version of the circuit was horrible. The other reason i don't think this was a worker is that again he used the same odd values, which puts the push/pull transistor version as the latest and functional circuit.

Hoppy	Re: Akula0083 30 watt self running generator. « Reply #544 on: 2014-04-08, 20:03:30 »
Group: Guest	I'm still pondering as to why Akula fitted the 'Stop' switch. Although this works OK, the current consumption stays much the same and unless the ferrite E-cores are held / glued together they clonk loudly when the switch is operated. Any thoughts on this?

verpies	Re: Akula0083 30 watt self running generator. « Reply #545 on: 2014-04-08, 20:45:10 »
Group: Professor Hero Member Posts: 3363	Quote from: Grumage on 2014-04-07, 21:43:25 I am finding that by rapidly altering (back and forth) my frequency pot on my PWM I seem to be able to lengthen my ringdown on switch off ?? Where is this "ringdown" measured? Are you referring to the switch-off of the MOSFET/IGBT or of some mechanical switch or power supply?

lost_bro	Re: Akula0083 30 watt self running generator. « Reply #546 on: 2014-04-08, 20:50:45 »
Sr. Member Posts: 281	Quote from: Hoppy on 2014-04-08, 20:03:30 I'm still pondering as to why Akula fitted the 'Stop' switch. Although this works OK, the current consumption stays much the same and unless the ferrite E-cores are held / glued together they clonk loudly when the switch is operated. Any thoughts on this? Good day Hoppy; You mentioned that the stop switch funcitons OK as it kills the output when 'closed'. and above; the E-cores vibrate and 'clonk' together when the switch is 'closed'. The clonking sounds like a big clue, just like the Russion forum member mentioned that after about 4 days his core shattered as it was self running........... Maybe could it be just the reverse of this and that when the stop switch is 'closed' it is recycling energy back to the circuit instead of (or) at the same time diverting energy from the LEDs and this is why they turn off? Maybe the Stop Switch when open is when the Self running aspect of the circuit ceases to be, and when the Stop switch is closed, it is in self running mode? Maybe the coils need to be adjusted so that the output from L1 & L2 is 'optimized' so it can feed back just a part of the current and NOT divert all the energy from the LED array, so it can stay lit. Just a few thoughts. take care, peace. lost_bro

Matt Watts	Re: Akula0083 30 watt self running generator. « Reply #547 on: 2014-04-08, 21:04:55 »
Group: Guest	Quote from: Hoppy on 2014-04-08, 20:03:30 I'm still pondering as to why Akula fitted the 'Stop' switch. Although this works OK, the current consumption stays much the same and unless the ferrite E-cores are held / glued together they clonk loudly when the switch is operated. Any thoughts on this? Seems to me you have something you can work with. Probably best if you don't allow those six surface faces to get damaged. Maybe use a thin layer of tape and make some notes in regards to signals. My personal preference would be to place a tiny dot of silicone on the center face of one side and let it cure. This should remain flexible enough to allow for some movement of the cores without damage. Hopefully the bobbin is tight enough it doesn't allow the cores to wobble side-to-side--you may want to add some thin shims. To be completely honest, having not seen up-close an Akula device running, I'm not sure what exactly we need in that gap, if anything. The list verpies presented earlier enumerates the variables we will have to deal with.

Grumage	Re: Akula0083 30 watt self running generator. « Reply #548 on: 2014-04-08, 21:16:55 »
Group: Moderator Hero Member Posts: 2740	Quote from: verpies on 2014-04-08, 20:45:10 Where is this "ringdown" measured? Are you referring to the switch-off of the MOSFET/IGBT or of some mechanical switch or power supply? Dear Verpies. I have my (mains isolated) scope probe on the junction of C4 R2 with ref to circuit ground. And I am very basically dabbing the Pos from the 12 V battery to charge C11. Dear Matt. I have found that a single piece of paper between just one outer leg of the cores creates the best ring !! Hope this is of help? Cheers Grum. --------------------------- Nanny state ? Left at the gate !!

Hoppy	Re: Akula0083 30 watt self running generator. « Reply #549 on: 2014-04-08, 21:39:02 »
Group: Guest	Quote from: lost_bro on 2014-04-08, 20:50:45 Maybe the Stop Switch when open is when the Self running aspect of the circuit ceases to be, and when the Stop switch is closed, it is in self running mode? Maybe the coils need to be adjusted so that the output from L1 & L2 is 'optimized' so it can feed back just a part of the current and NOT divert all the energy from the LED array, so it can stay lit. Just a few thoughts. take care, peace. lost_bro Hi lost_bro, If there is any self-running to be had, then its most likely to be a case of getting the right core / winding conditions. Something that could confuse experimenters is that LED's have persistence in illumination and given the fairly large capacitor C3, my LED's stay illuminated at a very low slowly diminishing level from the cap's charge for a couple of minutes and can look quite bright in a darkened room.

Pages: 1 ... 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 37 ... 48

« previous next »

Home Help Search Login Register

Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u

2024-05-03, 06:26:53

SMF 2.0.18 | SMF © 2021, Simple Machines
XHTML
RSS
WAP2