|
@Itsu, @Grumage, @Happy
This is a loop system where there cannot be any room for snags, delays, flux bounce, What ever L1 produces has to impart and return right away to C11 or its dead. Any delay will kill the effect. There is no other way because this is supposed to work in a loop meaning there is no time to play fancy flux games in the core. The circuitry just won't wait around for any delays.
First thing is the primary you guys are using of 15 turns. Way to many turns. Most flybacks I have used I would not put more then 5 turns primary to keep the full pulse as close to the core as possible instead of diluting the pulse through three layers of primary that also are on two layers of L2. You don't need any more then 5 turns plus if you need a little more inductance, 10 turns would be the maximum you would ever want on the L1 like Akulas' core photo that I mapped previously.
The other variable is the core type. There should be no center gap in this at all. If there is a gap you will transfer via a flux bounce but that will take too long for looping. L2 cannot sit around waiting for that bounce to occur because L1 is already getting its next pulse by then so there is waste and cancellation. There cannot be any more activity then pulse/receive in this system and everything is designed exactly for that.
The problem with gaped cores is you loose a good 80% of the Al value. The Al value is the only thing required to just keep looping and looping inside the core for this to work. Remove 80% of that and you just have a tight trampoline that likes bouncing people around but without any time to do that its just wasting energy.
Then, if we look at this scientifically, the T1000 mode has no basis in anything logical. No one can explain why such a wind is required or even explain what the T1000 wind is creating inside the cores besides what I tried to explain in my previous post. Everything right now in the Ecore is wrong. So why are we all working this thing with a T1000 wind. Of course, I got it, this wind is so exotic, so out-of-the-box, hmmmmmm, more like out of this universe in the "Wishful Thinking" galaxy.
The T1000 wind method is seen in air coils like the TK coil types, but not when you are using the added influence of an e core.
To prove this, just run the circuit, scope the L2 then stop the circuit, remove one turn of L1, then try it again to see if L2 is reduced, if not, remove another turn of L1, etc, so you need to work backward and remove as much of the L1 as possible. This will slowly increase the pulse applied to the first 5 turns of L1 that is nearest the core or in your case over two layers of L2 (which I think is premeditated stagnation). If you see no change in L2 by the time you are at the last 5 turns, then remove the 5th turn and so on. You will quickly realize that the problem all started when L2 went under L1. What a waste.
If I were you guys, I would strip the damn things and rewind them properly or find a new ecore to work with.
Start 5 turns for L1 starting top left going down and stop at the gap, leave enough length of L1 wire to make 5 more turns plus a length more to connect to the circuit. Wind 2 layers of L2 starting at bottom right going down and leave enough wire to make two more layers. The best will be if the L1 wire is thicker then the L2 wire (just like Akulas' coil). You basically want to be near the same weight of copper on each side. Once you have 5 turns and 10 turns, connect the core to the circuit even with the longer wires, or, if you want, make a slit in the non turned length to connect it closer to where the turns stop. Run the circuit, check L2 output.
Now add a few more turns to L2 and check it again, if it increases output, add 2-3 more turns to the L1 now on the second layer. Check L2 output. If it does not increase, remove the 2-3 turns. Add more turns to L2 and check again and as long as L2 increases with more turns, keep adding more turns to L2. You can always try 2-3 more turns on L1 and check again to see if L2 increases but I doubt it because after the first 5 turns L1, as soon as you start the second layer of L1 you start diluting the first layer of L1 which is not wanted at all. So you keep adding turns to L2 and check if it increases until you get to the point where adding any more makes no difference or little difference. Remove those turns and stop there. That is now your Ecore with winds. That's what this device wants as a transformer. Nothing exotic, just fast and straight transfer. If you only have a gaped core, use it but think about getting a non-gaped core as well because you will have to do this on that non-gapped core as well and interchange them both just to cover all the bases. It is possible that the gaped and non-gaped cores will require a different number of L2 turns.
I am sure once this is done, the system will start purring along near loop mode. The only other factor will be to find the right value for R3 to match the real output conditions of L2 so that C11 gets ONLY just the right return to keep the loop going.
For now, forget about spikes in the circuit. Once the coils are well balanced and running, that problem should wind up being eliminated with a smoother running ecore.
Just ask your self, with such a furty little core and very good strong driving pulse, why do I need so many primary turns? hehehe
I figured out the finer criteria of the circuit. The game is simple but the rules are unforgiving. Via L1 push the output to L2 high enough so D6 cannot add to L2. This way more of D5 goes to load and if L2 can go even higher on its own, create a reverse threshold effect to D6 so L2 also adds a very strong ripple to the D5 output to the load. There is no other way I can see this work. It is a great system to test with.
If L1 is right on the first layer of the former, when Q1 shorts it out and you scope L1, you should be seeing high spikes and not the wave forms seen so far. I think the stable waveform you are seeing is because L1 is just being (*&&&?%*(*&) mufflered or dampened by the distance is has to impart through two layers of L2 then to the core. Imagine when L2 is energized, now L1 has same two layers distance to pass plus the fact that the two layers are energized, this isolates L1 from the core even more so you are working the core in a very suffocated way.
I promised to stay on mute mode but I needed to convey this now cause I am ordering my e-cores ETD-29-16-10 N87 gaped and non-gaped plus a few others (bigger ones) to have on hand but while they arrive, I did not want you guys to waste any more time with the T1000 method because I know for sure it is a waste of time and it is creating bad effects on the circuit operation.
All the best.
wattsup
---------------------------
|
Author
Topic: Akula0083 30 watt self running generator. (Read 975192 times)
