Effects of Recirculating BEMF to Coil

[rave154]

a lil help with circuitry

Hi all,

see the attchement, i have the circuitry set up as the diagram, but i want the switch-1 "transistorized" so to speak....to give the following set-up...


PWM pulse - mosfet - switch-1(transistor)

positive - ON - OFF

zero - OFF - ON

in other words, when pwm is positive voltage the mosfet is 'in' circuit but the switch-1 is open and when the pwm goes to zero the mosfet opens up and switch-1 closes.

im guessing but i think i need a PNP transistor?.... but i have little idea how to wire it as ive never worked with PNP before....not that im an expert with NPN ( an idiot actually )

i have a couple if BFY51's.but not sure if they will handle the flyback voltage...

Help very much appreciated,

Thanks,

David. D

[rave154]

Hi all,


as youre no doubt aware, the ainslie circuit and the GOTOLUC circuit with which GOTO & myself have been playing around with in various ways ( ive been doing the playing, GOTO has been doing the serious stuff ! )......these two circuits are strikingly similar...

so, ive just observed another odd effect.

using a neon in place of the "output" bulb ( as i refer to it, see the recycling thread for a diagram & definition ), with my bedini running ( which simply charges a cap and gives me access to a higher voltage source for the GOTO circuit )....with this running BUT the negative lead of the PWM circuit NOT connected.....the standing volts on the cap is around 200V.......if i then connect the negative lead of the PWM.....there is a good 1 second pause before the neon lights up...if i then disconnect the negative lead of the pwm the neon flashes.

also, with both circuits running, neon lit up...standing volts on the cap is 105V....if i then slowly insert a core into the coil, made of welding rods.....the neon dims and with the welding rods fully inserted the neon is at its most dimmist with the volts on the cap at around 130V.

i just thought i would mention this, just in case its important or triggers any thoughts that might be useful.

Ahimsa, David. D

[Joit]

Datasheet can help you in much cases for the Transistors.
A NpN is usual simple to set, just Plus to Kollektor Emitter is mostly Ground,
and for the Base you can look for the specific Volts there, but usual a few 2-3 Volts are enough, You can put a Pot there maybe or a Resistor.

[Raui]

The voltage isn't higher. Luc is using a 39V source supply, and the kickback pulse going into the 3 x 12V batteries (that are in parallel) is going to be slightly over 13.2V or so.

So, which is higher, 39V or 13V ?

.99

I stand corrected. However wasn't the kickback pulse being fed back into the electromagnet and not the battery?

-Raui

[poynt99]

I stand corrected. However wasn't the kickback pulse being fed back into the electromagnet and not the battery?

-Raui

I am fairly certain that the flyback diode is now feeding the parallel batteries as opposed to being connected back to the coil's +'ve end. That is the "standard" way to charge batteries and capacitors with inductive kickback, so I assume this is what Luc is doing.

However, Luc should confirm that.

.99

[gotoluc]

I am fairly certain that the flyback diode is now feeding the parallel batteries as opposed to being connected back to the coil's +'ve end. That is the "standard" way to charge batteries and capacitors with inductive kickback, so I assume this is what Luc is doing.

However, Luc should confirm that.

.99

Yes .99,

that would be correct.This is not what I was demonstrating at fist but now with the battery between the re-direct back to the positive side of the coil this seems to make it the classical inductive flyback circuit

However, Ive never had this good of an effect in the past.

Luc

[gotoluc]

Hi everyone,

I made a quick change to the circuit this morning. I removed the 3 batteries in parallel on the flyback side and replaced it with a 30,000uf capacitor to give a large storage so we can tune the duty cycle and get our 12.95vdc with load (bulb) attached to have a comparison to the previous test 11.

It was so fast to see the results that even after 5 minutes it was obvious to me that there is a real humongous efficiency bonus using the batteries.

Video link: YouTube - Effect of Recirculating BEMF to Coil test 12

I had to stop the test after 1 hour and 40 minutes since the voltage on the source batteries started making large drops to 35vdc and climb back to 38vdc and back down again. I don't know what is up with that but anyways it's clear that there is an astronomical efficiency boost using the batteries in parallel on the inductive flyback side.

I hope others can replicate this so we can advance to an even more efficient way to use this effect. I'm now realizing that if I would of tested this method in my resonance experiments I may of had some better results. So many things to test now.

Talk to you all soon.

Luc

[rave154]

GOTO,

as i said in the rosemary thread, i think a possible way to go from here is some kind of hybrib "tesla switch" combined with your circuit to get the best of both worlds :-)


have a good trip,

Ahimsa,

David.D

[sucahyo]

Forgive me if this had been answered before.

What is the resistance of both coil in video bellow?

YouTube - Effect of Recirculating BEMF back into Coil test 1


I think if the coil-collapse-current powered coil have much less resistance than the main coil, the power will be greater since I think there would be more current passing trough the child coil than bouncing back in the main.

I think vibration happen because different polarity between main and child.

[rave154]

Hi all

in my attempts to figure out just how much "oomph" does the flyback spike have, i just performed the following test..

PWM switching a mosfet across which is a240V->24V transformer, the mosfet is across the secondary.

using the basic GOTOLUC diode placment with a cap in line, DMM measuring the cap.

but also from the primary again a diode as per GOTOLUC circuit and DMM measuring the cap.

Also, analog ammeter measuring current draw into the main coil.

running the circuit for 20 seconds, gave the following results..

battery voltage(feeding coil ) 13.0V

current draw from battery 1.75mA

CAP-1, 15uF (main coil ) voltage 46V

CAP-2, 220uF voltage 7.5V

joules in CAP-1 = 0.01587

joules in CAP-3 = 0.00618

total in CAPs = 0.02205

joules used in battery = 0.455

ratio of input to ouput 4.8%

now either im doing something terribly wrong, OR.....flyback spikes arent worth hardly anything when stored in capacitors.

David. D

[Mario]

Hi Gotoluc,

first of all thanks for sharing all the results of your hard work and your detailed videos. I have a question regarding video 8. Do you get the same result without the magnet on top?
The reason I'm asking is that I've been working for a long time on Bedini solid state circuits and have tried pretty much everything I could think of.
In one setup I did I had one circuit charging a HV cap which was the source of the next ciruit, ranging from about 100 to 300V. So basically the second part is what you did, a flyback circuit with a HV supply(actually a buck converter). The switching was done by HV mosfets at a frequency and duty cycle dictated by the coil dimensions and load, the latter being batteries or a big cap with a lamp across.
With "dictated by the coil and load" I mean where it naturally wants to run at. I have done many selftriggering setups in the past so after a while you get a feel for how to set a circuit like this for the most efficient frequency and duty cycle, simulating a selfriggering system withouth the drawbacks of a trigger winding.
Anyway, the circuit was very efficient but not over the hill, although it had very sharp switching.
Since we basically did the same thing except for the magnet, and you got that kind of a result I can't but wonder if you get the same without it...
Again, thanks for your work

regards,
Mario

P.S. What is the wattage of your bulbs? You say 12V but not how many amps or watts.

[gyula]

David,

Would you write here the type of the flyback diode you actually used for charging the pulse into the capacitors? Also the type of the MOSFET and did use an NE555 type for driving the MOSFET's gate or something else? Also the frequency of the pulse?

thanks, Gyula

[rave154]

GYULA,

diode 1N5408

MOSFET BUZ350

NE555 driving the mosfet

frequency 5.55 kH

duty cycle 8%-10% (had to approximate it by 'eye')

[gyula]

Thanks.

Can comment two things, your diode is a general purpose rectifier and as such it is "lazy" for flyback pulses. Use any types that include "fast or ultrafast" designation in their data sheet, like for instance this:
http://www.datasheetcatalog.org/data...ild/1N4936.pdf but there are many other types, use at least 400-500V reverse voltage ratings and 1-2A forward current ratings. Such diodes has a Reverse recovery time of max 150-200 nanosec, most are under 100ns.

The other thing is your MOSFET has got 1500pF-2000pF gate-source capacitance and the NE555 is able to drive (with its specified 100ns rise and 100ns fall times) capacitive inputs of 15-20pF max only, otherwise these rise and fall times will be slowed down. This involves also a slower magnetic field collapse at switch-off. The best remedy here would be to use a dedicated MOSFET driver IC for driving your MOSFET, such as like here:
IC MOSFET DRIVER 6A HS 8DIP - TC4420CPA but also there are many other makes too.

If you tend to choose these faster components, then your BUZ350 should also be changed because you will very likely get flyback pulses of some hundred volts and in case you do not load your output, then the 200V drain-source voltage rating will not be enough and failure happens with it.

rgds, Gyula

[rave154]

GYULA,

i actually happen to have a pair of 1N4937's...so i will recreate the test later to dau using those. :-)