A green and clean environment is a human right!
Projects by members => Projects by members => Sebosfato => Topic started by: sebosfato on March 06, 2017, 06:57:04 am
-
I had an idea that i feel need to be investigated....so here it goes...
imagine a transformer having two primary winding and one secondary winding... the secondary being in the middle of the two primaries what decouples the two primary a little
the two primary will have oposite fields so they will be pulsed at the same time for start than they must be sequencialy turned of
the idea is that if they are pulsed at same time their magnetic fields will cancel but when they are sequence turned off one first and next the other it will create an ac voltage across the secondary, one half cycle per primary turned off
as the coils are oposite in magnetic field they should saturate with current very fast, allowing both to have their time of collapse..
the main idea is that we can possibly charge both coils at same time and as we sequence turn them off they will have a sharp flux change
to achieve this we need a power supply and two independent switches
the timing circuit could be made with 555 timers and some inventive spirit
the two pulses must start in sync than one will terminate, and the other remain a little and than it terminate too and than there is a time with both off for the collapse to end for start over again..
so maybe one timer being the master clock and the second being a monostable
-
when the first switch open a clamped voltage will appear across the other primary and secondary... on the other primary the voltage will be clamped to the source voltage level since its already connected to it... when the second switch opens a higher level voltage should exist but it will also be clamped by the diode that exist in parallel with the switches..
so the output in secondary should be proportional to the turns count
however i´m not sure the current will be conserved...
tests need to be done
we need a circuit for generating this kind of pulses...
(http://www.ionizationx.com/index.php?action=dlattach;topic=3178.0;attach=15076)
the duty cycles should be 33% for t1 66% for t2 and 33% t3
of course maybe if for example t1 could be reduced to 1% than t2 would represent 49% and t3 the other 50% or something like that
-
You could instead ground the center tap, then connect the positive outputs of a flip/flop - or relay - to the a and b inputs of the primaries. This is called "splitting the positive".
-
You could instead ground the center tap, then connect the positive outputs of a flip/flop - or relay - to the a and b inputs of the primaries. This is called "splitting the positive".
Wouldnt it be just the same?
i usualy prefer to switch on the low side because is much simpler...
-
I think i just found a way of doing it... similar to what i did in the fracture cell ckt..
Basically is one 555 timer astable operation and two comparators making 2x 0 to 100% duty cycle variable frequency generator
connected to two high voltage and amperage igbts having diodes and some tvs for safety in parallel with it
the simpler the better
-
the simpler the better
-
Hi Fabio,
You said to put your secondary between the two primaries. Is your objective to produce bucking fields, with splatter potential energizing the secondary, or do you just want to produce a sine wave from chopped DC?
-
I had this feeling that the primaries must be coupled more to the secondary than to each other however as they make part of the same core they will be coupled anyway...
my objective with this would be to create a different kind of voltage output in the sense that when you pulse both oposite polarity coils at the same time no induction will result as you see on the drawing, however when the first primary is interrupted a positive voltage will come out from the secondary and when the other primary is interrupted a negative voltage will result...
my guess would be that it could maybe generate a huge spike too i´m not sure how the waves output will be exactly but i tried to show in the drawing a prediction...
i´m not certain about the spike because once we interrupt the current in one primary the other primary already have also maximum current thereto no change in flux actually.. the change in flux comes only when we interrupt because when we activate them together they result in no flux... its prety much like having a high turn ratio of one to the secondary number of turns... but not exactly as we can simply turn half of the turns off and so on sequentially ...
-
Good luck in your quest for discovery. Keep us posted if you try something.
-
i worked at the schematic for the freq gen so far... now i need to do a pcb and build the test
-
today i got some ICs for this project... i got the lm219 and lm393 for the comparator part
for the timer i will use a ne555 i have many here..
for the driver i will use a ir4427 or a couple of tc4420
for the switch will be a couple of igbts 1200v capable of 15a
i will add Tvs in parallel with the coils for a voltage of 880v along with a diode and up to a current of 5 amps or more depending on what i have here available...
to increase the current ability of the tvs you just need to add more in parallel otherwise they simply explode and than the transistor burn right next...
usually i also have a resistor in parallel with the tvs around from 220 to 500 ohms high power and not soldered because solder melts with the resistor geting hot and all burns as well
I will set the coils like if i was designing for 5khz and so the circuit will go from maybe 2khz up to 16khz 3 octaves or less...
-
ckt for it
-
i did this circuit
the input voltage is 12v
for the timer part i used a 4.7k resistor in series with a 10k potentiometer 10nf capacitor..
on the comparator part i used a two 10kohm resistors connected respectively to the vcc ground and to a 10kohm potentiometer so its central tap output from 4 up to 8v
basically i want to send the 555 timer ramp that goes from 1/3vcc to 2/3 vcc into the non inverting input of the comparator and sending a voltage to the inverting input within the same range for generating a couple of pulse width variable frequency signals
-
this datasheet is a * joke
check the pin numbers it says the package i´m using have the 1,2,13,14 NC while the 1 and 2 is output and ground respectively
-
so you found the problem?
-
well i tried to find each datasheet for it and i found one only one that may have the answer...
it says if the voltage supply is higher than 9v the v- should not be tied to the ground... and its something i did... i will try to disconnect and see if it works
http://pdf.datasheetcatalog.com/datasheets/270/52732_DS.pdf
-
i´m happy to tell you that the problem is now solved... i found that this chip is open collector output so it need a pull up resistor on the output... 1k will give 12ma using 12v supply.. i had to substitute the chip because i think i burned the first in the way..
the circuit works perfectly and generate waveforms in sync with different pulses widths. .
for that i had to add a uf4007 diode to the charging path from pin 3 to the resistor capacitor network on the 555 timer..
i found out also something new and interesting.. i was observing some ringing on the output of the mosfet driver and i tried to add resistors everywhere but i found that adding a small capacitor as load on the input side of the driver killed the ringing immediately... i tried 1nf .. but i will try some smaller caps too .. this ringing would make the end of the pulse cause false trigering of the mosfet or igbt... i thought of the capacitor because i thought the ringing was coming because the igbt was not yet present but surprisingly even adding a capacitance to the output the ringing wont stop only if add on the input part...
-
part list and schematic for who is interested...
Here you see the full schematic and part list... i call this circuit Dual Sync Pulse Width Frequency Generator..D.S.P.W.F.G., because is just what it does..
the capacitor that filters the ringing is in parallel with a 1k ohm resistor.. and it cause a slight delay in the turn on of the driver.. but is not important for what i´m trying to do... it has this delay because the resistor connected to the vcc 1k takes a time to let the capacitor voltage rise above the threshold of the driver.. but its ok since it prevent ringing totally! i have to find some smaller caps to try if still cut the ringing while take less delay
the output with igbts arrangement i will post later... with some waveforms...
it works pretty well and now i just need to wind some primaries to test the theory!
the pcb i made with the tracks on top so i dont need to make holes i just solder the components over it..
-
Are you trying to make the alternating pulsed DC like I have done in the attached image?
-
Hello Tony
The idea of this circuit is to create two pulses that start at same time with same frequency but have independent variable pulse width
My idea is to get this pulse combined in two split primaries being in opposite phases
The objective is to try loading a coil with current but having zero magnetic flux on the charging time
I'm almost done to start the tests but didn't turned it on yet