Author Topic: Retry N1001  (Read 6133 times)

Hidden and 24 Guests are viewing this topic.

Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
Re: Retry N1001
« Reply #64 on: June 24, 2018, 19:53:29 pm »
the tvs i found too luckly i had 200  pieces saved for this time!

i have the 1,5ke400c that has the following characteristics ... @4amps its is maximum allowed power dissipation and is going to have 574 volts across it

therefore is easy to understand that all you get to do is to get series and parallel  to your wish to get desired output

if you let 1 amp per diode is ok to not over heat it so if you plan pulsing 30 amps in the primary as a peak you need 30 diodes in parallel for 400v and 60 diodes if arrange in two series so it would go up to 900volts or 1000 easily

1 diode per amp is over kill.. we could probably use only 15 in parallel and  will be ok the important is to use a fuse or circuit breaker in the case of current get over some limit because if the transistor burn imediately the stepdown transformer will fry!!!! i´m thinking about using car fuses,,, not sure if its not too slow..

my input voltage will be from 0 to 30v so it will be able to protect the transistor if the voltage dont overcome 1150 v or so

this tvs allow for collapsing voltage very fast as they will maintain the loads voltage to a certain level whatever is the current so its impedance constantly change to keep the voltage high reseting the core very fast

there are more powerful tvs but i never found here in brazil to buy

in parallel with the tvs i add resistors to help on the protection too!

the resistance depend on the load impedance and they dissipate quite a lot of power

for example if the primary is consuming 30 amps for 30volts the impedance is 1 ohm the resistor should be set to not exceed the voltage of the transistor so in the case 33ohms times 30amps is 990v so if you add a higher value resistor it may not be helping that much however the tvs will save the day 

notice that this ratio is the ratio of the collapse voltage because is the voltage that will be impressed on the primary and proportionally at the secondary because the current in the primary is being stopped

during this pulse off cycle the power of the pulse transformer will be directed to extract electrons from water manly because it will force the cell cathode to a high positive voltage while the anode receive a negative high voltage spike provoked by the charging chokes (flyback action)

 



Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
Re: Retry N1001
« Reply #65 on: June 25, 2018, 03:02:28 am »
i already started nothing will stop me now until i get it working or homeless

i started stacking the tvs to get a good heat sharing among them

i´m doing stacks of 10 in parallel that should handle nicely up to 30 amps 450v  since each of the diodes handle max 1,5kw

next step is to cut a piece of aluminum to make the base of the drive circuit, power supply, pulse transformer and choke..

i´m trying to find a opto isolated igbt driver to get the circuit isolated as maximum as possible to be able to perform ionization easier

between the circuit ground and earth ground go  a capacitor and hv diode in parallel allowing electrons to be discharged to ground when the capacitor is not positively charged...



Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
guests dont see the attachments
« Reply #66 on: June 25, 2018, 15:59:25 pm »
i already started nothing will stop me now until i get it working or homeless

i started stacking the tvs to get a good heat sharing among them

i´m doing stacks of 10 in parallel that should handle nicely up to 30 amps 450v  since each of the diodes handle max 1,5kw

next step is to cut a piece of aluminum to make the base of the drive circuit, power supply, pulse transformer and choke..

i´m trying to find a opto isolated igbt driver to get the circuit isolated as maximum as possible to be able to perform ionization easier

between the circuit ground and earth ground go  a capacitor and hv diode in parallel allowing electrons to be discharged to ground when the capacitor is not positively charged...


for the primary side i used 40 diodes in two series of 20 in parallel
i used a copper foil to couple them electrically and thermally on both sides and took advantage that they were on their original package to solder the foil with easy
if they have a different temperature they wont share the power dissipation evenly so one ends up burning
this should handle up to 40 amps at 1000v PIV easily
for the foward current this diodes handle up to 200 amps single pulse 1ms

i found a nice optoisolated driver available to buy called fod3184... there are others better but here this is the best i found

WITHOUT THE OPTOISOLATION THE CIRCUIT WILL BECOME ELECTRIFIED

i noticed that the optoisolation is required for one reason! to decouple the electronic circuit components from the pulsing side... the optoisolation must be able to handle maybe up to 2kv depending on the isolation capacitor used between ckt ground and earth ground

this capacitor will store positive charges left on the cell during next pulse on, after electrons extraction to help ionization by providing the positive charges back when the pulse turn off 
« Last Edit: June 26, 2018, 13:20:13 pm by sebosfato »

Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
missing components
« Reply #67 on: June 25, 2018, 20:34:19 pm »
hello

today i got the fod3184 to drive my igbts optoisolated to allow the positive charge left on the cell to return to the isolated ground capacitance... this capacitance is in parallel with a diode that allow the electrons to be discharged to ground earth but when positive charges come it will get charged using electrons from ground to counterbalance the positive charges left on the cell.. . the potential rise will only depend on the capacitance.. so is good idea to start with a considerable value to accomodate the charge without rising the voltage in a manner that will kill the optoisolation..

i would keep within 2kv 100nf limit to start testing and i´m thinking of adding a tvs in parallel to limit it to this voltage
 
is important to have the current wave form on the oscilloscope to get track of the amps peak...
also is good to have a voltage reading of the igbt voltage to be safe about whats going on
every component burn is a big step backwards specially having no money to get more... i guess this components are here since about 2 years now it was a time that i had some money spare and decided to get components for the future testing... if i didnt had this components here i would probably get a long time until testing...

now i need to buy some 14 AWG wire and some 18AWG for the coils and pulse transformer

as i´m planing to have 10amps discharge to start with my primary will need to handle 30 amps as the pulse transformer is 1:3 but the switch will be seeing 40 amps or little more as the secondary is in series with the primary

the charging choke work as a load during pulse on! helping concentrating the power for later use during the collapse

the lower is the voltage applied the lower can be the inductance of the charging choke for the same amps so the greater is the speed of the discharge

so decreasing the coil size or frequency increase power out

 increasing voltage and or amps  increase power output

but increasing the coil increase the time of discharge and reduce power


Today i also buy a diaphragm pump to use with my filter it claims to be 100psi 6.8 bar 3amps 12v i hope it works ok because i got the cheapest one i pay like 20$ on it

so i would really like if you by any chance want to collaborate with me because i´m buying all this stuff with credit card but i´m not at a good situation to spend on this components.... of course i´m trying to use all i have to avoid having to buy more stuff

hopefully i live in a big house and i accumulated all the prototypes and spare parts i bought long this 13years researching

so to donate please follow this link

https://www.paypal.com/cgi-bin/webscr?cmd=_donations&business=sebosfato%40gmail.com&lc=BR&item_name=Sebosfato+Probe+Crowdfunding&item_number=1&no_note=0&currency_code=BRL&bn=PP-DonationsBF%3Abtn_donateCC_LG.gif%3ANonHostedGuest

as a counterpart i can make some videos with the progress of the tests and explaining better what we are doing here..


thanks so much for your effort in changing the word!

If you enjoy this website you can also contribute with it and other members that are working hard at this technology all of them so far as i know by their own cost and time! Be aware of that! Even if you just read and not a scientist or not able to do experiments you can help who can and follow the progress and be part of something great!

In attachment i added a excel sheet showing my calculations

i will explain here

the idea is to use inductive discharge

when we charge an inductor we charge it with current

1 Henrie is equal to L = 1 Volt * 1 Second / 1 Amp

this mean that if you have 1 henrie inductor and you apply 1 volt for 1 second you will have 1 amp on it ... the current increase linearly if dc is applied

so you have also 0.5 Joule because energy E  = 1 Amp ^ 2 * 1 Henrie / 2
 
so my calculation assume 50% pulse width and i manipulate the frequency and voltage and current desired to get the inductor value i need to match that condition

i made it this way because it change the parameters dynamically allowing me to discover the behavior of parameters

when this energy is discharged it is discharged into the cell in 100 nanoseconds scale and will depend on the cell capacitance and choke inductance

so from the energy i got the maximum voltage for resonant discharge

and also calculated the voltage and time for discharge considering water just a resistance

this is simply taking the amps and multiply by the resistance V=I*R and the current on the inductor must get thru the resistance

this allow to see if your really charging the capacitance or just forcing the resistive side!





 
« Last Edit: June 26, 2018, 02:40:02 am by sebosfato »

Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
Using Full Power
« Reply #68 on: June 26, 2018, 13:19:34 pm »
I was thinking about the usage of the coil in parallel with the circuit and after making some more drawings i discovered something increadible

i found that is possible to use the power of the pulse transformer not only to collapse the field and perhaps cause ionization (that would be too low of a load) but it can also be part of the high energy discharge circuit and guess what

i was thinking about the wiring the primary in series with the system but it takes its isolation out complicating more although doable.. the reason i wanted to do that was because i was not thinking that another coil could be added in series with the discharge circuit to get the power out of the core during the collapse so saving the protection tvs from dealing with all the power that didnt make thru the other side

i have now exactly meyer drawing except by adding one coil in parallel after the diode being this coil be coupled with the lower choke while the upper choke is in the same core of the pulse transformer

this way the cicuit must work because all the hell of power it can absorb it can discharge in the water and thats the idea

i found also something else

leaving the pulse transformer an isolation transformer (not connecting the primary) turns unnecessary the optoisolation and the isolated ground and its complexity ... since the circuit of water is floating... however is required a manner to ionize the gas coming out and it may be achieved with the primary but connecting in an isolated manner thru a diode... the only thing that can pass here is electrons going to earth (if earth is connected on the negative side of the power supply but)

if that is connected at the positive side of the battery or power supply, the electrons are forced to go there recharging the battery or powersupply with water electrons

another isolated coil can be added on T1 for this work connected from circuit ground to the cell tube electrode where the electrons are going to be theoretically bombarded to create a potential during pulse collapse to extract those electrons ...

or it could be an external circuit etc..


something else that i was elaborating is a transmission line snuber to add in parallel with the tvs on the primary side... but i´m still wondering how to use it..

another way to do it is to add a small transformer in parallel with the tvs but it must be designed to have a secondary with the turns factor to allow the pulse to return to the source in the form of a high current pulse when the tvs see 800v for example ... however is important that the primary in this case (of this small transformer ) have a decent inductance to not saturate and not create a much lower resistance than the tvs or pulse will not collapse well...

meyer said Keep it simple stupid so probably is better to use the tvs only and maybe a resistor and varistor only!
« Last Edit: June 26, 2018, 14:02:42 pm by sebosfato »

Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
matching impedance making sense
« Reply #69 on: June 26, 2018, 18:31:05 pm »
Matching impedance was told to us to be the key...

well theoretically it is indeed the way to discharge the energy completely without reflection and at one cycle but there are a bunch of impedance matching  for different things

for example matching the transmission line impedance will cancel reflections

matching the impedance of a resonance tank with a resistor will bring the Q factor to 1

so what is about this matching

Lets put it simple

when we get coils carrying current they must be connected to coils with the same current going otherwise it will not discharge at the same time and the coil with higher current will send another pulse into the other coils

i will give you an example of a 1:3 pulse transformer

the primary have 30 amps the secondary 10 amps if we connect them in series to discharge the primary will discharge forcing 30 amps into the other coil where there was only 10

the idea is to maximize the discharge of both cores simultaneously to get all energy at once!

to do that is important to match the coils

to do that is not complicated basically the coils must be similar in size and shape and have the same current on

for example primary L1 =1 secondary L2 before diode and L3 in series after diode dots summing = 3   

and charging chokes the other coils C4 (in parallel after the diode) and C5 (between C4 negative side and cell ) also are similarly  dot summing

so during discharge they will experience the same current and the discharge will be as fast as possible and complete..

now if the discharge matches the water resistance than the impedance is matched with water also and there wont be reflections!

for that is important to determine the impedance of the discharge of the coil...   but i do not think it is necessary!
« Last Edit: June 26, 2018, 18:47:24 pm by sebosfato »

Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
Design
« Reply #70 on: June 26, 2018, 19:06:47 pm »
the coils design i will do here to keep records

i will be using a couple of cores with the same characteristics

starting with the primary

first lets find a core that can handle a good amount of energy for the discharge

we know that is better to use lower voltage and higher amps to get better speed response by needing a lower inductor for it

number of turns = Applied volts / Bmax / Core Crossection Area / Frequency / factor of 2 (because is only dc pulse)

So if i plan to apply 30v the bmax is 0.3 the are is 1inch ^2 and frequency of 5khz
remember 1 inch is 0.0254 meters
the turns ends up = 15 turns and is pretty reasonable since this one  is a big CC core i have from thornton.com.br is th NC100 lets say we are going to add a small gap and consider the bmax to work on a lower level of flux we can get up to twice the turns with no problem.. so lets say 30 turns for the primary keeping 1 volt per turn during pulse on

30 turns means 3 meters of wire and it must have a low resistance to not provoke a voltage drop on the primary that is greater than 2% so it must have a resistance of 20miliOhm lets see what wire is good for this

so if i want 30 amps on the primary it must have a low resistance for ... i think that adding 4 x 18awg wires in parallel is a good idea or you can use a 12AWG ... using wires in parallel is not a bad idea as it raises the frequency you can work for 18awg is 17khz and 12awg is only 4khz


the secondary will have 90 turns so 9 meters of 18Awg wire is around 200mohm and at 10 amps flow give a dissipation of 19w
 


the bad thing about using a too big core for it is that it can cause problems with to much energy accumulated but i guess its easier to achieve the effect with a powerful pulse!

i still have those small cores that Steve kindly send to me some years ago that is going to be used for a smaller version too for sure....

« Last Edit: June 27, 2018, 02:01:14 am by sebosfato »

Online Hidden

  • Global Moderator
  • Hero member
  • ****
  • Posts: 3178
sorry if you think this is another replication
« Reply #71 on: June 27, 2018, 00:35:18 am »
well this is design from zero

i´m sorry for those who dont believe we are capable to get there with our own legs and ideas

today i´m reviving labview as it is a great way to monitor some of the circuit parameters like the temperature of the circuit... how many watts going in ,,, the voltage ,,, the impedance...

i was also thinking on developing a simple peak metter for amps to detect the amps in primary

i will be posting the code for it here in attachment... soon,,

please donate would be very nice to receive your donation

i need to get some wire for this transformers as i dont have much here...

something like 50$ donation would be very helpful.

thanks 


i´m using for the current meter acs712 and an arduino nano because it has more analog inputs than the esp but i will add an esp32 to it so i can transmit the data via wifi so i can get the readings...and to store the webserver that will show the readings... the problem is that the sensor is analogic and is 5v so as arduino but the esp is 3v so i must send the data digitally from the arduino to thru a level converter to the esp so i can get the data transmited wifi.. 
this is a 30A hall sensor module that only need to be connected in series with the power supply... i can also be connected to the oscilloscope too...

this arduino will measure the volts applied the amps in to start with.. it will than calculate the watts... also indicate the frequency and temperature meter to check the temperature of igbt or coils etc i can even make it count the pulses if i want
calculate the duty cycle and other stuff

the esp also have other inputs


i will keep it simple at start to not waste time but will let the code on the html files and

« Last Edit: June 27, 2018, 05:45:08 am by sebosfato »