PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-11-26, 08:25:23
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 [2] 3 4 5 6 7 8 9
Author Topic: DSRD pulse generator  (Read 200788 times)

Group: Professor
Hero Member
*****

Posts: 3498
I had a play tonight, i don't have all the bits, i tried using a 555 timer set for the min mark/space i could get, unfortunately i could only get down to 1uS i
You could use the attached cheap nanosecond monostable to produce shorter pulses (20ns - 800ns).
You can drive it with the output of the 555 Astable or any other 0-5V waveform in the kHz range.

P.S.
The older 74HCT02 is weaker but it works, too.
   

Group: Administrator
Hero Member
*****

Posts: 3960


Buy me some coffee
Thanks Verpies & Welcome.

That's a nice simple circuit  O0

Peter
   

Group: Tinkerer
Hero Member
*****

Posts: 3055
« Last Edit: 2012-11-06, 03:18:08 by Dumped »


---------------------------
For there is nothing hidden that will not be disclosed, and nothing concealed that will not be known or brought out into the open.
   
Group: Guest

I have used a 74LS123 (monostable) for a full and easy control of the pulse width in the range 50-550ns, independently of the frequency. It works well. You just need an external RC circuit, see the datasheet.
The difficulty now is to amplify enough the TTL level to drive the mosfets. It's painful to fight their input capacitance. Despite a supplementary stage, I have not exceeded 250v pulses on 52Ω load while I got 600v with my previous setup. The best value for the pulse was 500ns.
More work needed  :(. My goal is to reached 10KW. I'm convinced it's feasible.

   
Group: Guest
It would seem that the technology is well developed.

Not a bad solution...  18kV => 75Ω  = 4,43 MW :D
But what's the price?

Quote
Another technical treatise.
Insights from the recent past.

Thanks for the links. I have tested 1N4004 without success (very weak pulses) but perhaps a different circuit is needed.

   
Group: Elite
Hero Member
******

Posts: 3537
It's turtles all the way down
It would seem that the technology is well
developed.


Another technical treatise.

Insights from the recent past.

Nice finds Dumped!

 The treatise has an interesting "physics" explanation of the the devices studied.


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   

Group: Professor
Hero Member
*****

Posts: 3498
The difficulty now is to amplify enough the TTL level to drive the mosfets. It's painful to fight their input capacitance.
The UCC27511 drivers are cheap strong and effective down to 15ns.

According to the Pg.4 of the Belkin Document, the transformer should saturate when the voltage in the capacitor on the secondary side is at its maximum and the current in the secondary winding is crossing zero.
« Last Edit: 2012-11-07, 15:34:16 by verpies »
   

Group: Administrator
Hero Member
*****

Posts: 3960


Buy me some coffee
So much power gets wasted trying to drive the fet gate, it would never make a viable low power driver, maybe we would be better to try a transistor driver if possible

pulsing 6 Amps to the gate just seems to suck to me.
   

Group: Professor
Hero Member
*****

Posts: 3498
So much power gets wasted trying to drive the FET gate, it would never make a viable low power driver, maybe we would be better to try a transistor driver if possible
pulsing 6 Amps to the gate just seems to suck to me.
But 6A is not much power at 12V and several nanoseconds at low duty cycles.  Remember that after the gate is charged it does not draw any current anymore, yet the MOSET keeps conducting between the source and the drain.
Also, the charge stored in the gate could be returned to the power supply when the MOSFET is turned OFF.  Not many engineers try to recover that charge because it's too much hassle for little reward.

BJT transistors require base current during their whole conduction period, so I can't see how they would be more efficient.
Those UCC27511 MOSFET drivers are cheap, quick, strong and readily available and they can drive BJTs, too.
   

Group: Administrator
Hero Member
*****

Posts: 3960


Buy me some coffee
Don't get me wrong, fet drivers are a great way to drive a fet, i have been using them for a number of years with the pulsers i have built, but a cop 2 device would stand little to no chance of self sustaining using a fet driver, even if it was only switched on for a very short time.

For instance i can inject 25nS pulses into a coil that's being tested from a 50V supply and draw a mean current of 3mA, where as the fet driver is drawing a mean current in the 10's of mA when running at a few khz, in some cases i am not going anywhere without a cop>10 because of the wasted power i used to drive the fet.

Now i dont have a transistor circuit so i cannot say and my theory is not as good as you guys, but if i drive the base for 25nS do i still need to inject 6A @ 18V to turn the transistor on, if it's less then it's better from my point of view.

offcourse none of this really matters until we start seeing free energy in the first place and we then need to start looping, my logic says design as low power as possible ready for that day because that day will come sooner.
   
Group: Guest
The UCC27511 drivers are cheap strong and effective down to 15ns.

I just read the datasheet. It seems to be really the solution, thanks. I wasted two hours today for unsatisfying results, so I will probably order one.

   
Group: Guest
...
because that day will come sooner.

I hear this from at least 10 years and Tom Bearden repeats it since the 80's!
Don't put the cart before the horse.
It reminds me politics, preparing for the big day that never came.
"Stand up, damned of the Earth
 Stand up, prisoners of starvation
 Reason thunders in its volcano
 This is the eruption of the end.
 Of the past let us make a clean slate
 Enslaved masses, stand up, stand up.
 The world is about to change its foundation
 We are nothing, let us be all."

This is the final struggle.    :)

   

Group: Administrator
Hero Member
*****

Posts: 3960


Buy me some coffee
I did not say free energy will be coming, just that if it did it would come faster by using less to create it in the first place.

It is a nice chip if you have surface mount rework gear and are able to make pcbs  or are you going to manually solder wires to the legs ;D

It 1.45mm x 2.75mm with 6 legs.

I lost my patience with sm fet drivers when i kept getting EMP pulses wiping them out during one of my experiments, i ended up reverting to dip.
« Last Edit: 2012-11-07, 21:59:36 by Peterae »
   
Hero Member
*****

Posts: 2735
@Peterae
Quote
I did not say free energy will be coming, just that if it did it would come faster by using less to create it in the first place.
It's already here, now consider all the time spent here and the results you have, then consider 3x6 solar cells on Ebay at $265/kw versus what you have spent on FE research. The only thing we have to understand is that if we invest our own time wisely then the payback period can be a matter of a few years ..... then it's all free.
I don't think it will be very long before what we are doing here will become a mute point as the cost to build or buy an exotic FE device may be double the future cost of solar which is dropping like a rock. It's actually kind of funny that the price point that some of the FE sellers are at is almost right on par with solar if you DIY, now which do you think would be a better investment?.
Personally I see no difference between FE and solar FE as the source is most likely one and the same.
AC


---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   

Group: Administrator
Hero Member
*****

Posts: 3960


Buy me some coffee
Yes indeed solar is coming down in price, it's close to £1/watt here in the UK that is £1000 per 1kW if i install it myself, offcourse the law says you cannot install it yourself as it is not accredited to be connected to the grid, so to be lawfull here, i would need to disconnect from the grid to use my solar unless i pay the £8000-£14000 they are charging to install them for me,so there are a few problems.

During our cloudy winter months which seem to be about 7 months of the year, they just don't work well enough, then you need roof space that is south facing for them to work well, and then they don't work at all at night, so then you need to buy expensive storage batteries to store the excess for night use, i have looked at it very closely.

If it was summer here all year and we did not have nights even then to meet my family needs we would need 3-4kw of panels, but i would first need to demolish the house and rebuild it, because i have a west facing roof.
If you live in a flat you are screwed, if you don't have £6000 for the controller, batteries and solar panels then you are also screwed, other countries the situation maybe different.

If i stuck up a 3kw wind turbine, which would be huge my neighbours would probably form an orderly line outside my house with bags of tomatoes, so that is out as well.
   
Group: Guest
There is a free energy thought!

Run all your external drainage to a cistern converted to a galvanic cell.

Install a very large and obnoxious satellite dish and wind generator (non-functional).

Power your house from your new battery energized by the acidic fluids of the veggies/fruits thrown by the neighbors. I would need to install bullet proof sheathing, here.

Problem solved  O0
   

Group: Professor
Hero Member
*****

Posts: 3498
For instance i can inject 25nS pulses into a coil that's being tested from a 50V supply and draw a mean current of 3mA, where as the fet driver is drawing a mean current in the 10's of mA when running at a few khz, in some cases i am not going anywhere without a cop>10 because of the wasted power i used to drive the fet.
For 25ns the gate charge time will take significant part of the total pulse period. That energy delivered by the gate charging cannot be calculated by multiplying 18V by 6A because both such current and voltage does not exist at the same time at the gate.  The gate waveforms are triangular or trapezoidal.  It's better to treat it as energy delivered to charge a capacitor (E=0.5CV2 ).
At 25ns the BJT has significant parasitic capacitance too, so a detailed study would have to be made to compare two specific BJTs and MOSFETs at 25ns.
However the DSR Diode forward pumping time is in the range of several 100s ns, and at those time the gate charge time will constitute a small part of the total pulse width.

   

Group: Professor
Hero Member
*****

Posts: 3498
Here is an article about people that were trying to manufacture a Silicon Carbide (SiC) DSR Diodes for picosecond range switching.
   
Group: Guest
Thank you verpies for the analysis on switching efficiency.  If gate is conduct by voltage and base conduct by current, then we can compare how much energy spent for each type?

gate = .5CV^2 where V is full conducting voltage
base = I^2Rt  where I is full conducting current and t is pulse time

Gate energy is a set value while base energy depends on pulse time.  If we take the ratio of gate/base or reverse, we can see which one is more efficient.  For non recoverable switching energy, base would approaches zero if pulse time approaches zero.  For recoverable, gate energy is almost zero.  However, if gate is being charge like a capacitor, then there is a specific time it takes to charge up (if we use too high voltage, it may break?).  We can put a spikes into the base to have short pulse with low energy.  I think this method use little energy for switching.  Jonnydavro seems to mastered this.  But it seems the energy put/stored into the system must be many time greater than the energy use for switching if OU is possible. 

http://www.youtube.com/watch?v=UXI7B0j0Ajc&feature=related

I like the concept of solar panel.  What if every coil or conductor is a receiver?  Similarly it absorb energy just like a solar panel.  I speculate the reason energy does not build up in a wire naturally because it just pass right through it.  I'm thinking of a theory which initially or continually excite the atoms so it can capture those energy.  If atoms vibrates violently, it has more chance of being hit by incoming energy of any wavelength. 

   
Hero Member
*****

Posts: 2735
@peterae
Quote
Yes indeed solar is coming down in price, it's close to £1/watt here in the UK that is £1000 per 1kW if i install it myself, offcourse the law says you cannot install it yourself as it is not accredited to be connected to the grid, so to be lawfull here, i would need to disconnect from the grid to use my solar unless i pay the £8000-£14000 they are charging to install them for me,so there are a few problems.

Here in Canada bulk cells bought in volume on Ebay can approach the $0.25/watt mark at which point we must build our own panels which is fairly inexpensive and easy. As well I build all my own electronics which relates directly to the topic here in that a Buck/Boost converter could be made inexpensive if very little materials were used. This means a low turn/low resistance, high power MPPT converter switching KW in the nanosecond range instead of slow switching bulky converters using long conductors and inferior core materials.
It should be noted that most all the manufacturers give data from ideal conditions which never actually occur in actual use which is another way of saying the real numbers will never be even remotely close to what they claim.



---------------------------
Comprehend and Copy Nature... Viktor Schauberger

“The first principle is that you must not fool yourself and you are the easiest person to fool.”― Richard P. Feynman
   
Group: Guest
In my new setup, a 74123 monostable generates 50-550ns pulses from the signal of a function generator which is used only to fix the repetition rate.
Then a NPN 2N3553 amplifies the pulse so that the voltage is now compatible with a MOSFET input.

I have put a new stage between the 2N3553 and the final IRF640. It's a MOSFET BUK152-100A BUK452-100B mounted in source follower to drive the IRF640.
By adjusting the pulse width and L1 (see post 18), the DSRD pulses reach 500V/52Ω, i.e, near 5KW. It's not my previous record of 600v but the signal now is better, the pulses driving the IRF640 are more clean. I clearly see on the oscilloscope the step of the charge of the diode. Then the back emf happens at about the same moment as the DSRD pulse. Probably the back emf helps the diode to quickly release its charge. I think it's interesting to not block the back emf, it enhances the process.

« Last Edit: 2012-11-09, 10:02:26 by exnihiloest »
   

Group: Administrator
Hero Member
*****

Posts: 3960


Buy me some coffee
Very impressive whats your source resistor value, you mean BUK552 not 152 i think. good move  O0

The BUK552 is a logic level fet do you still need your pre driver transistor because i have been driving this fet from a pic micro with good results.
« Last Edit: 2012-11-08, 22:25:56 by Peterae »
   
Group: Guest
Very impressive whats your source resistor value, you mean BUK552 not 152 i think. good move  O0

The BUK552 is a logic level fet do you still need your pre driver transistor because i have been driving this fet from a pic micro with good results.

I made a typo for the 1st digit and last letter, sorry. This transistor comes from a recovery of waste materials. It's a BUK452-100B. Here is the datasheet: http://www.datasheetcatalog.org/datasheet/philips/BUK452-100B.pdf.
I took it because it was the MOSFET of my collection with the lowest input capacitance, while being compatible with the level of required power. For example it allows a better pulse shape than an IRF630 that I tested before.

For the today program, I will test many diodes and double-diodes in flat cases like IRF640 transistor. I had forgotten them and tested only my cylindrical diodes with axial terminals. I hope to find other working models.

   
Group: Guest
None of my diodes beats the 5404 for the DSRD effect.
Several give pulses around 2/3 of these from the 5404:
ERC38, BYW51, BYV32, FMUG2Y, YG912S6, STPR1020, FMU-G2YXS

Most are fast recovery diodes (15<t<50ns). But there are also fast diodes that present a reduced DSRD effet, or very reduced. For instance, the S3L gives pulses of 1/3 of the 5404 and nevertheless has a recovery time of 35ns. The datasheets are useless to choose the best diodes.

Only one diode is not far from the 5404: RG2.

   

Group: Administrator
Hero Member
*****

Posts: 3960


Buy me some coffee
Great work ex thanks for your time on testing these diodes.

Looks like i better order some 1n5404 diodes  O0

Now i am thinking of modifying your drive circuit as follows
BUK552-60A (I have these) driving the gate of a PFET to ground, then put the transformer primary in the source of the PFET IRF9640.

I will use your monostable to drive the BUK552-60A

Does this sound like it may work.
   
Pages: 1 [2] 3 4 5 6 7 8 9
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-26, 08:25:23