OverUnity Research

Here's my documentation and experiments regarding a pulsed Bifilar Coil using a variable Width and phase controller.

I was following Spherics Bifilar coil setup and looking for any interesting effects.
It was found that at a certain phase delay of a variable amount of nS a new pulse would appear which is a larger amplitude than the driving pulses.

At some points in the phase scan, the controller seemed to be subjected to extreme RF noise and was causing my display chips to read incorrectly, i was also getting false triggering of the Fet's causing excessive on periods, but during this Loud Explosions could be heard inside the wire from upto 5 feet away, the coil was 22 turns on a wooden former See Video Kick1 below.

Lab Bench Picture Showing setup for some of the following tests.
(http://www.overunityresearch.com/index.php?action=dlattach;topic=37.0;attach=163)
Experiment 1
Pulse Discovery.
Using my First Generation controller.


Experiment 2
I found that using both coils of the Bifilar coil when put in parallel and driven by both fet modules causes phase delay periods of great RF Interference, this later went on to the Explosions in the wire experiment, and also driving a single coil wire with both fet outputs connected.


Experiment 3
This time i used only one coil of the bifilar connected to both fets.


Experiment 4
Same as Experiment 1 but using a coil with half the length of wire used in Experiment 1


Experiment 5
Same as Experiment 2 but using a coil with half the length of wire used in Experiment 2


Experiment 6
Same as Experiment 3 but using a coil with half the length of wire used in Experiment 3.


Experiment 7
Same as Experiment 5 but using 2 less turns on the Coil, i am now trying to study the pulse and when it occurs and it's amplitude when i have varying wire lengths.


Experiment 8
Same as Experiment 7 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.

[/youtube]

Experiment 9
Same as Experiment 8 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.

[/youtube]

Experiment 10
Same as Experiment 9 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Experiment 11
Same as Experiment 10 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Experiment 12
Same as Experiment 11 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Experiment 13
Same as Experiment 12 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Experiment 14
Same as Experiment 13 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Experiment 15
Same as Experiment 14 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Experiment 16
Same as Experiment 15 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Experiment 17
Same as Experiment 15 but using 2 less turns, studying the pulse timing and amplitude with different wire lengths.


Summary of data in Video 7-17
P = PULSE TIME IS AT PEAK AMPLITUDE
DN = DISPLAY READING NOISE OR GARBAGE
FP = FISRT APPEARENCE TIME OF PULSE
DP = TOTAL DISSAPPEARENCE OF PULSE
-----------------------------------
Video 7

DN 127-130
DN 134-136
DN 143-145
FP = 141
P 160 =       80 VOLTS
DP = 176
DN 161-170
FP = 179
P 207 =       82 VOLTS
DP = 255
DN 243-244
----------------------------------
Video 8

DN 122-130
DN 132-133
DN 141-143
DN 147-153
DN 161-165
FP = 178
P =208 =       90 VOLTS
DP = 255
---------------------------------
Video 9

DN 120-127
DN 131-132
DN 134-135
DN 141-143
DN 146-153
DN 156-163
FP = 175
P = 209       90 VOLTS
DP = 255
DN 238-240
---------------------------------
Video 10

DN 66-71
DN 119-125
DN 129-130
DN 140-141
DN 151-153
FP = 174
P = 208       90 VOLTS
DP = 250
DN 238-239
-------------------------------
Video 11

DN 65-70
DN 116-156
FP = 174
P = 208       85 VOLTS
DP = 255
DN 216-224
DN 236-237
-----------------------------
Video 12

DN 45-46
DN 64-67
DN 113-151
FP = 185
P=202       76 VOLTS FIRST SIGN THE PULSE HAS SHIFTED
DP = 237
DN 214-222
DN 230-234
-----------------------------
Video 13

DN 58-69
DN 80-88
DN 110-127
DN 140-144
FP = 168
P = 200        58 VOLTS
DP = 233
DN 225-226
----------------------------
Video 14

DN 45-49
DN 56-68
DN 81-88
DN 106-113
DN 117-130
FP = 167
DN 216-231
DP = 234
-----------------------------
Video 15

DN 41-49
DN 55-65
DN 105-110
DN 116-124
FP = 161
P = 185      44 VOLTS
DP = 235
DN 213-218
DN 233-233
----------------------------
Video 16

DN 38-51
DN 101-107
DN 115-121
FP = 154
P = 185      38 VOLTS
DP = 238
DN 207-215
DN 227-225
-------------------------
Video 17

DN 40-49
DN 63-63
DN 104-109
DN 119-121
FP = 159
P = 184      42 VOLTS
DP = 236

DN 227-231


Experiment 18
27 Turns of multistrand wire on one quarter of my OTPU former, Just testing a new coil to get a baseline.


Experiment 19
Using the new 27 Turn coil i am now studying the current


Experiment 20
False Triggering of fet gates during RF bursts

http://www.youtube.com/watch?v=TPab8_yxv_w

Experiment 22
Scoping the gate- source pins of my fets

http://www.youtube.com/watch?v=MJKFNpmfhq0

Experiment 23
to pulse a bifilar coil

http://www.youtube.com/watch?v=ZaYOMWmcfHo

Experiment 24
Pulsing a bifilar Coil

http://www.youtube.com/watch?v=pRyat3GR7Xc

Experiment 25
same as 24 but 1 coil is short circuited while pulsing the other

http://www.youtube.com/watch?v=MaQlZctApQg

Experiment 26
Pulsing a monofilar coil phase delayed with ferrite rod

http://www.youtube.com/watch?v=haylq5NAe_U

Experiment 27
Looking at psu voltage to see how it effects the pulse

http://www.youtube.com/watch?v=EPtROTLJ3Jc

Experiment 28
1 meter wire looped on bench then scanned

http://www.youtube.com/watch?v=O-zDZOnJ0Is

Experiment 29
same as 28 but using solid core wire.

http://www.youtube.com/watch?v=sH-XCD04kPM

Experiment 30
same as 29 but using litz wire.

http://www.youtube.com/watch?v=EchXggR4DQM

Experiment 31
Switching back to OTPU Former

http://www.youtube.com/watch?v=clUf7vQeVzA

Experiment 32






The Explosion in a wire Video
A Monofilar coil was connected and driven from 2 fets, 1 fet was pulse delayed by 151nS, The explosions could be heard from 5 feet away from inside the air core 22 turn coil, the coil was being driven from a low voltage power supply.

Hey Peter,

I don't see a link to a video.

EDIT:
I see it now.

I had similar "explosive sounds" between teh termionals of my ignition transformer when running a dual series spark gap (10kv).  The was a loud detonation sound about every 15 seconds at what appeared to be steady intervals.  Sounded like a rifle shot.  I placed a piece of 0.25 inch thick polyethylene between the terminals and the noise ceased.  The loud sound was not accompanied by any sort of light or  flash, just the sound.

Hi Grumpy

I am only using 30 volts as a psu source to drive the coils. the noise was inside the wires, i havnt posted this video yet, but it was up on OU at some point so you may have seen it already.

Yeah i am just starting to post some of the videos, i think there was about 50 of them in total but many were duplicate experiments, so it's going to take me a while to weed out the rubbish, so may end up with about 10 or so

Peter

Quote from: Peterae on 2009.12.19, 14:20:13
the noise was inside the wires

interesting

if you have time to reproduce this, wrap some sort of tape aroung the coil and see if the noise goes away or changes.

Hi Grumpy

i used my fingers to hold all 22 turns in place to see if i could alter the loudness of the noise and it didnt change, until i pressed with all my might and then it stopped totally until i altered the delay phase to find the spot again.
I went on to build another controller this one was in a thick metal can.
The idea was to try and stop the RF noise falsely triggering my fets, and this worked to a degree, but the problem was that i couldnt reproduce the effect anywhere near so good.
So it would appear that a sequence of random pulses from my fets would every now and then hit the right sequence to produce the effect.It is worrth looking carefully at the yellow trace this is the current from the psu, and see when an explosion occurs and how much current is drawn.

If a magnet was placed near to the coil, even without the explosions occurring i could hear hissing and low level crackling from the wire.

One other thing i did do another video of the explosions that didnt work, although i could clearly hear the explosions from the wire as before, the electret mike was placed to close to the coil and it was effected so much being close to the coil, it didn't work properly.

Peter

put a single layer solenoid over your bifilar coil and apply DC to it.   

Just Recently i have been studying something that blows the crap out of my fets, and it's getting very interesting if not expensive.

The problem occurs when i use my controller to pulse a piece of wire, i have tried this wire in different configs and it seems to happen most of the time.

The other day i was pulsing my Don smith primary. i had no flyback diode, i was pulsing with 40nS pulses using 2 fet channels with about 100nS phase difference, and did a sweep of the frequency, the current was minimal but was climbing slowly as the frequency increased, until i hit a certain frequency and suddenly the coil started drawing a 10 fold current, luckily i was scanning the frequency quiet fast and it went through this high current setting and came out the other side without damaging the fets, nearly wrapped my moving iron current meter round the end stop though LOL.

So i went back slowly and reduced the psu voltage to the coils and there it was again current suddenly rose 10 fold again and this time it destroyed both of the fets.

Now i have seen this a few times before, when i found the big pulse before, i would find a phase delay setting that would cause my cpu to crash and cause the fet stages to trigger without the fet driver activating them, this caused more current to flow because the on time had increased, this lead to the crackling in the wire experiment i videoed.

Today i made a non inductive solenoid coil Tesla style, i set my sig gen to 40nS and master frequency to 1khz, i scanned the phase, starting at 0nS, now at 1khz 40nS turn on time on both channels draws under 1mA, as i scanned up in phase i hit 1012nS and my current meter wrapped around the end again, my cpu crashed and both fets were instantly damaged, both read short circuit but were not warm to the touch.

Now there is something going on here, there seems to be a point where and em or rf burst emanates from the setup and is causing both fets to turn on hard and then sure destruction occurs, but how can such an em or rf field appear when i am driving with less than 1mA.
Tomorrow i will replace both fets and use resistors in series with the supply to try and limit the peak current to the coils.

Sounds intresting....

I would start to drive with tiny batteries that are not able to destroy the mosfets.  :)
Or even try it with some thermoionic valves :)
Did you measure if it is a case of retriggering in this case?

Marco.

Quote from: Peterae on 2010.01.12, 18:50:02
Just Recently i have been studying something that blows the crap out of my fets, and it's getting very interesting if not expensive.

The problem occurs when i use my controller to pulse a piece of wire, i have tried this wire in different configs and it seems to happen most of the time.

The other day i was pulsing my Don smith primary. i had no flyback diode, i was pulsing with 40nS pulses using 2 fet channels with about 100nS phase difference, and did a sweep of the frequency, the current was minimal but was climbing slowly as the frequency increased, until i hit a certain frequency and suddenly the coil started drawing a 10 fold current, luckily i was scanning the frequency quiet fast and it went through this high current setting and came out the other side without damaging the fets, nearly wrapped my moving iron current meter round the end stop though LOL.

So i went back slowly and reduced the psu voltage to the coils and there it was again current suddenly rose 10 fold again and this it destroyed on of the fets.

Now i have seen this a few times before, when i found the big pulse before, i would find a phase delay setting that would cause my cpu to crash and cause the fet stages to trigger without the fet driver activating them, this caused more current to flow because the on time had increased, this lead to the crackling in the wire experiment i videoed.

Today i made a non inductive solenoid coil Tesla style, i set my sig gen to 40nS and master frequency to 1khz, i scanned the phase, starting at 0nS, now at 1khz 40nS turn on time on both channels draws under 1mA, as i scanned up in phase i hit 1012nS and my current meter wrapped around the end again, my cpu crashed and both fets were instantly damaged, both read short circuit but were not warm to the touch.

Now there is something going on here, there seems to be a point where and em or rf burst emanates from the setup and is causing both fets to turn on hard and then sure destruction occurs, but how can such an em or rf field appear when i am driving with less than 1mA.
Tomorrow i will replace both fets and use resistors in series with the supply to try and limit the peak current to the coils.

is your power supply current limited?

can you add a lot of resistance or a choke to limit it?

energy comes from the environment in everything - all electrical circuits - magnets - everything - you are just releasing it.

deliberately tune off the peak and you cna explore it better

move the fets far from the coils and shield with thick plastic  (like cutting boards)

Hi Grumpy

The fet modules are shielded in metal boxes anyway.

now thinking about it the power supply is a variable 0-125v 0-1 amp i was running the above tests at 25v so the most it could deliver i 1 amp, i am using irf840a which is an 8 amp device, although i dont use a heat sink because it was designed to pulse at a max of 63 nS which is the max setting of my controller, so never really thought i would be turning on long enough to generate heat.

It appears to be something other than current that is damaging the fets? the only thing i can think of is a very high voltage pulse which is at a higher voltage than the device can manage.I am using fast flyback diodes on this setup.

Peter

Hi Peter,

interesting stuff! The last vid sounds very energetic. Does what comes through on the youtube vid reflect what you hear by ear fairly well?

the fact that it works with monofilar seems to suggest it might be some remenance  effect of the aircore, maybe first pulse stretches the net, second pulse kicks some free stuff through at just the right time.

I presume you will be trying to collect through coupling perhaps and measure OP. it's going to be real hard to analyse the in/out powers of these chaotic signals, probably easier to try and harvest through diodes into cap and watch its V risetime. while at the same time monitoring supply power through filters. But perhaps your scope has the power to grab and calc true RMS powers over non inductive shunts? I suppose your Hantek could always export the data for deep analysis on computer.

Quote from: Peterae on 2010.01.12, 20:40:01
Hi Grumpy

The fet modules are shielded in metal boxes anyway.

now thinking about it the power supply is a variable 0-125v 0-1 amp i was running the above tests at 25v so the most it could deliver i 1 amp, i am using irf840a which is an 8 amp device, although i dont use a heat sink because it was designed to pulse at a max of 63 nS which is the max setting of my controller, so never really thought i would be turning on long enough to generate heat.

It appears to be something other than current that is damaging the fets? the only thing i can think of is a very high voltage pulse which is at a higher voltage than the device can manage.I am using fast flyback diodes on this setup.

Peter

a metal box is an EM shield - try something dipolar and thick - another option may be a double layer Faraday cage and keep an charge between the two layers (like a capacitor - this puts a dipolar field around the fets - Hell just through some palstic and aluminum foil to make a quick capacitor tube around them.

Total "current" is the sum of all types of current - and there are several types such as polarization current, displacement current, convection current, etc.

FET's can fail if the oxide layer at the gate is exposed to over 10v (something like this - I can't recall the details)

do you have any transils?

overvoltage - like avalanche would not damage the FETs unless there was a lot fo current

so, the FETs just measure as shorted and no visible signs of damage?

EDIT:

The gat oxide failure is called SEGR (single event gate rupture) and the other mode of failure is called SEB (single event burnout) and "shorts the source to the drain"

SEB is typically caused by energetic ions, cosmic rays, etc. where the energetic particle path casues secondary breaksdown (multiplication) in the device leading to failure.

I am not saying that you have solar flare or cosmic ray particles flying around, just that you may have this failure mode.

EDIT2:

shield with mica board - it scatters the stuff (it's bifringinent)

I hear you about zapping FETs Peter, after a while you can get through quite a bit of cash whis way, it all adds up!

This is a VERY nice pulse amplifier scheme:

http://accelconf.web.cern.ch/AccelConf/e02/PAPERS/THPLE025.pdf

Motorola MRF182 FETs swing the grids of YL1056 Tetrode tubes. The novel thing about this design is the 2 parallel stages, basically one stage does the overall shape (where the bulk power is) and another stage gives the high freq part to give rise and fall times of 2ns. The stages hi and lo cuttoff freqs are matched to roll together so that the response is flat from DC to about 50MHz.

Of course the above is a very powerful amp at 5kW RMS and we could redesign using cheaper lower power parts (we could use 1 nice FET and two modest tetrodes per channel) and I´m guessing a ballpark figure of €50 per channel to get <=5ns rise and fall at near 50W RMS per channel, of course peak pulse currents could go much higher thanks to the tubes.

A very nice amp for the bench! can reproduce anything you throw at it from any high impedance source like an IC or micro output etc. Just one amp to cover everything and having a tube on final would make all silicon before it pretty safe. The ultimate coil banger amp!

edit:
I realise this would be a project in its own right, probably take a few months (on and off) work in a dedicated thread.

Thanks for the suggestions Grumpy, i will measure the fet and report what and where the shorts are, the fet failed cold instantly with no smell or sign of destruction anywhere.

I am not going to be able to get to my bench at work as i am snowed in again, we only got back to normal yesterday.

@Fraser
Thanks for the Valve stage, i was talking with marco about building one yesterday, so this doc will come in real handy.

The Explosions in the wire experiment was done a long time ago now, the noise in the wires did sound like explosions, and could be heard upto 5 feet from the bench with the ear, it was very difficult to use a microphone to record the sound, if it got too close it stopped the microphone working, interestingly it was very hard to set the experiment up to get the noise that loud, many times i had the sound but it was quieter but if a magnet was placed near the coil the noise would get louder and you could feel the explosions in the magnet if it was held by the hand.
You couldn't see the coil physically move, if enough pressure was applied to the coil you could stop the noise, but in other experiments i had done, i think this was because each coil turn had to be exactly spaced for the effect to occur at a set phase delay, alter the coils spacing slightly by applying pressure would lead to the effect occurring at a slightly different delay phase setting, from this observation, if this was indeed the kick then i was begging to understand why SM had glue everywhere, it was to stop movement of each coil turn, otherwise he would have had to remake his iron delay coil for a different delay setting everytime the coil was knocked slightly.

EDIT: oh something else that was interesting was, i had run the test and made the video in the evening and the left it for the next day, but when i powered up in the morning it wouldn't make the noise, but when i came back from work later in the day to have another go it started fine, i am annoyed it didn't try this a bit more, afterward having thought about i wondered if the effect was only occurring at night, but didn't do enough tests to draw a conclusion on this.

Anyone trying to replicate this, you get a good indication if you are close by holding a magnet close by and listening for the crackles, then you can adjust your parameters and try moving the coil further away to see if the effect is stronger.


Lets build this valve output stage, i need one badly right now.

Peter

I wanted to add that "vacuum polarization" causes changes to electrons shells that are several orders of magnitude larger than "radiation reaction" (which are usually small).

Before you discount this as "unrelated":


http://physics.fullerton.edu/~jimw/general/radreact/

and

http://cdsweb.cern.ch/record/966452/files/PhysRev.95.1048.pdf

On your dieing FETs....

Take at look at my diode text pic under my workbench-reference material. If the FETs have silicon in them you could be seeing that effect during reverse bias.

Hi Peter,

i saw your post on overunity.com; very interesting sound / phenomenon.

Did you ever found out how your FETs got destroyed?

Quotei will measure the fet and report what and where the shorts are, the fet failed cold instantly with no smell or sign of destruction anywhere.

Regards Itsu

Hi Itsu
They always failed cold, i think because it was all related to the anomalous pulse having a very sharp steep rise, i might of been creating an infinitely high voltage, and it was this voltage that was causing breakdown of the fet, in most cases the average current draw was only a few ma but suddenly the fet would fail, i should go back and try the new 1700V SIC's which is what yfree wanted me to do, you will notice that my yellow ring infact resembles a McFreey ring but without the acoustics.

I always had a feeling there was a magic series of harmonics that make unlimited power appear, hence the switch to whitenoise, the idea being once in a while the right random harmonics would release the power and my idea pretty well proved true because an anomalous 80 watt pulse would randomly appear as suspected, it only showed using copper plated iron, i tried pure copper ring and never could see the 80 watt pulse.

Cheers
Peter

Did you ever place a magnet near the coil while the explosive noise was occurring?

yes i have a video somewhere where i made a bifilar flat pancake coil glued down on a cd, i then start the crackling and throw a small square neo magnet on top, it vibrates to the clicking noises, also if a neo magnet is held in the hand near a crackling coil you feel the magnet tugging as the clicks are heard.

EDIT found the video
https://www.youtube.com/watch?v=sehl2bW_Dpg

I wonder why the magnet always gets thrown outwards as if theres centrifugal force acting on the magnet.??

I think a radial force created. 

While magnets responded with the greatest reaction, I found that pieces of plastic and conductive materials would be tugged toward the Brooks coil that I used.

Fascinating Peter,

i have checked all the video's, but i fail to see what the setup is.

You mention in post #1:   "pulsed Bifilar Coil using a variable Width and phase controller".

Can you be more specific on what you used and how it is set up?
I understand you have 2 MOSFET's being driven out of phase, but with what? ("a variable Width and phase controller"  what is that? Homemade?).
Do you use MOSFET drivers?  Is there a diagram of this setup?


Sorry for the questions,    Thanks.    Regards Itsu

Quote from: Itsu on 2015.09.17, 11:05:02
I understand you have 2 MOSFET's being driven out of phase, but with what? ("a variable Width and phase controller"  what is that? Homemade?).

Your DG4162 or DG 4102 is such a controller.

Quote from: Itsu on 2015.09.17, 11:05:02
Do you use MOSFET drivers?  Is there a diagram of this setup?

Good question.
I think you'd need to use MOSFET drivers because it is impossible to quickly pulse a MOSFET gate in nanoseconds from a signal generator only ...even if it is one of these RF MOSFETs with pF gate capacitance.
Also, a MOSFET driver constitutes an additional barrier and protection from these killer pulses.  You do not want your signal generator to be damaged by one of them.

I wonder if the magnet moving to the outside is related to the virtual inertia effect that Harold Aspden discovered.  (First article in attached pdf)

Hi Itsu & verpies

Yes for the delay i use a digital monostable i designed using 255 tap 1nS delay chips, 1 chip delays 0-255nS the other creates the pulse width drive to the fet driver 0-255nS, there are circuits in my work bench.
I do indeed use a fet driver again circuits are in bench i will dig them out tomorrow, off to bed now ;)

PS i would not use any expensive equipment to test my setup, emp pulses are there for sure and do damage, even my scope got damaged at one point, luckily the manufacturer told me what i had to change.

There was a video some years ago made by JDO who designed a DDS generator that Giantkiller uses, anyway he did a scan and it blew up, for the same reason you see in my videos as i understand it.

This was the first digital monostable i designed, it used 2 8bit banks of dip switches to set the delay and pulse width, i also built a little controller which plugged in place of the dip switches so i could sweep in nS with this board, that can be seen with the 7 seg display in the early videos above.
http://www.overunityresearch.com/index.php?action=dlattach;topic=217.0;attach=1209

you can see my typical fet driver circuit in this diagram, i always use a 18V regulator to supply the fet driver, i used MCP1406 & 1407 drivers as they turned out to be better than the TC4429 in the diagram.
http://www.overunityresearch.com/index.php?action=dlattach;topic=217.0;attach=1215

To ensure the safety of expensive signal generators an isolated gate driver (http://www.analog.com/en/products/interface-isolation/isolation/standard-digital-isolators/adum3220.html#product-overview) can be used.

Optos are too slow for this purpose but other companies make fast RF and capacitively isolated (http://www.low-powerdesign.com/130509-article-pin-compatible-digital-isolated-gate-drive.html) drivers - they are suitable for this purpose, too.

Quote from: verpies on 2015.09.18, 16:14:30
Optos are too slow for this purpose but other companies make fast RF and capacitively isolated (http://www.low-powerdesign.com/130509-article-pin-compatible-digital-isolated-gate-drive.html) drivers - they are suitable for this purpose, too.

Thank you Verpies for the link.  I never knew such devices existed.  Coupled with DC/DC converters to keep the power rails isolated from transients, these capacitive isolators hold great promise for much higher frequency drive boards.

Here's a schematic that shows the early setup for explosions/crackling
http://www.overunityresearch.com/index.php?action=dlattach;topic=31.0;attach=11221

Schematic of fet and driver used
http://www.overunityresearch.com/index.php?action=dlattach;topic=31.0;attach=11223

Video showing basic micro changing the dip switches and thus sweeping in 1nS steps
https://www.youtube.com/watch?v=Ub5XhPJ2Y3A#t=27

Video showing thin enamelled copper wire jumping around in the presence of a magnet, quiet a complex coil wind
https://www.youtube.com/watch?v=O8GgiEeH1W4

Some interesting tests and data
http://www.overunityresearch.com/index.php?topic=136.msg1090#msg1090

I'll echo Matt Watts thanks to Verpies as well!

The Low Power Design page has a link to their numerous
Articles (http://www.low-powerdesign.com/designarticles.htm#ALT) which is indeed a Treasure Trove of Technical
Gems!

A Seeker of Knowledge could spend hours, if not
days, studying the various Design Articles on some
very interesting subjects.

There is also a list of White Papers which may be
read/downloaded (http://low-powerdesign.com/SiLabs-PDFs.html) following a short registration.

Quote from: Matt Watts on 2015.09.18, 19:11:37
Thank you Verpies for the link.  I never knew such devices existed.  Coupled with DC/DC converters to keep the power rails isolated from transients, these capacitive isolators hold great promise for much higher frequency drive boards.

Actually, for a while now, I've been working on a universal MOSFET driver based on these devices and DC-DC converters, for the CARA circuit and its synchronous rectification.

Quote from: verpies on 2015.09.18, 21:19:56
Actually, for a while now, I've been working on a universal MOSFET driver based on these devices and DC-DC converters, for the CARA circuit and its synchronous rectification.

Hopefully you saw my attempt of a universal driver (http://www.overunityresearch.com/index.php?topic=2594.0).  With your help I'd be happy to upgrade this to something everyone can take advantage of.  Probably use some sort of pluggable MOSFET arrangement for driving whatever load is needed.  May need to think about how best to drive something like SIC MOSFETs that pull the gate below zero.

Quote from: Matt Watts on 2015.09.19, 15:50:58
Hopefully you saw my attempt of a universal driver (http://www.overunityresearch.com/index.php?topic=2594.0).  With your help I'd be happy to upgrade this to something everyone can take advantage of.  Probably use some sort of pluggable MOSFET arrangement for driving whatever load is needed.  May need to think about how best to drive something like SIC MOSFETs that pull the gate below zero.

Good day All
Couldn't resist posting some pic.s of my SiC_Pack.  It is designed to pulse Six Cree C2M0080120D mosfets in Series.  This is the equivalent of a 7kV2 switch, with complete galvanic isolation (GDT with 12kV2 30awg magnet wire).
Still in testing phase, but working extremely well. Have attached some scope shots as well.  This was primarily designed for nano-second pulsing, but as you can see it will (with the appropriate GDT) produce a clean 50% duty cycle as well (30ns rise/fall times at the gate).
Generally I can obtain <1ns skew between channels (mosfets) at anywhere from a 50kHz to over 2MHz range.
I am currently using an Arduino based signal generator which interfaces with a hardware hacked AD9850 dds. The hardware hack allows for control of the duty cycle of the AD9850 dds by the arduino as well.
The gate signal floats a -4 to +20 volt signal. The positive offset is adjustable on the fly and the negative offset is part of the design (must change out components to modify).
This was designed as part of another project which requires HV nano-second pulses.

take care, peace
lost_bro

Very impressive lost_bro  O0