This is it !

[gmeast]

I did the calculations for the resistor and it comes out to 60 ohms... It literally is resistance (power into heat) that would be provided by the driver if it wasn't provided by the resistor between the out and the MOSFET gate. Literally the resistor is meant to turn the power into heat before it gets into the driver and turns to heat there, burning up the driver.

How are you going to find a transistor that can switch fast enough to use a transistor instead of resistor to dissipate the MOSFET gate charge? I think that would end up being squishier than using the resistor.

The specs on the sheet for switch times take the resistor into account, I'm sure, otherwise they would burn them up in testing.

Hi 7imix,

I agree. A resistor of 60 Ohms is a very 'stiff' mother. Thanks for spending time on the calc's. Like the sheet alludes to, it's 'meant' to help with heat dissipation. A 60 Ohm resistor will degrade the rise and fall times minimally, but I bet it will benefit in the IC's cooling. I wonder if there is a 'best specie' of resistor for this ... there must be. Either way, this group will assuredly test all possibilities and arrive at the best choice.

We now have 16 times our base resolution (tested). I think we may have 256 times our base resolution soon. I'm shooting for that, but that extreme may be unnecessary.

Later

[Farmhand]

Hi 7imix, I believe what you say and i'm wondering how the transistor in the driver chip can switch so fast.

Anyway 60 ohms sounds pretty good. Maybe 47 ohms will do.

Cheers

[7imix]

Here's my current build of a single phase pulser. It's producing really nice pulses. I want to add a couple of transient suppressors before I hook it up to a coil though.

[7imix]

By the way, my driver is getting hot with the out to gate resistor in there, but it hasn't burned out... Yet... :-)

[gmeast]

resistor

By the way, my driver is getting hot with the out to gate resistor in there, but it hasn't burned out... Yet... :-)

Hi 7imix,

After all of the resistor thing, I have been advised by a very reliable source to nix the resistor ... don't use it. And we though we were smart. Sorry.

Later

[Farmhand]

Another way ?

Hey guys I was thinking there must be another way. If the driver cant do it, that is. Another way to ground the gate charge.

I would have thought originally it would have been grounded back through the driver chip. But now i'm not sure.

Oh well, I have some small powdered iron rings ( for choke test ) but i'll have to surgically remove them from thier host circuit. Shouldn't be too difficult.

I'm down one board since one is a dedicated desufator now so i'll whip another one up on a solderless board.

Ummm OK here's a silly idea, what about a Quad input logic chip to ground the gate charge. Maybe connect the gate to one input the ground to the output and a very small cap between the other input and the out of the driver with a signal diode between the inputs to the logic gate. Make any sense at all ? Feel free to tell me it make's no sense, I won't be offended.

Crazy huh.

Cheers

[gmeast]

finally finished

Ok everyone,

I finally finished it ... my toroid. Each primary has 50 turns ... and I don't like the length of my leads ... too short I think ... about 12". Now the secondary has 116 turns, and by the formula for Hydroxy, I should only have 39 turns. It's easier to UNwrap turns than it is to wrap turns ... if I need to.

Here:
http://02d1852.netsolhost.com/radian...hed_toroid.JPG

Later

[FRC]

Did you See

Did you see Aaron shutting down the Water Sparkplug thread and starting a new Water Sparkplug 2. Seems someone has patented his design also. So beware how much info you give out.

FRC

[Farmhand]

Mine's ready to roll !

OK my Donut is iced ! I used plenty of wax a mm or two covering the windings for bump protection.

http://9xfdag.bay.livefilestore.com/...005.JPG?psid=1

Don't have any more transformer tape yet, so I just wrapped it with insulation tape for now.

Cheers

[gmeast]

information

Did you see Aaron shutting down the Water Sparkplug thread and starting a new Water Sparkplug 2. Seems someone has patented his design also. So beware how much info you give out.

FRC

Hi FRC,

Thanks for the advice. Some details will be kept in close company for now, and besides 'generating or self-charging electricity' has already been proven and shown by Bob Boyce using this method.

Everyone please remember:
This thread is about a replication and it has turned into an effort to 'replicate' using different circuitry, but really doing the 'exact same thing' with ONE variable being changed ... the controller ... the controller is a sub-set. If it doesn't work out, then we'll use a microprocessor. BOB is 'protecting' his controller from thieves and preventing 'stupid' people from hurting themselves.

I was completely unaware that this forum is somehow a watchdog to prevent original thinkers from going ahead with intellectual protection. There's nothing wrong with protecting something you have worked hard on and that involved original thinking. A patent only gives an inventor a time-limited monopoly to commercialize their 'gadget'. If a patent remains 'un-acted on', that patent CAN be taken away and given to another person who WILL 'act' on it. At the same time, ANYONE can build ANY patented or unpatented invention for their own use ... they just cant build and sell any. It's always been that way.

A big energy provider like a utility company that may 'want' to steal and patent 'your' creation cannot block ANYONE from personally using it because they somehow want to force you to buy their electricity. If that were so, then there would not be any solar powered or supplemented homes or businesses ... and YES ... it seems at times like some local ordinances might be playing in the hands of 'special interest (utilities)', so those must be examined on a case-by-case basis, I guess.

When I was installing Solar in San Diego, the (then) SDG&E tried to get the PUC to approve an "EXIT FEE" for anyone going on solar. The fee would have been a huge amount of $$$. SDG&E justified the 'fee' because they said ' ... that's revenue that would have been OURS from that rate payer ... and now we're not going to get it anymore...'. WTF! The PUC laughed in their collective corporate faces ... actually to everyone's surprise, but thank goodness anyway (a PUC is a double-edged sword in my opinion).

Anyway, sorry for the RANT. These issues get me going. Yes, thanks FRC ... we'll be careful.

Later

[gmeast]

iced Donut

OK my Donut is iced ! I used plenty of wax a mm or two covering the windings for bump protection.

http://9xfdag.bay.livefilestore.com/...005.JPG?psid=1

Don't have any more transformer tape yet, so I just wrapped it with insulation tape for now.

Cheers

Hi Farmhand,

Yours is much prettier than mine ... dammit!

Later

[Farmhand]

Haha lucily for me looks aren't everything, if things needed to look pretty to work not much of my stuff would be operable. I figured since I was using insulation tape I might as well make it look like a dohnut traffic light.

Green coil is first yellow second and red third, I am hoping the red bit makes it go a bit faster.

I take it my crazy logic gate idea for grounding the gate charge make's no sense. I'm not surprised it is "out there " a bit. Haha


Some pics of the desulfator I boxed up. This is the smallest amont of power I could achieve and still have an effective desufator. Thanks to Seamonkey for the circuit. Much appreciated.

Input current
http://public.bay.livefilestore.com/...001.JPG?psid=1

Output current, Output current on a drained battery is more 55 Ma. Very efficient circuit.

http://public.bay.livefilestore.com/...005.JPG?psid=1

Waveforms. 10 x probe 0.5 volts per div.
http://public.bay.livefilestore.com/...002.JPG?psid=1

http://public.bay.livefilestore.com/...006.JPG?psid=1

Circuit in box. I wound the coil up to 750 uH.
http://public.bay.livefilestore.com/...006.JPG?psid=1

Cheers

[gmeast]

Haha lucily for me looks aren't everything, .................................................. .................................................. .................................................. ......................................
I take it my crazy logic gate idea for grounding the gate charge make's no sense. I'm not surprised it is "out there " a bit. Haha

.................................................. .................................................. .................................................. ..................................

Cheers

Hi Farmhand,

Nice boxed circuit. On the gate grounding: I'm not sure I understand. Are you saying use a transistor run by the gate in synch with the MOSFET gate driver to help it along in getting rid of the gate charge? ... uuumm ... did I say that right? There would not be any component conflict I don't think. Is this what you meant? Dealing with JUST the grounding would reduce the load to the driver IC by half ... I think ... if fast enough.

Did you take note of a very important statement in the PM to me ... "...if the [{(*"'CHOKE'"*)}] is working properly ... " ME NEED INPUT ON CHOKE THINGY ...please!

And also Farmhand, I fiddled with 3 and 4 phases today. I can handle just about any 'pause' of any duration between any of the phases. This is so sweet. As you noted ... power density and frequency are directly related, therefore I think I need to up the circuit density too so we can slow WAY down for low power testing, yet maintaining the short, sub-saturation pulses to the coils. This is important I think ... when we lower the frequency, I don't think we want to increase the pulse width. At some point we will begin to saturate the coils/core and that, I think is a no-no ... on the 'edge' ... maybe ... over the edge ... no think so.

@7imix: if you have the time could you show me what you would do to the SR circuit to turn it into a NAND gate circuit ... or did I misinterpret an earlier statement you made? I'm planning a final circuit now. What I am going to do is design around the concept of a "master" Logic Module with Phase Modules that 'plug' into/onto a parallel digital signal bus. It can expand to 'n' modules (using buffers past a certain number of phases). Maybe the phase modules should all have buffers to begin with ... I think so ... but there's no problems with the four I have right now ... gotta' think on that ... it's theoretical design vs practical design ... and power considerations too ... buffers use power.

Thanks in advance 7imix ... if you can find the time.

Farmhand, you made an indirect suggestion when you said "... the driver will work out ... ". The driver is a circuit variable. All that circuitry can be independent from the 'physical board' (maybe). If we can come up with a properly designed driver circuit then we can make that part of it on its own small chunk of PCB with proper ground planes on the correct sides of the PCB, reaching just to the right place(s) under the IC like the app notes say. Anyway, I'm just thinking at a hundred miles per hour right now. I'll shut up.

Later guys.

[Farmhand]

No thats not quite what i meant there i'll have to make a drawing. I'm terrible at explaining things, half the time I don't even know if i'm using the right terms.

7imix is right the transistor to ground the gate charge does make the spikes squishy at high frequency definately, I looked and seen it. But at lower frequencies it's very good. My idea is to replace the PNP transistor with a logic gate, same principal. The output of the logic gate to where the emitter of the PNP go's (ground), one input to where the base of the PNP usually go's(PWM out) the other input to where the collector of the PNP usually go's (the mosfet gate). When the pulse hits both inpus it's grounded. Not sure if there is a type of logic gate that will do that or if it would work.

Won't be long , i'll edit the drawing it in here.

Like this, if this won't work there must be a componant that can do that up to Mhz. 7imix may have already mentioned it but it's hard for me to know when I don't understand everthing thats said. If so sorry .

http://koynpq.bay.livefilestore.com/...001.JPG?psid=1

Umm looks like I left out part of the internals of the mosfet.

Cheers

[7imix]

Greg, sure, I'll rework the logic to use NAND.

About the gate charge thing, I wonder if putting a heatsink on the dip packaged version would help.

Farmhand, I just don't know enough about the idea of using a transistor to drain the gate charge to comment.

It would probably work with a high speed transistor. Whenever the MOSFET is off, the gate drain transistor should be on.