Archive for the “Mad Scientist – Boo” Category

Any technology, sufficiently advanced, is indistinguishable from magic.


Comments 1 Comment »

Here is a design for an overhead light fixture (see the Creative Commons license notice below).

I’m not really sure I’m going to save any money with this.  I’m spending a lot for a very nice lamp, but economically, I’m not sure the energy and light bulb savings will be worth it.  But then again, that’s not the only reason to do anything.

This lamp will replace the aging standard fixture in the ceiling of my study in a house built in 1965.  This existing fixture is made for two Edison base light bulbs, and switched at the wall near the entry door.  Bo-rrring.

The idea is to stop the 120V wiring at the switch.  There I’ll put in a small DC power supply and a circuit inside the wall box with a special wall plate for the switches mounted on the PC board.  A three wire cable (ground, 24 volts DC, and a 0 – 5 volt square wave) will connect to the overhead fixture.  The lamp will be made of 6 pentagonal printed circuit boards connected together on the edges like the lower half of a dodecahedron (12 sided solid).  Each board will have 5 Cree XLamp MX-6 high-brightness LEDs, 6500 Kelvin (cool white color), for a total of 30 LEDs.

Read the rest of this entry »

Comments 2 Comments »


I am beginning to have need of additional lighting in my house.  A few years ago, my eyes started to show age, and this is not a trend I expect to reverse itself.  It’s the 21st century, I’m bored with snapping on and off light switches as I move from room to room, so I’m building a few lamps I can use.  These will be especially helpful in stairwells, but also in my basement.


  • Motion activated – this is pretty easy, if nothing in the area is moving, it probably doesn’t need to be illuminated.  This also implies adjustments for sensitivity (will it trigger for me, a cat, or a mosquito) and delay (how long does it stay on after triggering).
  • Ambient light sensor – no point turning on a lamp when there’s already enough light
  • Energy efficiency – uses pennies worth of juice each year.  Ideally, this works during a grid power outage, and/or is completely self sustaining through auxiliary power generation devices like solar panels, a micro-windmill, or a windbelt humdinger.  It should be bright enough to keep me from tripping over laundry, not for reading.
  • Solid state – no maintenance required other than occasional dusting.  Zero recurrent costs.

Create a device which triggers one or more high-brightness white light LEDs.  Device will use low voltage DC power supply, centrally located to supply multiple devices throughout a house.  Device will be installable using a circular hole saw through existing drywall, power supplied through in-wall wiring.

Read the rest of this entry »

Comments 3 Comments »

I put another quart of Slick 50 in my car’s engine.  It’s a brand of motor oil that is saturated with polytetrafluoroethylene.  I did it when I got the car at 10K miles, and here I’ve done it again at 67K miles, which is about right.  I’ve always reapplied it at 50K mile intervals, and I think that’s what is recommended.  I drive a vanilla 2001 Volvo S60, no turbo.  It is blue.  No, it didn’t come with an “Abba” CD.

Basically, you change your oil… drain the old oil, replace the filter cartridge, and put in the right amount of new oil minus one quart.  Then with the engine running, you pour in the quart of Slick 50 oil, put on the oil cap, close the hood, hop in, and drive for about 30 minutes.  This shears the Teflon and the rest of the oil together, and the Teflon gets burnished into the metal of the inside of the engine.  I’ll let it ride for a regular oil change interval (which in my case is like 6K miles), then replace the filter and oil again.

I’ve been using the stuff for 30 years.  I’ve never not been impressed.  Back when I had a pretty old car (in the 80’s I had a 1968 Ford Mustang with a 289 cid V8), the effect was just shy of miraculous, and I’ve got friends who also have seen pretty amazing results.  The internet is full of claims and counterclaims on Teflon motor oil, I’ve always only used the original Slick 50.  I also looked up the quoted magazine article by ‘Consumers Digest’, using a microfilm reader as this was before the invention of the World Wide Web.

I did the research, I gave it a try, and it’s terrific.

I do believe in the engine additive self-fulfilling prophecy… when you put some snake oil product into your car, you unconsciously alter your driving habits in tiny ways which make the claims of the product come true.  Get some macho guy pouring a pint of DDT Carburetor Ointment in your gas tank and hell… it’ll even make your girlfriend’s tits seem bigger.  I don’t use a lot of that junk, all I’ve ever seen any of it do is glop up my spark plugs, and I don’t need a pseudo-sexy sports car to get happy.  (Wait, maybe I do.  I saw a ’68 Jaguar on the road recently.)

Okay, so it seems to me that on my 30 mile 75MPH drive I didn’t have to push down the gas pedal quite as far as I previously had to push it to go the same speed.  Nothing definite, nothing scientifically provable, no… I didn’t measure the amount of travel with a micrometer… it just didn’t seem like it needed as much gas to go.  Improved gas mileage has been the most noticeable effect for me, the Mustang got 21 MPG before and more like 25 MPG after I used it (yes, back in the early 80’s I got better mileage than I do now).

I’m not selling the stuff.  I don’t care if there are tons of articles on the ‘net that say it can’t possibly work.  I’ve used it, I know it does, so there, neener neener nyeah.  I’m saving my gas receipts and writing down the mileage, so I’ll have some hard data on the gas mileage.  There won’t be any way to prove that the oil is what caused it, but *I* will know.  I’ll post the results here when I do the math.

Drawbacks?  I sometimes wonder if my engine will ever wear out so I have an excuse to buy a new car.

To the makers… Gentlemen, thank you for this wonderful product, once again, it’s paying for itself.  For me.

Comments 1 Comment »

Notes on phase-lock-loop circuitry of a Meyer resonant water fuel cell

Here’s the schematics.  They are improving daily.  Note: all POV-Ray images on this site are just for laughs and are inherently inaccurate.


“ZeroFossilFuel” has a fabulous idea in this video: (also see ).  I have not tried this yet, but the more I look at it, the more I like it.

Picture Right: Lawton style gated pulse circuit, 4″ toroid, 3/8″ ferrite rod, E-core ferrite, new meter

Referring to WO 92/07861 ( ) , the ‘A27’ chip in Figure 7 is equivalent to a 4046 PLL chip. The write up on Figure 8 is particularly interesting. Note that ‘A31’ chip looks like a 555, there are many of these throughout the drawings. “The scanning circuit o Figure 8 scans frequency from high to low to high repeating until a signal lock is determined.” : it runs a siren! I had thought about this as a way to initially discover the resonant frequency on start up. The next statement is incredibly enlightening: “The ferromagnetic core of the voltage intensifier circuit transformer suppresses electron surge in an out-of-resonance condition of the fuel cell.” The pulse monitor tap on the VIC circuit toroid in Figure 1 gets cleaned up like in Figure 9 and feeds the PLL control.

Figure 1 Figure 7 Figure 8 Figure 9

I was having trouble figuring out how to pickup and feed back the ‘resonance’, this is helping a lot.  The original Puharich design also addresses this.

09-AUG-2008 The preliminary circuit design is nearing completion.  I need to get breadboarding!

12-AUG-2008 I’m getting some help on the forums at .  Files and schematic picture updated.

17-AUG-2008 I had to split the schematic into two parts, the frequency counter is now on it’s own board.  The free version of Eagle wouldn’t handle one big board, not that I could handle it either.

25-AUG-2008 Files updated.  There are new driver transistors on the MOSFET, and the board layout is pretty close.  Currently breadboarding the Frequency Counter.  This is still DRAFT.

1-SEP-2008 Files updated.  The Frequency Counter is complete, PC boards are ordered and should be here in a week.  Now breadboarding the PLL circuit.  The Resonance Scanner is working, but sensitive.  VCO-IN is responsive to 1 volt – 4.9 volts, but can go to VCC.  The top quarter volt is VERY reactive.

15-SEP-2008 With the frequency counters built, and me ramping up my metalworking capability, I’m concentrating on the PLL now.  The best values for the PLL chip itself seem to be 0.22uF and 10K, with a 1uF lock detect cap.  I might make a ‘scan speed’ select on the resonance scanner, it’s a lot easier to adjust the scale and shift of the output when it goes fast, but I’m worried that the equipment won’t be able to handle too quick a scan speed.  I’ve got the dwell side of the 556 making a standard astable square wave, and feeding it into the PLL SIG-IN line.  When power is applied, it scans once, finds it, and locks.  It’s working well, I can ‘adjust’ the frequency to simulate skewing and it keeps track.  I’ve also worked out the driver circuit after a few different tries.  I’ve settled on a push-pull totem pole design which I’ve tested up to 100KHz with the “ST8NKy” MOSFET, this is working very well.  Schematic updated. I’d like to add a lot of goofy cool blinky lights, but I’m running out of real estate.  I’m considering a ‘hybrid’ design which uses surface mount components for lots of little indicator LEDs, they aren’t required for circuit operation but look nice.

16-SEP-2008 I’ve decided to split the safety circuits off to a daughter board, like the frequency counter.   As sad as I am to admit it, some people’s kids just aren’t going to use the safety circuits, and removing them from the main board will give me lots more room for blinky LEDs.

18-SEP-2008 Schematic updated.  PLL capacitor 0.22uF is good for 20Hz – 35KHz locking.  0.1uF is working for ~200Hz to 120KHz, but I need to do more testing on this range.  Sometimes it won’t re-lock after losing it at high frequency, but will initialize and lock if power is cycled.

21-SEP-2008 Lots and lots of blinky LEDs!  I’ve added 5 different color LEDs to the board now: red = fault, blue = pulse, green = lock, yellow = dwell, and white = gate.  I’m also adding in an LM3914N chip to drive a bar array of LEDs to the scanner voltage.  Thatt should be interesting.  It will do a sawtooth back-and-forth of the LEDs, this will allow limited adjustments to be made without an oscilloscope.  Lots of lights!  I’ve ordered 10K millicandela (so like, 10 candle) LEDs from , so a big Hello to them and the China Young Sun LED Technology Co., Ltd.

24-SEP-2008 On the breadboard… I have the resonance scanner circuit working, the dwell disable signal is working, the PLL is interfaced to the scanner and dwell with logic gates, the MOSFET driver works, the pulse pickup circuit kinda works. I’ve got a store-bought dual coil choke on the pulse circuit and the MOSFET driving a resistor + LED + capacitor + the other side of the choke.
Given this primitive testing setup, I’m feeding the MOSFET output back into the pulse pickup circuit.

The ‘scanner’ display looks pretty cool, and will allow fine tuning of the scanner without a scope. That could be critical to operation in some iffy combinations.

Power on: it starts up and locks. Sometimes it scans a couple of times, other times the lock is instantaneous. Changing the value of the feedback capacitor makes it unstable (smaller caps are more stable too)… for a while… then it seems to mellow out. Freq rises slowly while doing this, which I think is some type of slow magnetization of the choke core, not significant.

I have a magnet I took out of a hard drive, pretty powerful and polarized right across the flat center on one side. I hold this magnet near the D-core of the choke and the frequency goes up . I tend to pull the choke coil out of the breadboard about this time, but if I hold it down, I can hold the magnet very close to the top… and it goes fast . When the magnet gets too close or I click it onto the coil core, it loses lock and won’t regain until I hold the choke and pull the magnet off. I’m guessing the magnet blocks the transfer of the pulse in the coil, so the lock is lost and the scanner switches into circuit. When I pull the magnet away, it scans a couple of times and locks. I’ll bet I’m demagnetizing my magnet too.

Bottom line… I can modify the environment, this changes the frequency of the pulse, and the PLL keeps lock. It’s doing what it’s supposed to do. The frequency counter is a big help with this :-) As
far as I’m concerned, I’ve got a circuit that mostly works. I need to work out the final design of the pulse pickup amp, and it’s ready for the proto shop.

I don’t have all the LEDs hooked up (just gate and lock for now), but I’ve tested the transistor driver
for them and they’ll work. I’ll put the whole thing in a blue translucent NEMA box and it will glow .

I’ve discovered McMaster-Carr. I have good stuff on the way, delrin, nylon, stainless steel, pressure switches. I need to get outside and get to working on the test tube, I just can’t seem to get motivated to get off the breadboard now that I’m having some luck with it. I got a really nice 12VDC brass water valve from . I’m about to send the safety circuit design off for prototyping.

24-SEP-2008 (Later) Schematic and POV-Ray updated.  I cut some tubes with the new saw, it’s… scary.

27-SEP-2008 Made test cell. Tired.

28-SEP-2008 Delrin sucks, no way to glue it, but it will be useful one day with mechanical sealing.  Remade holder out of some of the 2″ clear PVC.  It looks great!  Schematics, files, pics updated.  I added a nice trick with one of the spare 4066 gates, now the scanner display gets brighter when the PLL is not locked, and dimmed when it locks.  I ordered more from Digikey, some very nice switches and knobbies for the front panel, a bunch of forgotten resistor values, and the relays for the safety board.  When I can test the optoisolator circuits with the relay, I can send the safety board design off for prototyping.

03-OCT-2008 Files and images updated.  It’s very close.

06-OCT-2008 Okay, so the circuit is close, but everything else is still far away.  It’s not delivering the voltage through the power transformer, and I’m confused.  I’m trying distilled water in the test cell and getting almost nothing.  Yes, I actually see tiny bubbles, but with the amount of juice I’m giving it, I should be melting wires.  Very very low current, and I can’t raise it even with the dwell gate defeated.  I blew an oscilloscope probe, I think I overvolted it, but I’m not sure how.  This is confusing.  I need a proper VIC transformer, that’s for sure.  I have a small heat sink on the output MOSFET, and it doesn’t even get warm.  Maybe I already cooked it, but I don’t think so.  I have to look at this another way and go back and reread the old tomes.

I tried using the Triad power transformer as the step-up, it just doesn’t seem to be drawing much from the MOSFET.  I probably need to (minimally) heat sink the MOSFET properly, but the current draw is very small, I can’t get it to draw more than 200mA, and the regular breadboarded PLL circuit draws 100mA or so of that.  The other side of the transformer shows like 20mV (MILLI volts), maybe that transformer is just a lot screwier than I thought.

I’m using a 7.5 inch ferrite rod bifilar (two wires) choke with another few turns (third winding) as the pulse pickup, that much seems to work, but the polarity of the pulse pickup seems to be the big factor on that part of the circuit.   This choke measured like 1.75mH a side when I measured it before with the cheap meter (now with blown fuses and mostly burned out).  With the pickup coil loosely wrapped over the ferrite one way, it works great, the other way, the response is very awkward in that a signal only appears when the dwell is off.  That’s strange, I have to play with that some more, at least it seems to be working.  With the diodes across the +/- inputs to the pulse op-amp, anything bigger than about 1.5 – 1.75 volts (forward bias of the MUR800E diodes) should be clipped.  Output from that op-amp is between VCC and GND.  I can see that output pulse on the scope, it works with the dwell and a little bit more after the dwell cuts.  I’ve noticed that it’s rather stable oscillation (unless the pickup coil is reversed), for all I know, it’s working perfectly and picking up too much garbage electrical interference from computers, oscilloscopes, video monitors… the other electronics in the room.  I’ve already noticed that with no pulse coil (open leads), it readily locks onto a 60Hz line signal out of the air.

At any rate, I’m becoming convinced that the circuit is doing what it’s supposed to do, and not much will need to change even if I get the right inductor(s).  If I determine that there won’t be anything I could add to the circuit to make up for any deficiency, I may send the PLL PCB off to get made anyway.  I’ll probably add in the manual override again, but that’s about all I can think to do.

I’ll take some pictures soon, maybe that will help.


Pictures of Oscillation Overthruster output measured with a Tektronix 475 oscilloscope.

1189.  Top trace is dwell, about 25% at 878Hz (active low) taken at TP3 in the schematic (2V,0.5ms). Bottom trace is MOSFET gate (2V,0.5ms). MOSFET is driven by 9 volt supply, dwell is from one side of a 556 dual timer running on 5 volts.

1188. Top trace is dwell, same as above.  Bottom trace is the pulse output and PLL pin 14 SIGIN (2V,0.5ms).  Notice that pulses are detected after the dwell has shut off.

1190.  Close up of one dwell cycle with PLL VCO output.  Top trace is dwell (2V,50μs).  Bottom trace is PLL VCO OUT at TP6 (2V,50μs).  Here you can see the VCO is following the pulses.  Frequency counter at the VCO OUT says about 30K, it varies from 31K cold to 29K warm.  About 33 pulses in this picture per 878Hz dwell cycle is 28974, that’s about right.

1191.  Measurement at the test cell using 1.75mH dual choke, blocking diode, but no VIC transformer (1V,50uS, GND at center)  Input DC Amperage is 90 – 120 mA, sorry my cheap meter won’t do better than that right now.  The circuit normally uses that much juice, I don’t know why it’s not pulling more power.

Obviously, there’s no bubbles.  No hydrogen here, move along.

Note: I just saw something strange.  I noticed that after I turned the power off, the test cell had some voltage still on it.  I thought the scope was just decalibrated, but on GND it was at center.  Flipped back to DC, it was still showing 1.75 volts DC offset, same as in the picture above (the line at center of the waveform).  Huh?  I disconnected one of the cell wires, and the voltage is still there.  As much as I can figure, the cell is pulling some DC from the scope, and when it gets to the forward bias of the blocking diode, it stops.  So I short the cell, and it bounces back to about 0.6 volts.  I turned the circuit on, and voltage rose slowly to around 1.8 volts DC offset.  I turned the power off again, and it stays around 1.7V.  It looks like it’s sinking, very very slowly.  It’s a little weird.  It’s actually acting like… a capacitor?  Naw, that couldn’t happen.  Okay, I disconnected the scope, shorted the cell, reconnected the scope (0 volts), and then unshorted the cell.  The voltage started rising.  It’s acting like a capacitor and charging from the oscilloscope probe leakage.  Very cool, that’s a nice clue.  If I’m right.  It’s been a long day, I might be hallucinating.

This might explain a few things, there’s not even enough impurities for the water to conduct at all, so I can’t pump any current through it.  Maybe tomorrow I’ll try tap water.  Here’s a picture of my desk, don’t make me regret this:
Left to right: roll of solder, oil can of ‘Tap Magic’ cutting fluid, one-tube-set test water fuel cell (4 inch tubes), connection box with rod inductor choke, Tektronix 475 oscilloscope, yellow cased ampmeter behind banana connectors and wires, breadboard with 7046 PLL circuit, roll of Scotch 92 kapton tape, frequency counter (30579 Hz), stereo boom microscope

16-OCT-2008 I’ve done some rewiring.  The MOSFET and regulator diode are heat sinked, the test cell is hooked up.  It’s taking amperage, not a huge amount amount, but I’m still using distilled water.  It locks too easily, I’m thinking I might add the divider chip.  For some reason, the scanner circuit is invading the locked signal and knocking it out of lock.  I’m under the impression that once locked, it should really stay that way.  It falls out of lock regularly with the scanner, usually at the top when the frequency is highest.  I need it to stop scanning when locked, maybe I can tie the lock signal into the reset on that side of the 556.

26-OCT-2008 Okay, that’s doing something.  Now the scanner stops when the PLL locks, and this has a big effect.  The Triad still isn’t putting out the voltage that it should, but I’m getting a few tiny bubbles, what I might call an hour-old Alka-Seltzer.  I hooked up the cell directly (no VIC transformer) but with chokes, and I saw some hot spot frequencies.  There was one around 44KHz, but that spot was not very stable and usually jumped to 22Khz or 11KHz fairly quickly.  Sometimes the scanning would rest at 22KHz, but more often between 10 and 11 KHz.  This is interesting in that these are (reasonable) harmonics of each other and so this would tend to indicate something good.  Sometimes the circuit would jump to these points on it’s own, but more often I would cycle the power and these occured on first scan.

I’m trying the full circuit now, with reversed chokes.  If I run it with the dwell disabled, it soaks 1.82 DC Amps total circuit.  The control circuitry uses about 100 DC milliAmps.  The secondary of the Triad power transformer I’m trying as a VIC transformer is 0.92 Henries inductance according to my half burned out VC9808+ “Sinometer”.  Both sides of the choke measure 1.8 milliHenries.  The dwell at 150Hz and “ON” about %40 draws about 0.9 DC Amps (1.20 AC Amps).

When running this way there are some spikes in the +5VDC power supply, they occur on the MOSFET on times, as would be expected.  There are thick pulses during the gating times 50 millivolts high, and thin spikes at 0.1 volt.  This is ugly, it probably deprecates performance of the PLL.  I may experiment with powering the control circuitry completely separately.  This would involve cutting a fat trace on the safety board, but nothing a dremel won’t handle.

I’m also using the divider (CD4040) chip.  The chip divides the PLL Comparator input from the VCO output by 2^(the number of the pin), so Q1 = 1/2, Q2 = 1/4, Q3 = 1/8, et cetera.  This is providing a manner to adjust something… but I’m not sure what.  The frequencies change, things move around, but not really sure if it’s helping.  It will probably be a nice option to have a switch on the front of the box for it even if it’s a LOT of wires, so I’m thinking it might be permanent.  Note that Meyer’s circuit used 4017 decade counters, and so he could divide by 10, 100, or 1000.  Not sure if I want to try that.

I just tried using the JW Miller / Bourns choke instead of the 1/2 rod choke.  It seems to be working well, the circuit does everything the same but the frequency is much lower.

Creative Commons License

Copyright (c) 2008 MDVE.NET.  Some rights reserved.
This Oscillation Overthruster design is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License .

Comments 15 Comments »

The Voltage Intensifier Circuit (VIC) is the term Meyer used for the final components of his water fuel cell.  These particular parts have caused a lot of consternation in the water fuel cell research community, as the physical design and electrical values of the components vary greatly, and can be manufactured and used in many ways.  Here I’ll discuss my thoughts and efforts in this area.  Please comment on this!

Here is a set of simplified and unfinished circuits showing different ways to connect a water fuel cell (click to enlarge)

The first drawing is the minimalist example of the MOSFET drain-source circuit (driver circuits for MOSFET gating will not be discussed here, those can be very convoluted and the one I’m using for the Oscillation Overthruster is working just fine).  In this circuit, a MOSFET gates a positive potential to ground.  Nothing else is really needed.  However, this leaves a lot of room for stray voltages from various sources, some will damage the MOSFET if left unchecked.  A more “complete” version is shown in the augmented MOSFET output circuit, which includes
* a regulator diode, which ensures that total potential is never reversed
* a clipping diode, which “shorts out” reverse currents across the VIC primary due to magnetic feedback
* a “PTC” device, this is a “resettable fuse” which protects the MOSFET from over current at the price of a small amount of resistance
* a zener diode, this “shorts out” reverse voltages across the MOSFET drain/source.  Some MOSFETs (like the STP8NK100Z) have internal zeners which perform this function.

Many would argue that some of these components would not be very beneficial, and would even hurt the purpose of the circuit, I do not dispute this.  In particular, some experimenters would not use the clipping diode as clipping the buck feedback from the VIC primary will definitely affect performance.  I’ve added the PTC myself, as the cost of these devices is far cheaper than MOSFETs, and turning the dwell up just a little too high is very easy.  It’s all perspective, if you have a warehouse full of 20-year-old junk 10GB SCSI hard drive controller PCBs with BUZ11 MOSFETs all over them, then by all means… burn out as many MOSFETs as you want.  If you’re paying $5 apiece for them, a 40 cent protection diode starts to look pretty good.  Diodes, zeners, and PTCs that handle the wattages and voltages used here are not expensive.  Some experimenters are using MOSFETs that I cannot find for under $25.

On the Oscillation Overthruster design, I’ve allowed a lot of freedom.  Some components are not required, others can be shunted with bare wire, and I’ve allowed two sizes of PTC for mounting on the PCB.  All of the heavy transistors are TO-220 types and can be socketed and (MUST be) heat sinked.

Next on the schematic, I show the quintessential Meyer VIC circuit.  The bottom choke is adjustable, either through a wiper arm in contact with the wires along the side of the coil, or through multiple taps along the choke which can be switched manually.

The other drawings on this schematic show various ways to actually connect the water fuel cell to the MOSFET.  You can use a VIC transformer, or not.  You can use chokes, or not.  The blocking diode is recommended, but is probably also optional (try and find out!).  If you use a single choke device with two windings on it, I think it’s called a “bifilar” choke (I think it’s a term Tesla used?) and you can hook that up two ways.

Okay, lots more pictures now…

Smaller inductors

Left to right: roll of 1/2″ wide Scotch 92 kapton tape, small ‘E’ core with plastic bobbin, Epcos D-core 47mH 1.3A choke (Digikey 495-2790-ND), cheap measuring scale, JW Miller / Bourns 8116-RC 50mH 2.3A choke (Digikey M8911-ND), 3/8 x 4 inch ferrite rods

POWERLITE C-cores with scale, label from Elna Magnetics.

Left to right: wrapped AMCC-100, two parts of AMCC-320 sitting on oiled anti-corrosion “gun” paper (note light rust on end), AMCC-100 set (this set is varnished with Corona Super Dope)

These are very strange things.  They are made of nanocrystalline amorphous metal in the form of extremely thin ribbons of iron alloy wrapped to make the oval, then cut through the center.  By extremely thin, I mean on the order of 25 micrometers.  These are by Metglas and I got them through Elna Magnetics.

POWERLITE C-cores – AMCC-100 closeup in sun.  Notice the edges of the thin iron ribbons are somewhat visible on the edge.  This is painted with Corona Super Dope.
Ferrite rods in the sun, see the shiny sparklies? Not really, it didn’t come out in the picture, but they do sparkle.
Two are 3/8 inch by 4 inches, the one on the right is 1/2 inch by 7 1/2 inches.
Toroids: large yellow T400-26D (4 inch diameter) wrapped with Scotch 92 dielectric tape and a few turns of 20 AWG enameled copper wire, and Amidon ‘pulse’ toroid, pretty much the largest one they sold. The Amidon toroid is smooth and hard, makes a nice ring when struck lightly.

Honestly, I’m not sure if I can ever use these, they’re just too hard to wrap.  I can make a jig I suppose, but my attempts at locating a reasonably priced toroid winding machine were futile.

Left: various colors and gauges of copper magnet wire.
Right: 3 BUZ11 N-channel MOSFETs in the middle of a power supply PC board from a 10GB 50-pin SCSI hard disk drive, circa 1994.  These will be harvested and reused.  You thought I was joking, didn’t you?  I just have this one, not a warehouse full :-)
TRIAD VPS10-8000 power transformer, front and back.

This is a 115VAC/230VAC to 5VAC/10VAC 80 watt power transformer.  The ‘primary’ is made to take wall voltage 115/230 volts, and step that down to 5 or 10 volts.  As shown, both sides are hooked in series, so it’s wired to take 230VAC and put out 10VAC.  I use it backwards, so my input is 12V pulsing and output is… well, messy… but usually a few hundred volts with some very high spikes.  Pulses go off the scope at 50 volts/division with 8 divisions.

Off topic: get a load of the cat hair stuck in the solder flux on the bottom right tab on the back.

Ferrite rods, one raw, one wrapped with dielectric tape and then 2 windings (and again covered with dielectric tape) plus one small ‘pulse pickup’ winding with a relatively large gauge

I’m suspicious that this is too large, it measures 1.75 milliHenries a side.  When I wrapped one of the 3/8″ rods, it measured 130μH a side, which seems more likely to be closer to a good value.

Not pictured: two sheets of 12″x12″ teflon, one 1/8″ thick and one 1/16″ thick.  These are for use as insulators on the ferrites or the C-cores.  They are white and square and boring, so I didn’t take a photo of them.

So I’m trying to get the TRIAD transformer to work. It’s not cooperating. I was hoping to use relatively simple off-the-shelf parts to get this working, that’s why I’m considering it at all. I don’t think this stuff is so hard that you can ONLY do it the SAME WAY, but I might be mistaken. Lots to try before giving up on that.

When I hook it up, the voltage just goes to naught. With the circuit off, I can verify very low resistance conductivity through the coils, I can verify the diode conducts only one way, everything seems to be just fine… but the actual voltage present on the water capacitor is a few millivolts. I’ll hook this up again and take another picture.

Quick edit: the rewiring is working, not sure what to make of it yet.  Basically I’ve temporarily eliminated the hookup box I’ve been using so I can just plug wires into the breadboard.

Comments 4 Comments »

Not everyone is going to be interested in the safety circuit.  I have mixed feelings about that, but there isn’t a way to force the issue.  Table saw manufacturers can’t keep people from removing the safety guard on the saw, this isn’t any different.  I’ve removed the circuit from the main PLL board and put it on it’s own board.  This also allows a lot more real estate on the main board, but I must stress the safety… splitting water is a hazardous business.


I’ve worked out the design and the board, I’m working through the logic to make sure that it’s viable.  It’s an ‘active live’ design, any failure causes a ‘fault’.  A fault should occur if:

* wires are broken
* the optoisolators get burned out

And the more normal faults:

* low or high water level
* low or high pressure

Low water causes a fault, and also switches a relay for a pump or valve solenoid.  Very low or high water, wow, that shouldn’t happen, especially high water.  High pressure, yeah, definitely turn it off for a while.  Low pressure: normal when the system is first starting, but what happens when there’s a blowout?  You wouldn’t want this to just keep on running with a hole in the plumbing, so low pressure causes a fault.  To get the system running, there’s a “pilot” momentary switch which defeats the low pressure fault, this is just like the pilot switch on a gas appliance, you hold it down for a minute while the system starts, then you can let it go as soon as there’s enough pressure to keep it running.

Parts list

Qty Value Device Comment/Digikey Parts
1 0.1uF poly capacitor, 5mm lead-in voltage regulation C2
1 1K 1/4 watt resistor, 10mm lead-in R3
1 1N4003 regulator diode or 1N400X
relay coil clipper
1 1N4148 small signal glass diode or 1N914 D1
1 2N2222A NPN transistor
TO-92 ( or TO-18 )
relay driver Q2
1 2N2907A PNP transistor
TO-92 ( or TO-18 )
relay driver Q1
5 4.7K 1/4 watt resistor, 10mm lead-in 5V pull-up resistors R1, R4, R5, R7, R10
1 74AC02 IC: Quad 2-input NOR 14-DIP 74AC02PC-ND IC1
1 100uF electrolytic capacitor
8mm with 3.5mm radial leads
voltage regulation C1
4 330 1/4 watt resistor, 10mm lead-in LED current limiters R2, R6, R8, R9
1 G5V-1 Omron G5V-1 RELAY SPDT
water pump switch
1 LTV-847 Quad Optoisolator 16-DIP 160-1370-5-ND OPTO1
1 SPST Normally open SPST
momentary contact switch
for “pilot” S1
4 Fuse clips for ATO fuses F067-ND F1, F2
1 1A ATO Fuse for logic circuit F1
1 10A ATO Fuse for cell circuit F2
1 MF-R040 PTC Resettable Fuse
400mA hold, 800mA trip
logic circuit protection

28-SEP-2008 Added fuses and PTC.  Board will fit 90mm spaced holes.  Schematic, files and pictures updated.

01-OCT-2008 I got the relays, tested that part of the circuit, and it’s good.  The ATO fuse clips look really good, I’m happy with the way that turned out.  Files and schematic updated, this is the version I’m submitting for manufacture to Sierra Proto Express .  It’s been accepted and is “in progress”.  6 boards should be here 09-OCT-2008.

Creative Commons License

Copyright (c) 2008 MDVE.NET. Some rights reserved.
This design is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.

Comments No Comments »

The frequency counter is a complex side circuit to the Oscillation Overthruster, so I’m putting it in a new post.

This is the final design.  There were a few mistakes in the prototype, the 1st and 3rd LED digit strobes were switched, a few capacitors that didn’t have the right lead spacing, and the logo was bad… ProtoExpress didn’t silk screen over the hole in the solder mask.  The parts are renumbered and mistakes have been corrected. (1 MB)

Eagle 3D POV-Ray of Frequency Counter board

The POV-Ray image has some inaccuracies and missing parts, I’m just learning how to use Eagle 3D.

This circuit takes a square wave signal (0 – Supply Volts) and displays the frequency on six 7-segment LEDs, for a range of 0 – 1MHz.  It counts the pulses occurring in one second, and displays the count over the next second.  It needs a 5-15 volt regulated power source, which will be supplied by the Oscillation Overthruster.  The zip file includes the Eagle PCB CAD files, Gerber files, the bill of materials, a few PNGs of the board and schematic, and where to get the full Creative Commons license.

I tested the circuit using a 5 volt supply, but all the ICs are rated for over 15V, so a regulated 12V supply and square wave input should work just fine.

This was designed using Eagle Light version v5.20 for Linux.

Partlist exported from frequency_counter.sch

Qty Value Device Comment/Digikey Parts
5 1K 1/4 watt resistor, 10mm lead-in R2, R3, R4, R5, R7
1 1M 1/4 watt resistor, 10mm lead-in OK anywhere from
910K to 3.3M
1 1nF poly capacitor, 5mm lead-in for LED driver strobe C1
3 2N2907 PNP transistor
TO-92 ( or TO-18 )
for LED digit select Q2, Q3, Q4
2 4.7nF poly capacitor, 5mm lead-in for pulse shaping C5, C6
1 4.194304MHz crystal, 5mm lead-in (49UA) X007-ND Q1
1 10K 1/4 watt resistor, 10mm lead-in for pulse shaping R8
1 22pF capacitor, 2.5mm lead-in optional, if needed C4
1 82pF capacitor, 2.5mm lead-in optional, if needed C3
3 100nF poly capacitor, 5mm lead-in for VCC to GND filtering C2, C7, C9
1 100uF electrolytic capacitor
8mm with 3.5mm radial leads
for VCC to GND filtering C8
1 4093 IC: Quad 2-input NAND
Schmitt Trigger 14-DIP
296-2068-5-ND IC3
1 4521 IC: 24 stage Frequency Divider
296-14158-5-ND IC4
2 14543 IC: BCD to 7-Segment
Latch/Decoder/Driver 16-DIP
2 14553 IC: 3 Digit BCD Counter
2 LTC-4724JR LITE-ON Common Cathode
triple 7-segment LED
160-1545-5-ND LED1, LED2
2 220 (x7) 7 (or 8 ) 14 or 16-DIP isolated
LED current limiter,
can use 7
1/4 watt resistors
R1, R9

I recommend for manufacturing.

Accuracy: assuming the crystal passes it’s +/-30ppm rating, that would be:
100% * ( 2^22Hz +/- (2^22Hz * (30 parts/1 MHz)) ) / 2^22Hz , right?
100 * ( 2^22 + ((2^22) * (30/1e6)) ) / 2^22 = 100.003
So that’s +/- 0.003 % maximum error.  Sounds pretty good to me.  I’m not really sure that the ( +/- 30 parts / 1MHz ) constant is the correct way to represent the error factor.  The counter chip also latches on falling edge, but the master reset on level (from what I can tell on the data sheet) so there may be some loss due to the length of the master reset pulse formed by R7 and C5.

Creative Commons License

Copyright (c) 2008 MDVE.NET. Some rights reserved.
This design is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.

Comments 1 Comment »

Water Fuel Cell forums:

Very good comprehensive page at PureEnergySystems PESWiki:

This is a pretty good document: .  Good detail on the stainless steel alloy and conditioning, but how much is actually necessary?

Independent study paper via Aquapulser: “Chapter 10” – see comments by Tad Johnson on 10-132 about key elements to successful operation, base schematic for pulser on 10-70.

Good page on 555 timers:

Good page on driving transistors from logic signals:

Over Unity forum:

Bob Boyce’s hydroxy Yahoo group:

Meyer patents and papers:

Comments 4 Comments »

The inductors are the most mysterious part of this circuit for me.  TTL was being invented when I was learning electronics, and as a result my understanding of radio electronics is very weak compared to digital electronic application.  So to make up for my shortcoming, I’m overdoing it.  I’m buying a lot of ferrite coil cores in various physical forms: I have a few rods, a few toroids, an E-core, some dielectric tape, and a lot of magnet wire.  I’m also buying a pre-wrapped coil to try, Digi-key has a few that looked interesting and were cost competitive with the raw unwrapped ferrites.

The E-core seems to have the most potential for a Voltage Intensifier Circuit (VIC) transformer in that it is much easier to wrap.  It has a plastic bobbin that can be wrapped on a jig, then the core snapped into place around it.  Rods are easy to jig too, but toroids, which would seem to be the best design, are of course the hardest.  It’s going to take me a long time to wrap a 4 inch toroid with 200 windings of 20 gauge, followed by as many wraps of 28 gauge as can be handled, but at least 600 per the Meyer patent.  I’ve barely started, with this toroid I want to make a 5 winding setup.  I might try the E-core with these same gauges, but only as a VIC transformer without the chokes.

Inductor junction boxI wrapped a 3/8 inch diameter 4 inch long type 33 rod with dual strands of #22 wire, that one is pictured here.  It measured 190 uH (microhenries) per side.  1.75 mH per side bifilar coilI just wrapped a 1/2 inch diameter 7.5 inch long type 33 rod with two strands of #24 wire, that came out to be 1.75 mH (millihenries) per side, almost a tenfold increase.  That coil I taped and wired into the inductor junction box, and draws 9 mA (milliamps) current across a 4.7K resistor load when using wide open DC (no pulse) from the Lawton circuit.  The waveform is extremely clean on the scope, it’s impressive.  If I unhook everything and measure the inductance with a meter across both sides of the coil, it shows 7.0 mH (milliHenries), which might suggest something if I knew enough about it :-)

Some two-winding chokes from Digikey04-AUG-2008 I got a couple of store-bought chokes from Digi-key today.  I’ve measured them with my cheap meter, they are surprising.  Here, I thought I was cool wrapping my own rod coil, these store bought suckers put me to shame.  Of course, none have been given trial by fire yet, so we shall see.  I’ll list them by rating, what I measured per side, and what I measured with them hooked up serially in differential mode (crossed over).  In common mode, the inductance cancels itself out and it measures zero.

Digikey M8911-ND (toroidal, above) $11.97 rated 50mH, 2.3A measured 69.7mH / side differential 276mH
Digikey 495-2790-ND (D-core, below) $3.74 rated 50mH, 1.3A measured 43.4mH / side differential 172mH

Right now, if the chokes were working or not working I wouldn’t know the difference. But soon I’ll get to hook these up and find out.

I also just hooked up a neon sign transformer.  It’s not in the circuit properly, I’m just pumping the pulse through it, but it’s putting out a LOT of voltage.  Neither of my meters will show the voltage, and with no load, it goes off the scope at 400 volts AC.  With the 4.7K resistor as load, it’s more manageable, and it sings the pulse frequency quite audibly without a speaker attached.  I will have to hook it up properly, I think it’s going to work.  It’s kind of scary.

Comments 4 Comments »

Copyright © 2010, Multi-Dimensional Visual Echo. All rights reserved.