Frequency Generator Chips
and Knowledge Duty Cycle & Invertors
Frequencies output by Main Frequency Generator Board
The attached document Stanley Meyer’s Gated Pulse Frequency Generator Functional Description is the result of analyzing the circuit for this card. It is another step in my understand of how the electronic works and what it actually does. Like the one I did for main frequency generator card it lists all the inputs and outputs and shows what the card does to them. I am doing this as Ronnie mentioned he did not understand how system worked until we looked at all the modules in detail.
The card has one main function it sets the pulse width for all the frequencies signals coming from the Main Frequency Generator and those on the Accel input (the gate function) and allows the width to be manually controlled from pot on front panel. The circuit sets a minimum width that cannot changed. Stan has stated changing pulse width effect rate of gas production.
It does provide switch to turned off cell and provides input to allow other modules to do this.
As far as I can tell it does provide the GATE (no signal space) seen it the VIC signal. I thought given the name of the card it was added here but it is not. I think gate here was meant to set size of opening (pulse width) like a gate in fence.
NOTE: If you plan on building these cards you need the front panel controls.
I built a bread board version of the Gated Pulse Frequency Generator and hooked it up to the Variable Frequency Generator board I had built earlier. I then ran a series of test to see how it actually works.
I included test results in my original analysis document (2 pages of test write up and 2 pages of O-scope screen shot to show what output looks like and how in changes when adjusting control POT. I replaced original document with new version and renamed to show that it now includes test results.
This circuit primary function is to generate a Gate wave train with it own frequency and period. The main function of the POT is change these values.
They both change at the same time. Its does NOT pass through the input signal from Frequency Generator board in fact the wave train changes very little from different input frequencies. I checked from 5Khz signal to one a 5Hz.
For these tests I did not change base frequency in the Variable Freq Gen board as I wanted enter to made changes in test configuration to make comparison of change I made easier to see.
So still looking for place where the carrier frequency and the Gate signal are combined at this point I beginning think where Gate signal enters the VIC as one of the Frequency Generator outputs goes to K10.
I did have some problems caused by using bread board but even those turned out to be useful as it gave me more information on how board functions and some of its limitations. I documented all this in test results.
I am glad I am building and testing these circuits as I am finding analysis I made while close, it is not good enough. Testing the board show me exactly what does as I can see the output on the o-scope and also see what happens when I make changes to inputs and changes using local controls.
finally found where the gate signal and the carrier signal are combined.
It is done in the Phase Lock Circuit input/output you can see this in Figure 7 of WO 92/07861 at top you see in boxes A and G which are both inputs to Phase Lock Circuit what I have been missing it is also in output to the Cell Driver it is even Labeled "Gated Signal to Cell Driver Circuit".
This is only place I found that shows the 2 signals being input logic gate together.
Found this as I was rereading this WO document after building a couple of the frequency circuits and getting ready to build analog circuits.
I was checking to see where all their outputs go as some of drawings do not show this clearly and some of the redrawn circuits have dropped the labels and just show connections as lines.
Also to be honest I did not expect it to be in Phase Lock Circuit which I have only glanced at so far, though it does look very similar to Variable Frequency Generator Circuit.
I was surprised how almost all it now makes sense as I can see where the circuits do what is described. Writing my own description of what the frequency and analog circuits do was a big help in that understanding.
I have posted those for frequency circuits with test results in this forum and have drafted ones for analog circuits which I will also post after I build and testing them so I an clean up error in my understanding before I post them.
I was going to post image of Figure 7 but need to figure out how to do that.
I have not been able to find what frequencies Stanley was using as input to the VIC so decide to analyze the circuit diagram for the Main Frequency Generator Board.
It's a simple circuit but my electrical engineering skill are very rusty and finding the data sheets and an use article with what information I needed was not very straight forward. But I think I have figured it out. I wrote the attached document to describe how I think it works. Short answer is 4 frequencies are generated on this board.
Left to right from the 555 through each of the 7490 chips the frequency get divided by 4. So highest is on left lowest on right which matches up to numbering on switches. Assuming the center range of the pot is being used the 4 frequencies are:
1.2KHz, 300Hz, 75Hz and 18.75Hz.
The Pot on the 555 is only adjustment in systems. But there are loops on board for O-scope probe to see output of each step.
The attached document shows all the inputs and outputs and what I think is being done by each element and why. It maybe over kill but it just the way I think and what I wanted to know. The 7490 was the hardest as most of the data sheets talk about it being a counter but in this case it is being used as a divider (still counting but results is division) there is a link to article that explains why in the document.
Most likely easier way to do this but a least now I know what Stan was using.
Parts came for Adjustable Freq. Generator today so I was able to get the whole circuit working and did some testing.
Found a few errors in the analysis I posted and corrected them and posted updated version with o-scope pictures of the output of each of the stages. Found that the 90 % time pulse out of 555 Timer is turned into 50 % square wave in the 7490 which also divide by 10 not 4 (another mode I hate data sheets).
The square wave is good news at that is a good input into analog wave generations and it is unlikely Timer output will be used as frequency would be to high.
I am doing this as I never seen this done in anything I could find and I also wanted to know frequencies and timing Stan used.
Ronnie said he did this as well and I believe him but he stopped posting before he got to describing what he learned. Though I think he was heading that way as there were a few references in some of his posts.
Air gap and or dbd barrier
I Believe what he refers to is to remove the dead short, which to my best knowledge is "made up" of the very water itself that which is filling up the cavity between cell electrodes
(I.E between the anodes and the cathodes, be it in the form of tubes or plates).
And one way to accomplish just that would be to invoke some good old Faradic brute force water electrolysis, thus "coating" the electrode walls with gas bubbles of sort, which "removes the dead short" given the electric insulating properties of gas, in order to sort of pave the way for some Meyer goodness to be applied to the WFC instead, creating gas the Meyer way as opposed to the brute force way. KNOWLEDGE IS POWER = WATER POWER
Actually Ronnie provided several steps to do this.
Tune the system in air 1st at low voltage as when you reach operation level there will gas in the cell generate a low voltage wave train that never goes away Vo in diagrams He stated Stan never said you start over from zero. This voltage fill pulse that in normally zero.
He pointed to circuit that does that its somewhere on one of the main cards but same circuit used else where. He found by studying all the control cards and circuits. Stated he build his own circuits to do this.
slow increase voltage by increasing frequency do this a steps
state you should start see you gas at around 2 volts if you don't your system is not tuned correct and you will wasting avail energy.
If you are in resonance by 6 volts you also have a problem as you are wasting potential on upper end
Never touch the cell for 2 reason you most like will get a nasty shock and will also kill the charge on well and will need to recondition it again.
I am also puzzled by people not talking more about several of the must do things that were in Ronnie's threads. I extracted key things out his threads and a few others as well so I could find them and reread them without wading though the whole thread again.
People commented each of item when he posted them but I do not see them in other must do lists.
I have reread what I copied several times and I keep getting a better understand of why it was built the way it was. Based on Ronnie's comments I believe people do not spend enough time getting the cell conditioned before starting to produce gas in volume.
Having said that Stan does not say how to do that only that it is a required step.
3) In conditioning discussion he talked about not using all the available power - start creating bubbles to late is bad (said should start around 2 volts and not enter full gas production until near 12 volts)
He also said that if you bump the tuning the slightest bit at 12 volts, you will fry your VIC.
It won't be an amperage type of destruction, instead it will be an arc-over that burns through the insulation--high voltage stuff. Hence the need for thick triple layer mag wire.
The procedure is actually pretty straightforward: Low voltage, center tune, up the voltage a tiny bit, center tune and so on until you finally have it precisely dialed in.
Once dialed in, you can power up to full voltage from a cold start and it should run.
Just repeating what I was told as I have no first hand experience of my own.
Ronnie also showed me a simulation run and the values have to be spot on to work.
believe you are 100% correct on the gas bubbles on the electrodes. I recently put up a youtube video where I measured the resistance of water between two sets of electrodes... when you reduce the surface area by 1/2 the resistance of the water increases more than 2X!
This does not mean shorter cells it means dbd barrier di electric barrier . coating
Ronnie also told us that you have to get the process making gas bubbles first...
When those bubbles begin to form on the electrodes they are essentially reducing the surface area of the electrodes (I even have scientific papers that make the same conclusion), without that occurring the waters resistance is not high enough to allow the cell to charge.
I recently did some testing on Stan's VIC- but only with the primary and secondary coils... It's very obvious when you do those tests you see that the VIC has very poor voltage regulation- and that the load has to be very light (I think over 100k ohms (leakage current) but will have to do more measurements and calculations) to generate high voltages.
If you haven't tried Multisim I would highly suggest it. It does a good job at simulating Stan's control and driver circuits. You can get the student editions pretty cheap and its an invaluable tool for circuit design and simulation.
Interesting test - results constant with several of Ronnie's comments
1) he stated that Stan's had said smaller cell's were better never said why
2.) In his discussion on independence balancing he showed why ten cell helped as it added more resistance
3) In conditioning discussion he talked about not using all the available power - start creating bubbles to late is bad (said should start around 2 volts and not enter full gas production until near 12 volts) resistance of the cells will be a factor in this
I looked at Multisim and played with examples. If I still had my UC Davis email address I could get student rate (I should still have but turned off by mistake when I retired). However Hobby rate is not bad and only reason I did not purchase it was I was looking at circuits with IC and could not figure out how to added them. Later I realized for some you could just add the logic gates inside still may be hard to do for more complex chips.
Then I looked at voltage circuits should be a big help there harder for me to figure out what exactly they are doing. I was always more interested in the digital side of things.
I plan on build some of these circuit to verify what they do and so I have something to play with. Actually on boxed by electronic parts so see what I had before buying things.
Sounds like a project for a proof-of-concept buildoff, in that such a circuit "only" should be serving the tuning protocol of such a manual tuning procedure for a given set of, let's say plates then, or to be more precise 2 plates and nothing else, with both set in stone plate areas and gap (I.E distance between plates) submerged in deionized battery water, powered by an equally set in stone dumbed down VIC kinda transformer populated with set-in-stone coils/inductors placed on some OEM core which in turn are powered by a manually tuned standard 555 PWM Circuit operating at 2-4-6-8-10-12VDC for this very purpose.
I was rereading Stan/s WO 92/07861 and found I must have an error in the attached file which is now fixed. I had swapped values for R1 and R2 when calculating the frequency values output from 555 timer. The work order said output should be over 10KHZ and with values swapped I had only over 7KHz.
With correct values it actually puts out over 12HHz. This also changes the output values of the other 3 stages. Divide by 4 still applies.
Other big difference is the duty cycle is not near 50% but is over 90% for all frequencies. I also believe this matches better with wave trains other have been testing.
Since I wrote this analysis I have been looking closer at signal path for other modules and can see where the output of this board is feed into the input of other modules and one of those inputs goes to the Analog Voltage Generator through the Digital Control Means card. Ronnie is one of his posts stated that signals were all sync. in the GMC unit. I beginning to understand how that was done.
So it looks like the Frequency carrier goes by one path to one side of the primary coil and the analog frequency goes to the other side of the coil the analog voltage generator and control path. NOTE: Using the 4 selector switches on the front panel the two frequencies do not have to be the same and most likely are not.
If I understand one of Ronnie's posts one use of these switches could be to increase the voltage on cells during the conditioning phase. Re: "Understanding How Stan Meyers Fuel Cell Works" « Reply #444, on November 4th, 2016, 06:14 PM »
If you read WO 92/07861 closely you can see this described and also the role the feed back circuit plays in all this. I believe it servers two roles it automatically searches for resonance again with any change in water and I also believe Stan's set a limit using it to keep system for damaging coils.
I finish building the circuit did some initial test today as 7490 chip arrive. Found another error in my analysis The 7490 actually divide by 10 not 4 as it uses a clock signal on pin one to do this. I also found that the output signal is a square wave with a 50 % duty cycle. This makes a much better input into the analog wave train. Also it unlikely that the pulse from the 555 timer will be used to drive the analog wave as frequency would be too High.
I also found an error in the circuit diagram. It shows the led to be connect to VDD that does not work right as it was always on. Need to be connected to ground as it gets +5 V from the inverter output. Works correctly when wired that way. I have updated the document to reflect these changes. I also include O-scope screen shots of each of the 4 stages signal negative and the final output from the inverter stage which makes the signal positive.
I have ordered the parts for Adjustable Gated Pulse Gen and have done initial conversion of the circuit diagram to layout with actual pin locations which are different. When doing this I have also found required chip supply lines have been left off diagrams is some cases. Not a big deal as it makes circuit cleaner but something to be aware of and to check.
Description of the Main Frequency Generator Board used in Stanley Meyer’s Water Fuel Cell Simple explanation of the function of this card it provides 4 separate frequency to other components.
A base frequency and 3 others that are direct divisions of the basic frequency which is generated on the card using a 555 Timer. As configured the center frequency output from the timer is 1.2k with a 90+ % duty cycle and a period of about 830ms.
Selection of which frequency to use is done manually using switches on front panel of this card. The output labels on circuit diagram below show up as inputs on other Stanley modules. The card is (PCB K2) or and labeled Module K2 in Stanley Meyer’s functional diagrams.
Re: Andrija Puharich's work suggests glass dielectric
As a have a working version of the Stan’s control circuits, I decided to test this. I set K2 (Variable Pulse Frequency Generator) to generate a 3khz signal. Using the switch settings, I can then select 3khz, 300hz, 30hz or 3hz as an input to the K3 (Gated Pulse Frequency Generator) and K8 (Analog Voltage Generator). Both K3 and K8 are fed from the same output source on K11 (Digital Control Means) and are labeled M and M1 on circuit diagrams, so I have been testing both using same frequency.
When I fed 300hz into the K3 I can get a 300hz gate, but it has very small range and I cannot get a 50 % duty cycle output more like 15-20 % max duty cycle. While I had notice there is no gate at 500hz and had reported this in my original testing report for this board I was not sure if 300hz would work correctly at least for the values provided for K3. I can get a signal all the way to the primary coil interface but lose most of all the adjustment in gate size that Stan appears to be using to change rate of gas production.
I am not saying Nav’s theory is wrong only that 300hz does not work with the listed values for K3 that I used to build the circuit. In a more reason thread The AM signal and how it works. Carrier and modulation Nav discusses how either 50hz or 60hz can be used and methods to test this independent of Stan’s control circuits.
While I have done most of my testing with a 50hz signal when I was doing the 300hz tests I also used 60hz and got also the same results as for the 50hz signal.
I still need to test K14 (Pulser Indicator Circuit) that input should not change the shape of the signal going to the primary coil just the frequency of the modulation, as it mainly effects the PLL function of CD4046B.
This part of testing will be put on hold for now as there is not much more I can do until I get coils built.
My current plan is try to test coils using outline by Nav in his AM signal tread as it should be easier to control variables, as one’s in Stan’s circuit are very touchy and I keep turn wrong knob.
Will take some digging but in almost all cases I used the values from the Stan circuit diagrams from estate drawings marked up by Don. Where I did not uses parts Stan did were in switches and pots. I used Standard items. In most cases I found items in Amazon it not there on ebay and a couple of items I when to Newark for a few things I could not find anywhere else as you can buy small amounts from them but you do have to pay shipping. That was only place I could find the 5 pole switches (4 in 1 out).
Turns on Amazon I bought 20 amps 5 each cheaper than just buying the ones I needed same for resistors, capacitors and some IC like the 555 timers. I even found the part on the pulsing circuit that others could not find on ebay. It was surface mount and I had to buy and adaptor so I could mount it on regular board did that as I was trying to stay a close to Stan's parts as I could.
I not sure if I still have the parts list that I used I know I wrote them down when I was build boards so I knew what I needed to order. I can look.
This is a N channel mosfet driver. you can use this for an arduino or from a 5 volt frequency generator. you need a 12 volt power supply and a 5 volt signal an irf640 handles 15 amps. mosfets are trigured by voltage. the irf640 triggers with roughly10 volts
This scr driver can also be used on this frequency generator. the opto coupler is triggered by 5 volts from freqncy gen or a arduino or pc the scr is made to chop a rectified ac power source. such as a wall outlet, with a full wave rectifier on it. the board here has built in most of the componates to make it function
7414 smit trigger
i use these in place of the 7404 inverter chip. Daniel Palacios from south america
Instead of the analog wave train I expect I kept getting a flatline voltage. This is from the 2N3005 (Q4 in K9 the Voltage Amplitude Control circuit).
I retraced the signal through the circuit, and everything matched until I got the final output instead of analog wave train, I got a flatline voltage. In my initial testing I did not have the output from the emitter of the 2n3005 connected to the switch and the last 1uF capacitor. The capacitor is the cause of the flatline. Remove it and system works. With it in you can still move the voltage level up and down but AM wave is gone. It does not make sense to have gone to all the work of creating the AM signal then remove it before sending it to primary coil.
It is possible this capacitor was intended to smooth out the noise on the AM wave as it has a lot of noise, however if this is the intended purpose 1uF is the wrong value. However, if you want to smooth out noise on a 12-volt source then it would do that.
I am continuing to check the other boards that feed the 5KHZ side and have started to gather some of the material I need to make coils. Looks like a should find the ferrite core first so I know what size and shape to make coils and bobbins. I know the dimensions for the areas for the wire but need to know size of hole inside bobbins.
Front and back of finished boards mounted power comes in on left through a 12-volt connecter and there 2 LM317s one for 10v and on for 5v that feed bus bars on back. Common ground through out system. Boards powered from the bus bars. Bus bar on the right will connect to primary and feedback coils.
The scope shots show the output of the 2n3005 with the one with capacitor in circuit and one with it removed. This is the signal going to one side of primary coil. The yellow trace is the 50Hz signal input used to create AM wave. I have it on screen to provide a good sync for the scope.
K9 AM wave output
K9 Output with cap on output
Continued testing with check the High frequency side. Verified boards are working and then checked inputs to both sides of primary coil. Channel 1 Yellow is from Cell Driver K4 and Channel 2 Blue is from Voltage Amplitude Control K9 on Analog side.
Both scope probes are hooked up prior to diodes as I wanted to see input to primary coil. At this point I am just using a 10-ohm resistor to provide a restive load.
The first picture shows this test setup.
As there is are no coils in the system, I have hooked the output of K21 (G) back to what would be the input (H) to the board from resonance sensing pickup coil and Pulse Indicator Circuit K14 (black jumper in picture).
I took three pictures of the signals with different gate settings. Large gate, 50% gate and small gate. Gate size was set using pot on Gated Pulse Freq. Generator (3).
Last picture show the output when the Switch on the Voltage Amplitude Control board (K9) is set to “OFF”
I would not pay to much attention to frequency listed as the scope locks on the gate pulse to determine frequency.
It should also be noted that each of the signals is re fenced to same system ground point.
Stanley A Meyer Test setup across input to Primary Coil.
Stanley A Meyer Small Gate
Stanley A Meyer 50 Per cent gate
Stanley A Meyer Large Gate
Stanley A Meyer Cell off with Sw on AM board
After writing this and looking at screen shots the 2 signals look the same just at different levels.
I begin to wonder if something had happened to analog signal. So, I disconnected it and went back and rechecked output of board using setup I had used to check analog path.
I set scope for CH1 yellow to be 50Hz reference from K2 CH2 blue to be analog signal which was the triangle wave train, so it was working.
I then hooked it back to output strip but left CH1 hooked to 50Hz reference and got the following signals. I again varied gate size on K3. Switching off analog gave flat line.
This was the analog signal I expected to see. Did not expect pulses on top but was glad to see them. But they are there as all inputs to primary coil are hooked up including diodes.
I am beginning to see why people say scope shots do not make a lot of sense. Would not have seen the analog wave in this form if I had not changed scope sync reference.
Note: I used same ground reference for all of these screen shots.
Stanley A Meyer Analog signal with small gate
Stanley A Meyer Analog signal 50 per cent gate
Stanley A Meyer Analog signal large gate
As using a separate sync source gave me such a good view of the analog signal,
I decided to look at the pulse stream the same way. I went back to K2 and tried all four switch settings on one of the other switches, the 500Hz output gave me the best view of the shape of the pulses.
Picture was taken at about 50% gate setting. You can also see effect of gate changes in this view. This was taken just before diode going into Primary coil like the other pictures above.
At this point I am not sure what the output modulation frequency is as the gate messes up that reading on my scope.
I will need to go through K21 again and check center frequency I am sure it was changed when I mounted board and I have not yet tried to reset it.
In this testing I was check that all the boards worked together, and primary goal was to see what signals looked like at input to primary transformer. I think I did that.
Stanley A Meyer Close up view
of pulses with 500 hx reference
I tried to set the center of the CD4046B to 5khz.
This is kind of a pain to do with the gate active as scope syncs on the signal with gate so you cannot see the true frequency.
To do this I bypassed the circuit that raises the gate level to 12-volt levels. This leaves the gate signal present, but it never goes high so gate in not present in signal. (See report on K21 for more detail on this).
I may put a switch on the board that has the circuit to raise it gate level so I can select the bypass mode so I can set center level.
Once I set the gate level low, I used the Manual Freq. Adjust Pot on K22 (Resonant Scanning Circuit) and the Freq. Adjust Pot on K21 (Phase Lock Circuit) to set center frequency on CD4046B to 5khz.
You must set the pot on K22 to a high enough level to make adjustment to pot on K21, so you set center level this high. I checked output of pin 4 of the CD4046B to do this.
While I was at it, I also verified that the levels out into switch from the divide by 10 chips where correct and they were. So, I have 5khz, 500hz, 50hz and 5hz out of the switch.
My goal was to capture the input to primary coils as a known 5khz center.
I have not yet built primary and secondary coils, but I do have a 1 to 1 60hz isolation transformer so I hooked that up across the 10-ohm resistor and diode so I could see signal out of primary and secondary coils.
The screen pictures below show signal with pot setting above and gate active at 50% duty cycle.
Signal across 10-ohm resistor alone
Signal across 10-ohm resistor with transformer connected
Signal output from transformer
Signal output form transformer showing frequency sweep every 8 seconds. As I mentioned in K21 report this sweep is present not sure why.
Have been thinking about this for a while and wonder if it is from way the sweep signal in K21 is setup (see K21 report for details).
The K21 signal looks like a piece of a triangle wave just one period with a flat line in between each period.
Yet the data sheet for CD4046B shows this input to be a triangle wave.
It is possible that the valves for components controlling the 555 timer on K21 are incorrect as this sets the timing of the pulse that creates the sweep wave.
Most likely will look at this some more as I had not looked at K21 input requirements when I did original K22 to testing.
Next step is to build and test some coils. Have already started collecting parts but do not have much experience with this.