Mosfet Heating Circuits

[Harvey]

Hi Gad,

IIRC, That 0.0033uF was the value originally specified. It is an arbitrary value selected to arrive at the target frequency.

So you can try different values here. The larger the value, the lower the overall frequency. The smaller the value, the higher the frequency up to the operating limits of the 555.

You can set your two pots to their center point and then try different cap values to find what frequency works best for your resistor. (Didn't I do an analysis on that to find the resonant frequency for you ) I can't recall exactly, so much time has passed. But then once you have your cap at the desired center frequency, then you can use your pots to adjust the ON and OFF timing (duty cycle) from there.

[gadh]

Hi Gad,

1. Evidently you are not getting the large current spikes that Glen did in his data, or if you are they are so narrow that the LED's light is invisible. We can increase the resistance of the parallel resistor, but we run the danger of impacting the circuit function with the increased resistance at that node. What is your peak voltage at the MOSFET Drain?

GAD: in my last test the peak was 500v.

Why did you remove that cap?

GAD: it was blown by my wrong connection to the 24v batteries instead of 12v...

4. The 50K pots spec'd in the original schematic provide a much larger range of adjustment and are used to get the timing of the pulses in the range needed by your specific Inductive Resistor. Every custom resistor is different and so you need to find what works best with your resistor. The two pots control the on time and off time of the gate pulses. And of course the overall frequency range is interdependent on them both. Not the best design, but that is what worked for the SA team and so we sort of stuck with it. You can use your pots if your frequency is where you want it.

GAD: i did all your modifications except the 2 x 50k pot. but i see only flat lines on the scope... i'de better check the wiring again. also i do not have 0.0033uf (3.3nf) cap. so please suggest a replacement in the uf range (a more standard value...)

[Harvey]

Wrong Link

I just noticed that my link to my suggested modifications post was not going to the right place

So that is fixed now, and here it is again:
Harvey's Suggested Modification

I don't know how that happened, I must have clicked on that link 5 or 6 times after posting it and worked each time, but this time it went to some other post.

Aaron, has the database been compromised? Can you compare the locked thread to the backup?

[Harvey]

Hi Harvey.
few issues to discuss:
1. after applying your idea of the leds, they wont light up. the voltage on them is only 16mv. we need to come with a better solution. we need at least 2v for the leds.
when removing the 0.1R (i use 10 x 1R in parallel) only the red led light up. voltage on it is 2.3v.
2. how can i measure the power consumption of the 555 circuit, assuming not connected to the MOSFET - i want to check if its improved after applying your changes to the circuit. when i connected multimeter between the 555 'out' pin and the 100R pot. it measured zero current (either in AC/DC mode). only in the 10A range i succeded to measure and in DC mode it showed 50ma. why it could not measure in the 'ma' range ? is that right ?

3. when i removed the 100uf cap. - the waveforms did not change at all. what is its role in the 555 circuit ?
(when i connected 470uf instead of it - the same waveforms also)

4. if i want to apply your changes to the Ainslie circuit (555 part), i see i need 2 x 50kohms pot. can't i use the ones i already bought - 2+10 kohms pot. ?

Hi Gad,

1. Evidently you are not getting the large current spikes that Glen did in his data, or if you are they are so narrow that the LED's light is invisible. We can increase the resistance of the parallel resistor, but we run the danger of impacting the circuit function with the increased resistance at that node. What is your peak voltage at the MOSFET Drain?

2. The power consumption will be that value of voltage x current at pin 8 of the 555. You may have blown the internal fuse in your meter for the mA setting. You could test this by measuring the current flow through a bench resistor - use Ohms law to determine the correct value. Example: If your meter has a 200mA fuse, set the mA range to measure 100mA and measure the current of a 12V battery through a 120 Ohm 2W resistor. The meter should read 100mA. If it doesn't then the fuse is probably blown.

3. The 100uF cap is a filter cap on the supply line and is there to help remove ripple on that line. The 555 is very durable against supply fluctuations because it is triggered in thirds of the supply line, not by a specific voltage. Therefore 2/3 of 12V will produce nearly the exact same timing as 2/3 of 5V using the same timing parts. However, if the ripple on the supply is large enough, the ripple itself will come through the Pin 3 because it is rail driven. It will not be seen as a timing change, but rather an amplitude change on that pin if it exists at all. Batteries are really good filters, better than 100uF caps Why did you remove that cap?

4. The 50K pots spec'd in the original schematic provide a much larger range of adjustment and are used to get the timing of the pulses in the range needed by your specific Inductive Resistor. Every custom resistor is different and so you need to find what works best with your resistor. The two pots control the on time and off time of the gate pulses. And of course the overall frequency range is interdependent on them both. Not the best design, but that is what worked for the SA team and so we sort of stuck with it. You can use your pots if your frequency is where you want it.

[gadh]

Hi Harvey.
few issues to discuss:
1. after applying your idea of the leds, they wont light up. the voltage on them is only 16mv. we need to come with a better solution. we need at least 2v for the leds.
when removing the 0.1R (i use 10 x 1R in parallel) only the red led light up. voltage on it is 2.3v.
2. how can i measure the power consumption of the 555 circuit, assuming not connected to the MOSFET - i want to check if its improved after applying your changes to the circuit. when i connected multimeter between the 555 'out' pin and the 100R pot. it measured zero current (either in AC/DC mode). only in the 10A range i succeded to measure and in DC mode it showed 50ma. why it could not measure in the 'ma' range ? is that right ?

3. when i removed the 100uf cap. - the waveforms did not change at all. what is its role in the 555 circuit ?
(when i connected 470uf instead of it - the same waveforms also)

4. if i want to apply your changes to the Ainslie circuit (555 part), i see i need 2 x 50kohms pot. can't i use the ones i already bought - 2+10 kohms pot. ?

[Harvey]

Hi Gad,

If we had a flatline DC ON condition we would expect a current flow of about 1.9A through the 0.1R and diode in parallel (24V / 12.35 Ohms [load + mosfet + CSR + {diode || shunt}). So, if the diode will conduct with 0.19V drop across it's junction, then we can expect that 1.9A to be split between the 0.1R shunt and the LED.

I wont go into all the technical near logarithmic current to resistance characteristics in the LED, but suffice it to say if we allow too much current to flow in the LED, it will try to hog the path and will fail. So my above concept is based on the operating range of the circuit and observed maximums in Glen's tests. Your results may vary depending on the devices you choose, and without a series limiter, the diodes could fail if they are unable to handle the current we allow to flow through them.

Since my choice of resistor value was based on the maximum current shown in Glen's data the target current in the LED was about 12mA at that maximum and that relates to about 12mW. However, I did those calculations based on a 1.2V drop for the LED at maximum circuit current. With such a large variety of LED's out there and the different dynamic resistance curves involved, it is difficult to specify an exact setup without putting constraints on the experimenters. I would rather say that this is the concept and you will need to experiment with your devices available to you to see what works best. Because the diodes are in reverse parallel, they protect each other from reverse voltage breakdown. Many LED's require over 2V just to begin conducting. (See chart Technical LED Color Chart) but we want the voltage drop to be as low as possible to prevent restriction to the rest of the circuit.

Experiment and see what works for you, the object is to get a visual representation of the current in both directions and the beauty of the LED over a galvanometer is that the LED is more responsive to high speed transitions. But if the LED's prove to be problematic, the Galvonometer would still be useful.

50mW low Vf LED's should be fine

[gadh]

Hi Guys,

I checked to see if I had a drawing of this - please bear with me on this, because when I design things, I do 99% in mind before I ever put it to a drawing and it is so vivid that after so much times passes I cannot recall if I ever put it to paper or not. I cannot find the schematic I remember, so chances are it is stored in my memory.

I did find the text describing it dated Saturday January 23, 2010 6:16:11pm in an e-mail to Glen as follows:



As you can see from the description, it is really very easy to make, just 3 parallel components between the CSR and B(-). If you want me to draw it I can. The objective is to adjust R1 to get a mostly "Green" light which is a visual representation of a negative mean current flow.



Edit: Wasn't that much time, so I just did it
Enlarge Image

Hi Harvey.
what should be the power rating of each of the LEDs ? 50mw max ?

thanks
Gad

[Harvey]

Bidirectional LED's

Hi Guys,

I checked to see if I had a drawing of this - please bear with me on this, because when I design things, I do 99% in mind before I ever put it to a drawing and it is so vivid that after so much times passes I cannot recall if I ever put it to paper or not. I cannot find the schematic I remember, so chances are it is stored in my memory.

I did find the text describing it dated Saturday January 23, 2010 6:16:11pm in an e-mail to Glen as follows:

Hi Glen,

I did a little math on the LED stuff. The maximum current spike that have seen in the data was around 12A if I recall correctly. In most circuits where we use LED's we have a constant voltage and we use a dropping resistor to limit the voltage applied to the LED. This works because the LED junction clamps the voltage across it to the forward voltage drop value - typically 1.2V. Current usually does not begin flowing through the diode until a specific voltage is reached across it. Depending on the device this could be as low as 0.3V or as high as the threshold which for some jumbo LED's is 2.4V or higher. We want the lower current, lower voltage variety.

Here is the application:
Two LED's back to back and a shunt resistor of 0.1 Ohms parallel to those two. So you have a package of 3 parts all in parallel. If we are moving 12A of current through that shunt (and in this case it is a true shunt) then there will be a 1.2V drop across it. Most of the current will flow through the shunt while a small portion will flow through the LED. The actual ratio is dependent on the "on resistance" of the LED. We are looking for a ratio of about a thousand to one here. But it can be varied. If 1.2V produces 12mA of current in the LED then we can think of the LED as being a 100 Ohm resistor in parallel with the 0.1 Ohm resistor which would be 1000:1. But realistically, most LED's are rated around 30mA or 60mA or higher. We sort of have to try it and see.

This 3 part pack is simply inserted in the ground lead, somewhere between the battery and circuit - perhaps at your switch. One LED is for (+) current and the other is for (-) current. When you get the battery to stop discharging, there should be more (-) current than (+) current and it will be evident by the intensity of the LED. One will simply glow better than the other.

I suggest a Green LED for negative and a Red LED for positive current. After I get the paper finished, I may experiment with this here. I have dozens of LED's.

Best Regards,

Harvey

As you can see from the description, it is really very easy to make, just 3 parallel components between the CSR and B(-). If you want me to draw it I can. The objective is to adjust R1 to get a mostly "Green" light which is a visual representation of a negative mean current flow.



Edit: Wasn't that much time, so I just did it
Enlarge Image

[b4FreeEnergy]

Hi Harvey, Gad,

Since I was never sure I got into that PMOO either, that LED indicator would indeed be interesting. (I didn’t even know it existed) Harvey, could you post the schematic of that thing over here?

I might clean the dust of my Ainslie circuit and give it another try…

Cheers,
Bart

[Harvey]

Hi Gad,

The configuration I used in my tests is as shown:

GND - BAT + === - BAT + === Inductor === MOSFET === CSR* === GND
GND=== 555 --^

So the 555 is connected across the battery connected to GND. The 555 can only support supply voltage up to 18V. It is quite likely that you have somehow fried the 555 if you exceeded that value, and possibly the MOSFET as well because the Maximum Gate to Source voltage is +/-20V and if the voltage drop across R1 was less than 4 Volts you would have exceeded the maximum for the MOSFET.

The configuration shown, isolates the CSR from the 555 currents so that the measured currents in the CSR only relate to the load.

If your MEAN voltage across the CSR is more than Glen's when you are using 24V, then you probably have not yet pinpointed the "Preferred Mode of Oscillation" (PMOO). I worked up a special LED design to help experimenters find the PMOO visually and passed that on to Glen. I don't know if it was ever tried by anyone except me . . . I think I have a video of that, but I never uploaded it to you-tube. That was right about the time things got ugly and I lost any ambition to carry on. But if I can find that drawing, I post it here - it may be helpful in pinpointing the PMOO.

I'll get back with that as soon as I can.



*CSR = Current Sensing Resistor

[gadh]

24v test results - still no gain

Hi Harvey and all.
24v battery test result - no gain (COP ~ 0.6).
i used now a rechargeable 9v battery (200mah) for the 555 circuit, and joined my 2 12v batteries for the load circuit.
i noticed that during 1 hour test - my 9v battery was completely drained ! so its no help using it (since many of my tests can last more than 1 hour) and i have to solve this energy drain ASAP...

then i tried to connect my '+' and '-' power connectors of the 555 circuit to one of the 12v batteries also - and this action fried the connectors !! i noticed then that the '+' of the 555 circuit is somehow connected to the 12v 'load' battery '+' endpoint - but the only possible connection can be the MOSFET !
is it right ? maybe my circuit is ruined ?

Also, i cannot get a better waveforms - i need to get the shunt waveform below zero (gnd) as much time as i can (in order to see a gain), but the avg. voltage is ~200-300ma (by my eye). any ideas how to get this improved ?

[Harvey]

No Modification Necessary

Hi Gad,

The only reason the modification was mentioned was because of the excess current in the timer circuit.

It is not crucial to producing the thermal output that Glen was demonstrating.

Let's start with the 24V and work from there

[CatLady]

Hi Harvey, some questions:
1. who are the "SA group" ?
2. what is the purpose of your modifications to the circuit ? - if:
a. to get the 2.4khz + 3% duty cycle - i got it already with my circuit (Glen based)
b. to reduce the current leak of the 555 or/and prevent its over-heating -its not my first priority. I intend to focus on getting a gain first, then improve it by your suggestions.
if you think some modifications are still crucial to finding gain - i'll use them right away.

The answer to : 1. The original South Africa group. I believe it was Donovan, Buckley, and some guy that killed himself...I don't remember his name right now.

[gadh]

Hi Harvey, some questions:
1. who are the "SA group" ?
2. what is the purpose of your modifications to the circuit ? - if:
a. to get the 2.4khz + 3% duty cycle - i got it already with my circuit (Glen based)
b. to reduce the current leak of the 555 or/and prevent its over-heating -its not my first priority. I intend to focus on getting a gain first, then improve it by your suggestions.
if you think some modifications are still crucial to finding gain - i'll use them right away.

[Harvey]

Harvey , please let me see if i understood your modifications correctly:
1. replace R5 with 1.5kohms resistor or with another pot. ?
2. R5 connected to RST leg. R1 (pot.) connected to OUT leg. do you mean that changing R5 will affect R1 sensitivity ?
3. replace all components (R/C) according to your "red" circuit modifications or just R5 ?

(you need to bring your explanations "down to the ground" for me...)

Hi Gad,

Sorry for the delay.

1. A wire connects Pin 7 to the junction of R4 and R6. Move that wire so it connects Pin 7 to the junction of R6 and D2. Note that the connection between Pin 7 and R7 remains unchanged.

2. Replace R5 with a straight wire (remove R5 and put a jumper in its place).

=======================

The component values in my suggested modification are those values that allow the circuit to work within the original Quantum parameters. Please note, that when Glen tried those values, he was unable to get the circuit to behave as claimed by the original article. His arduous experimentation resulted in the schematic shown in the very first post in this thread. Therefore, apart from 1 & 2 above, the schematic in the first post of this thread is the one that should be used if you wish to replicate Glen's work.

If your goal is to replicate the original SA claims, you will need to contact that group for assistance as their schematic and procedure is flawed and cannot work as claimed. Only they can provide the true values and procedures used. My suggested value modifications were provided to show that their stated parameters (2.4kHz Original Parameters) could be achieved with the simple correction of the capacitor and resistor values as shown. The SA group later admitted that the Quantum schematic was in error but never offered a correct schematic. Therefore we were forced to reverse engineer the device with a "best guess" approach and a lot of time and energy on Glen and Aaron's part to get it to work at all. My experience is that the SA group is purposely close lipped regarding their schematics, parts, apparatus and data. They are not sharing these things with the Open Source community for some reason.

. . .do you mean that changing R5 will affect R1 sensitivity ? . . .

No.
Removing R5 increases R4 Range, not R1 sensitivity.
With R5 in place, it is impossible to bring R4 all the way to supply rail thus limiting the frequency range.

Moving the wire as stated above, will reduce the current through the discharge circuit and remove the need for R5 at all.

What is needed in R1 is greater resolution, not greater sensitivity. We want less sensitivity and greater travel on the adjustment. For example, you can have two pots in series, both with 20 turn resolution. The one pot can be your R1 100 Ohm, and the new one (R1b?) could be 1 Ohm. Now you have better control using the 100 Ohm for course adjustment and the 1 Ohm for fine adjustment. My earlier suggestion of a length of Nichrome could have a 1 Ohm value to become R1b with a slide action rather than 20 turns. It is a low cost alternative to finding a 1 Ohm 20 turn pot.

Cheers,