COP 17 Heater | Rosemary Ainslie

[FuzzyTomCat]

Hi Fuzzy. Yes we did use a heat sink. Hope that helps.

EDIT by the way - I saw that link earlier. What do you think? Jibbguy thinks it's another example intended to sucker us into believing it with the actual intention to point out that we're idiots to believe he claimed overunity. Not sure of the energies involved and he's not telling. Typical.

Hi Rosemary,

Thats good about the Heat sink on the Mosfet but I'm going to insulate mine with something thin and heat tolerant, the temperature appears not to get that high anyway .....

As for TK from what I've seen unless he's getting help on this last "invention" of his ...... he is a very talented individual and could If he choose to direct someone in the wrong direction, could and be quite good at it with the apparent knowledge he possesses

Glen

[kent_elyue]

From the values I've witnessed in this circuit, you should be fine with the IRF842. I have seen spikes above 750V with different things I've tried, but your Avalanche would probably handle that fine as long as you allow at least 1µs in between Avalanche (let it recover). Nothing I've seen, even during aperiodic mode, comes close to that interval. Get a good heatsink on it just to be sure, but with 1.1 on resistance, it won't get as hot as the IRFPG50 for the same current. Probably why its rated for higher amperage. It will be an interesting thing to see if the 1300pf input capacitance makes a difference in the aperiodic mode - if it even occurs.

If you do not have a scope, try putting a small flashlight bulb between the ground of the battery and your circuit. You will find adjustments that let it glow more than others. When the bulb experiences a reverse current it will glow brighter. Place a diode in series with it momentarily to see if the lamp dims when it loses the reverse current. If it doesn't, then you are not clipping any of the waveform below ground with the body diode. Even so, if you can get a good temperature rise in the resistor without burning out the bulb you may be getting something worth putting a scope on. Keep us posted.

Thanks! The bulb and diode combination is a great idea. Because of my lack of high-tech test equipment I have often had to get creative simply to be able to confirm or rule out certain possibilities and conditions. The little grey cells don't work as well as they used to, so this is great. Thanks so much!

FWIW, many years ago my electronics teacher told me to get into the test equipment field. He said that technology can only advance as fast as our ability to measure it. Oh! how I wish circumstances had allowed me to be able to follow his good advice.

-kent

[witsend]

Hi Fuzzy. Yes we did use a heat sink. Hope that helps.

EDIT by the way - I saw that link earlier. What do you think? Jibbguy thinks it's another example intended to sucker us into believing it with the actual intention to point out that we're idiots to believe he claimed overunity. Not sure of the energies involved and he's not telling. Typical.

[kent_elyue]

Hi Kent,
...
Hope this helps, Thanks for your parcipitation in this venture !!
...
The Single Pulse Avalanche Energy is low but it may work, but without a scope it would be hard to know.

That's a big help. Thank you very kindly. I wasn't sure what else I should be looking for.
-kent

[FuzzyTomCat]

@ Rosemary and Harvey,

Rosemary did you use a heat sink on the Mosfet in your original models ?

Harvey have you seen TK's new video of a modified RA circuit ?

YouTube - Electric OU: Properly Switched MOSFETS Produce More Power Gains than the Ainslie Circuit

I noticed in the opening shots that TK's Mosfets are insulated from the heat sink and in this PDF .....

A Combined Single Pulse and Repetitive UIS Rating System

There is a statement on Schematic page 3 "Obviously the heat sink is more of a problem than the UIS capability"

Could the heat sink be adding something unwanted or a problem in the circuit ... I'm just one for small details and that rock that no one looked under.

Best,
Glen

[eternalightwithin]

I agree with poynte99. Funny I was thinking about this late last night. I had the feeling that if one wants to get any appreciatable amounts of power out of this thing, we're going to have to change the 555?

What about a VCO? You could take a noise source in avalanche mode and attached it to a schmitt trigger. Add a shift register, say 12-bits, and a DAC. You got your self a random signal generator.

That clock signal would be attached to a comparator and... how do you say it charge pump? Attach "charge pump to your mosfet.

More discrete than all that crammed into your 555

[poynt99]

One other suggestion for tapping any excess energy off the load resistor is to wind a secondary on the load resistor itself using high temperature wire (just in case). Feed this secondary output to the bridge and connect as the previous diagram to feed energy back to the source battery.

.99

[poynt99]

.99
So it goes with fringe science and unconventional engineering.

In my career I've used diodes as thermal sensors, reverse biased transistors as zener diodes, 20 amp 2N3055's with a zener from base to emitter as 20A Zeners, Capacitive reactance for transformerless power supplies, and a host of analog and digital peculiarities used in a way to maximize a benefit from their anomalous behavior. Sometimes it is the unexpected reaction of a part that gives us exactly what we need for a given application. Look how useful activated charcoal was for mouthpieces in the early telephones.

This is fairly conventional stuff Harvey. I've done much the same here over my years. Almost anything goes when it comes to discrete components. It's when we start "dicking" with IC's, and particularly application specific IC's such as the 555 that we run into problems or unpredictable behaviour when we deviate substantially from the manufacturers recommended basic operation. If we're placing current limiting resistors in the supply of a chip where it is never called for or required if used properly, then we should revisit the suitability of that chip for that particular application.

There is nothing wrong with driving a Mosfet with a 555. If you look at the application notes for the HEXFET series, you will see that there is a large variety of ways to do this. One of the methods specifically shows the TTL totem pole output driving the gate directly and using a pullup resistor to help with the on current. In the 555 that arrangement is improved substantially by a darlington arrangement that negates the need for any pullup resistor. With 200mA of available current, the 555 can drive a large array of MOSFETS directly with no worries of the Miller Charge being an issue. As regards driving an inductor, it all depends on what you want the inductor to do. If you want a square wave, then the MOSFET is the way to go. If you want a sine wave, then IGBT's are a better bet. The 555 can give you both drives with some passive parts.

You can DO almost anything you want in electronics. It comes down to wise choices. I too have driven MOSFET's with devices other than MOSFET drivers. I've driven one with a CMOS inverter chip. The point is there are better ways. In this case clean drive is not the goal, so anything is valid as long as the desired type of drive is achieved.

I'm not sure what it is that you are 'encouraging me to do' here as far as the desired result goes. Are we on different pages? My desired result is to accomplish the same wild frequency splatter and negative spike energy that Rosemary did in her device. Your suggestion doesn't sound like it would accomplish that in the least.

By result I am mainly referring to how the mosfet is being driven, and duplicating the D-S wave form. The wild 555 oscillation is obviously unpredictable and difficult to achieve. If the goal is to achieve this same output from the 555 but in a predictable engineered manner, then I am suggesting that the drive from the 555 be characterized as best as possible, then emulated as close as possible by other means, such as a HF wave form in burst mode or FM mode, or some combination...whatever it takes to get a similar result, if indeed the drive to the MOSFET (and resulting wave forms in the circuit) is the "magic" which makes the "odd" measurements possible.

So in short, I would encourage you to closely characterize the aperiodic oscillation and then try to achieve it by some reliable and predictable means. Then run this through the circuit and see if it appears to give similar "curious" results you've been seeing with the 555 as the MOSFET drive. Then you have something to go forward with. If it does not pan out to give similar results, then it is back to the 555 and trying to make it stably "unstable".

I've looked at your COP > 1.5 and I may have an isolation transformer here to test that on. I think we can simply replace the entire resistor with the transformer. Naturally, we all know that we are not going to have a 100% efficient transformer. But at least we should be able to get at least that if this circuit does what we hope it does. I don't know how the output will tell us what you want but I'll think on it.

I'm not sure replacing the load resistor with the transformer primary is the way to go if we are trying to test the original Ainslie circuit. It becomes quite a different animal then, but go for it. Indeed placing the primary across the load resistor changes the circuit quite a bit also (but so does a flyback diode). Just trying to suggest an alternate tool to test for COP>1. The goal of the output is basically just tapping the excess energy (assuming there is some) in order to return it back to the source. For many, this is the ultimate test. If the device can run itself indefinitely, then without question the device is operating at COP>1. This really is a variation of using the flyback diode to return some energy back to the source.

.99

[witsend]

We've been disconnected since 6.00am. Cable theft. One of the consequences of our 3rd world conditions is that it's viable to steal copper from our utility supplier. But that theft is now encroaching into our cities and suburbs. Extraordinary.

In any event - I read through Harvey's analysis of Poynt's clashes with me - or is it mine with him? The truth is I'm getting way too used to saying what I want instead of what I aught. Yet again apologies Poynt. This is getting tedious, but I solemnly undertake to do better. And it's not you who's the obsessive compulsive. It's me. I think the term is projection.

Harvey - you're our very best find. Not only that first class brain but all that wisdom. We need some urgent cloning - that way you could share the burden and maybe get some 'off duty' time.

[Harvey]

.99
So it goes with fringe science and unconventional engineering.

In my career I've used diodes as thermal sensors, reverse biased transistors as zener diodes, 20 amp 2N3055's with a zener from base to emitter as 20A Zeners, Capacitive reactance for transformerless power supplies, and a host of analog and digital peculiarities used in a way to maximize a benefit from their anomalous behavior. Sometimes it is the unexpected reaction of a part that gives us exactly what we need for a given application. Look how useful activated charcoal was for mouthpieces in the early telephones.

There is nothing wrong with driving a Mosfet with a 555. If you look at the application notes for the HEXFET series, you will see that there is a large variety of ways to do this. One of the methods specifically shows the TTL totem pole output driving the gate directly and using a pullup resistor to help with the on current. In the 555 that arrangement is improved substantially by a darlington arrangement that negates the need for any pullup resistor. With 200mA of available current, the 555 can drive a large array of MOSFETS directly with no worries of the Miller Charge being an issue. As regards driving an inductor, it all depends on what you want the inductor to do. If you want a square wave, then the MOSFET is the way to go. If you want a sine wave, then IGBT's are a better bet. The 555 can give you both drives with some passive parts.

I'm not sure what it is that you are 'encouraging me to do' here as far as the desired result goes. Are we on different pages? My desired result is to accomplish the same wild frequency splatter and negative spike energy that Rosemary did in her device. Your suggestion doesn't sound like it would accomplish that in the least. I've noticed that when I get a scope lock, the temperature drops on my little resistor. Probably because the duty cycle drops or the negative excursions fall off. But its obvious that the aperiodic operation is a very necessary component of this phenomenon. That resistor heated up much faster than it did tonight with straight battery power across it. So there is something to this. My thinking at the moment is that the negative spike creates a high Q mode that compounds it on repetitive cycles and increases the phantom current that never gets registered in the sense circuit.

I confirmed this evening that the sensing resistor is necessary for aperiodic action in my circuit. Without it, my SE555 is very stable across all adjustments. In my case I used a one ohm resistor directly off the B(-) for the whole circuit, and then later doubled that up to 2 in series. I cannot duplicate the effect I had the other day, it gets close and the small resistor heats up to about 39°C, but it all stablilizes and holds there. Later I will try the higher voltages. My FET runs cool. My 555 gets hot. There is no doubt that energy is passing through that chip and in my case, it is not the ill engineered discharge path, because I am using my 'proposed changes' version wrt that path.

I've looked at your COP > 1.5 and I may have an isolation transformer here to test that on. I think we can simply replace the entire resistor with the transformer. Naturally, we all know that we are not going to have a 100% efficient transformer. But at least we should be able to get at least that if this circuit does what we hope it does. I don't know how the output will tell us what you want but I'll think on it.

[poynt99]

The variation in 555 operation comes as no surprise between manufacturers and lots. This is even more the case when using devices like this in an unconventional manner such as by inserting series supply resistors (not saying you have that), and driving high capacitive loads that are switching inductive loads themselves.

Driving a MOSFET directly with a 555 output is a poor design choice from the start (esp. if switching an inductor), but since it is being used that way here, one should not be surprised if weird and unpredictable operation ensues.

I would encourage you try driving the MOSFET with a high frequency generator in the range found when in aperiodic mode and see if you get similar heating in all the components as before. I highly doubt the 555 is contributing any special ingredient here, but rather it is just operating "badly" due to poor component choices and/or subjecting it to an application it can't readily nor properly handle.

Drive it accordingly as you find with your tests (which is why I asked about drive levels). This might come quite close. Two other options to get closer yet could be to apply the HF drive in burst mode or to apply an FM type drive.

.99

[Harvey]

Harvey,

In the aperiodic state, what are the G-S and D-S voltages like? Is the mosfet getting enough drive for full saturation and depletion?

What's the range of the frequency when it's aperiodic?

.99

Interestingly, it depends on the 555 chip. With the NE555 I had, it was full scale switching, like Aarons early waveforms. With the LM555 it produces two different types of oscillation. In one case we get a perodic series of HF pulses in between the timer pulses, and those HF pulses are those sharp pointed ones I show in the video that resemble a full wave rectified AC signal. Those do not seem to come full scale but IIRC, around 8V or so? Then the other one is a full scale oscillation that takes over the entire timer operation - the timer runs full scale on its output in the full wave rectified version of the waveform and drives the IRFPG50 accordingly.

The SE555 is completely different. If I do not put the cap in place between pin 5 and pin 3, that one will self oscillate prior to the timer pulse even if the IRFPG50 is out of the circuit. But this one is hard to get into aperiodic mode. I will have to revisit this one for that.

You can see a definite loading of the gate signal just prior to the aperiodic oscillation. But I would have to measure all that for you to be accurate.

[Harvey]

I drew back once in answering your posts, but obliged lately in hopes of making a connection. No Go. I simply can't communicate effectively with you. I am also growing rather loathsome of the insults, labels, and consistent and flamboyant convolution of my posts and their intended messages.

Adieu.

.99

Perhaps I could be a liaison here.

Sometimes we tend to impugn motives into written words that are not there simply on past experience. We may see sarcasm where it is not intended or we may feel stabbed where genuine offers were extended. This can lead to comments that are intended to reciprocate what we think has happened to us. Different personalities respond in different ways.


I'm sure Rosemary is still feeling the effects of the unwarranted attacks on her claims and test data. And among those feelings are all the scrapes and bruises related to the various hoops she has been asked to jump through so as to satisfy the curiosity of those unwilling to accept what she has already shown. Any configurations that seem to be designed to alter the proper function of the system will simply come across as more jibes and stabs.

Poynt on the other hand, is anxious to learn something new and protects himself behind the comfort of the skeptic persona just in case there truly is nothing new to be learned so that he can save face. When 5 of us replicators start showing increases in COP, then he will change his energy toward finding out why instead of assuming that Rosemary has made a mistake. But in the lack of such evidence thus far, he seems bolstered in his stance that she has made a mistake. To find closure, he must prove to himself how and when the 'mistake' was made. This drives him push the issue and devise tests that in his mind will give him the closure he seeks.

Rosemary and Poynt both want the same thing, confirmation of COP > 17. But instead of looking for ways to accomplish this goal together, the tend to work at odds with each other.

I think the solution here is for Poynt to draft a simulation that shows the negative return I mention in my ramblings above. The part where I discuss changing our point of reference to the drain and monitoring the battery current from each pole against that reference. For this to work, the negative spike at the drain must swing well below ground for several of the ringing cycles. Also we need a measurement of the power dissipated in the load resistor. If I am right here, a net reading of the B(-) current will be insufficient to explain the power dissipated in the load and the simulation should show that. This would be constructive

[poynt99]

Harvey,

In the aperiodic state, what are the G-S and D-S voltages like? Is the mosfet getting enough drive for full saturation and depletion?

What's the range of the frequency when it's aperiodic?

.99

[Harvey]

Just a quick blog here:

After I signed off from all my ramblings I went to experimenting to see if I could get the negative swing in my circuit. I tried some other inductors and stuff just enjoying the hands on that I have been removed from for so many years. I was looking a bit closer at that 'punch through' I showed on my video and was really trying to pin down how and where that was coming from. Clearly, it occurs at a time that the 555 is supposed to be off. Tracing it back from the gate, and expecting to see it diminish the farther I got from the gate, I found just the opposite. Uh oh...I'm recorded as stating that this can't be coming from the timer , but here it is clearly jumping out when the timer should be off. Why is it there?

My first thought went straight to the ground on this thing - press fit wires in old oxidized breadboard contacts etc. Lot's of room for 'noisy' connections. That all check out clean. Something was telling this chip to fire up the output and then before it even gets on, it was being told to 'unfire' it. But what could do that? This chip has a TTL flip flop in it and the SET / RESET inputs to it have to conform to very specific levels in order for it to flip. So I took a close look at the input waveforms to pins 2 and 6. The were clean. Where is this spike coming from? So I looked at pin 5, it is decoupled to ground through a 0.01nF cap. It clearly had the spikes and this is why it is switching the flip flop. Where are they coming from...the cap? I pulled the cap out, they persisted. The only other way they could get here is via the Vcc line. I could put a big filter system on the Vcc...

So I put the same cap from pin 5 to pin 3. Problem solved, the negative feedback cancels the effect. Very unconventional approach, but it worked.

So... I see that I have a 5 Ohm 50W 5% resistor here. It has a number on it - RCLALN50 with a date stamp of 8507. I wonder what that would do in the circuit. Oh, look there are two there, they look similar. So I put them in series. The waveform looks similar but slightly smaller in amplitude on the spikes. All of a sudden the thing goes into aperiodic oscillation and I just monitor things a bit. Whoa, that little thing is getting hot. So I take a reading. 42°C on one end and 34°C on the other... and climbing. IRFPG50 is cool to the touch, it reads 29°C same as the wood on my desk - ambient. I slide the thermometer laser along the body of the resistor heatsink...it is a gradient. Now I have 45°C on the hot end and 35°C on the cool end. Wierd, must be a flaw in the internal resistance. So I measure the other one... weird, its ambient. A closure look and I realize the second one is a precision 0.1 Ohm 1% 50W. So I turn the 5 Ohm around expecting the gradient to reverse. Same same. I repositioned the device to see if the gradient was related to the matt it was sitting on - nope. This is a thermal conduction matter and since both ends of the resistor have identical alligator clips and wires it must be related to how the energy flows in the device. I'm sure someone reading will know why the heat migrates to the higher voltage side

I was seriously considering running a full test on this and was about to start taking notes and video when it dropped out of aperiodic operation - on its own. I adjusted the gate resistor, changed the pulse width etc and messed with it for about an hour and finally gave up. I could get into operation briefly and then it would drop out. It was late, moving to 3AM IIRC so I set it aside until tonight. So I will see if I can get a better response now. I may try 24V to see if it improves things as last night it was running on only 12V (or less - I suppose I should measure that, been using it for a few days now).