COP 17 Heater | Rosemary Ainslie

[Harvey]

Well, I'm back on line (for now) the power keeps surging here - millisecond brown outs, but not enough to reset the computers yet - just enough that I can see the lights dim and here the AC motors skip a beat or two.

Spent most of the day catching up. Enjoyed Jibbguy's post regarding the internal multimeter functions and have to agree with him there. But at the same time, I find myself agreeing with .99 if an older style analog Galvanometer were used in place of the current sensing resistors, or specifically right between the B(-) terminal and all other components. In fact, one on each battery terminal would show if there were a 'leak' in the circuit. Swapping them between multiple tests would eliminate any calibration concerns. Nevertheless, either we trust the 3054 or we don't and I for one have no reason to doubt its results. My far inferior B&K Precision 1476 has a vertical precision of 0.02V/div with each division on the graticule subdivided into 5 equal parts. A well focused trace occupies about 50% or 2mV of screen just for the trace width. Therefore, the smallest accurate ripple that can be measured would be about 3mV, just enough to tell it is not a flat line. So even with my cheap scope, a 5mV signal would be a trustworthy signal. That being said, I can grab hold my scope probe and easily produce 800mV of AC energy from my fingers while the ground wire simple floats at the Scopes Chassis potential (which is tied to an 8 foot long 1/2 inch diameter copper rod driven into the earth right outside my office not more than 4 feet away. The 'Noise' envelope on that signal measures at about 200mV across a 10K resistor with various harmonics and a measurable fundamental that produces 11 cycles for 10 div set on 1 microsecond per division, or 1.1MHz. You can measure this 'Noise' by placing your probe connections across a 10K resistor and touching the probe end while not touching anything else. I put an alligator clip jumper wire on that point and placed the other end on my tongue to enhance the signal to the measured value. Now, we must stated then, conclusively, that the free energy 'noise' must have produced 200mV / 10K current in the resistor. That has to be 20 micro amps right? So, 4 micro-watts of power right there, free for the taking, even if it is buried in the 'noise'. One million resistors on a chip, touch one end to your tongue and the other to 'ground' and you have yourself a 4 Watt heater that runs on 'noise' (ok, you may need a million tongues for that to work, but hey ). It is beyond the scope of this post to ascertain where the energy originates from - whether it is biological, RF, Spatial, Terran or otherwise. The point here is that it is there and it is measurable.

Is this enough to trigger the IRFPG50? Not according the specifications. The Fig3 on page 1117 of the Designers Manual shows zero drain current at 4 volts gate to source. But it then shows 500mA at 4.5V. Does the IRFPG50 do any strange things below 4V? I would have to say yes. The turn off time for the device is 130ns. So if the gate is fully charged and then depleted, it takes 130ns before the drain current ceases. This means that there is some residual charge dissipation occurring that causes it to stay on, and some of that is between four and zero volts. Down in that range we are really in uncharted territory and attempting to use the device in a way that it wasn't intended. So can the device self trigger down there? I think there is enough resistance in the circuit to prohibit it, but a floating gate pin is not the way to find out. The wirewound pot makes a good RF antenna for whatever frequency it is tuned to and that is pumped into the gate. So Aaron may have a keen little RF feedback oscillator when he removes the timer from the circuit. I would expect a little more conduction through the IRFPG50 if that were the case however. It is more likely that the noise we see is descriminated RF passed through the body diode and internal capacitance. It may be micropower, but it is still power and the scope reads it. This is somewhat evidenced by the signal which looks fully rectified (I would have expected to see halfwave instead)

Well here it is Sunday Evening and I haven't even fired up the circuit since Friday and I want to get it into oscillation - hopefully tonight it will cooperate.

BTW - Glen, your scope is waaaaaay fast enough to evaluate the original 140KHz aperiodic oscillations. I just hope my 10MHz scope is fast enough for these 'noise' oscillations

[witsend]

Hi Fuzzy. Just a quick point. You notice that Aaron got his oscillation with an entire disconnect of the switch - having first established an 'oscillation' at a given frequency rate? Have you tried this? It seems that, once established, that resonance no longer relies on the switch. But if you do this as well - it's above and beyond. I'm just so grateful that you duplicated the test at all.

Incidentally - I agree with you regarding the voltage drop over the battery. Some is inevitable over time. Goto also references this. My own experience is that batteries do tend to 'lose' a certain amount of charge when they're at rest. So there's a definite consensus here. It's just that I am not sure that the battery can actually 'recharge' unless the voltage below zero is greater than the battery voltage itself. But I'm not sure. This is because the oscillation also points to the fact that the current flow could be doing that clockwise/anticlockwise thing that I fondly believe is happening. In terms of this current first flows from the postive to the load, to ground. Then, during the off period it flows from the load through the battery - through the MOSFET's body diode - back to the load.

Anyway - it's all very intriguing.

Edit - David - Just seen your post. Delighted to see that you're also duplicating. Many thanks.

[FuzzyTomCat]

Yes, Radio shack has a NE555CP 555 timer. CMOS version, low power
They also have a NE556 "Dual Timer" chip with 14 pins.

Anyways I'm about to start my COP 17 Heater
Thanks TomKat for the IF50 Mosfet.

I'm doing the circuit updated as of August 9th. Didn't see the August 10th rendition will I was home with my parts. I ran into some potential problems with exact parts so I'm running it by you guys/gals while I'm wiring up my "smoke generator"

They didn't have 0.25Ohm resistors so I got a 0.47Ohm 5watt one instead. They had some 10 Ohm 10Watt resistors but I went for the 8 Ohm 20 Watter instead.

Do you think these two component changes will be a problem or can I still tune it with the 2 10K Pots I have?

Sincerely,

David

Hi David,
The correct version of the RA COP>17 Heater circuit is the Revised : August 12, 2009 one COP 17 Heater | Rosemary Ainslie

This is the shopping list for just the components -

1 (1) NE555 timer ( several )
2 (1) 10k 10 turn potentiometer
3 (1) 5k 10 turn potentiometer
4 (1) 2k 10 turn potentiometer
5 (1) 100 uF capacitor
6 (1) 0.01 uF capacitor
7 (1) 0.001 uF capacitor
8 (1) 110 ohm resistor 1/4 watt
9 (2) 330 ohm resistor 1/4 watt
10 (1) 1N4007 diode
11 (2) 1N4148 or 1N914 diode
12*(1) 10 ohm wire wound resistor 100 watt
13*(1) IRFPG50 Mosfet
14 (1) Mosfet Heat Sink
15*(1) .25 ohm resistor 3 to 5 watt
16 (3) 12 Volt lead acid batteries ( two matching for 24 volt bank )
17 (1) .047 uF capacitor

* - Items 12, 13 and 15 were hard to find ordered items
------------------------------------------------------------------------
I would not deviate from these components because they do work and produce a lot of heat plus this would be a common build for all that are participating if any improvements were to be made or found useful to the group build.

I'm glad you got your Mosfet and good luck with the build, if you need help in locating Items 12, 13 and 15 just PM me, there is several on-line places to get them

Best,
Glen

[eternalightwithin]

Yes, Radio shack has a NE555CP 555 timer. CMOS version, low power
They also have a NE556 "Dual Timer" chip with 14 pins.

Anyways I'm about to start my COP 17 Heater
Thanks TomKat for the IF50 Mosfet.

I'm doing the circuit updated as of August 9th. Didn't see the August 10th rendition will I was home with my parts. I ran into some potential problems with exact parts so I'm running it by you guys/gals while I'm wiring up my "smoke generator"

They didn't have 0.25Ohm resistors so I got a 0.47Ohm 5watt one instead. They had some 10 Ohm 10Watt resistors but I went for the 8 Ohm 20 Watter instead.

Do you think these two component changes will be a problem or can I still tune it with the 2 10K Pots I have?

Sincerely,

David

Hi Glen,

I don't know how important it is to have the same 555 as I have... NE555N. I bought them at radio shack. I learned from Peter that there are low power versions. I didn't know the existed.

I may be using more power than I need and might get better gains with a lower power 555. OR, there might be something to using the higher power one that enhances what is going on. I could try a lower power one and see what happens.

[FuzzyTomCat]

Hi Fuzzy. Seems like you got the same 'cooling effect' that Aaron saw - except over the pot which was also Aaron's result. But you've got a more general cooling as I think Aaron got heat on his switches. EDIT And an evident net loss to the battery? Is this consistent with the voltage you measured across the load?

May I say that this has got to be the tidiest set of results I've seen yet on the forum. Many, many thanks for this and for doing the test. Hopefully Harvey will add his comments here.

Hi Rosemary,
This was strange when I was taking my measurements because when my air conditioner is off in my upstairs living space in my home ( 13' x 32' ) the temperature rises slowly in the summer and it appears the ratio between Ambient Temperature compared to the Mosfet Temperature and 10 Ohm Resister Temperature stayed similarly always cooler. The temperature of the 555 potentiometer always warmer as the Ambient Temperature went up ...... this has to be the result of the circuit design somehow and how to show power gains or losses with this phenomenon.

The voltage loss is from the measurement across the battery with my DMM a net loss of what looks like .01 Volts +- in three hours with the circuit working and the cooling / heating thing, I attribute the first recorded losses in the battery voltage as maybe a fluffy area for the type of lead acid battery I'm using, it's not the same size or type as the one Aaron used much smaller. The replicated test circuit using the presets only ( with no potentiometer tuning ) on a tempered glass surface for insulation with my 10 ohm resistor elevated the potentiometer's on top of a folded piece of 22lb paper for less heat dissipation through the glass.

The next step to see if it is possible to tune the circuit with what equipment I have and see if the results can be better ....... but parts of the wave form looks like a DNA helix coil in person and is pretty fast for my 150 MHz scope almost like looking through the wrong end of some binoculars and driving a car.

The potentiometer presets are so very important that are posted from Aaron without a 250 MHz Analog Oscilloscope minimum real fine tuning may be hard.

Best,
Glen

[witsend]

1] Time (24 hour)
2] DC Battery Voltage
3] Ambient Temperature Degrees F
4] 555 Timer Temperature Degrees F
5] Mosfet Temperature Degrees F
6] 10 Ohm Resister Temperature F
7] 555 Potentiometer Temperature F


[ 1 ] | [ 2 ] | [ 3 ] | [ 4 ] | [ 5 ] | [ 6 ] | [ 7 ]

22:40 | 12.52 | 78.3 | 77.8 | 76.8 | 76.4 | 79.1
22:50 | 12.52 | 78.6 | 78.3 | 77.2 | 76.3 | 79.4
23:00 | 12.52 | 78.1 | 78.3 | 76.8 | 76.4 | 79.9
23:10 | 12.52 | 78.3 | 78.3 | 76.9 | 78.4 | 79.8
23:20 | 12.51 | 78.1 | 78.3 | 77.6 | 76.6 | 80.4
23:30 | 12.51 | 78.8 | 78.6 | 77.8 | 77.3 | 80.4
23:40 | 12.51 | 79.0 | 79.1 | 78.3 | 78.3 | 81.3
23:50 | 12.51 | 79.4 | 79.4 | 78.8 | 78.3 | 82.2
24:00 | 12.51 | 79.8 | 79.9 | 79.1 | 78.6 | 82.3
00:10 | 12.51 | 80.1 | 80.3 | 79.7 | 79.7 | 83.1
00:20 | 12.51 | 80.3 | 80.4 | 79.8 | 79.4 | 83.3
00:30 | 12.51 | 81.3 | 80.5 | 80.1 | 79.8 | 83.5
00:40 | 12.51 | 81.5 | 80.8 | 80.1 | 80.1 | 83.8
00:50 | 12.51 | 81.7 | 81.5 | 80.5 | 80.1 | 83.6
01:00 | 12.51 | 81.5 | 81.1 | 80.1 | 80.3 | 83.4
01:10 | 12.51 | 81.5 | 81.3 | 80.3 | 80.4 | 84.2
01:20 | 12.51 | 81.3 | 81.3 | 80.6 | 80.6 | 83.6
01:30 | 12.51 | 81.5 | 81.4 | 80.7 | 80.4 | 84.3
01:40 | 12.50 | 81.7 | 81.3 | 80.6 | 80.4 | 84.2

Now for some number crunching from anyone that is interested in finding the results ....

Glen

Hi Fuzzy. Seems like you got the same 'cooling effect' that Aaron saw - except over the pot which was also Aaron's result. But you've got a more general cooling as I think Aaron got heat on his switches. EDIT And an evident net loss to the battery? Is this consistent with the voltage you measured across the load?

May I say that this has got to be the tidiest set of results I've seen yet on the forum. Many, many thanks for this and for doing the test. Hopefully Harvey will add his comments here.

[FuzzyTomCat]

Hi everyone,

Here is my test of the battery draw down and temperature results for the "Ainslie-Murakami Negative Dominant Waveform Generator" Replication build PDF FILE

Parameters -

• Gate resistance: 53 ohms
• On pot resistance: 32.8 ohms
• Off pot resistance: 293.9 ohms
• NE555N power adjustment pot resistance: 93.1 ohms
• NE555N timer duty cycle: 55%
• NE555N timer frequency: 915 kHz
• Actual run frequency: 1.17 MHz

Oscilloscope - Tektronix 2445A 150 MHz 4-Channel

Channel 1 ground is on battery negative side of 0.25 ohm shunt and probe is at 555 negative rail side of shunt.

Channel 2 ground is on battery negative side of 1 ohm shunt and probe is at mosfet source side of shunt.

Channel 3 ground is on battery negative terminal while the probe is on the positive terminal.

SCOPE WAVE FORMS

Digital Multi Meter - Fluke 87

Non-Contact IR Thermometer - Cen-Tech 93983

Battery - Long #WP3-12 12V 3Ah Sealed Lead-Acid Battery

Test Data -

1] Time (24 hour)
2] DC Battery Voltage
3] Ambient Temperature Degrees F
4] 555 Timer Temperature Degrees F
5] Mosfet Temperature Degrees F
6] 10 Ohm Resister Temperature F
7] 555 Potentiometer Temperature F


[ 1 ] | [ 2 ] | [ 3 ] | [ 4 ] | [ 5 ] | [ 6 ] | [ 7 ]

22:40 | 12.52 | 78.3 | 77.8 | 76.8 | 76.4 | 79.1
22:50 | 12.52 | 78.6 | 78.3 | 77.2 | 76.3 | 79.4
23:00 | 12.52 | 78.1 | 78.3 | 76.8 | 76.4 | 79.9
23:10 | 12.52 | 78.3 | 78.3 | 76.9 | 78.4 | 79.8
23:20 | 12.51 | 78.1 | 78.3 | 77.6 | 76.6 | 80.4
23:30 | 12.51 | 78.8 | 78.6 | 77.8 | 77.3 | 80.4
23:40 | 12.51 | 79.0 | 79.1 | 78.3 | 78.3 | 81.3
23:50 | 12.51 | 79.4 | 79.4 | 78.8 | 78.3 | 82.2
24:00 | 12.51 | 79.8 | 79.9 | 79.1 | 78.6 | 82.3
00:10 | 12.51 | 80.1 | 80.3 | 79.7 | 79.7 | 83.1
00:20 | 12.51 | 80.3 | 80.4 | 79.8 | 79.4 | 83.3
00:30 | 12.51 | 81.3 | 80.5 | 80.1 | 79.8 | 83.5
00:40 | 12.51 | 81.5 | 80.8 | 80.1 | 80.1 | 83.8
00:50 | 12.51 | 81.7 | 81.5 | 80.5 | 80.1 | 83.6
01:00 | 12.51 | 81.5 | 81.1 | 80.1 | 80.3 | 83.4
01:10 | 12.51 | 81.5 | 81.3 | 80.3 | 80.4 | 84.2
01:20 | 12.51 | 81.3 | 81.3 | 80.6 | 80.6 | 83.6
01:30 | 12.51 | 81.5 | 81.4 | 80.7 | 80.4 | 84.3
01:40 | 12.50 | 81.7 | 81.3 | 80.6 | 80.4 | 84.2

Now for some number crunching from anyone that is interested in finding the results ....

Glen

[Joit]

Btw, what are all this calculations actually worth, when the Direction from Current is actually opposite,
and negative Values mean positive Values.

We are talking here about Spikes actually, not Frequencies, and we know it now often,
that it can not be measured well with any Meter, or exactly, how some seems it need to have it.

And well, sorry, when i cant show the same Results with the same Setup, because
i dont have the same Resistor from the same Brand, with the same Wires from this time
and the same Batteries and exactly same Wires for the connections.
All what left, is, to tune it, best as someone can instead pedantic looking over it.
It doenst help, to only have a Plan, everone have to walk the Way by his own.
Its better for them anyway, to wait, till they can buy it at Walmart.

Just a lot of wasting Time and blowing out hot Air in here again ?
I doubt anyway, that this Guy will ever rebuild it, and if,
someone here, what do care, if he do or not?
To do these measurements now is actually the Point, where you do rebuild it by your own, and start testing it by yourself,
in case, you are really interested, BEFORE you start discussing about.
Maybe just better to dont feed this Troll.

[poynt99]

DC Voltmeter Averaging Test

Hey Luc,

If you're out there and you're so inclined, this would make a great video demo. You've got all the gear to do it and I know you'd do it justice

Let me know if you're interested and I can help you with any questions. PM me at here or at OU.

.99

[poynt99]

Obtain an understanding of how a DC Voltmeter works and you will see that what I've said is true.

Better yet, prove it to yourself with your own meter. Anyone with a function generator and DMM can do this test. Using a square wave output to your meter, you will need either an offset adjust or duty cycle adjust on your FG, (or preferably both) in order to play with the settings to observe how the meter responds accordingly. A sine wave output will work also, but you will need an offset adjust on the generator. Go as high as 1MHz if you like, or higher. With a perfectly set 50% duty cycle, zero offset wave form, the meter should read 0.000V. Tweak either the duty cycle or the offset and observe the meter readings. A pulsed DC output will work also with this to give you the average voltage, but it will never go negative.

Once again, the purpose of this test is to determine the direction of the net current from or to the power source. It is not to determine the actual value of the net current, only its direction.

The purpose of using a DMM set on DC Voltage for this test is to average out the positive peaks and negative valleys, not to capture them. The DC Volt meter does this quite well because the averaging function is built in. The result is a net positive or negative voltage, which if measured across one of the current shunts, will indicate whether the current has a net flow into or out of the battery. This is an important measurement to verify the scope's indication of a negative net current into the battery.

.99

[jibbguy]

YOU do the test yourself, and see the obvious (i don't have to, as 17 years in the test and measurement industry provided me 16.9 years of experience enough to know it would be a worthless proposition):

You will have two choices for a DMM: Set to "RMS", or "DC"... Right?

The "AC Current" or "DC Current" settings are just RMS or DC voltage readings off of the internal shunt... Is this correct? BTW the internal shunt resistance must now then be considered a part of the circuit... The circuit's performance will be affected by the action of observing it (sounds positively "Quantum" lol).

> There is no such thing as "negative RMS" (which blows away the whole point of the measurement). So if trying to verify "dominant polarity" ("DC bias offset" value), RMS reading are out.

> Trying to read high F pulses with the "DC" setting will provide a meaningless jumble that changes every single time the display updates... If the DC bias offset was significantly above or below ground (it is not) it might serve as a very rough check.. But there are so many error factors involved that such subjective data from trying to compare how many "plus" and "minus" readings would be instantly discountable... By people like YOURSELF; if they weren't trying to beat a dead horse.

There is a type of signal conditioner called "Pulse Envelope Amplifier" used as a "front End" to an instrument that can give a DC-like line representation for a high F pulse signal (...showing only the Peaks, in a sort of "Peak and Hold" line). And this can also often be done as a post-capture math function to raw DC data. However, it always reads "positive" like RMS; and therefor is also not useful here.

If you want a better determinant of the "dominant" polarity of the signal, the best method is to Average the raw sample data. This IS THE ESTABLISHED METHOD, and whether you "like" it or not, it is still the proper way it is done in these cases. So it always comes back to the calibration of the Scope...

And i can tell you right now, that Tek's accuracy will defeat anything you could throw at it Have fun.

_________________

EDIT

Thx Fuzzy, my pleasure M8.. It is the very least i can do as brave and talented souls like you and the other great folks who Replicate are actually pulling the weight around here

[FuzzyTomCat]

OK we go over it again...

This is the problem: DMM's, and even bench multi-meters, are known to not accurately measure such fast non-sinusoidal and non-repetitive waveforms... Even if they have "True RMS" written on them.

Since this problem is fairly well known and often used as a favorite debunking argument, presumably you must know this yourself, so it is odd to hear you requesting it.

The reason they cannot accurately measure such "spiky" and non-repetitive waveforms is because internally, separate RMS circuitry is used... This RMS circuit physically conditions the signal before sending it to the A-to-D and display. The circuit has limitations in F Response; and some assumptions are made within its design for expecting repetitive signals, which are then not always true for non-repetitive. And when in RMS mode (the only DMM mode that could hope to meaningfully read any of these signals except the "Temps"), the result is always "positive"... There is no such thing as "negative RMS". And of course if left set to "DC" you would get meaningless constantly changing numbers and polarity... Wholly dependent on the display's update rate.

Sometimes in a well-balanced and stable circuit, you can "cheat" and put a cap & resistor across the terminals of the DMM set to DC and get a poor man's "Averaging" effect just as a rough check, but any messing around like that WILL affect this circuit that is balanced "on the head of a pin" impedance-wise; and is out of the question here: As this is NOT a stable and "well balanced" circuit, BY DESIGN!

...And this instability and extreme susceptibility to changes in impedance is also the reason why some detractors' yelling on the other forum for "using a capacitor instead of a battery" is absurd... the oscillation effect being studied in the first place would almost certainly be ruined by all that capacitance in the circuit (adding capacitance is the most common method for getting rid of "unwanted" astable oscillations in commercial electronics)... Lol that is what they want anyway; as those pesky little oscillations are obviously driving them crazy

And of course when trying to read current in-line with a DMM or analog Meter, voltage across an internal shunt resistor is used...So all the above restrictions apply. Analog panel Amp Meters that use inductance to move the needle are heavily reliant on operating within a frequency range... And would not be accurate in this case either.

Any Hall Effect Current Clamp-Ons i've ever seen have very serious restrictions in Frequency Responce... Way too slow here. The only way to accurately deduce current for this circuit is across a shunt; with either the scope to do the math on-board, or the math done on PC-stored raw samples. And that, as Aaron has pointed out, is how the Pro's do it (whether they use a digital storage oscilloscope, or a fast-sampling PC-based data acquisition system, the results would be exactly the same).

Scopes that have on-board measurement do not use an internal physical circuit to condition the AC signal into a flat "DC-like" representation before measuring it.... They use math to figure both Average and RMS based on the actual "raw" signal... By using calcs based on "area under the curve" that are computed in a rolling average. This is why they are the MOST reliable and accurate method in these situations.

This is exactly how Aaron has done it and we should applaud him for it

Hi Jibbguy,

I'm in total agreement with your statement and I think you would agree that one of the best DMM on the market the Fluke 87 cannot in any way capture fast non-sinusoidal and non-repetitive waveforms, this is the one I'm using and this is no "CHUMP" Digital Multi Meter bought from some hardware store. The proposed testing method using the 200ma DC mili-ammeter in the circuit ( mine is 400ma) is nothing but a waste of time and effort.

I only wish is that I have a faster scope so I could put this nonsense to bed for once and for all.

Thanks for your involvement in this adventure

Glen

[poynt99]

The DMM average current test is completely valid for determining the direction of net current. Do not be mislead Aaron.

I use good and cheap DMM's set on DC Voltage to test for net current flow down in the mV input range and they work perfectly well, and this at 1MHz input frequency. Square waves/pulses, 50% duty and otherwise.

Do the test for yourself and you will see how well it works. Again the goal here is to determine net current direction, not the value per se.

.99

[jibbguy]

That goal of a second source may be "worthy", but your suggested methods of accomplishing it certainly are not. Only another scope or sufficiently fast data acquisition system could perform this check. That's not opinion, that is fact. Whether you like it or accept it or not.

And say Aaron agrees to double-check it with another scope... Then once that is done, and the results seen to be the same... What's next: A third scope..?

STATE IT NOW, so we don't see an endless list of contrived drivel from you.

[poynt99]

Do the average current test as I have suggested using a DMM and the results shown by the Tek scope will either be supported or unsupported. If the latter is true then the results must be in question.

If the DMM reads "0.000" even on DC milli-amps or milli-volts, then either the net current is zero, or the meter's integrating A/D is too slow to provide enough signal to resolve. In either case, the meter reading would then have to be dismissed.

All we are looking for with this test is a solid "0.001" or "-0.001" (or better).

The goal is to provide a double-check against the scope reading.

.99