COP 17 Heater | Rosemary Ainslie

[Harvey]

The math is clear and engineering rules are very well known. In fact I was engineering some very similar circuits for quite some time now and I think I have a pretty good grasp of what's going on in various parts of circuit. If you use such inadequate components they will cause unnecessary power losses and you can't beat that no matter what you try. Sorry, it's the fact.

Anyway, I'm currently buried under a ton of paperwork for a commercial project so I doubt I will be going through 132 pages of discussion. However you're right I should leave you people alone figuring out things for yourself. In fact it's perhaps better that you learn on your own mistakes than having someone pointing them out for you.

Off I go.

Hi Lighty,

Most of us here are seasoned adults and apart from strong beliefs based primarily on life experience we are open to constructive criticism. Granted, there are a few that would base their belief structure solely on the mathematical models and conventional application by 'tried and true' results specifically designed to remove 'unwanted spurious effects' - and they can seem a bit egotistical and pedantic in a puerile fashion while in there own minds they consider themselves pragmatists with much to share. But for the most part, if you understand their maturity level you can pretty much skim over the ostentatious pleas for attention and glean some of the value from what they have to offer.

Someone seasoned as yourself with practical hands on experience with these types of things would be a refreshing change from derogatory norm we often get from the current critics.

Please feel free to share anything constructive you wish to share - I for one am always open to new views and perspectives, and heaven knows the mistakes I've made over the years - (wish I had put a dollar in the bank for each one of those ). So, if your here and would like to share something, please feel free. If it's something we've heard before, perhaps it needs repeating

Cheers,

[lighty]

@Harvey

First of all there has been a number of people contacting me via PM to give me their support. Thanks but I don't need any support because I'm not in any kind of conflict with anybody. I was also not offended or bullied by anybody least of all poynt99. I don't like poynt99's way of rude communication and he seems a way too eager to prove himself. However that doesn't mean I'm in any kind of conflict with him. I choose to ignore his offensive tone. There are people here who know what I do for living and what kind of projects I was involved with in the past and who are some of the people I was involved with. That's enough for me. I'm secure with my life and I am not inclined to get into discussions which serve no other purpose than to prove who is "better". Scientific argument is one thing but if you cannot get on the rational level of discussion with somebody than it's easier to ignore him or if it's easier to you then to put him on the ignore list. It's as simple as that.


As for the rest I did spend an hour or two to produce better driving circuit (optically isolated with Schmitt trigger and totem pole driving) and with simple recovery snubber circuit. Aaron has the schematic so it's up to him to share it now or after he checked circuit.

I also did some quick and dirty simulations to check everything out and it should work just fine. Interesting thing with simulation is that it differs significantly from the measured results when it comes down to the resistive inductor but I suppose that's because of inadequate model used. I used inductor model- set it to 30uH with equivalent series resistance set to 0.2 Ohm and of course it produced voltage of about 655V (peak). Then I set equivalent series resistance to 10 Ohm and it produced voltage peak of about 216V. Well, that was to be surprised because the resistance of the inductor itself would dissipate some of the energy by it's very nature. The interesting part is that Aaron told me that he got spikes of over 1000V. That's where real life measurement differs from simulation models. Also in the latter case impulse duration was about 400ns while in the former it was about 1.2uS. That puts lower resistance inductor at about three times as high peak voltage value with about three times shorter impulse. Again this is depending mainly on the simulation model used but it's interesting.

As a solution for unpredictable behaviour of simulator model one could measure real life resistive inductors under various conditions and if their behaviour is non-linear one could plot a curve and make look-up table for simulation in Spice (I'm using ProSpice simulation engine based on Spice 3f5, I don't use PSpice- mostly because Proteus simulation program can also simulate microcontrolers running code).

Ah and one last thing. When optimal frequency, duty cycle and power levels are established for a stable operation then a more optimised energy recovery snubber circuit can be calculated and that is an engineering skill on it's own. Simplest energy recovery snubbers like usually used "on the scene" are usually efficient up to 50-60%. More complex and calculated energy recovery snubber can go up to 80%.

In any case I have to go now. My bike is waiting and it's one of the rarely sunny days of this period of year so off I go.

[poynt99]

Quotes from Poynt.

It seems, Poynt. that you have elected yourself as some kind of critique adjudicator. I, as an amateur, know that the most critical value here is to determine the amount of energy delivered by the battery. The inclusion of no more than one shunt at -B. on the ENTIRE CIRCUIT is needed for this. Harvey's request that Aaron add more shunts was for purposes of determining the wave shapes during the switching cycles. Why do you not undestand this? Frankly - I'm concerned that you should so insistently require the extra shunts when you and everyone reading this thread know that only one shunt is required for determining the acutal power drawdown. The rest were simply included to allow Harvey, and not you, to determine the energy drawn and dissipated during the switching cycle.

Rose it has nothing to do with any objection I have regarding getting a total power draw, in fact I support it. This is being blown way out of proportion. Yes total P draw can be obtained this way, but for the sake of letting the 555 operate properly it is not the greatest idea. The very same thing was being accomplished when Aaron had a separate shunt for each--the powers are just added then. As I said, it's what ever you guys want to do regarding the shunts and circuitry in general--whatever works for you.

Where had you 'indulged' anyone for goodness sake? You were dancing. You were parading your requirement for 'niceties' and nonsense. All that was needed here was some effort to stress Spice parameters and 'explore'.

I took it upon myself in good faith to help Harvey out with his simulation. I did all that was asked, and I took it one step further asking for clarification as to the desired outputs because quite frankly what was being asked for clearly showed lack of knowledge both in using SPICE and in how the circuit operates.

I might add that I was chastised early in this project for even using SPICE (the posts can be quoted), and it's quite comical now that SPICE is not only being championed, but used to go way beyond what is possible in reality, and further away from the original Ainslie circuit. The value I used and Harvey was content with was 864uH, which is 100x the actual value.

I tried my best to get what Harvey was after, and since he was not forthcoming with the information I needed to complete the task, it was not done. Quite frankly, what he was asking for was nonsense, and I'm certain that one of his "discoveries" was that this is indeed so. I look forward to seeing his simulation...still waiting

Harvey is the only active member of this thread - thus far - who has proposed confidence in your objects regarding your interest in this thread and this forum.

Perhaps, along with the sparse support from yourself. And this is why as a positive gesture I indulged Harvey in his request for the simulation, not pre-judging where it may go. I was curious where he was going, but evidently he did not want to reveal enough to make it possible. Giving me latitude on values was not the issue, I already exercised that. The issue was measurements and measurement points. Harvey does not have a good grasp on this topic regarding SPICE, and I was trying to ascertain some clues from him as to what he REALLY wanted. I think I was going out of my way in performing this exercise. Strange that the same courtesy was not afforded to provide the information I needed.

How dare you refer to any contributor's efforts as 'techno-babble'. You need to withdraw this statement and apologise.

I'd prefer not to have to go back through all these pages and cite the countless times I've been slighted. A "contribution" is subject to point of view.

I resent the snide sarcasm. it's inappropriate in any context. I, for one will be reporting you to a moderator and I hope that other contributors will follow suit.

I think I have said this before but I will repeat: There is no need to "get me kicked off this topic or forum". Just simply ask me to leave and it will be done.

You have dismissed Lighty - whose expertise I, for one, would be most glad to enjoy. You are now insulting Harvey? I suspect you feel that by criticising him you pull yourself to some sort of level with him? How absurd. Harvey is a star player here. You're not even in the reserve. You know what the difference is here Poynt apart from his rank expertise? It's in the sincerity of the search. You self-evidently want this side to lose and any results that you disclose will thereby be discounted or dismissed - even at best. You have now over-played your hand.

I have not dismissed lighty. The only issue I had with him is how he entered this topic without prior knowledge of the history or where it was going and began to present all the solutions to problems that "didn't exist" because the circuit was already set as required (as poorly as it is in its present state). My impression is that lighty leans to the classical point of view regarding this circuit and its potential for OU (correct me if I'm wrong lighty), and if this is the case, he only becomes an asset to the classical side in terms of proving the convention of the circuit and its operation. In other words, his input here will only likely lend credence to the classical point of view that myself and a few others maintain.

.99

[poynt99]

The ammeter would have to placed between the HEXFET source and the battery. Therefore, if an ammeter can work there accurately as a single device, then so can a single current sensing resistor.

I have no qualms with this as I already stated. The circuit has been changed from the original. There is now only a single battery vs. 2 for example. The original circuit also does not account for the 555 circuit power, and hence only one shunt was used for the MOSET Source. The 555 was tied directly back to the battery negative. Aaron added a shunt to account for the 555 power which is fine, but tying the 555 to the MOSFET shunt instead is where I take issue. No big deal ok? You folks can do whatever you want.

I gave Poynt99 complete latitude to use the simulator as he sees fit to produce the desired results. I can only conclude that he cannot engineer the proper parts values or arrangement to produce those results. He has already shown willingness to deviate from the actual schematic by inserting an unnecessary resistance in series with the power source (evidently he thought he had to model extra source impendance) and he chose to use a pulse source rather than modeling the 555 circuit. Well, in his defense, I did the same thing at first - it is easier - but not a true simulation.

Again, it was not the latitude on the values that is at issue, it is what exactly Harvey was trying to measure. Why Harvey has such a difficult time understanding this is puzzling.

Inserting a resistance in the power source is not "unnecessary". If Harvey was astute he would realize this.

Using a pulse generator is valid and Rose has stated this a few times. If Harvey has achieved aperiodic oscillation with his 555 circuit then congratulations!, Sincerely. I did not try it yet (and I might), but unlike him, I have a great deal of experience with it and am not surprised at all that it is possible in SPICE. Indeed I surely would have shared that if I had made the attempt and achieved it.

.99

[poynt99]

Rose,

When I say that I am not surprised that SPICE is capable of achieving the "aperiodic oscillation", strictly speaking that is precisely and only what I mean.

That does not mean SPICE will indicate overunity. Aperiodic oscillation and overunity do not necessarily go hand-in-hand, especially in SPICE. Oscillation means just that--oscillation, albeit a strange or unstable one, but that is all it is in terms of how SPICE will treat it.


.99

[witsend]

Poynt I just don't know how to answer you unless I see some apology and an acknowledgement of a need for this. It seems that rather than address this issue you're simply adding more fuel to the fire. You have been insulting to Aaron, Harvey and the objects of these tests. Surely you see this? And surely - by now - you realise how much Aaron is trying to protect the standard of postings here?

[lighty]

My impression is that lighty leans to the classical point of view regarding this circuit and its potential for OU (correct me if I'm wrong lighty),

You are wrong. I worked with "non-classical" systems that had a number of similarities both regarding the some underlying principles as well as used materials. I cannot go further into that topic so let's just leave it like that.

You are quite right in the sense that it would be wise to first reproduce the original experiment. However that also means that same resistors should be used (same types and manufacturers- not just same inductance and resistance values) which is currently not possible so the experiment is differing from the start. Also, I assure you that cutting down losses can only do good to this particular circuit. Regardless of whether one choose to use tightly timed circuit (adjustable frequency and duty cycle) or drives it into self oscillation resistive and reactive losses must be addressed.

You are also quite right that one should use scopes with completely isolated inputs in order to prevent possible influence of common ground as well as using battery operated scopes in order to prevent incluence of physical grounding.

That being said, you're quite rude and regardless of your knowledge you will eventually offend people and lose their ear. It happened before and it will happen again so my advice is to tread carefully. You cannot speak in everybody's name and with such condescending tone. You never know who is the person to whom you're talking and how can that person help.

As for the rest I will continue to help behind the scene as I usually do. No need for anybody to PM me.

[Groundloop]

Group,

My first test with my RA replica did not turn out well. The ON time
was so high that the transistor heated too much. At 8 volt input
the circuit pushed 10 watt into the resistor. I'm going to change
the component values as pr. quantumschematic6.jpg and try again
to get the low 3,7% duty cycle ON time. My 10 Ohm 80 Watt resistor
has an inductive value of 6,5 μH (micro Henry). I have updated my
web site with the new revision B circuit. The "old" pcb can still be used
as is. Only change is that the 1N4007 do not need to be soldered into
the board.

Aaron,

If you want to try a "cleaned up" version of the circuit, just ask GOTOLUC
for a pcb. I hope he still has some left.

Regards,
Groundloop.

[poynt99]

Hey Groundloop.

Thanks, you've provided yet another verification that the circuit does not work as advertised regarding the 3.7% duty cycle. Hopefully changing the circuit values and/or making some slight modification on the PCB can make it work at the proper duty and frequency.

Also, could you please post the part number of your resistor? That is the closest I've seen yet for the original resistor used in RA's tests.

Thanks again.

.99

[poynt99]

You are wrong. I worked with "non-classical" systems that had a number of similarities both regarding the some underlying principles as well as used materials. I cannot go further into that topic so let's just leave it like that.

Ok, enough said I think. Now we know where you stand.

That being said, you're quite rude and regardless of your knowledge you will eventually offend people and lose their ear. It happened before and it will happen again so my advice is to tread carefully. You cannot speak in everybody's name and with such condescending tone. You never know who is the person to whom you're talking and how can that person help.

Busting into a thread the way you did is rude. I don't believe I'm speaking in anybody's name other than my own, and generally I'm quite a nice person until I'm treated poorly, at which time the gloves are off if pushed far enough. I stand up to the level at which is necessary in my own defense. Thanks for the advice but it's not required.

Anyway, I'm quite tired of and not really interested in all this babble. Let's try to stick to the facts, the experiments, and analysis of the results. I've got a video to complete.

.99

[Groundloop]

Point99,

The 10 Ohm resistor: GRF20/100

The firm: Danotherm Electric A/S

Found here:
Tubular Resistors; Vitreous Enamelled - Danotherm A/S

Data sheet here:
http://www.danotherm.dk/files/datash...rresistors.pdf

>>or making some slight modification on the PCB can make it work
There is no need to change my PCB as I stated in my post below.
Only the component values must be changed.

Regards,
Groundloop.

[poynt99]

Thanks GL.

Here's the complete part number according to their web site:

GRF20/100L 10R. Cost about $30 ea.

I see they offer SPICE thermal models

.99

[Groundloop]

Poynt99,

Yes, the firm offers thermal models for PSPICE.

They also do:

snip
"Custom designed resistors and assemblies are
available on request.
Details like wire configuration, creepage distance
and inductance can be specified by the customers.
The choice between more than 50 sizes guarantee
our customer that the best resistor configuration
can be found within our programme."
end snip

GL.

[Aaron]

Circuit proposed by Lighty

Here is the proposed schematic from Lighty. I have not had time to try it yet but I will try the recovery part of the circuit with some of the diodes I do have on hand. I have to make an order for a few different mosfets like the one shown or a IRFP450 and some diodes. I have everything else.

If anyone tries this out before I get the parts to do it myself, please post the results. Being able to optically isolate the load side of the circuit from the timer makes for easier measurements for sure. I tried some 555 variations with a H11D1 but I didn't get it to work like I hoped. In any case, this circuit is driven completely different than just a 555 with opto isolator.

D1 & D2 should be Hyperfast diodes.
D2 could be an SCR triggered by a zener or something to allow the
cap to build up to even higher voltage and the cap could be higher
capacitance as well if you want a cap discharge circuit dumping the
recovery back to the front battery like a Bedini system so it would be
like an Ainslie-Bedini hybrid.

D3 & D4 should be suppression diodes that are bidirectional such as 1.5KE400CA for example. AND these are optional as long as D1, D2 and C1
are working fine and are in good shape.

The circuit is set to run the load at 12volts.

If you want to modify it for 24volts, double the value of R3 & R4.

Q2 & Q4 is optional. If you don't want that, then just take the connection
from the Q1 collector straight over to R6.

And of course put the current sensing resistor at the battery ground. Use
as low of resistance as possible. I've been using 0.25 ohms because that is
what Rosemary used. Peter gave me a 0.05 ohm resistor that I used
originally but I'm not sure where it is at the moment but I will switch to that
when I find it.

[Aaron]

Control Data

Here is some data from one of the most boring tests in the world!

Anyway, it is using a BK Precision DC Power Supply 1711. I hooked up
an inductive resistor and increased voltage 1 volt at at time and over
10 ohms, it makes for easy measurements.

The most important data is the WATTAGE and ABOVE AMBIENT column.
But after spending all that time just waiting for the temps to stabilize
a few hours a day over a few days, I just had to make the graphs and
everything - I didn't spend all that time to just post 2 measly columns of
data so there you go - 1 entire page of high tech graphs and data!!

So with this identical resistor on the Ainslie circuit, at whatever stable
temp the resistor is at, I can simply look at how many C above ambient
and know about what wattage is supposed to be necessary to make that
heat.

The temp in this room is extremely stable so the ambient is almost always
about the same.

If you are running the Ainslie circuit and get 30C over ambient for X watts
and the watts are about the same shown by the control to get that temp,
DO NOT ASSUME your draw down will be 1:1 automatically. The Ainslie
circuit will outrun the control battery even if the wattages are shown to
be the same.

But of course this control data is very specific to specific temps. So
whatever your Ainslie circuit is running at, you'll need to do the control
test to match that EXACT temp to know the wattage necessary to base
your resistance requirement on the control battery to draw that same
wattage.

Here is a link to the spreadsheet and I'd recommend everyone do the same
for their own resistor on a control supply for quick reference as you're doing
tuning tests.

Control Data Spreadsheet xls format

Screenshot of data: