I guess the concept of a temperature-compensated reference voltage is a brand spanking new idea for you, old chum.
Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Monday, May 27, 2013 12:39 PM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? Come on Andrew. You need to research your old books on control theory. Where are you mounting that fixed voltage output thermocouple you speak of? The last time I checked the output depended upon the temperature to which it is subjected. Since your thermocouple will change readings as the device heats up or cools, your reference will drift likewise. This will force the controlled device to head toward one direction or the other. This will be anything but constant old boy. Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Mon, May 27, 2013 3:22 pm Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? What? In the control regulation, everything is represented as either a voltage or a current (because it's, like, electronics, duh). Normally, temperature comes out of a thermocouple and is thus a voltage. The reference voltage, to which the actual temperature voltage is compared in order to generate an error signal for regulation, will be a fixed voltage representing the set-point temperature, as would be output by that thermocouple at the set-point temperature. Why is this so hard for you to understand? Your characterisation of the ease of regulation of a system with intrinsic positive feedback is grossly over-simplified. Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Monday, May 27, 2013 11:44 AM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? I do not follow how the set point can be the operating temperature. How is this inputting to the comparator? Are you proposing some external heat source which remains constant at that temperature? For a loop to function it must have a reference that does not change with the controlled parameter. If this is not the case, then the temperature will drift toward one of its limits. The beauty of positive feedback is that this type of behavior is exactly what you desire. As long as you can reverse the drift direction periodically you are in control. Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Mon, May 27, 2013 2:23 pm Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? Sure, the reference would be the set point, and that's simply the operating temperature. Notionally you set this as high as possible, consistent with materials integrity and the ability to regulate a strongly intrinsically-positive feedback system (the device itself). The idea is that you end up with Kp*Kn <= 1, where Kp is the intrinsic positive feedback gain (>1, and becoming higher at higher temperature), and Kn is the negative feedback gain (<1, representing the characteristics of the active cooling system). Of course, it's more complex than that due to first (and higher) order time differentials, and an integral term due to stored heat energy, but that's the basic proportional rule. Designing that would be fun. The most fun I had in my 40+ engineering career was designing industrial robots. Right now, I'm looking for a new job. Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Monday, May 27, 2013 10:49 AM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? Rossi keeps this information secret. It is unfortunate that he does this, but that is his nature. I would love to see a number of measurements associated with his material, but all questions of that sort are blocked due to IP concerns. It is frustrating to be kept at arms length from such important and history making knowledge. You mention active cooling in the context of negative feedback and I suppose that might be somewhat applicable. Systems can be stabilized by adding an overall negative feedback loop around the process but in this case I do not see how any form of reference temperature can be used to generate an error signal for correction. Do you detect a reference upon which this loop would act? Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Mon, May 27, 2013 1:30 pm Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? OK, thanks for the info - I had not seen those reports. Certainly it is in general expected to happen if it's known that the reaction rate increases with temperature. So the trick with active negative feedback (cooling) applied at higher temperature is that this technique holds the promise for much higher COP values. Indeed, an excellently engineered device promises to be very hot, to be under complete temperature control, and to perhaps to generate double digit COP values. Assuming that at some point Rossi licences this technology, the thermal control and the temperature operating point look like they would be key market differentiators. Do we have data as to how low the temperature can go, and still maintain over-unity COP? Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Monday, May 27, 2013 10:18 AM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? The earlier posts by Rossi on his blog mention many cases where thermal run away happened. Most of these were when he was developing the earlier versions of his mechanism. The fact that thermal run away can occur has been common knowledge for a very long time. Anytime a positive temperature coefficient is present thermal runaway is possible under certain conditions. Power transistors are a prime example of this when they self destruct unless the heat sinking is adequate to reduce the thermal resistance so that the positive feedback loop gain is below unity. Rossi has a similar problem to deal with. In his case, he is using what is normally a problem to his advantage to improve his COP. Without this help he would have a far lower COP. You get a COP of 1 for free, and much beyond that might result in unstable operation. Even operating at a COP of 3 has risk of thermal run away. Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Mon, May 27, 2013 1:03 pm Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? Glad we're back in sync. Although there's definite evidence for thermal runaway 25 years ago with P&F, with Rossi's kit I'm not so certain. In fact, I don't know of a single example. He only got the meltdown when he applied continuous power at a level far above that which he uses now. Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Monday, May 27, 2013 7:00 AM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? I suppose that it would be easier in person to discuss this issue, but that is not available. Yes, we are on the same page regarding the positive feedback threshold leading to self destruction. I refer to what you mention as active cooling of the system. We have discussed this in vortex on several occasions in the past. I think that it is a winning idea, but so far I have not detected Rossi putting it into his design. It appears to be a technique that would allow Rossi to force the loop gain back to below unity at an elevated temperature that would normally be beyond recovery with heat input modulation alone. This should result in a downward retreat of his temperature excursion and looks very promising for high power operation. Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Mon, May 27, 2013 1:35 am Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? We are totally at cross-purposes here; if we were in the same room, this crap wouldn't happen. So here's the deal. I'm considering the scenario whereby we operate the heating system to bring the device just past the stable temperature; further heating results in thermal runaway (at least, that's what's claimed for Rossi's device - it actually melted down due to the application of constant heating, but whatever). To keep the thing stable when it wants to apply positive feedback to itself, we need to apply negative feedback. And hence I began to discuss and describe characteristics desirable of an active cooling system. You dig? ----- Original Message ----- From: David Roberson To: [email protected] Sent: Sunday, May 26, 2013 5:22 PM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? But, we are talking about the ECAT. It operates by using positive feedback to get high gain. You are the one that mentioned a negative feedback system that achieves the same thing. That is not comparable. Stable operation of negative feedback systems is trivial. Think of taking a tunnel diode and keeping it within the negative resistance region without heavy resistive loading. The problem is similar to that which Rossi faces. Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Sun, May 26, 2013 8:14 pm Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? Of course I'm talking exclusively about a negative feedback system!! The positive feedback purportedly occurs internally to the device itself. Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Sunday, May 26, 2013 5:09 PM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? No, there is a large difference between a negative feedback system and a positive feedback system. Tell us how to make your temperature controller hold a constant temperature with positive feedback and a loop gain of greater than 1. If you do, you might find that it matches my model. Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Sun, May 26, 2013 8:05 pm Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? See my follow-up on this. There's always going to be a tracking error, no matter how sophisticated the regulation algorithm. I think the prime objective here is not to have absolutely constant temperature per se; rather, it's to guarantee that thermal runaway cannot occur. Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Sunday, May 26, 2013 5:00 PM Subject: Re: [Vo]:Re: Constant temperature Operation of ECAT? How many of these controllers use positve thermal feedback to keep the sink at a constant temperature? Dave -----Original Message----- From: Andrew <[email protected]> To: vortex-l <[email protected]> Sent: Sun, May 26, 2013 7:52 pm Subject: [Vo]:Re: Constant temperature Operation of ECAT? Seems to me that if active cooling control is used as the only control input, thus satisfying the "unplug it!" sceptics (and I'm one of them), then it only has a chance of working if there is good thermal contact and good thermal conductivity and substantial enough heat capacity in the active cooling implementation. I don't know why this is supposed to be hard. Gaming PC's of the high-end variety use this all the time. Prompt temperature feedback to the cooling pump is all that's needed, plus a simple PID controller. This is very well-known technology. Andrew ----- Original Message ----- From: David Roberson To: [email protected] Sent: Sunday, May 26, 2013 4:44 PM Subject: [Vo]: Constant temperature Operation of ECAT? My model demonstrates that constant temperature operation of the ECAT is not going to work under normal conditions. The relatively high value of COP when temperature control is used depends upon operation in a positive feedback region. This can be thought of as related to the question that always arises about why the device does not supply its own drive and therefore run continuously in SSM. Once the loop gain becomes greater than 1, the device will tend to move in the direction that it is currently heading. This allows it to heat up to a relatively larger temperature than that due to the drive alone. When rising in temperature, the device begins to put out additional heat, more with time. The trick is to turn the process around at a good point before it goes too far. The best turn around temperature is well defined and shows up as a tendency for the device to continue putting out power at a constant rate with time. Unfortunately, this exact point would be impossible to achieve while maintaining control. It is a balance between how long you want the temperature to remain nearly constant and the risk of loosing control. Rossi chose a relatively safe turn around temperature for the last test which caused the COP to drop below his desired value of 6. I suspect he chose this because a COP of 3 well demonstrates that the process is real and also has enough margin to keep the device safe from melt down. I think I would have done the same under the same constraints. Dave
