Dave, It seems that your model of heat conductivity leads to a system equation that's a linear first order differential equation, if I'm not mistaken. That's a tractable system to deal with from a simulation and control point of view, and as such lends itself to numerical optimisation techniques.
Andrew ----- Original Message ----- From: David Roberson To: vortex-l@eskimo.com Sent: Saturday, May 25, 2013 9:36 AM Subject: Re: [Vo]: ECAT Drive PWM Issues Fran, my model takes into account the rate of heat transfer out of the device by using a parameter that simulates a thermal positive feedback loop. And, as you suggest this depends greatly upon the rate of heat generation with temperature and the thermal resistance that it delivers that heat into. Another way to think of this effect is to consider what would happen to a block of active material which is surrounded by a perfect heat conductor. In this special case, any additional heat that is generated is immediately absorbed by the conductor and can not raise the temperature of the block. This would be a stable condition and the COP would be low. Now, if you modify the surrounding heat conductor by increasing its thermal resistance then any newly generated heat from within the block would result in an increase in its internal temperature in a positive feedback manner. The resistance can be increased until it reaches a point such that a tiny incremental input of heat to the block results in a temperature increase of the block that causes additional heat generation slightly larger than the initial increment. Rossi appears to operate above this resistance point when his device has the desired performance. That was a lot of words and I suspect is not clearly written. The meat of the description is that there will be a temperature that depends upon the heat sinking where the device becomes unstable and begins to proceed toward melting. My model suggests that this is the temperature above which Rossi should operate his device to achieve good COP. The model further indicates that you can maintain control of the device while operating above this point as long as you reverse the process before a second temperature trip point is reached that leads to run away. It is important to realize that operation within this region is unstable unless a drive waveform is applied with the proper characteristics. In the radio world this type of device would be referred to as a negative resistance component. Rossi must be relying upon the energy generated in this mode for his large gain. The hard part is to keep the ECAT from getting out of control since he is operating on a sharp balance to obtain good COP. I am not modeling any process that occurs beyond the two temperature trips that I described since operation above the second one is destructive. Operation below the first temperature point results in a COP that is too low to be useful. I have included energy loss due to a 4th order radiation process in some of my runs, but so far I find that control issues occur before this has significant effect. I believe as you do that operation with a heat exchange fluid will be easier to control. This also allows Rossi to adjust the flow rate which could be used to modify the thermal resistance factor and thus total loop dynamics. For example, he could raise the temperature at which the core become unstable thereby compensating for different core activities. My model operates upon the average behavior of an ECAT type device. It assumes that the design has been developed by good engineering processes. If the design team allows the system to harbor inconsistent heat transfer such as would occur with too many and too large in size hot spots, then there is no control technique that will work effectively. I suspect that much effort will center around making sure this issue is handled. Dave -----Original Message----- From: francis <froarty...@comcast.net> To: vortex-l <vortex-l@eskimo.com> Sent: Sat, May 25, 2013 7:16 am Subject: re: [Vo]: ECAT Drive PWM Issues Dave, I think you we are both in agreement with the initial post of Ed’s thermal analysis, http://www.mail-archive.com/vortex-l%40eskimo.com/msg80803.html but it does not mention the difference between the destructive test in open air and the unit in normal operation which is constantly bathed in a heat extracting fluid.. are you modeling this in your SPICE calculation? The thermal circuit in the destructive test only has air cooling to keep the runaway at bay and represents a softer – more fragile target for the waveforms to temporarily exceed while I think the reactor in heavy heat sinking mode would have much higher tolerance for controlled PWM excursions into areas that would be considered runaway if not for the steady drain. Fran [Vo]: ECAT Drive PWM Issues David Roberson Fri, 24 May 2013 23:30:52 -0700 I was adjusting my spice model of the ECAT when I decided to determine how important it is to keep the device operating within the normally unstable region at all times. Here I refer to the unstable region as that operation range where the ECAT would tend toward over heating unless under control. There is no end to the questions which keep arising as to how heat can be applied in the proper format to keep an unstable device operating under control when it is capable of putting out more heat than required to drive it. And, the ECAT tends to operate best when the COP is equal to 6 which clearly is within this mode. One day this will be accepted. For now, I want to mention that it is important to keep the ECAT operating near the ultimate thermal run away region. If the device temperature is allowed to drop too far before the drive returns then the COP degrades significantly. And, as is somewhat demonstrated by the waveforms shown in the recent report, the length of time that the temperature hesitates at its greatest level is determined by how by Coupon Companion" id="_GPLITA_0"close to that ultimate run away temperature the device operates. My test runs demonstrate that the ECAT needs to be operating at a maximum temperature near to its ultimate thermal run away point and that the variation in output temperature needs to be maintained low by timing of the PWM drive. Both of these requirements should be met if the ECAT is to deliver the desired COP of 6 and remain stable. My spice model offers good guidance even though it can only approximate a real device. Dave
