Circuit Analysis of the E-Cat QX with DC Exitation (Donald Anderson)

The following post has been submitted by Donald Anderson.

A photograph accompanied by much discussion in LENR blogs has been interpreted without regard to electrical circuit theory of voltages around a closed circuit.  The confusion is, I believe, because I believe that the E-Cat QX is an energy-producing device once excited, and that Andrea Rossi will withhold some knowledge until full demonstration and disclosure before the end of this October.

Consider the following statements which have been made by Andrea Rossi regarding the E-Cat QX, either in one of the papers published with he and Gullstrom in Arxiv on July 18,2017 or in response to questions in his blog:

  • The QX can produce a combination of heat (to 2600C), light, and electricity. If the total is say 20 watts (same as watt-hours per hour to the EE), it may include perhaps 10% electrical energy, perhaps 20 or 30 % light, and the rest is available as heat through a heat exchanger surrounding the lamp.
  • In the most recent paper, and as reported earlier, the energy source provided to ignite the “plasma” can be pure and simple dc, for example two 12V batteries.
  • The thermal output, by measuring temperature rise in a heat transfer oil in a heat exchanger surrounding the QX, and hence including absorption of any light, is stated to be 20 watts.
  • This heat transfer appears to be the temperature rise in 1.8 seconds.
  • Start-up occurs in perhaps a minute or so, and shut down can be in seconds.
  • Voltage measured across a one-ohm resistor is 0.105 volts, so the measured current is 0.105 amps or 105 milliamps. NOTE THAT THE SIGN OF THIS MEASURED CURRENT IS NOT OBVIOUS IN THE PHOTOGRAPH SHOWING TWO VOLTMETERS MEASURING ONLY ACROSS THE ONE OHM.

As an analog, suppose I were to construct a system involving a voltage source of 24 VDC, a one ohm resistor, and a tiny dc motor/generator attached to a tiny gasoline engine.  When used as a starter, if the motor has a very low impedance, nearly 24 volts would start to spin-up the motor.  The current would be limited to 24 amperes, and would decrease as rpm increase.  Given a fuel and spark, in perhaps a minute the current would drop as rpm built until-lo!- the engine starts at say 1000rpm.

Now assume that the engine is governed to idle at say 1100 rpm, and this the “starter”is viewed as a “generator”, producing 24.105 volts DC at that rpm. .  Yes, the one ohm resistor does have a current of 105 milliamps, but it measures the recharge of the battery!

I offer thus a possible explanation in accord with all released information for the E-Cat QX  listed above:  The QX is like a fluorescent lamp or arc lamp which cold-starts in a time of the order of minutes when excited at 24VDC, with a maximum current perhaps approaching 24 amps given the ballast resistor limit.  When producing energy once “ignited”, it produces 24.105 volts and recharges the battery.  Just as an automobile draws very large starting current, the generator replaces that energy in the battery.

Compare this to the purported argument presented by Rossi that the energy from the battery continues to be V^2/R or I^2 x R, about 11 milliwatts.  This suggests a COP, or thermal output of 20 watts divided by at 11 milliwatts, of over 1800.  Rather, once ignited, the QX has a net negative energy input and is in the self-sustaining mode (SSM).

Donald Anderson

  • Gerard McEk

    Hi Donald,
    I have been promoting the electricity generating QX since the second Gullström/Rossi paper for more or less the same reasons as you do.
    In my view can only this explains the curious way of how the input power is determined.
    I believe that if the reactor has a very high resistance while it is off, you need a high voltage start of say 200 V or higher. That can be generated with very simple things from DC voltage like with an interruptor and the bobine in a car. Obviously, if the metallic powders allow a low resistance between the electrodes, then maybe no high voltage is needed.
    Because the voltage of the controller drops to zero, the generated voltage is not oposite to the voltage of the controller (like the EMF in you example). Once in operation the voltage over the reactor is 0.1 V, so the internal resistance of the reactor causes an internal voltage drop equal to its internal generated voltage. I believe that Andrea Rossi once has admitted that the generated electricity is acting like current source. If that is the case than the external resistor can maximal be 200 ohm.

    • Dr. Mike

      In my opinion it doesn’t really matter if the E-Cat QX is generating a current or has a positive or negative voltage drop across it during its powered or its self-sustained mode. The system has a controller that is supplied with 24 volt dc source and is able to 1) supply a voltage necessary to start the device, 2) maintain a 0.1A dc current once the device is turned on, and 3) supply some unknown high frequency energy during the “powered” operation of the device (that would not show up on the dc voltmeters measuring the voltage across the 1 ohm resistor). Rossi will not reveal what amount of energy is being supplied to the device during the powered phase of operation (or during the start-up operation) in his demonstration in October so there is no way to calculate the COP of the device. The system COP could be easily calculated if the dc current supplied by the 24V to the controller is measured during both the powered and SSM period of operation of the device (with also measuring the time period of each phase of operation). IMO this will not happen.

      • Gerard McEk

        If AR will not allow a proper measurement of the input power, the whole demonstration is useless, dr. Mike and we are just as far as 5 years ago.
        If he uses a battery to drive it, the DC current and voltage of this battery supply should be measured. If he uses mains supply, the power from mains should be measured with a proper power meter.

        • Dr. Mike

          Agree 100%!

        • LindbergofSwed

          But it will be ssm for 67% of the time.

          • Gerard McEk

            Yes, so? Energy-in is cumulatief power over time, just as energy out. COP is EnergyOut/EnergyIn.

        • Wholewitt

          Many digital scopes (DSO) can measure power by using a current probe on one channel and a voltage probe on the other. It can handle fairly high frequencies usually limited by the response of the current probe and not only shows the waveforms but calculates the power in real time. This is what should be used.

          • Gerard McEk

            I agree, but I am not sure how accurate they do this. The main difficulty with power measurement is that the phase measurement (between I and U) becomes extremely sensitive at higher frequencies. Measuring distorted shapes, especially at higher frequencies is difficult. The better power meter suppliers have limits for harmonics at 20 kHz or so.
            Therefore, measuring power should be done at the low frequency side.
            If Andrea Rossi uses GHz pulses for stimulating LENR, no normal power meter would be able to measure that, unless you are aware of it and selected power meters for these frequencies (if you can find one for these circumstances).

  • Donald Anderson

    I taught circuit theory using voltage sources and current sources. Perhaps the QX once ignited generates a current flow of 105 mA to recharge the voltage source of 24VDC. This is 2.5 Watts, very close to 10% of the total energy generated (22.5Watts, thermal and electrical). Of course Rossi should be pleased if one can “start” with less than 24 amps for a short time, with the battery then being recharged. After all, a car battery usually lasts for about 5 years used to start the engine several times a day!

  • Frank Acland

    Mike Phalen asked on the JONP if AR could tell us how many Watts the Control system consumes. His reply:

    Andrea Rossi
    July 23, 2017 at 9:22 PM
    Mike Phalen:
    No and it is not relevant to the COP of the E-Cat QX. By the way, I can say that such consume mainly is due to the cooling system and the heat can be recovered. We have a strong overheating of the electronic components due to the high temperature of the reactor that by conductivity through the copper cables reaches the circuitry. It has been a struggle against the time: how to exchange heat soon enough and efficiently enough to avoid the burn of some essential component. We burnt a lot of control systems before finding the solution, because the normal thermostatic systems were useless, due to the time lapsed between the set and the real exchange of enough heat to save the components. It seems stupid, but it was not. We tried without success all the traditional systems, but for one reason or another, they were either not efficient or too “dirt” ( like immersing all in dielectric oil). Most of our unreliability issues came from this problem. At the end we found the solution.
    Warm Regards,

    • Dr. Mike

      However, the power used by a controller will determine the system COP. Even if each controller can operate 100 E-Cat QX’s, it is not evident that the amount of power used by the controller won’t be proportional to the number of units being controlled. So unless the controller power is independent of the of the number of units being controlled, or the controller power is much less than 1W, it is relevant to the system COP calculation.

      • Omega Z

        100 QX’s output x 20 watts would be 2 Kilowatts. In the past, Rossi’s controller was estimated I believe at 100 watts by those running the test(Hot cat). Actually, they didn’t say that specifically. It was just the energy unaccounted for and they assumed it was the controller consuming that energy as it din’t feed out beyond the controller. In effect, Rossi knowing this, and that he can operate 100 QX’s with it, all that is important for these tests is the COP of the QX. Everything else only matters when it becomes a product that is yet to be defined.

        • Dr. Mike

          !00W for a controller to operate 100 E-Cat QX’s outputting 20W each would yield an acceptable system COP of 20. However, the controller power and any other power needed to circulate a heat transfer fluid is what is required to calculate the system COP. The system COP is what would be important for a commercial device.

          • Chapman

            I am not arguing against your point, but you are only considering the commercial viability of the unit based upon secondary power/control requirements.

            That is all well and good, and a proper observation, but the first issue is the proof of an over-unity reaction, and that is specifically the power in vs power out of the QuarkX alone, ignoring the power supply and control consumption.

            At the heart of the matter is the question “Can a LENR reaction actually tap binding energy and produce more power than the reaction consumes?”. Once PROVEN, then we move on to questions as to whether it can be operated in a manner that allows us to UTILIZE that gain.

            So again, you are right as far as the Big Picture goes, and what can go to market, but even if the QuarkX is shown to produce an observable power gain from the fuel, yet the control circuitry is either outrageously expensive, or requires a supercomputer, or consumes by itself more power than the Quark can generate, it will STILL be the first absolute proof of the LENR reaction’s validity.

          • Dr. Mike

            I agree that it would be great to actually measure the COP of the device, but I just don’t see that happening in Rossi’s demonstration. He will not put an oscilloscope across his device or across the output of the controller (or across the 1 ohm resistor). We will have to take his word as to the input power the device is receiving. I was hoping that there would be nothing proprietary about the battery current, which might mean Rossi would show the output current of the battery with a sampling resistor and an oscilloscope. Now that Rossi has stated the power to the controller is not relevant, does anyone believe his demonstration will include measurements to show the total power/energy delivered by the battery?

          • Chapman

            I sympathize with your consternation.

            The truth is that if he feeds his control circuitry from a fixed DC power source, he can have a Watt meter at that point to track the full power delivered TO the control circuit and reactor system with no need for a scope, and provide the needed data without any chance of exposing downstream ac secrets.

            He can argue that the control system power consumption is irrelevant to the internal reactor power gain, but that leaves the possibility of a hidden AC component from upstream directly contributing to the output and going unreported by the DC Voltmeter on the bias resistor being used to prove reactor current. Now, we have seen that at every other opportunity he says he ignores potentially favorable values to force the most conservative estimates, so accepting a lower provable COP in order to keep the signal profile secret seems like a no brainer, and would eliminate all doubts. A demonstrated minimum COP of only 10 would still make everyone flip out and crack the champagne, especially when we know it is a sloppy “guaranteed bare minimum” gain orchestrated to conceal said signal details.

            I get it that you are NOT concerned about the AC component simply to expose the secret majic frequency, but rather because any real power metric MUST cover all DC as well as AC components. Hence the scope.

            But a strict absolute raw power reading at the primary DC source will give everyone just that. And the resulting COP can be touted as a “and that is INCLUDING all the loss in the first generation control system, which we can now work to optimize and thereby increase the COP beyond even the miracle level you see before you today!” kind of thing.

          • Dr. Mike

            My assumption in that the controller will draw one current when the device is in the “power” mode and a much lower current when the device is in the “SSM” mode. If the controller only operates in these two modes and the Watt meter is capable of measuring the total energy delivered to the controller while switching between its two operating modes, it should be possible to calculate the system COP (with an accurate measurement of the output power). I don’t think we will see a power meter hooked up to the supply power. I hope that I am wrong!

          • Chapman

            And that is pretty much what I was betting you would say. That is, that you were not demanding the scope for the sake of peeking at the specific AC waveform parameters, but merely holding out for a valid and comprehensive demonstration of the true power involved, and that any OTHER method that clearly demonstrates power at all points, AC and DC alike, would be perfectly acceptable.

            It DOES seem like a very simple issue to address, and it would certainly eliminate a lot of after-test doubt and second guessing like we had with past tests. It is on par with Bob’s observation about nitrogen in water messing with calorimetry. It is not a disqualifying thing, but why not take it into consideration and incorporate it in the test if it will raise the general confidence level, cost nothing to implement, and NOT compromise trade secrets? At this point, you would think that if Rossi was going to put himself through another of the rectal probes that seems to inevitably follow each public demonstration, he would be taking feedback and tailoring the test to ensure it was a home run!

            Let me be real clear here… Rossi owes me/you/us NOTHING. I am not saying that he NEEDS to do these things to satisfy ME. I am just observing that he sets himself up for frustration by holding demonstrations that are predestined to be perceived as suspicious, incomplete, and of questionable validity. If he does not want to go through the motions and do it right, just don’t do one – and tell everyone to stuff it! That would be a less stressful path FOR HIM than to do another demo that everyone is going to pick apart and question. Me? I love the guy, and I don’t need him to jump through hoops. In fact, I have more fun trying to deduce what he is doing than I would if he just published everything outright! Were would be the fun in that?

            But the purpose of my post was simply to let you know that I recognized you were making a valid point, and what your motivations were in making it. Not that I feel that my opinion matters to the universe, or that it somehow validates your thoughts, but sometimes it is just nice to know you are not screaming into the void, and that there is at least one person out there who got the point you were making. 🙂

          • Dr. Mike

            Thanks for your response. Just so you know, I’m all for Rossi maintaining his secrets until he has them patented and therefore can easily see why he would not want an oscilloscope connected across the device, the output of the controller, or the 1 ohm resistor. That’s why I believe we will end up having to take his word on the input power. (Not very encouraging for anyone that has read much of the pre-trial evidence from the recent court battle with IH.)

            I agree that it is more fun for those trying figure out what is going on in his device than if he just told us. However, I’m not sure that Rossi has a clear understanding of how his device really works based on published theories to which he has allowed his name to be attached. Also, any paper Rossi makes public can not contain any unknown information about his device if he intends to include that information as part of a patent claim. I believe that is the reason why there just doesn’t seem to be any real information in the published paper. I don’t expect to see any useful information on the E-Cat QX until Rossi has a patent issued. Also, if improvements are made to the device, we really shouldn’t expect to here about those improvements until after they have been protected with additional patents.

          • Axil Axil

            My assumption is that there is an LENR reaction activation field that consumes very little power, When the field is activated, the LENR reaction is active, and when the activation field is off, the LENR reaction terminates.

            Fabiani states:

            “Over the years we realized that the reaction needs more stimuli than only heating. Everyone thinks that thermal stimulus is enough but that’s just the beginning. It’s not enough for maximum efficiency. It’s the base, the synthesis of the reaction. But the reaction has almost behaviors as of living matter, and it has responses as a function of the stimuli. They can be of many types other than thermal. And these are the ones that trigger, let’s call it the most fun part of it, allowing excellent gains in terms of response to the stimuli.”

            “With the failures, I found myself having to believe in it. Why? Because when something fails, you see the behavior of the object. The next time you adjust it, then you see that it behaves very differently. And then you realize that it is something unique. We have it all filmed, which still cannot be disclosed. We have photographs of creatures that emit pure light that have completely melted the reactor down, all in a very quiet way. You just turn off the stimuli system and the reaction is switched off. It’s impressive.”

    • f sedei

      Heat control has been a problem from day one. Has ceramics played any part of the control solution?

  • Donald Anderson

    My analogue was a single motor/generator which could start an engine with a voltage source and current flow, multiplied together, giving positive energy input. Once the engine is running, the current flow reverses and hence the input energy becomes negative. Thus, the QX once excited will have a voltage higher than 24 volts and power flow is out of the QX. I predict that the COP (say just thermal power out divided by power from the mains or from a battery) will be a negative large number.

  • GiveADogABone

    If you believe that the QX operates as a Hydrogen filled, gas discharge tube, the sweet spot for running the QX in a stable fashion is at (i1,v1) or maximum current at minimum voltage, hence minimum resistance and maximum CoP of the gas discharge tube). If you want AC power generation between i1 and i2, then you need a DC bias current that has to be supplied by a controller :-
    negative differential resistance (NDR), dv/di<0

    Current controlled negative resistance (CCNR, open-circuit stable,… "S" type): In this type, the dual of the VCNR, the voltage is a single valued function of the current, but the current is a multivalued function of the voltage.[77]In the most common type, with one negative resistance region, the graph is a curve shaped like the letter "S". Devices with this type of negative resistance … and gas discharge tubes like thyratron tubes,

    The maximum AC output power is limited by size of the negative resistance region (v1,v2,i1,i2 in graph above)
    A thyratron is a type of gas-filled tube used as a high-power electrical switch … Thyratrons can handle much greater currents than similar hard-vacuum tubes. Electron multiplication occurs when the gas becomes ionized, producing a phenomenon known as Townsend discharge. Gases used include… and (in special high-voltage applications or applications requiring very short switching times) hydrogen

  • Donald Anderson

    To GiveADogABone: There are four quadrants for I and V. If the QX is physically symmetrical, only need to look at two. What is different for the QX is that when it goes excited, the quadrant with same positive voltage but negative current defines the operating point. Wow!

    • GiveADogABone

      Rossi finding and using NDR. Awesome!

      The ‘S’ curve starts from the origin (I and V are zero) with something like a straight line (constant resistance). Then the gas breakdown starts at V2. More current (and this is a current driven device) drives the resistance down to V1. The best explanation that I have found so far is :-
      In plasma physics it is well known that the S-type NDR is related to the appearance and disruption of a complex space charge structure (e.g. double layers, multiple double layers, etc.) [1,2,3], whereas the N-type NDR is related to the spatio-temporal dynamics of such a structure [4,5], or to the onset of lowfrequency instabilities [6,7,8].

      A negative resistance requires an active component in the electrical circuit able to act as a source of energy. In plasma systems, this component could be a selfconsistent double layer existing at the border of a fireball. The double layer works as a nonlinear element of circuit able to convert thermal energy into electrical energy, creating all the conditions necessary for the appearance of the S-type NDR in the current-voltage characteristic of a plasma conductor.

    • GiveADogABone

      It is worth noting that this is not a voltage driven device. If you supply a low impedance voltage source the current can have three values and the gas tube be unstable. That is why the power source MUST be a current source which Rossi sets to 0.1A.

      By contrast, type ‘N’ devices need a voltage source.

  • sam

    Andrea Rossi
    July 26, 2017 at 6:38 AM
    Today we are making substantial improvements to raise the power of the apparatus that will be presented in the demonstration.
    Warm Regards,

  • GiveADogABone
    AC electricity generator using NDR from an ‘N’ type gas tube
    Circuit diagram provided with battery power supply and ac load

  • Ted Rygas

    Ted-Z: (now with the schematic)… sorry for repeating.
    My view on Quark-QX:
    1. The device is a high temperature version of a Xenon-strobo–lamp, operating “almost continuously” and combining the “flash” effect with LENR. It produces high currents in a plasma containing Nickel and Hydrogen. Xenon strobo-lamps are known as having “zero” resistance in the pulses. The pulses are very short, about 20 nano-seconds, but the currents are very high, for example 2000A (the back-EMF can provide that). Please note that there are claims of nuclear transmutations in high currents by other scientists (Balutov and several others).
    2. Xenon is most likely replaced with Caesium or Rubidium, as these elements are less likely to escape from the hot enclosure of the “Rossi’s LENR lamp =Quark-QX”.
    3. The primary circuit (24V) is most likely “chopped” at high frequency to create bask-EMF. The back-EMF goes into a resonance in the secondary circuit containing a capacitor and a coil (inductance).
    4. For the replicators: addition of LiCO3 could create some carbon monoxide in the H2 atmosphere and promote volatilization of nickel (through the formation of nickel carbonyls), thus making the operating conditions easier, for example 1500 deg. C.
    5. For some reason, the posting system (DISQUS) is not accepting my drawing (jpeg or bmp), so I have to describe it verbally:
    – primary circuit: 24V DC, 1 Ohm resistor, plus a “chopper” (in-line), connected to the Quark-QX electrodes.
    – secondary circuit: capacitor plus inductance, connected to the Quark-QX electrodes.
    – tertiary circuit: rectifying bridge, one side connected to the Quark-QX electrode and the other side “after the coil” ===> Quark-QX-electrode—coil—the second side of the rectifying bridge.
    For the schematics you can email me at
    ((( This information is “PRO PUBLICO BONO”.)))

  • GiveADogABone

    An explanation of NDR at the atomic level :
    December 30, 2016
    Negative Resistance with a Single Atom

    Negative differential resistance (NDR) refers to current decreasing as voltage increases, contrary to a normal resistor. The phenomenon is useful in electronics, and now a research team has demonstrated a reliable form of single-atom NDR and has explained in detail how it works.

    They found that current mostly increased with increasing voltage, except over the range of about 1.2 to 1.4 V, where they observed NDR for the dangling bond.

    As the voltage rises further, the Fermi level becomes low enough that an electron occasionally jumps from the lower dangling bond state. This energy state refills much more slowly because its energy is far below that of the conduction band, and the upper state cannot refill until the lower one does. The computer model showed that this “refill delay” leads to NDR, because the flow of electrons is held back further as the voltage increases.

  • GiveADogABone

    QX output – 20W of heat, light and electricity.
    We need to generate a current 0f 0.1A at a few volts directly from the QX.
    How to do it?
    1: Thermoelectric (Seebeck Effect) generators
    2: Solar PV cells
    3: AC to DC electricity conversion.

    Then the QX demo box has no electrical power input and no CoP, even if it needs external power to start.

  • Abd Ul-Rahman Lomax

    Chopped DC isn’t what is implied by DC The paper purports that the power supply to the device is DC, with 100 mA current, measured using a DC voltage drop on a 1 ohm resistor. There could be far higher AC currents, that could be merely the DC component, an offset.

    So if there is complex power being applied, the discussion in the paper is highly misleading.

    Bottom line, the paper does not disclose crucial details, so analysis (other than revealing what is missing and possible internal contradictions) is highly speculative and generally useless.