E-Cat QX Presentation Thread #3 — Full Video of Event, Comments from Mats Lewan, Measurements By Eng. William Hurley

The video of the event has been published by ecat.com here: https://www.youtube.com/watch?v=Nz0Z94Ix-kc

The following comments have been sent to me from Mats Lewan. Since the film of the presentation is yet to be produced, this helps provide us with some more details of what was measured today.

‘I think the demonstration today went well, with some limits that depends on what Rossi will accept to measure publicly. The problematic part is that the voltage over the reactor could not be measured, which would be necessary to calculate the electric power consumed by the reactor. In the calculations made by Rossi and Eng. William S. Hurley, who oversaw the measurements, the power consumed by the 1-ohm resistor was used as input power instead, assuming that the plasma inside the reactor has a resistance close to that of a conductor, thus consuming a negligible amount of power since the voltage across the reactor would be very low.

The dummy measurements that I insisted to do, which can be seen in the slides, consisted in replacing the reactor, first with a conductor, then with a 800-ohm resistance. Using a conductor gave a similar electric situation as when the reactor was running—a voltage across the 1-ohm resistor of about 0.4V, slightly higher than the 0.3V measured with the reactor in the circuit. Using a 800-ohm resistor instead of the reactor the voltage across the 1-ohm resistor was about 20mV. At that point we also measured the total voltage over both the 1-ohm and the 800-ohm resistance together, basically the output voltage of the black box power supply, which was then about 12V. That is consistent with the 20mV voltage across the 1-ohm resistor (it should be 15mV).

The 800-ohm dummy was used since at the current flowing through the circuit with the reactor – about 0.3A – the reactor would have consumed 72W electric power if it had a resistance of 800ohms. However, this would have meant that that the voltage across the reactor would have been about 240V. The dummy measurement with the 800-ohm resistor indicated three things: 1. The voltage from the power supply only reached only about 12V as a maximum. 2. If the reactor had a resistance of that order of magnitude it would have resulted in a much lower current, about 20mA, than the one measured with the reactor in place. 3. The power consumed by such a resistance in the circuit would only be about 0.3W.

From the two dummy measurements we can also conclude that the black box power supply adapts its output according to the resistance in the circuit, or rather, to how the reactor behaves. The dummy resistance that most closely replicated the situation with the reactor was using a conductor, indicating that the reactor really behaves as a conductor. This means that the power consumed by the resistance and the reactor together was about 0.3 x 0.3 = 0.09W, as used in the calculations, but that in fact, the reactor consumed a negligible part of that power, resulting in a COP that could be in the order of tens of thousands or more.

We can make another possible conclusion. Since the voltage across the 1-ohm resistance, while using a conductor as a dummy, was about 0.4V, the total output voltage from the power supply at normal operation could be about 0.4V, indicating that the voltage across the reactor could be at the order of magnitude of 0.1V, and the power consumed at 0.3A would be about 0.03W. This would result in a COP of about 1,000. (I don’t remember the value of thermal output power from the reactor, as calculated by William Hurley, but I think ot was about 30W).

Having said this, it seems strange that the power supply, even if it is a complex design, is such that it needs significant active cooling, resulting in a total system that has a COP of about 1 or less at this point. On the other hand, this is what Rossi explained to be one of the challenges in further development of the system.’


Mats also sent me the following:


Duration of the measurement period: 1 hour: the measurement has been made after the apparatus has reached a reasonably constant temperature
amount of water pumped through the reactor: 1 000 g
Water temperature at the input of the reactor: 21 C
Water temperature at the output of the reactor: 41 C
Delta T: 20 C
Energy produced: 20 x 1.14 = 22.8 Wh/h
Measurement of the energy consumed ( during the hour for 30′ no energy has been supplied to the E-Cat) :
V: 0.3
OHM: 1
A: 0.3
Wh/h 0.09/2= 0.045
Ratio between Energy Produced and energy consumed: 22.8/0.045 = 506.66

Instrumentation used for the measurements:
Oscilloscope Tektronix TBS 1052B
K probes Omega supplied and calibrated by Prof. Bo Hoistad of the University of Uppsala
Water pump Prominent. The water pumped for 1 hour has been poured in a plastic container seat on a scale to measure exactly the water passed through the E-Cat.
Temperature Data Logger: PICO Technology
The scale to weight the water passed through the E-Cat has been supplied by Eng. Mats Lewan of Stockolm

William S. Hurley
Senior Engineer- Endeavor
Los Angeles

  • psi2u2

    Thank you, Mats Lewan!

    • I was about to touch the metal parts connecting the reactor, but Rossi stopped me. Knowing that 240V would have been needed across the reactor if it had a resistance that would consume enough power to account for the thermal power, I thought that I could feel if there was such a high voltage and not just millivolts. But maybe Rossi know about the ignition spikes and realised I could be hurt.

      • Bob Greenyer

        He has insulation on the the connection – it is clearly HV.

  • Hank Mills

    Thank you for the detailed explanation, Mats! Clearly, it seems that in *any* of the scenerios you describe, the Quark produced a truly anomalously high COP.

    Some folks on another site have suggested that there are periodic voltage spikes visible on the oscilloscope that potentially could have transfered power to the reactor that was not measured or included in the input calculation. Can you address this issue? Was the input measured in such a way that all the input voltage variations were detected, calculated, and factored in? If so, I’d say this demo likely shows something amazing!

  • psi2u2

    “Having said this, it seems strange that the power supply, even if it is a complex design, is such that it needs significant active cooling, resulting in a total system that has a COP of about 1 or less at this point. On the other hand, this is what Rossi explained to be one of the challenges in further development of the system.”


    • Vinney

      One controller can control multiple Ecat QX devices, but can it control 1MW version at same power usage.
      I mean, apart from the incremental 0.3watts per Ecat QX, would the EM pulse and RF generator consume the same as per 1 or 3 QX modules.
      It consumes a lot at present, because it is powering only 3 QX devices, but at 10,000 QX modules, that consumption would be negligible.

  • WaltC

    I believe that’s a typo. He meant 0.09W. It shows up correctly further down– in the calculation made in the update portion.

  • Sorry to be simplistic. Is the conclusion good or bad? Two statements seem to conflict
    1) ….This would result in a COP of about 1,000 – GOOD?
    2)…. resulting in a total system that has a COP of about 1 or less at this point – BAD?

    • Frank Acland

      COP and System COP are two different calculations.

      COP (3) is the ratio of power in to power out of the E-Cat reactor(s).
      System COP (2) is the ratio of power in to the Control Box, and power out of the E-Cats
      Rossi admits in this presentation that they have had big problems with keeping the control box cool, and that most of the energy used in the control box is for cooling.
      My conclusion: Amazing result in terms of the QX reactors themselves; much engineering needed to deal with the control box issues. Rossi states he is dealing with companies to help with the control cooling.

    • Bob Greenyer

      If the COP were 500+, this device could EASILY be powered by an attached thermoelectric-generator, something I believe Rossi has experience in.

      So why not close the loop?

  • Engineer48


    Do you know the time base of the scope?

    Some of what is seen on the scope, the high freq oscillation in envelopes, could be back emf from the reactor.

    Both the smaller cube like ECats and the larger Tigers had a circuit to monitor back EMF from the reactor, so Rossi being a creature of habit would probably carry that ability to directly monitor the “Heart Beat” of the reactor forward.

    • Time division 100 us, voltage division 100mV. I think you can see that in the pictures.

  • Vinney

    The full un-edited video is on YouTube.

  • Matts, you’ve done and are doing a great job over the years but, drawing on all the hands-on experience you’ve so far been allowed to have with all the various versions of the E-Cat, to what extent can any given experiment be “replayed” in its totality. In other words, drawing inspiration from both aircraft “black boxes” – the flight data recorder the cockpit voice recorder – it is possible, in most cases, for crash investigators to solve the sad puzzle of what happened.

    So, how comprehensive is or has been the data-logging for the E-Cat? If not necessarily in this instance but when the R&D/performance certification/safety testing is underway? If I were in charge (and armed with a reasonable budget), I’d have synchronised, high-speed, multichannel data logging being captured by a pretty grunty computer. I’m talking channels for everything you can name – multiple monitoring channels for temperature, voltage, current, water flow, voice, video, you name it. There’d be a lot of Analog-to-Digital converters being used.

    At the end of any run, the analyst would be able to return and re-examine any and all the parameters for any instant of time. And, by the sound of it, we might need to be talking microseconds to catch those spikes. As well as all that, I’d have a number of radiation detectors (for neutrons, alphas, gammas, etc) also being constantly logged.

    • This is of course what you would have done in a scientific experiment, but as I said in the presentation, this wasn’t a scientific experiment but a demo by the inventor who decides what information to provide. Also, when Rossi makes his own experiments he has a more detailed control of measurement conditions. This was a demo.

  • AdrianAshfield
  • Engineer48

    We have IGNITION!


    I suggest the signals that go below the base line are Input drive pulses and those oscillations that do not go below the base line are back EMF signals generated by the QX reaction.

  • Bob Greenyer
    • LesioQ

      Spectral results would give the answer, with thermal glow being a smooth graph.

  • Bob Greenyer

    The control frequency / waveform is on ready for the spark as can be seen in ripple on scope

    I think the spark creates a plasma channel (near zero resistance) which is sustained by the waveform that is then cut dead periodically to collapse the plasma

    1. the sustain waveform is put on (and seen as ripple on scope since no conduction path, just noise)
    2. then spark – plasma channel formed and sustained by waveform, 100kHz sounds fair
    3. then the power is cut (or reversed?) momentarily, plasma collapses
    4. Goto 1.

    https://uploads.disquscdn.com/images/d08525571df38f6b87aadc07f8b375d49f50a276745385246e2915a6c476981f.jpg https://uploads.disquscdn.com/images/e0435325099629254f5582fe0fdd4c61d51285333e7395c8cccbcb4c78c410a2.jpg

    • Gerard McEk

      We must be very careful on what we see on the scope. I agree with you about the start pulse. The following control wave can look very different if it is sampled at the right frequency. Sampling at a too low sample speed would distort the shown waveform completely.

  • CWatters

    What’s all the nonsense about the power supply needing 60W of cooling? Mats suggests it’s because heat from the reactor is conducted back down the wires! Oh come on really?

    • Andreas Moraitis

      Maybe he uses a standard power supply with a built-in fan that was originally designed for heavier loads, or for installation in narrow containments (such as PC casings).

    • I later said, during the presentation, that Rossi no longer claims the heating problem is due to heat through the wires, but an internal heating problem in the control box. Fulvio Fabiani, who has built the original design of the control system, confirmed this, and said that it would need investments to and resources to build a control system that eliminates this problem. I agree that this seems strange.

      • Bob Greenyer

        If the device really does use a fraction of a watt, a competent engineer should be able to design a cool running control system in a matter of days if the process is understood.

        Perhaps it is breeding Charge Clusters and they are conducting down the feed wires and massively heating the control hardware, like that Russian group that sends 25kW down a 8um wire.

  • Frank Acland

    I found this exchange between Mats and Andrea very interesting (at around the 47 minute mark)

    Mats: The control box, which is quite large now, how small do you expect it to be:

    AR: This is a very interesting question which connects with the industrialization. We have a very big problem with overheating, so this box, most of this box is designed to cool down, and most of the energy consumed by this box is to cool down. There is a big technology to cool down inside here, because it is not so simple as this appear, because we have a time at our disposition to put down the temperature that is very short, it is not that easy.

    I put in contact a cold thing with a warm thing it cools down — yes, but it takes time. During that time we have companies [components?] that burn and this has been a big problem. I am pretty sure, now we are in contact, I cannot say the name of the concerns we are in contact with …

    Mats: Corporation, you mean

    AR: Yes, corporations, but I am sure that with their power of fire, and with their technological capacities, we will have reduction in costs and dimensions that will be very remarkable. Now the dimension of the E-Cat is already very very small, because you consider that the cubic centimeter of which correctly Mats Lewan has spoken before, is the volume not of the reactor, it is the volume of the reactor plus the volume necessary to expand the heat exchanger. Because the real dimensions of the reactor are 0.08 mm of diameter, and 0.6 mm in length. This is the real dimension of the reactor. And with these dimensions and the consequent density of energy many things can be done. But basically this white box which is a basically an I-rator[?] this is to keep cold all the system, etcetera, can become, if we resolve the problem of cooling down as I am, it is just a matter of time, (no do not touch, please, Mats do not stay there, I am sorry — go to your place, thank you very much. At the most you can put your finger in the water to be sure that it is not cold) I was saying that this can be as big as a packet of cigarettes, but not just for one E-Cat, but for thousands of them. But to make this, you know we need an engineering capacity that currently we do not have, but we are struggling to get, but I have the right alliances.

    • Bruce__H

      In the first part of Mr Rossi’s answer he seems to be talking about cooling parts of the electronic control system inside the white box at his elbow. In the second part he is talking about the dimensions of the QX and it sounds as though that is the thing he is trying to cool. I don’t understand what these to parts of his answer have to do with each other.

  • As it was shielded by the heat exchanger you couldn’t see that, but from the side of the exchanger there was some light coming out and you clearly saw the ignition spark and some light after that. I think the light stops for the period when there’s no power supply.

    • as far as i could see during the spectrometer test there was a light through the water (inlet/outlet) during the whole on-phase (3sec) and not when off (4sec). And the ignition “flash” every 7sec…

  • I did enjoy the journey!
    The Royal Swedish Academy of Engineering Sciences (IVA) had no connection with this demo, apart from owning the conference center, but it’s a normal commercially based conference center where Rossi basically rented a conference room on normal conditions. Very nice and professionally managed place.

  • Andreas Moraitis

    I think it depends on a number of factors, not just overall power. On of them is the distribution of power over time. No problem if it happens continuously, but more difficult to handle if there are short phases where a lot of power is consumed. In such cases one would need a much more effective heat transfer. Other things that come into my mind are the dimensions of the components (smaller resistors will get hotter than big ones with the same resistance, for example), adequate spacing and airing. Finally, there is the possibility of current leaks – although this is a rather theoretical point.

    • Bob Greenyer

      Small or big resistors of the same resistance will still need to dissipate the same amount of heat, though admittedly big ones will be less likely to fail.

      • Andreas Moraitis

        Of course the amount of heat is the same, but it will be better dissipated by the big resistor due to its larger surface area.

        • Bob Greenyer


      • Abel R Massa

        A resistor fails when it is underrated for the given circuit. It is customary to derate by 50% to improve reliability by preventing it from stress leading to failure. So, a 1/2 W resistor is used in a circuit that only requires 1/4 W dissipation. I hope this is a little more precise.

  • Will be posted ASAP. It was cut out to make the rendering faster.

    • pelgrim108

      Thank you Mats for your work.
      Was there any talk yesterday about the spark/flash every 8 or 4 seconds.
      Do you think it was sparks or flashes?

      • I’m not sure how to define spark or flash when igniting a plasma—it’s not really an everyday thing to me… So I wouldn’t know.

        • pelgrim108

          I mean spark as in an electrical arc outside of the reactor. And flash as in only a sudden burst of light.

          • I wouldn’t know. It was not really discussed.

          • pelgrim108

            Can you confirm that the spark/flash was alternating between the left and right side of the reactor?

          • No, I actually only looked at one side.

  • It was a challenge to render the edited video, compress it and upload it in short time. It might not be visible at Ecat.com but I think the YouTube channel is theirs. The theory by Carl-Oscar Gullström was cut out to speed up the process. Will be posted ASAP.

  • Sorry, correct, that’s a typo. Frank, can you correct please?

  • As I said below, it seems to be high frequency (~100kHz), high voltage (at ignition, maybe 1kV), high velocity (adapting to the reactor behavior), which might be challenging combined.

    • We may be talking at cross purposes but, if the load on the power supply gets shorted (by a CC?), even for a few milli or micro-seconds, the power supply would be called upon to supply a huge current and it certainly would heat up.

      The simple fix would be to place an RF choke in series straight after the power supply outlet. Maybe also a ceramic capacitor (to earth) after that and then a second RF choke to choke off any “back emf”.

      So, in that final configuration we’d be talking a LCL “star” configuration but I’d try the simple “L” first, followed by the “LC”. Others might prefer the “pi” configuration – “CLC” .

      Hey, it’s been well over 40 years since I designed and serviced stuff like this but I can almost catch the smell of solder as I write this.

      • Send an email to Rossi with suggestions. I also suggested that he’d contact companies like Ericsson, experts very fast HF and high power systems for cellular networks. With the right expertise I’m sure that the control system will be optimised and minimised.

        • Yes, I could send an email but I know that Mr Rossi routinely reads ECW and occasionally directly responds to particular comments. And another big advantage of this forum is obviously that it has an element of immediate “peer review” (although not always constructive!).

          So, bearing that in mind:

          Attn Mr Rossi:

          Please note my latest post, above, and see what you think. I should explain that the chokes I’m referring to would need to be made with pretty thick wire and have not too many turns so as to minimise their DC resistance. The capacitor would need to be ceramic (and voltage rated above the DC level) because ceramics ones have a high frequency response. That said, there could also be electrolytic capacitors in parallel with the ceramic ones to exploit their very large capacity and thus excellent smoothing (although only at the lower frequencies).

          Last night I mistakenly called the best configuration the “star” one but I now remember that it’s correct term is “T Network”. So, I’d say that if you had a power choke on each side of a ceramic capacitor, you’d have it both ways: Any temporary short from the E-Cat would be blocked from shorting the power supply by the first choke and any voltage spike generated inside the E- Cat – trying to “counter-attack” the power supply – would be blocked by the second choke. The bonus of that would be the any such spike (or CC as some contend) would be forced to dissipate itself as (eventually) heat inside the E-Cat and thus improve its heat efficiency.

          All that might be a very simple and cheap solution that the nearest soldering-iron wielding electronics geek could breadboard between sips on his coffee.

    • Jerry Soloman

      Mats nice work as MC yesterday, also i did not know you also speak fluent Italian.

      • Grazie 😉

      • Skip

        He needs to. Otherwise he may find his house chilly when he returns at night.

    • Gerard McEk

      Yes, a plasma has a non linear behaviour. Usually one uses a current control to get it stable. I did not see current control behaviour with your short circuited and 800 ohm test, though. Maybe the controller is very limited in its voltage. 5 ohm would perhaps have been a better choice, but I say that now I have seen it.

  • Bob Greenyer

    There appeared to be a DC offset on the waveform.

    • Andreas Moraitis

      AR has always stated that the input is plain DC. One might speculate that the ‘AC component’ (except the spikes) was generated by the reactor itself.

  • Bob Greenyer

    If it was making Charge Clusters, then you can transfer 25kW down an 8um wire.

    • US_Citizen71

      He wasn’t making a suggestion of electrical transmission, but thermal so charge clusters are non-sequitur.

  • Frank Acland

    Yes, I made an error. It is fixed now.

  • Jerry Soloman

    Camilo Urbina Sums the Event up here:

    “I must say that I am not impressed by the demo, at all, — and I am just interested in the attendance as all there seem to know much more than us, the peanut gallery. — Alan Smith already gave a short impression on LENR-forum and he is cautiously optimistic, which is good to me, and he reckons this demo is a very limited value as it had no independent checks.”

  • That’s an opinion, not supported by measurements. Although the measurements were not optimal they all indicated the contrary.

    • Sumdum

      Sorry Mats, the measurements just don’t show anything conclusive. People have already suggested at least two ways that the system could be set up for deception, and many ways it could have been done differently ton be more convincing. It is not a convincing demo for anyone who was not already convinced. There was no useful purpose for this demo. It was a waste of everyone’s time, as many predicted it would be.

      • You don’t have to be sorry. Obviously there are tons of ways to fake the result. That’s not really what we’re discussing here.

        • What I’m interested in are mistakes and misinterpretations of the measurements.

    • Dusty

      But Rossi said that he needed around 50/60W of power do dissipate the heat of a device?

      The total heat produced was around 22W, the balance is still negative.

    • Thomas Kaminski

      From what I read about the Tek scope, it is capable of many measurements, including RMS values of displayed waveforms. When you were “reading” the measurements, what were you using for the values you quoted?

      • I was just estimating the approximate value of the maximum voltage of the waveform as shown by the graphics display on the oscilloscope.

  • Answered above.

  • Ged

    And a CPU needs quite a lot of cooling because of that (only water cooling lets a CPU run at its physical best because of heat). Which supports what Mats said.

  • To prove something you need to at least measure something. Otherwise it’s an opinion.

    • Dusty

      We don’t need to measure it, he says that he needs at least 60W input power.

  • Pekka Janhunen

    It seems that a simple way to fool in this arrangement would be to have 3 modes for the control box: one for the reactor (which would have been just a 800 ohm resistor with some bells and whistles to create the flashes), another mode for the straight wire dummy case and a third mode one for the 800 ohm resistor dummy case. Such scenario would be consistent with Rossi forbidding Mats from touching the wire, since it would have had 240 volts 3/7 of the time.

    That is not the only possible way to fool, since the energy generated was only about 80 kJ which is equivalent to burning only 2 grams of gasoline. Don’t know the size of the reactor, but maybe such amount of fuel would fit inside it.

    Like Mats said in the beginning of his intro, a demo is a demo, while a scientific measurement would be something else. It’s hard to pass the chemical limit in just one hour.

    That said, if one believes that Rossi did not fool, then the outcome was quite positive: high COP, stable operation, fast turn on/turn off. Close to ideal energy source, the only thing missing is lack of direct electricity output, and the control box engineering issue that Rossi mentioned.

    • Gerard McEk

      I am sure the COP was not as high as what was said. Basing it on the power consumed by the 1 ohm resistor is ridiculous. Luckily Mats did some dummy tests afterwards and based on that you really can calculate (see above) the power consumed by the QX, if you assume that the QX behaves in a linear way. However Mats thinks it does not, so I think we are as far as before we did the test.

      • Simone

        This calculation cannot be reliable since the control box was manipulated by Rossi just before the dummy test started.

        • Gerard McEk

          Yes that may be true. We do not know why Andrea seemed to tumble some switches inside the control unit before the dummy test. My assumption is that he switched off the HV pulse and the HF carrier frequency that is used to control the plasma (both issues he didn’t want to reveal).
          AR did it sneaky without mentioning it, which makes the measurement not credible. Mats wasn’t aware of it.

    • Andreas Moraitis

      When did Mats propose his additional test procedures? If he did it in the course of the event, there would have been no time for manipulations.

  • Ged

    Thank you, Mats, for pushing to get those dummy load measurements. They provide really important pieces of data for analyzing the actual device, and controls like that are something the field needs more of in general.

    • It is surely in nobody’s benefit that this confusion persists. Please could you Mats, working with the community, design a “test/demo” that will conclusively “prove” the LENR effect is taking place. And will deliver a definitive COP measure.

  • Gerard McEk

    Based on my calculations I conclude that the demonstrated COP is about 40. It is possible to calculate all the details based on the current measurements short circuited, with 800 ohm and the QX connected. I have made a spreadsheet that does that. Maybe Frank can make it accessible here.
    The calculation assumes that the controller is a voltage source with an internal resistor and that it behaves linear.

    • I don’t think that the control system behaves linearly. In some way it must detect if the circuit is open or closed (reactor plasma ignited or not). Once the reactor is active, from what I understood from Fabiani the control system keeps receiving feedback from the reactor and adapts to that feedback. I think the most indicative measurement is the one with the load short circuited, but I would be glad to see your spreadsheet.

      • Gerard McEk

        I sent it directly to you, Mats. Maybe you read that first.

      • Stephen

        Thanks Mats for your part in this demonstration. It took it up to the next level in my opinion it looked professional and great.

        I’m a bit behind and I still need to catch up with the official video. So I apologise if the answers are already there.

        I wonder if the 60W was more used for the power supply for the necessary signal processing unit rather than the power supply to the ECat it self. It seems to me quite possible given Fabiani’s experience in real time data processing and control I think. Is it realistic to think about this amount of power being required for that processing? Or is it more to do with the way the energy needs to be controlled stored and broadcast in the ignition spikes?

        Are the high voltage ignition spikes sent down the same wires as the DC current. Or are they some how generated at the Ecat from the supplied DC.

  • And I’m not really sure of Rossi’s test being correct. Someone might help us out here. Would a charged particle flow, like the one in the Van Der Graaff generator have this effect? Or would it have a separating effect on positively and negatively charged particles? Or no effect at all?

    Another question: If there’s a net DC current through the reactor, then either the positive particles outnumber the negative or move faster. Or have greater charge per particle.

    • Pekka Janhunen

      If one has plasma which is dense enough to glow, it must be quasineutral to high precision (that is, the difference between electron and ion number densities is much less than either of those densities). Because otherwise the charge imbalance would create a strong electric field that would blow the plasma apart. Numerically, it’s the smallness of epsilon0 (8.85e-12 As/(Vm)).

      I’m not sure why Rossi needs to prove quasineutrality in this case. Even if the plasma is quasineutral, the current is typically carried mostly by the more mobile species, i.e., electrons. The other species is there to neutralise the space charge density of the electrons.

    • Engineer48

      Hi Mats,

      I would add another question.

      Assuming 2 circular Ni electrodes 0.6mm in diameter and seperated 0.08mm in a H2 atmosphere of some pressure.

      What is the min breakdown voltage difference needed to form a plasma between the electrodes?

      Or more specifically is 12 volts enough to generate a plasma?

  • Leonard Weinstein

    The test protocol was very poor. The fan could have been powered with a separate power cord and avoid the fan power issue. Then the input power to the power supply box measured with a power meter. The fact that the power supply got hot indicates it had a significant added power use which has to be made small before any real progress can be claimed. The OUTPUT from the power supply could be measured with an rms power meter, so the true COP could be found without giving up any secrets. The COP has to be corrected for the 1 Ohm resistor power use.

    • Timar


  • GiveADogABone

    The internal resistance, as expresssed by (dv/di), of the operational QX is negative IMHO.

    I do not see anything from the demo that is inconsistent with the concept of a hydrogen-filled CCNR gas discharge tube being biased with a DC current. Starting requires a voltage spike and the DC bias current must be sustained to keep the CCNR operational.

    1.1 http://g3ynh.info/disch_tube/intro.html
    Gas Discharge Tubes – Introduction
    ‘Paschen’s law’
    For Hydrogen, the minimum breakdown voltage is about 300V @1 ((mm Hg) x cm). The QX electrodes are about 1cm? apart. That sets the approximate Hydrogen pressure at 1mm Hg.
    A gas discharge tube, once it has struck, usually has a negative resistance characteristic. This means that the product of the voltage across the tube and the current through it diminishes as the current is increased. Thus the current is effectively unlimited, and must be controlled by the external circuitry if it is not to lead to the destruction of the tube or the power supply.

    2.1: https://en.wikipedia.org/wiki/Negative_resistance
    Negative differential resistances can be classified into two types:
    Voltage controlled negative resistance (VCNR)
    Current controlled negative resistance (CCNR)
    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 include the … electric arc, and gas discharge tubes …

    2.2: https://en.wikipedia.org/wiki/Gas-filled_tube
    Hydrogen is used in tubes used for very fast switching, e.g. … Hydrogen (and deuterium) can be stored in the tube in the form of a metal hydride,

    2.3: https://en.wikipedia.org/wiki/Hydride
    Applications: Hydrides such as …, lithium aluminium hydride, …

    So, a gas-filled discharge tube containing LiAlH has a negative resistance region and can oscillate in the right circumstances. That would include the QX. A detailed read of
    is needed. e.g.
    A negative differential resistance device can amplify an AC signal applied to it if the signal is biased with a DC voltage or current to lie within the negative resistance region of its I–V curve.

    Negative conductance (current controlled) oscillator: CCNR (“S” type) devices, in contrast, require a high impedance bias and are stable for load impedances greater than r. The ideal oscillator circuit is like that at bottom right, with a current source bias Ibias (which may consist of a voltage source in series with a large resistor) and series resonant circuit LC. The series LC circuit has low impedance only at its resonant frequency and so will only oscillate there.

    Conditions for oscillation

  • Valeriy Tarasov

    Since the reactor has electrical contact with control system, it can be that the observed light flashes from the reactor have been accompanied by electric pulses as feedback to control system. If so, these electric pulses should be quenched in control system, cooled down ? And, maybe this is what was about generation of direct electricity from Ecat QX.

  • US_Citizen71

    From the descriptions and measurements made it seems the control box is a complex ballast similar to what is used for metal halide lights. If the current control box is a specifically designed magnetic ballast keeping it cool would be a bigger problem than keeping an electronic one cool. It sounds like Rossi needs a good design EE on his staff.

  • Gerard McEk

    Current is measured and over an 1 ohm resistor. What was missing is the voltage, but that can be calculated. The power of high frequencies can best be measured using a resistor and a proper RMS voltage meter.

  • Timar

    What I find particularly frustrating is that there is no reason whatosever why the fan couldn’t simply have been connected to a separate power supply…

  • CWatters

    I also think the test protocol was very poor, to the point we really have no idea how much power went into the reactor when running.
    All tests on the power supply (eg using a short and an 800R resistor) were carried out with the reactor switched off and removed. It’s not possible to be sure the power supply behaved the same way when the reactor was connected.
    For example suppose the reactor actually has a resistance of 255 Ohms and when connected turned on the power supply produced 76V with a 0.4A current limit. Then the actual current would be 76/255 = 0.3A and the voltage across the 1R would be 0.3V (exactly as measured). The power to the reactor would be about 76*0.3 = 23W (roughly that measured by the water calorimetry).
    Then when the reactor is turned off and disconnected lets say the power supply is voltage reduces to 12V.. When a short circuit is added the 0.4A current limit is triggered producing 0.4V across the 1R (again as measured). When an 800R is connected the current would be around 12/800 = 15mA (bit out from what was measured).
    The reduction to 12V could be done either by actually reducing it to 12V or by mark-space width modulation of the 76V.
    I might not have the actual numbers right because you would need to pump in more than 20W to see 20W in the water, but the point is we really can’t be sure how much power went into the reactor.

  • AdrianAshfield

    So…As forecast demo will not end the debate. No matter how good the experiment the skeptics will stick to their belief that Rossi is a fraud until working units are on sale.

    The measurement of energy produced was indisputable. The measurement of energy in was less so, but still valid. The unknowns included the waveform, the high voltage pulse to start the process and whether there was feedback from the QX that controlled the power pack. The waveform of the test with the reactor replaced by a conductor provides some clues. The large clumsy power pack reflects the complexity of this.

    That the power pack used 60 W cooling provides the skeptics with endless opportunities for mocking the demo, but I don’t buy that power was magically used to heat the water. This will get sorted out relatively quickly.

    The very small size of the QX was a surprise, particularly given the larger dimensions stated earlier. Not sure that these dimensions aren’t a result of Rossi protecting his IP.

    All things considered, it looks like more engineering development is still needed and it will be at least a year before any mass production. New technology takes time. Boeing took forty years to develop the V22 before one was sold.

    If the power pack was heating the water, how can he possibly make money from such a scam? This is not for the general public to invest in, but for venture capitalists who will be able to do their own due diligence – like measure the voltage across the reactor.

    • psi2u2

      Very nice summary, Adrian.

    • Mylan

      “No matter how good the experiment”. But then you describe the flaws yourselves. Following the Rossi story since many years full of hope that are shattered on a regular basis is just sooo frustrating. After all, it is still possible that it is all a scam. Or that Rossi is at least fooling himself.
      I think you can get investors even with a flawed test protocol. Some will be willing to believe.

    • Engineer48

      Hi Adrian,

      My concern was the scope voltage range was too small to fully see and record the amplitude of the plasma ignition pulses. Fulvio stated they were 12 vdc. That may be +-12 volts as they are seen to produce positive and negative pulses on the scope.

      Had the scope been set to show the full amplitude of these pulses, it may have been clearer as to the input power.

      I believe the 100kHz signal with it’s AM modulation was an output,, produced by the reaction and sensed by the control box such that if it was too low, implying the reaction was dieing off, another ignition pulse would be generated.

      It is clear that during the active period, many ignition pulses can be seen. Something must determine then they are produced as clearly the generation of periodic ignition pulses are required to sustain the reaction.

      • AdrianAshfield

        Hi Enfineer48.
        “My concern was the scope voltage range was too small to fully see and record the amplitude of the plasma ignition pulses.”

        As I’ve tried to explain already, Rossi is not going to show details that might help his competition, whether you like it or not. The purpose of the demo was to show the characteristics of the QX and it did that well enough.

        What people seem to be overlooking is just how large sn advance the QS is.
        Consider its small size, that it can be instantly switched on and off, that it has a COP 500 – 1000. It can operate at 2300 C.

        That is a major step forward. Proof of whether it really works will be left to the due diligence of potential investors.

        • Simone

          Adrian, the whole plot of the demo makes us think that the overall performance of the circuit was less than COP 1. If this is the case then he proved to have made nothing special in all these years.

          As mentioned by Bohem, the purpose of this demo was to undeniably proof that input power was lower than output power.. he could have kept all the secrets he wanted to do so, not a problem.. but he didn’t.

          Presentation was useless, badly made and with a purpose very hard to determine.

          I was hoping this technology would come to light sooner or later but if this is the case, he could have avoided wasting time make this demo and go directly into production phase.

          • AdrianAshfield

            Simone, Try reading my comment again.
            What you say make no sense at all. Rossi has spent a lot of time and money on this over the last year. If what you said was true this would all be a dead loss with no hope of making a dime.

            The purpose of the demo was to show the properties of the QX. Any investor could quickly test whether it works by putting an oscilloscope across the reactor. As repeated many times already, Rossi does not want the signal to be public knowledge and help his competition.

            Why do people like you write this rubbish without giving it any thought? What do you get out of it by running others down?

    • Rene

      No, Adrian, Rossi deliberately screwed around with the controller during Mats resistors tests to ensure the measured power would be low. It’s on the video. You cannot unsee that.

      • AdrianAshfield

        Think what you like. I think Rossi was concerned about his IP getting out.

        It makes no sense for Rossi to demonstrate a reactor that didn’t work. How could he possibly gain anything? Any potential investor could make some simple measurements to prove if it did and there are enough wrarnings that you can be sure they would check before parting with money.

        • Rene

          I think where we disagree is the motivation. You think it’s just IP issues while I believe that’s only part of the problem. I believe the huge problem is he *still* has reaction control issues. One constant throughout this seven year excursion about Rossi is that he loves to brag and overstate his successes counterpointed by his strong deceptiveness that belittles any problems he is stuck on. Rossi inventor is interesting. Rossi entrepreneur is downright deceitful.

          Yes, it makes a lot of sense to show off his QX if he has something that kinda sorta works. Then he can tell his investors he needs funding to get past some technical challenges. Then he points to the crowd to demonstrate how much interest people are showing, a great technique to convince those potential investors they need to rush to get in the line.

          Does his QX work? Maybe. He certainly made is practically impossible for anyone to determine that in the meeting and videos.

  • Frank Acland

    Mats got the pump, he might be willing to take it apart and see if it could have added some heat.

  • US_Citizen71

    Unfortunately not really my forte, my experience with ballasts is mostly as a user and what I have learned about them is from maintenance and repair. I understand what they do but I don’t design them. When I switched from magnetic ballasts to electronic ones there was a huge reduction in heat production and power usage. I was always afraid the magnetic ones were going to cause fires if something flammable like paper or cloth contacted them for too long due to the heat they produced. Magnetic ones are much simpler to make as they have few parts, once upon a time they were part of the base of the bulb in old mercury vapor lights, but they have never been very efficient.

  • alexro

    Lets suppose that the control box can power at least 10 QX (I think that it was also confirmed by Rossi that it can do this).
    The question is why Rossi didnt connected at least enough reactors to show real COP > 1?
    At the presented situation the COP is significantly under 1 which was also mentioned by Mats.
    If I am not wrong there was no input power measured. And we are not counting even the pump but I believe that pump can’t heat the water so much in any case. Why they are using pump even if they can use simple free waterfall?
    In any way the control box could feed possibly 100W and nobody will know about that.
    And it must be pretty easy to program that from certain time frame or just different load it will behave differently. As confirmed the control box is checking the feedback so for resistors it may not do the same thing.
    And what would need that much cooling in the control box? If cooling is needed then at least something that dissipates (to the air) few watts is present. I can’t find any reason why you will need any circuit that will waste so much heat to produce signal of few mW. The only possibility is that it is some extremely exotic device. But what it is doing actually is not visible on the oscilloscope.
    I doubt that Rossi would reveal any secret. So the waveform we can see is probably not too much important. It could maybe look very differently.

  • SG

    I thought Bitcoin was a good buy at $100 and also think that this demo moves the ball forward considerably.

  • US_Citizen71
  • Luca Meli

    I subscribe every word

  • Mylan

    I would rather interpret what Rossi said in a way that 60W of excess heat are removed. I don’t believe he really meant the fan had 60W, that would be enormous. You can blow away 60W of heat rather quietly.

  • Scott Beach

    CW: I do not know what kind of pump was used. A “peristaltic” pump like the one shown at the following link would not have added any significant heat to the water flowing through it.


  • 21:00 EST 25/11/2017 Youtube video stopped streaming.

  • Engineer48

    Just before Mats starts the tests with different resistor values, Rossi lifts the cover on the box and apparently switches a switch or two:


    Which to this observer casts a big shadow over the validity of the tests Mats conducted.

    • Andreas Moraitis

      Even if it was harmless (deactivating the HV pulses for safety, maybe), it makes the resistor test worthless. Switching between constant voltage/constant current mode would suffice to produce the observed characteristics. The important question is, when did Mats inform AR about his plans?

      • Engineer48

        Hi Andreas,

        We both know what Rossi did totally invalidated the tests Mats did.

        My question is why Rossi did not explain to ALL present what he did and why? Even with an explanation, the data Mats collected would still be suspect.

        Look fair enough if Rossi wanted to hold back data on the QX operational characterists and the drive voltage needed to ignite the plasma. Rossi should have rejected Mats request, saying such test would reveal confidential data.

        Making a change to the control box and not telling Mats and the other attendees is not how to win friends and make a positive influence.

        • Andreas Moraitis

          Surely. It looks like an attempt to voluntarily destroy credibility. What we still need to clarify is if there was a similar intervention when he switched to the 800 ohm resistor (and back to the QX).

    • Didn’t notice this. Not good.
      I discussed the dummy tests with Rossi two months before the demo. Shortly before the demo Rossi only wanted to do the short circuit test, not the 800 ohm, since the waveform would be altered in a way that would reveal sensible information, he said. I told him that in that case it was meaningless to do the test at all. He came back a few days later and said he would do it.
      Whatever adjustment he did before the dummy test, legitimate or not, it’s a serious problem. The input measurement was already weak as it was. Now I suppose the only conclusive test would be to repeat the measurements when he has a developed control system that doesn’t need cooling so you can measure COP of the whole system.

      • maxhart

        Thank you for your great work in the presentation.

        The only thing he needed to do was to add few more QX so that output power is higher than input power. With QX this should be pretty easy.
        Why he refused to at least measure input power with any wattmeter so we have at least some imagination? This can’t reveal any secret thing.

      • maxhart

        You have tried to touch some part of the reactor during the demo. Did you touched it? Do you remember part it was?
        I am curious about this because if there is high voltage he could be afraid so you wont get electrical shock. This would confirm that really there was high voltage.
        But if you touched it and didnt felt anything it could reveal that there was no high voltage at that time.

        • I commented on this before somewhere. Yes, I wanted to touch the connector of the reactors just for the reason you mention. That was my plan, but Rossi’s stopped me. However, I and others concluded that this was probably wise since the control system seems to use high voltage, maybe 1kV, to ignite the reaction, while immeditalely reducing the voltage once the reaction starts. Since essentially no current flows at ignition, this would not entail any consumed electric power, but it’s hard to asses.

      • Frank Acland

        This is AR’s comment on the JONP in response to a question about it:

        Andrea Rossi
        November 26, 2017 at 9:45 AM
        I checked what you say, it is between the time 2h 30′ and 2h 31′.
        I opened an air window to help the air circulation since when the E-Cat has been tiurned off the coling system was disconnected. After 2 hours of work and after the operation of the spectrometer without cooling circulation in the heat exchanger I wanted to help the hot air out.
        This has nothing to do with the measurements we were making.
        Warm Regards,

        • Bruce__H

          If Rossi is still in Sweden then perhaps he can show Mats Lewan the window that he opened.

  • Engineer48

    Sorry to say this but this is upsetting.

    Rossi is altering the control box just before Mats starts the resistor tests.


  • Rene

    Sorry to say that the amount of inventor meddling of the hidden control box totally invalidates the attempts by Mats to characterize the power delivered. So, basically Rossi sabotaged attempts to determine input power to the QX.

    Rossi was careless with the microphone so he left it next to the control box. Fortunately that gave us some insight to its workings, as you can clearly hear a relay going on and off on a 3/7 duty cycle. You can hear the closure tick followed seconds later by the opening tock sound. And you can see the scope waveforms synced to the relay closures.

    When that relay clicked on you see a hefty flash from the edge of the QX. That’s the high voltage starting pulse. We have no idea how high the voltage is or how long the pulse lasts because Rossi chose to keep that hidden. You can see sometimes see those spikes on the scope, completely off scale. Note also he deliberately turned off gating to force the scope trace to smear.

    Back to the starting voltage spikes, suppose that ends up being 250-1000V (just my guess based on flash tube trigger voltages, and that the trigger pulse is 100ms. That is depositing 25 to 100 Joules. That is a sizeable amount of energy, done once a second can easily reach the output power of that QX. By messing around with the controller during the time Mats ran his resistor tests, Rossi could have easily removed the source of the non-excess energy.

    Maybe the thing really works, but by hiding the input power and some sense of the energy deposited by the controller, it simply cannot be stated that it works.

    So, this demo did nothing to allay those conjectures. This demo is worthless except maybe to his new backers. He’s back to having controller problems, probably believes a bit more seed money and time will give hi time to fix it. At this point, I personally no longer trust Rossi at all.

    Definitely his lot is his own words: in Market Truth for he has burned all his bridges of credulity with me and likely others.

  • Rene

    In the vid Rossi hurried said that the heat down the wire problem was fixed (I utterly do not believe it), and that he had another problem with the controller electronics itself having a cooling problem. That is more believable but completely impossible to determine given the e-Cloak box.

  • Pekka Janhunen

    On JONP, Rossi keeps on insisting that the dimensions are 0.08 mm diameter and 0.6 mm length, and (I understand from other context) that it’s supposed to produce 20 W over 1 year. However, those numbers do not match up, for the energy density would then be a significant fraction of m*c^2. I suspect that Rossi is mixing cm with mm, as someone already suggested on JONP.

    Another thing is Stefan-Boltzmann radiation law. A D=0.08mm, L=0.6 mm cylinder at 2700K outputs only 0.5 W of thermal radiation, not 20 W. It might (?) output more if heat transfer occurs by conduction rather than radiation, but during the demo Rossi implied that it’s radiation. This is another reason to think that he mixes mm with cm. Also, cm dimensions would be more consistent with his earlier informal specs of QX.

    • can

      In the muted section of the new video Rossi said to Lewan that the plasma is in mm, and that it’s like a hair. He said millimeters multiple times and Lewan asked for confirmation too.

      • Frank Acland

        Frank Acland
        November 25, 2017 at 8:28 AM
        Dear Andrea,

        The video is very interesting. I do have one question about something you said on it:

        ” Because the real dimensions of the reactor are 0.08 mm of diameter, and 0.6 mm in length.”

        Surely it is not so small! Maybe you meant cm?

        Thank you,

        Frank Acland

        Andrea Rossi
        November 25, 2017 at 2:10 PM
        Frank Acland:
        It is not wrong.
        All the rest of volume is room for heat exchange.
        Warm Regards,

        • Matt

          One must assume that any work on it has to be done under the microscope.

        • Pekka Janhunen

          Yes I know that he insists on it.

          • Pekka Janhunen

            I mean, Rossi sometimes is blind to his own typos like this, even after someone tries to correct. Although usually he does correct, when someone points out an error. Whatever, but it’s a fact that the mm dimensions are not consistent with nuclear energy density and Stefan-Boltzmann radiation law, while the cm values would be.

        • pelgrim108

          A cubic millimeter has 330 times more volume than the inside of a E-CAt QX

          • artefact

            The fuel should be something the size 0.6mm of a gauge 40 wire.

          • ARM

            0.08cm = 0.8mm that’s gauge 20 wire sounds more realistic but makes the length shorter than the diameter

          • artefact

            It is said to be 0.08mm -> gauge 40 = 0.0799mm

          • ARM

            I agree amazing to get a better feeling it means stacking up seven layers of 49 cubes per layer of each cubic volume reported.

    • Stephen

      Hmmm just doing the calculations again

      If we have a surface area of 0.6 x 0.08 x 3.142 = 1.5 mm^2 for each ecat QX and 3 times that 4.5 mm^2 for the 3 Quarks.

      At 2700 K I get about 13.6 W. Which isn’t far off 20 W. Certainly not 1.5 W.

      If I use 2700 deg C (2973 K) I get almost exactly 20 W which is right on the money in this test if he meant the temperature in Centigrade. But it would imply continuous operation st this lower power.

      But this implies that at full power 60W the temperature would be 3900 K or 3627 deg C. And perhaps we would expect this if the 20 W power is achieved by having the ecat QX on 1/3 rd the time.

      So although I don’t get 0.5 W per Ecat QX it’s not fully explained yet I think.

      I wonder what I’m missing.

      • ARM

        shouldn’t it be 0.6mm not 6mm you used
        and so at 2700 deg C you will see almost exactly 2.0 watts
        Could you please review the dimensions and confirm what I am saying. Also, could you show the equation where power is proportional to fourth power of temperature T^4.


      • Obvious

        0.6 cm (gap) x 0.08 cm (diameter of plasma)
        So… circumference = 2piR; 0.08/2 (to get radius) = 0.04 cm
        2pi = 6.283; circumference = 0.4 * 6.283 = 0.2513
        Area is circumference * length = 0.2513 * 0.6 = 0.1508 cm^2

        emissivity = 1 (blackbody) so P = 5.67 x 10^12 (due to cm^2) * 2634^4 *A
        P = 0.00000000000567 * 48135233457936 * 0.1508 = 41.1 W


    • Omega Z

      ->”I understand from other context) that it’s supposed to produce 20 W over 1 year.”
      Pekka, I thought I saw where Ross said 6 months operation.

      We also must keep in mind that these QX’s were for the Demo. Production units may vary from these…

  • Engineer48

    We should also note that the 1st reading with 0 ohm resistor was negative 400 mv. Fauvio then reversed the scope leads across the resistor to make it positive.

    Point is with the QX operating there was an approx +100 mV Dc level with a +-100 mV 100kHz Ac signal with 4.5 cycle AM riding on the Dc.

    Then after Rossi switched 1 or 2 switches, the Dc level at 0 ohm went to -400 mV with no significant Ac signal.

    So 0 ohm = -400 mV & normal QX = 100 mV. Both a current flow reversal and the reverse current being 4x the normal current flow.

    • I think Fulvio was correct when he said he switched the probes of the oscilloscope, resulting in the negative voltage.

  • Remote control was obviously a potential issue, but I decided to give Rossi the benefit of the doubt on that point since any sophisticated remote control would be difficult to reveal anyway. In general there would of course be multiple ways to fake the result, so there must be a minimum of trust to attach any value to the measurements.

  • Matt

    The video is now missing sound, it seems to be faded out and later in again. Does not look like an upload glitch or something. Coincidentally the sound is missing in that very time frame where one could hear very clearly the clicking of a relay inside the box, which apparently triggers the visible spark at the QX.

    • It has been replaced by a version were the sound is cut off when no one is speaking to the audience, in order not to publish private comments. I think this is fair.

      • Matt

        Thanks for the explanation, this is fair and understandable.

      • Buck


        please correct me if I’m wrong about the following:

        I believe William Hurley’s employment has a typo and should be: The Los Angeles Refinery, Andeavor Inc.

        Link>> http://www.andeavor.com/refining/los-angeles/

  • LION

    This from: Alan Smith on lenr- forum.


    “Jed – You are wrong at least in part. Rossi is not interested in private investors at all. He wants corporate investors. Technical skills are a matter of opinion here, and a matter of who they are. I asked the refinery engineer who was on the platform what his interest in the technology was. ‘We use a lot of (process) heat’ was his response. Maybe they are potential investors, maybe not. I have no idea.

    After the demo was finished, there was a line of supplicants outside Rossi’s door. It was still there 4 hours later. And they weren’t there to complain by the way they behaved. Whatever you think, the Rossi magic is still working. And it doesn’t seem to be fading.”

    Thanks for all you do Alan. We are all looking forward to your report on the event and lots of great photos, well, after a good sleep and recovery.

  • pelgrim108
  • Thomas Kaminski

    My take on the presentation is that, although the test measurements leave a lot to be desired, one can infer the following:

    1). The controller periodically (once every 7 seconds) generated a pulse that seemed to ignite a plasma, since you could clearly see a flash of light from the end of the device holder. The plasma seemed to stay on (based on Mats Lewan’s observation that the device emitted light at times, different from the observed flash of light) while power was applied and off when power was removed.
    2). An AC signal was present on the waveform across the 1-ohm resistor, implying that the current waveform applied (or generated) had an AC component.
    3). Assuming that the device had an impedance similar to the 1-ohm resistor, very little electrical input power was used.
    4). From calorimetry, on the order of 20 watts of heat was generated, implying a very large COP.

    Carping aside, this demonstration shows that the Rossi effect is real. Anything over the F&P few percent excess energy is significant. Also significant is that a plasma is generated and that there seems to be an AC component to the excitation/generation.

    I will take Rossi’s explanation about the amount of applied to the controller at face value. If the device is responding to changes in the device on the order of a fraction of the cycle time of the 100kHz waveform observed, it could take significant computational resources to change control the device, generating heat. Also, a complex generator that can also apply a 1kVolt “spark” (estimated by Mats Lewan for the ignition phase) could be inefficient in the lab research version.

    For what it is worth, I estimated the time constant of the device to be about 206 seconds from the initial slope of the heating temperature curve. If this was heating water, the volume of the water heated was about 80 cc. That is not out of line with an estimate of the volume of the housing around the device.

    • Thomas Kaminski
      • pelgrim108

        In this video excerpt you can see the light clearly with the accompianing click sound.


        • Thomas Kaminski

          Look closely at 2:19:57. Even more interesting..

          • pelgrim108

            I will make an excerpt.

          • Thomas Kaminski

            Thanks for doing this. If you look closely at the end of the reactor, you can see both the ignition glow and a fainter glow after ignition that stops coincident with switching the controller off.

        • Obvious

          It may be that the horizontal clamps are a bit loose holding the QX, causing an unintended secondary arc at the end connections. I have a pair of those exact clamps (Powertec #20304) and they are rated by the manufacturer for 300 lbs clamping force. So one would want to be gentle clamping the QX between them. Too gentle, and the electrical connection might not be as good as desired.

    • Pekka Janhunen

      Do you mean it generated light for 3 seconds, then was dark 4 seconds, etc? Or that it generated light continuously until it was shut down?

      • Thomas Kaminski

        That the port used for the spectral light emission measurement emitted light, not just the “flash” observed at the ends. I think (but did not confirm) Mats said it was during the powered period and stopped emitting light when not powered.

        • Thomas Kaminski

          Actually, going back to the video at 2:19:57 (when Rossi finds out that he forgot to turn on the reactor for the spectral measurements), you can clearly see the “flash” to start the reaction along with a fainter glow that immediately ends when he switches off the reactor. I suspect that it will be a similar behavior during the 4-on, 3-off 7 second period test.

    • ARM

      water is flowing at 1 liter per hour = 1 liter per 3600 seconds = 1000 cubic centimeters per 3600 seconds = 0.2777777 cubic centimeter per second and based on that (206 seconds) x (0.277777 cc/second) = 57.2222 cc

  • Axil Axil

    I am about to write a post, This is a test to see if my posts are accepted.

  • I see that he is using a digital sampling scope and that could very easily be an alias waveform from under sampling https://uploads.disquscdn.com/images/3d302b1cb0b709e94f901777403c3eb9a554dc92a9bdc240ec0b2470fcd35979.jpg . note the large Spike shown in the image attached. in several places I see large spikes going both up and down above ground. I don’t believe this is providing any kind of an accurate measurement. combining this with his fiddling of the control system between resistor changes leaves way too many questions.

    • – A strong case for using a good, old (high speed) analog scope!

      But Robert, do you want to comment on the fact that it is now clear that Rossi is (also) using RF to stimulate the reaction? Not that different from your (Brillouin’s) use of what amounts to a high voltage spike inspired by an old TV time base (as I recall – correct me if I’m wrong or outdated)?

      Can I also take this opportunity to ask you where, in broad terms, Brillouin is relative to the E-Cat at this stage?

      • ARM

        The implication is that the scope is under sampling. However one of the most watched capabilities of a digital scope is its sampling rate. A 1 GHz sample rate can reproduce signals as fast as 500 MHz which is much higher than here. If anybody knows the make and model of this scope please let me know.

        • Obvious

          Looks like a Tektronix TBS 1052B

    • Engineer48

      Hi Robert,

      The QX is so COOL!

      Basically the QX is a Focardi reactor on steroids with an intense source of protons impacting the Ni surface with enough KE, thanks to the Ni being negative and several kV of E field to accelerate them, to penetrate into the Ni matrix and initiate a LENR reaction.

      Focardi will be smiling! He was right!


      BTW the low level signal is being generated by the H+ changes on one electrode and the electron – changes on the other electrode between the plasma ignition and sustain pulses.

      This output signal is in keeping with Rossi’s earlier statements that the QX outputs both Dc and Ac, which is now clearly shown on the scope.


      The AM of the signal may be due to the 3 QX reactors generating slightly different output frequencies and a hetrodyne generating the AM. Also note the signals are smuged or smeared on the scops, which may be the result of the 3 QX reactor output freq being slightly different.

      • Bruce__H

        I can’t see the signal declining between the non-clicking pulses. Where do you see evidence for this?

      • ARM

        I am not sure I understand your schematic. Dimension of L=0.6mm is the same as one electrode. My question is are showing 0.1mm as where the stimulation occurs namely in the middle of 0.6mm? That is 0.25mm + 0.1mm + 0.25mm = 0.6mm?

    • Omega Z

      Who provided the oscilloscope. I believe these things were provided by others. Only the QX reactors and control box were Rossi’s.

  • Engineer48

    It is clear when you view this video at speed 0.25 that there are 2 powerful clicking control pulses that 1st ignite the plasma and then some time later another clicking pulse terminates the plasma.


    Watch carefully as the grow continues, at a low level, after the 1st bright light burst. So we learn the clicking pulses can both ignite the plasma and also terminate it.

    This may be done by using reverse polarity on the termination pulse to drive H+ ions / protons away from the normally negative Ni electrode and to stop the LENR reaction by starving it of H+ ions / protons.

    I believe what we are seeing may work like this:

    Consider 2 x 0.08mm dia electrodes, 0.25mm long, separated by say a 0.1mm gap. 0.6mm overall length. One electrode Ni and one say SS. Seal both electrodes in a H2 filled high temp transparent tube.

    Apply -1kv to the Ni electrode and +1kv to the SS electrode for a short time, say 10us.

    Plasma forms between the electrodes, disassociating localised H2 into 2 x H+ protons & 2 x electrons.

    Protons are driven by the 2kv electrical potential to the negative Ni electrode, while electrons are driven to the positive SS electrode.

    Protons arrive at the Ni surface with some KE and are driven into the Ni matrix, initiating a LENR reaction.

    Electron and proton movements also create an electrical potential between the electrodes generating a back EMF, flowing current and energy out of the reactor after the plasma ignition pulse ends.

    QX control circuits monitor the self generated electrical output as an indication of reaction strength. When the reaction slows, as H+ protons are consumed, self generated voltage drops until at some point the control circuit generates another plasma ignition pulse, increasing the supply of H+ protons, increasing LENR reaction strength.

    From the scope images it is clear there are many ignition pulses, spikes on the scope that continue until the control circuit decides to stop the reaction for a few seconds.

  • Thanks Skip. I can tell you that I talked with some fairly competent people at the test who felt the contrary—that the demo had been very valuable in showing details making the process inside the E-Cat QX more understandable.
    But I’d love to take a beer with you at any occasion we might meet next!

    • Skip

      You have my greatest respect, if your sources have yours; I am encouraged.
      However, if I don’t experience it personally, its hearsay.


  • LION

    Hi Sam,
    naughty boys dont get rewarded, Mats was gifted the pump and the block, Andrea’s warmth and Respect for Mats is Genuine. It was only out of real concern for a Friend that the rebuke was so sharp, that the shock of it would to save him from coming to any harm. However I too appreciated the humor of it, the same thoughts passed through my own mind. Best wishes to you.

  • I have visualized both moment (“A.R. Before I open the air window, that is in the side of the control box in front of me”), where Rossi is definitely switching something (in slow motion) and afterwards when he ask Fabiani about the switch. Fabiani then said in Italian at 0048 min: “Andrea, you did it all by yourself. Get mad with yourself”.



  • Engineer48

    Good overview of gas discharge tubes:


    Note section on relaxation oscillation

    Plus understand a plasma or glow discharge breaks apart H2 molecules into 2 x H+ ions (protons) and 2 x electrons. Postive change protons are attracted or accelerated to the negative electrode and the negative charge electrons are attracted or accelerated to the positive electrode.

    Consider the left electrode is Ni and the right electrode is SS or some other non transition metal.


    Now consider swapping the electrode polarity 100k times a second with H+ ions moving away and then toward the left side Ni electrode. Imagine a jackhammer driving H+ ions deep into the left side Ni matrix 100,000 times a second.

    • Eng48—you seem to know a few things about gas discharge tubes. Do you know if there’s a well defined value of power consumption per length of such tubes? I got a suggestion that using that value, and entering the length of the QX reactor, would be a reliable method for estimating the maximum theoretical power consumption of the QX.
      I found this value in wikipedia: “33 to 82 watts per meter of tube length (10 to 25 W/ft) for T12 lamps”, but I don’t know if that’s a commercial value or a generic value valid for all kinds of gas discharge tubes.
      As the length of the reactor you could use the gap between the nickel electrodes as given in the Rossi-Gullström ArXiv paper—1.5cm. Or, conservatively, the whole reactor length, about 10cm.

  • Engineer48

    QX Proton Jack Hammer

    Accelerating and driving H+ Ions / Protons deep into the Ni Matrix to initiate LENR reactions, 100,000 times a second!


  • Deffoda Jonnyto

    To me the answer is very simple.
    after lifting the box and “opening the air vent”, Rossi is struggling to close it properly. he is looking around for something stuck that didn’t allow the cover to go back to its place.
    He asks Fabiani if he sees something out of place, and he says “it’s the switch”. than he flips the switch and Rossi is finally able to close it back to where it was.
    That switch is the only one accessible externally.

  • AdrianAshfield

    Well I’ve written a piece but its not been published yet. Tax and sex scandals are crowding everything else out.

  • Omega Z

    Rossi said he didn’t invite any journalists several times.

  • Omega Z

    Because in your world all positive posts must be Rossi’s alter ego. Only those that are not are real people,. Or perhaps, posts at JONP are logged only when Rossi approves them from moderation.

  • Omega Z

    When the reactor is disconnected from the control, it also disables the cooling fans.

  • Omega Z

    ->”The air window is in my side of the box. The window gives access to the circuitry.”

    Thus the reason Rossi didn’t want people behind the table. Some may recognize components which could give clues as to how the electronics work. This would be proprietary.

  • sam

    Andrea Rossi
    November 26, 2017 at 4:04 PM
    Andrew ( second answer ):
    You made me curious and now I had the time to watch with focused attention the part of the video you refer to. Please you too watch with attention the segment between 2h30′ and 2h32′: it is very interesting about your issue. Before I open the air window, that is on the side of the control box in front of me, Fabiani says to me absolutely nothing either before or during the action of opening the air window; after the action is finished and the box has been lowered in normal position, Fabiani says me to turn the switch, and at that point I go to the side of the box at my right, where the switch is, and make the switching action.
    Just to be clear and precise.
    Warm Regards,

    Andrea Rossi
    November 26, 2017 at 3:04 PM
    Dear Readers:
    It has been published on youtube the last part of the demo of Stockolm, with the theoretical presentation by Carl-Oscar Gullstrom: please go to

    November 26, 2017 at 12:30 PM
    Dear Dr Rossi,
    You said you opened an air Window to help air circulating, but Fabiani clearly tells you to push a Switch. It’s clear from the audio… Can you clarify?

    Andrea Rossi
    November 26, 2017 at 1:21 PM
    When Fabiani is saying me to turn the switch, he refers to the switch that turns off and on the power. We had to turn off/on the power when we passed from calorimetry to spectrometry, and when we passed from spectrometry to dummy 1 Ohm and when we passed from the dummy with 1 Ohm to the dummy with 800 Ohms. It is obvious that when we had to substitute the components of the circuit we had to turn the power off, make the substitution and, after that, turn the power on again. The switch is in the right side of the control box from my point of view, left side from the point of view of the public. The air window is in my side of the box. The window gives access to the circuitry.
    When I opened the air window, obviously, Fabiani said nothing, because in that moment I was not operating with any switch.
    Warm Regards,

    April 15, 2016 at 11:14 AM
    Dear Dr Rossi:
    There are so many people following your story, cheering for you, supporting you, praing for you, and yet the only thing you have given back in these last years are some ‘Rossi says’ that most of the time are followed by ‘someone says not ‘ . And I mean.. you don’t owe them anyhing, you are not supposed to satisfy the general public curiosity just in exchange of some sort of cybernetic support. But seriously, we are not talking about telenovelas curiosity, we are talking about people worried about the future of energy, the future of this planet, the future of their children, because hey, don’t try to give me the ‘all energy sources will be integrated’ joke, your tech. is kind of last hope.
    And I understand your worries about the competition, I understand the very long time needed to a such unconventional tech to be developed properly.
    But there are soooo many ways you could put an end to the skeptics chatters, that you could prove without any doubt that your tech does what ‘Rossi says’ . And yet you don’t do it.
    Once again it’s just Secret customers secret deals, secret results of reports..
    You are a good man I’m sure of that. Just think that it would cost you nothing to give something back to those that have been defending your name for this long time

    • Bruce Williams

      This is a brilliant message Andrew, thanks, I agree with all your sentiments. I hope that Andrea Rossi will give us a meaningful reply.

  • Bruce__H

    I am confused about the update to the original post on this thread. Why is it written in Rossi-ese but signed by Mr. Hurley? For example …..

    – Mr Rossi regularly spells Stockholm as “Stockolm”. This is the spelling we see in this update.

    – Mr Rossi often reports power in units Wh/h instead of the equivalent, but more usual, Watts. In the update Mr Hurley adopts Rossi’s Wh/h.

    – It seems that Mr Hurley is a senior engineer at Andeavor Corp., but he appears to misspell Andeavor as Endeavor just as Mr Rossi does.

    – Mr Hurley appears to write in slightly ungrammatical English just as Mr Rossi does, e.g., “The water pumped for 1 hour has been poured in a plastic container seat on a scale to measure exactly the water passed through the E-Cat.” Did anyone hear Mr Hurley talk? Is he American? I would expect less of a Rossi flavour to his writing if he is American..

    In sum. Since it certainly sounds as if Mr. Rossi wrote this report, how do we actually know that Mr Hurley really approves of it as his name at the end would indicate? Does Mats Lewan actually know if Hurley stands behind this report or has he just been told this by Mr Rossi?

    P.S. Has anyone located any information about a senior engineer for Andeavor named William S Hurley? I can’t locate anything (I am aware, though, that my internet investigator skills are not the greatest).

    • sam

      I heard him talk a little on a video
      and he sounded American.

      • sam

        Comment from LENR Forum



        18 hours ago

        The “Original Test” Rossi blog reader is a wonderful resouce! Here we meet William S. Hurley many, many times.

        A selection of his posts on the JONP blog …


        July 9, 2017 at 9:25 AM
        Me. Rossi,

        Congratulations! Please keep in mind our refinery heaters and boilers as an application.
        have a new name for the company, Andeavor, and now own 10 refineries
        across the USA. We supply over half the gasoline in California.
        Thank you and God speed!


        June 14, 2016 at 12:24 PM
        Mr. Rossi,

        You and your team have accomplished an incredible feat of discovery and engineering. Congratulations!

        Have you designed heat exchanger(s) for use in an industrial setting.
        I am thinking of feedwater heaters set up in a duplex design where one
        can be taken out of service. I am not sure what to do with the light,
        maybe solar collector. So many possibilities.

        God Speed

        June 3, 2016 at 12:48 PM
        Best Birthday Wishes!

        Success, Health and Love

        God Speed

        April 7, 2016 at 8:44 PM
        Mr. Rossi,

        They should be throwing a parade for you but instead try to steal every thing.
        I know you will prevail. You have many fans and supporters.
        God Speed

        March 30, 2016 at 9:48 AM
        Mr. Rossi,
        Sergio is smiling!

        Keep your container office for the museum.

        Congratulations, I had no doubt.

        God Speed

        March 2nd, 2013 at 10:21 AM
        Dear Mr. Rossi,

        If I want to buy a commercial 1MW ECAT for my refinery, where would I get it and do you have a price?

        Thank you and congratulations on your success.

      • Omega Z

        Hurley has been following Rossi’s work since at least July of 2011.

  • GiveADogABone

    ‘Measurement of the energy consumed ( during the hour for 30′ no energy has been supplied to the E-Cat) :
    V: 0.3
    OHM: 1
    A: 0.3
    Wh/h 0.09/2= 0.045’

    If the QX was running for 30 minutes in SSM mode, then during that period the CoP was infinite. When energy was supplied for 30 minutes the CoP was 22.8/0.09=253.33.

    Did anyone at the demo observe this 30 minute period of SSM and record it as such?

    • Bruce__H

      I think that the 30 minutes being referred to here is the non-stimulated part of the QX duty cycle. Over the hour of the demonstration the control box was actively stimulating the QX only about half the time (actually, according to Mr Rossi and according to the oscilloscope the stimulation phase lasts 3/7 of the duty cycle). One entire cycle lasts about 8 seconds so there were over 400 QX stimulation cycles carried out during the 1-hour demonstration.

      I actually don’t know what self-sustained mode means for this new plasma-based Ecat. I asked Mats Lewan if he saw the glow of the plasma last beyond the 3-second stimulation phase of the duty cycle and he replied that he didn’t think so. So it seems like the plasma collapses during the 4-second non-stimulation phase. And yet from what I can understand about the theory presented by Gullstrom the LENR heat production absolutely depends on the presence of the plasma. So how can you have self-sustained mode happening during the non-stimulation phase?

      I don’t get it.

  • sam

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies.