FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin


The Martin Fleischmann Memorial Project is a group dedicated to researching Low Energy Nuclear Reactions (often referred to as LENR) while sharing all procedures, data, and results openly online. We rely on comments from online contributors to aid us in developing our experiments and contemplating the results. We invite everyone to participate in our discussions, which take place in the comments of our experiment posts. These links can be seen along the right-hand side of this page. Please browse around and give us your feedback. We look forward to seeing you around Quantum Heat.

Join us and become part of the project. Become one of the active commenters, who question our work and suggest next steps.

Or, if you are an experimenter, talk to us about becoming an affiliated lab and doing your work in a Live Open Science manner.

Final calibration plots and master spreadsheet of the Cell#2 mounted with the oxide coated constantan wire (CuNi44).

After we modified the internal parts of the EU Cell, a constantan wire (CuNi44) - coated with an as yet unknown oxide layer - was mounted in parallel with a NiCr wire. We decided on a plan to calibrate the cell using this wire, this plan has been described in an earlier blogpost.

The philosophy was to do the necessary calibrations that would anticipate most of the behaviour of the cell during "THE RUN". The original papers from Celani were presenting results using a cell loaded with a gas mixture of 75% of dihydrogen and 25% of argon. We knew later that the reason for the argon was to "artificially insulate" the wire in the cell because the intrinsic temperature of the wire is the triggering parameter of excess heat production whereas too much current destroys the nanostructure.

Because Celani’s cell was leaking, we simulated this by varying the pressure (3.5b, 2b, 1b @ T_ambient). The other way of making the temperature of the wire higher without injecting more power is to use vacuum. An additional effect is that H2 tend to reform more at higher pressure, vacuum gives more opportunity for mono-atomic hydrogen to get back into the active material. So we calibrate at 0.5b and at 1b.

Celani mentioned that the USA cell calibrations may have shown excess heat during calibrations because of the use of a bare constantan wire that can slightly absorb hydrogen. The reason why we decided to use a wire coated with an oxide is that it would be less able to interact with hydrogen than a bare constantan wire would do, because the oxide would better insulate the wire from the various atmospheres used. This assumption was a bet more than a belief, our results tend to show this because the variation of resistivity of this wire immersed in hydrogen atmosphere is small compared to the previous measurement we did with the Cell#1 and the difference in emissivity from the wire is more dependent to the variation of ambient temperature than the nature of the gas.


However, the very first calibration using this wire highlighted less heat transferred to the outside the cell, where the thermocouple is located, for this specific gas mixture and pressure.

The first option is the following: starting at a certain quantity of argon per volume (mol/l), the argon absorbs more infrared than at lower concentration. This would spread the heat along the cell and distribute it as "IR heats more the gas than the glass", giving a lower value for ∆Tout. I have not found the IR absorption spectrum yet, but it sounds an interesting possibility.

Another option is more oriented on the micro structural effects that occurs when the current stops flowing inside the wire for the first time in the alloy. I still remember my (greatly hated) electrical teacher talking proudly about the "Ampérions", described by Ampère at his time, to explain how small unknown electricity "particles" are spinning chaotically inside conductors when we stop applying current in a wire. Maybe it is a path to follow, I don't know yet.

The final option is brought by Pr. Celani. He argues that during the very first calibration, the wire got loaded partially with hydrogen, thus giving excess energy on the next runs. What I don't understand is why ∆Tout would not be greater at higher temperature as we know this is the triggering effect? This really bothers me, but I trust his experience in his wires.
Another thing is that the supposed excess heat effect didn’t go away after three hours of baking the wire under vacuum. He gives this option as the best way to reset the wire before new calibration. Even resistivity went up after, why it would go back down during an helium run?
Please let us know what you think! I am boggled.

Anyway, all this uncertainties are forcing me to do the run with a pressure of 2 bar in order to have better reliability on the results based on the calibrations curves we have.
The good sign is that all the other calibrations are following the same tendency and are grouped into a 5°C range. This mainly confirms what Pr. Celani was saying about the borosilicate tube acting almost like a blackbody, absorbing most of the infrared generated inside the cell.


When we look at impedance variability, we have an unanswered question too. The Impedance, actually "dances", during the runs in a remarkably small way! The maximum variation is 2.4%, where we had more than 20% during the first loading process in USA. It also shows very good reproducibility during He runs. The lower value is always located at a value of T mica = 230°C. This inflection point is enigmatic and is likely related to metallurgical phenomenons. As I mentioned above, baking the wire made it recover its resistivity properties but it faded out during the following calibration run.


On the above plot the translation of one line on the left is due to the molecular nature of the gas used. The more insulating a gas is, the lower the value of ∆Tout will be. This effect increases with higher temperatures.

In essence, the wire has very unpredictable way of behaving and we can clearly have doubts on which baseline we should use for the calculation of excess heat until we understand completely the results we gathered. If you have a suggestion on which one we should use, let us know.
Because I am sending the active cell up to Nicolas next Monday, I am left with less time than I expected to do further tests and validate some aspect of the experiment I wanted to clarify. Especially reproducibility of the calibrations with the oxide coated wire.
It is very important that we have a good loading of the wire first, to maximise the likelihood of seeing the effect. Nevertheless, I am convinced that the loading and the run I will do today will be scientifically interesting.

More very soon to come as I have the results of the He calibration with the active wire. They seem to fit with the very first calibration, aka gas blend under 3.5 bar.



Over the past few days we have had a dialogue with Celani as he starts to really scrutinise our EU data and cell configuration. What follows is the e-mail chain, with corrected english for readability, as normal the originals are included as a PDF for comparison. But first, a video!

Date: Sat, Dec 8, 2012 at 4:22 PM

From: Francesco Celani
To: Mathieu Valat

Dear Mathieu,

Your data, published 6 December, is really intriguing and I am DEEPLY interested about it.

I would like to get the PHOTO of your reactor to determine the location of the thermocouple that is measuring outside cell glass temperature.

Moreover, it would be very useful to get the values of the temperatures INSIDE the reactor (MICA, internal cell, SS).

Many thanks for Your help,


Date: Sat, Dec 8, 2012
From: Mathieu Valat
To: Francesco Celani


Please do not share this link yet, it presents the cell environement.


The glass is 300mm long.


DATE: Sun, Dec 9, 2012 at 1:26 PM,
FROM: Francesco Celani
TO: Mathieu Valat

Dear Mathieu,

Thanks for the video.

The overall environment is very simple and clean:good.

Anyway, I was not able to see in detail where the OUTSIDE cell glass thermcouple is. It is very critical that it has good thermal contact.

Moreover, for cross check of the results, the temperatures INSIDE the cell could help a lot.

Please, send me such values of the gaph, e.g. pdf format.

  • To be honest, my "feeling" is that your data is correct but, because it is very important it must be cross-checked very carefully.

  • At the same time, the effect of Hydrogen absorption in the constantan (with related, even if little, resistance decrease and possible anomalous heat generation) has to be clarified also with your independent data.

  • To-day, at home, I will try to recover and study the very beginning of my experiments of wires #1 and #2.

  • In parallel, I would like to suggest that You contact your Colleagues in the USA who are using a quartz tube and ask them to look in great detail at these critical aspects.

In other words, there might be a risk that the calibration wasn't as "perfect" as supposed, because some there may have been some Hydrogen absorption since the very beginning.

It is planned another meeting in Italy, this coming December 14 and if it will be possible to get such information, ASAP, it will be very important for ALL of US (i.e. MFMP and myself).

My best to all of You,


Date: Sun, Dec 9, 2012
From: Mathieu Valat
To: Francesco Celani


The thermocouple giving T(glass out) is clearly visible at the end of the video I sent you.

Our thermocouples are coated with stainless steal. I use the elastic properties of this coating to keep permanent contact between the very end of the TC and the glass.

I do not use Thermo-resistant tape to hold these two together as you do, because it concentrates heat. See the attached picture. The whitest part that have a shape of square is the tape...

This explains why I have a value of ∆T(out-ambient) lower that yours...

I attached a graph showing the variation of resistivity throughout the calibration runs.


DATE: Sun 9 December 2012 22:38
FROM: Francesco Celani

Dear Mathieu, Dear Colleagues,

The effect of possible "concentration of temperature", according to my
tests, is not really significant.

* It is significant however, if the contact between the thermometer (SS screened thermocouple) and the glass surface is just a little point, the temperatures measured will be LOWER than actual. This is because the "SS body" of the thermometer is at room temperature and some of these “low” temperatures spread to the contacting tip.

In order to reduce such an effect, I put the thermometer HORIZZONTAL for about
5mm over the glass surface. Moreover, I added a little amount of
thermal conducting compounds (white colour). In addition, just to be
"safe", in the experiments performed up to the end of August, I added also
a little amout of kapton tape (almost yellow colour).

From the end of August to the present day, I changed to using a SS spring, like a collar, that holds a few mm of the thermometer tip against the glass.

* Again, it will be very useful to compare the temperatures with those inside the cell.

* Again, I think that the behaviour of temperatures, using the oxidised constantan, at the first run (H2-Ar) and later are really important.

Just to dream, IF YOUR RESULTS comparing the first and 2,3,4,5 measurements are correct, the problem of LENR, at least using constantan, are solved.

Thanks for your time,




So it would seem there are a few things to have in mind which we can improve on with the next round of wire testing after the up and coming test.

Calibration wire Hydrogen absorption (Potential for Pxs under estimation)

The potential for some hydrogen absorption / resistance lowering and excess heat generation in calibration runs. This will need further investigation and if significant, will need to be factored in for future experiments. 

Poor Outside Thermocouple Contact (Potential for Pxs under estimation)

The thermal contact of the outside thermocouple is critical so we might possibly under estimate the outside glass temperature due to heat loss from thermal gradients on the tip of the thermocouple. The higher the delta between glass and ambient the greater this underestimation would be. Therefore, the more the Pxs that may be generated, the greater the under-estimation. Better thermal contact methods should be investigated for their significance and the best solution deployed in future tests.

Location of Ambient Air Thermocouple (Potential for Pxs over estimation)

In addition, the position of the ambient air thermocouple is a little way below the position of the outside glass one. This was to shield it from convective and radiative heating by the cell. There may however be a vertical thermal column gradient in the room – particularly on long runs if the outside heatsink wall temp is low.

This could be investigated for significance with a low thermally conducting bar and the laser thermometer. Arrange so that bar covers from below the ambient air sensor to above the outside glass sensor and then allow the room to stand with a typical cell temperature for a few hours. If cannot be measure through window then open door a little carefully and take readings to assess potential thermal column temperature variation.

In a future experiment, having the ambient sensor at the same height but shielded behind a pin-pong ball should resolve any potential questions over thermal column.

Can you spot anything we should improve?


UPDATE 2012-12-11 20:24 CET: This is how the cell looks as we try to make it achieve a higher temperature after the  maxing out the power supply heating through the NiCr wire.  We are not trying to measure excess energy at this point, just achieve the best loading temperature.  We were limited to heating power by the smaller diameter of the NiCr wire used in this cell.

You can find here the Master calibration file that groups every measurements we have done on the cell#2.

If you don't have Excel, you can always use OpenOffice or LibreOffice with this file. Enjoy!

UPDATE #3 - Reading homework

This was published in 2008/9 as part of ICCF-15 but I feel it is something everyone should read. I was going to make a whole blogpost about it with some selected paragraphs of interest, but this will have to come later. However, it does explain why much of the New Fire advancements are coming from Italy and has papers from some of the key names before the current renaissance of the field. If you have little time, I would suggest you read the following

  • Sections 1 (in total) - Francesco Scaramuzzi
  • Section 2.1, 2.2, 2.6 - Francesco Scaramuzzi
  • Section 4.3 - Francesco Celani
  • Section 5.3 - Sergio Focardi and Francesco Piantelli
  • Section 6.3 - Ubaldo Mastromatteo

You will discover by reading these papers, much of it before the NanoParticle era and the replication of rare isotopic transmutation experiments that

  • Palladium Deuteride would never show an effect if the deuterium loading was below a threshold, hence the importance of trying to make sure we load as much as possible the Celani wire above.
  • Piantelli and Focardi had done the basics of the Ni+H systems with more basic Ni preparation and still reported achieving really significant Pxs - this explains why we, annoyingly, may have "got lucky" with our oxidised constantan wire in our calibration runs for the EU cell. See this key quote from page 172.

    "It was shown that nickel samples in a hydrogen atmosphere were able to produce additional power up to 50 W, after absorbing a given amount of gas, at temperatures in the range 150-400 °C, heated with power among 40 and 120 W. The energy production started after perturbing the system with power or pressure changes. In one case, before stopping the process, the system was kept for 24 days in stationary conditions with a power of 44 W corresponding to a total energy of 90 MJ produced. "

    On page 175, you can see the large volume of Hydrogen that was absorbed and would mean we would necessarily have to re-charge the cell to maintain pressure for and active run. Also on this page I quote:

    "At the end of the experiment the first one gave an energy production with a power equal to about 70 W to be compared with the supply power equal to 29 W (total 99W)."

    That is 241% excess - in a simple system. Hmm, this was in 1993, in fact, reading this - it is amazing that there are not many people reporting good results - is that all the power of scientific dogma?

    There is much to discover in that one paper that explains a lot of where we are today. I'll let you dig in.

In fact, I could go and point out all of the important insights in this document, but I need to prepare to go to Rome, so I would ask that you, the crowd, help pick out the juicy insights (there are very many) and quote paragraphs and page numbers for their relevance to what is happening this week in France and Italy. You may be surprised, the link is given below.

Cold Fusion Italy (till 2008)


Add comment

Security code


0 #75 Bernie 2023-10-11 21:25
What a information of un-ambiguity and preserveness of
valuable familiarity on the topic of unexpected emotions.

Visit my homepage ... cheap vpn subscription (www.yiquzx.com: www.yiquzx.com/.../
0 #74 Jayme 2022-10-19 08:29
Apart from Game 4, he didn't do it. With the constantly evolving features School of lust Free download: www.youtube.com/.../ this site, we provide you the ultimate selection and comfort to select and benefit
from the big selection of rummy game on dwell tables that are full of bounty prizes and eternal enjoyable.
This is an easy lure to fall into with boxed text: The creator (or GM) desires to ascertain the
important thing features of the scene - vines seem, ox panics, cart crashes,
disguised guards draw weapons - and the boxed format strongly biases you in the
direction of pushing all of that collectively right into a single presentation. Within the Art of the important
thing, for example, I speak about how these options of boxed text make it excellent for conveying what characters see
when first entering a room or location by clearly delineating the knowledge the gamers
should mechanically have from the rest of
the important thing. Make errors and your pay will probably be docked.
You pay a subscription to entry the entire library on-demand
rather than having to buy video games to keep.

This gives you full entry to any gaming store, but it's important to
pay in your subscription and your video games. If you’re a competitive gamer who
performs at the very best level, or just enjoy taking part in games for over eight hours at a time, it’s a
good idea to contemplate investing in a gaming monitor with flicker-free expertise.
0 #73 먹튀사이트 2021-11-07 08:20
Wonderful, what a webpage it is! This webpage
gives valuable data to us, keep it up.
0 #72 Esperanza 2021-10-11 03:43
There are some funn elements of constructing a enterprise — like picking out a name, creating our branding material, and connecting with leads.
0 #71 Lizette 2021-10-10 10:51
25% of U.S. buyrs consult social media earlier than shopping for a present.
0 #70 Mia 2021-10-09 10:16
Think about your customers’ neds whereas planning and designing your new site—this inherently leads to effective seo
0 #69 바카라 2021-09-26 05:03
Dead indited content, Really enjoyed looking at.
0 #68 web page 2021-08-14 11:41
Thiis website was... how do I say it? Relevant!!
Finally I have found something that helped me. Cheers!

Kasyno online web page: stellapolarisinitiative.com/.. ./... najlepsze kasyna internetowe
0 #67 купить диплом 2021-06-21 18:31
A fascinating discussion is worth comment. I believe that you should publish
more on this subject, it might not be a taboo matter but usually people do not speak about
these topics. To the next! Many thanks!!
0 #66 어 야동 봐야지 2020-12-21 09:34
하지만 많은 경우 회사가 액세스할 수 있는 경우 회원은
가족의 양육합니다. 앉아서보 30 분 필리핀 야동 실제 친엄마와 기준 세 번
과학적인 현실이다.워커
0 #65 KorySmall 2018-01-02 16:14
I have checked your page and i have found some duplicate content, that's why
you don't rank high in google, but there is a tool that can help
you to create 100% unique content, search for: Boorfe's tips unlimited content
0 #64 FirstChong 2017-10-29 17:44
I see you don't monetize your website, don't waste your traffic, you can earn extra cash every month because you've
got hi quality content. If you want to know how to make extra $$$,
search for: Mrdalekjd methods for $$$
0 #63 AlanG 2012-12-13 01:10
Here's a segment of the 12-8 calibration data, normalized to make comparing the curves easier.

The effect of T_ambient on the glass temperature is pretty clear, with a delay of about 2 minutes for T_out and 3 minutes for T_in. The dispersion of glass temperature data also suggests there is convective instability around the outside of the cell, since T_out seems to lead T_in for these short -term variations.
0 #62 AlanG 2012-12-13 01:05
T_ambient variability seems to be adding some significant measurement error. I'm working on a graph to illustrate this.
0 #61 Ecco 2012-12-12 20:01
@Ged: it looks like they're going to do that very soon, in a few minutes. See next blog post.
0 #60 Ged 2012-12-12 19:56
I agree with Ecco. The signal is much cleaner and clearer than ever, and the most conservative P_ex is showing a strong response. So, I too am eager to switch to the active wire and cross check.

I see we are now at 48 W in on the inactive wire. Once that is done, perhaps we should do the same power step routine with the active wire only.
0 #59 Rats 2012-12-12 19:51
Very interesting! Seems like there was excess power for about an hour but it's dropped off now.
+2 #58 Ecco 2012-12-12 19:15
Currently, 33W of input power seem to show a relevant amount of apparent excess power (about 3W as of writing, with the most conservative baseline). At these levels, couldn't power be safely switched to the active wire to cross-check this result?
+1 #57 David Roberson 2012-12-12 18:49
Ryan, I am using the history directly. There is typically 2 seconds of time between each event.

My curve fit routine does a remarkable job of matching the data when the power is stepped to a constant value. It is very evident that the outside glass temperature begins its rise about 30 seconds before the power reading shows an increase. I have been searching for a 30 second time constant explanation for a while now and I suspect that this might be the source.

Review the data for the time stamp beginning at 12/12/2012, 11:18 with the power reading 50.6236 watts blue and the glass temperature is 114.702 C. You should see that the temperature is just beginning to rise at this time but the power increase to 55.7 watts has not yet begun. It takes until 11:19 for that to begin to be registered. The actual time pass is 30+ seconds.

Do you see what I refer to?
0 #56 Ryan Hunt 2012-12-12 18:12
@ David - Are you looking at the raw history or the 30 second average? The power meter has been real responsive in my experience. You might be seeing an artifact of the averaging? A few rows of data showing what you are seeing would be a good thing to see.

Here is your generous contributions so far towards our $500,000 target, thanks everyone! : $45,020   Please Donate
See the current state of our booked costs here