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.

This is the most promising result I have yet written about.  I put it together into a google doc that is embedded here.  As usual, please let us know if I missed anything.


Many thanks to Celani, the rest of the MFMP facilitators as well as the frequent followers, commentators, and ad-hoc contributors.


Video explanation of results:

We put together a real quick video explaining the results in the report.  Enjoy.


Next steps:

  • in the EU, Mathieu is super close to starting up a very similar test with his cells in differential mode.
  • In the US, we will to heat up CuNi2 to get it to load, then watch for additional excess energy
  • Perhaps try to drive flux by operating in a partial vacuum and allowing the hydrogen to flow out of the wire.
  • Add some Hemholtz coils to put the whole thing in a magnetic field, per a suggestion from Dennis Letts based on his work in electrolytic systems.  Below is a rendering of a coil holder intended to hold 100 wraps of 26 gauge copper wire.



Update 9/5/2013

After looking at the data from Ecco's test, it appeared to Malachi that most of the leak was in cell A, so he tightened the threaded rods a little bit in case the leak was in the silicone o-rings.  It seems to have worked.  The graph below shows the Mica temperature in cell B, the pressure, and the resistance of the fully loaded CuNi_1 wire.  It's pretty cool how they all tracked so nicely.  The pressure does not have a general downward slope!  


Meanwhile, the CuNi_2 wire is starting to load a little more as we start to raise the current in it.  It is nowhere near as rapid as previous loading where we went straight to full power, though.  We're OK taking our time because Celani suggested loading them in 25 degree steps and waiting till the resistance settles.  Our steps are a bit smaller because I am not sure what to expect while the other wire is still on.

Our Helmholtz coils are taking shape.  We will test them in open air before we try to put them around the cells.  We'll post a picture after get them wound up and powered up.


Second Update on 9/5:  The value of motion detectors

Just after we upped power into the cell in our attempt to get the second wire to load cell B started going up to 7 degrees warmer than cell A.  This coincided with a good rate of resistance drop, too.  We got a little bit excited.

Then we realized that Wes had been in there working on the powder cells around that time.  We started to winder if he was causing an airflow disturbance.  When he came out, the temperature dropped.  We sent him back in and the temperature rose.  This is clear on the graph below including the motion sensor data.  

This is good to be aware of.  We have asked ourselves if this could possibly account for the good results.  While we can't rule it out totally, we also can't imagine a way the airflow could have been that far out of balance for that long.



Update 9/6/2013 - Magnetic effects on Cell A?

We set a magnet on the side of the cells several days ago on Sept 3.  

Cell A

Cell B



Today, I took it of of cell A to take a magnetic field reading with an app on my phone. I didn't want to disturb cell B. Then I put it back on in nearly the same location with the same magnetic pole up.  The magnetic had probably cooled down during that time.  

Shortly after that, the power to both cells was raised by 2 watts. Cell A rose very little from the previous temperature.

A bit later Malachi adjusted the location of the magnet very slightly (a few millimeters) to what he remembered was the location it had been in originally.
The resistances of the NiCr wires behaved very strangely. The temperature of the cell behaved very strangely. The graphs below show this.  


So the magnet would have been cooler and cooled off the sensor and taken quite a while to come back to equilibrium.  But why do the resistance act funny?  Especially on the NiChrome wires??  What will this mean when we put the Helmholtz coils in place?


Add comment

Security code


0 #43 Games 06 2021-02-16 03:16
Excellent post. I was checking constantly this blog
and I'm impressed! Very helpful information particularly the last part :
) I care for such info much. I was looking for
this particular info for a long time. Thank you and good luck.
0 #42 welho webmail 2020-05-21 00:17
If some one desires to be updated with newest
technologies afterward he must be pay a visit this site and
be up to date everyday.
0 #41 Lonnie 2018-02-17 13:47
Hallo zusammen, lesen i Blog von Zeit zu Zeit und ichh besitzen eine ähnliche und ich wwar nur wundern neugierig, wenn Sie eije
Menge von Spaam erhalten Feedback? Wenn dem so
ist, wie Sie verhindern, dasws Schutz voor es, jede Plugin oder
alles, was Siie können schlagen? Ich bekomme so iel in letzter Zeit es macht mich verrückt verrückt so ddass jeder Support wird sehr geschätzt.
0 #40 Lavina 2018-02-13 20:19
Hallo, Icch fühle mich i i saaw besucht Sie meinen Blog hier ankam zu
Rück der Lustt Ich?. versuchen beei der
Suche Fraen zu verbessern verbessedn mein Ich gehe davon! seine gut genug zuu nutzen machen einige von Ihre Ideen !!
0 #39 Cecile 2015-04-28 10:59
Users tend to be more responsive to pages and people that make
posts often and automation can hhelp to track the how many posts arre being made and when. Today, Facebook is
among the best if not the best social network that is visited
by the maximum number oof people every day. Mark Bush is an expert author of Telecommunicati on industry and contributes
his valusble thoughts for telecoms industry readers.
0 #38 AlanG 2013-09-23 04:16
Quoting Ecco:
I've been saying since last year that differences in wire emissivity could account for much of the excess heat calculated. Most of the rest might be due to random differences in air flow/currents around the cells, since they're not in a tightly controlled testing environment.

I suggested in April that selective finishing of the copper bands could give a useful approximation of a metal cell, and could be retrofitted without disassembling the current experiment. Simply apply a black finish to the inside of the bands and polish the outside, with a clear-coat to prevent tarnishing. This would more accurately include the radiated heat in the measurement of the cell. The suspected emissivity differences as the wires age, and the visibly changed reflectivity of the copper due to tarnishing over time would both be reduced or eliminated as unknown errors in the thermal data.

The unstable convection around the cells is harder to deal with. Something like a version of the CTC system large enough to hold the glass cells would be nice. Maybe the fan, diffuser and heater stack from the original air flow calorimeter could be adapted to a big tube enclosure.
0 #37 Edwin Pell 2013-09-19 04:19
I was looking at a Celani powerPoint and see the same ambient and external temperature oscillations. Very cool.
0 #36 Edwin Pell 2013-09-16 01:00
A 0.4 degree rise on the incoming air is amplified to a 3 degree rise of temperature_1_e xternal. The ambient leads by about 90 seconds.
0 #35 Edwin Pell 2013-09-14 16:38
This is the air incoming to cell B. Is the thermocouple isolated from cell B?

0 #34 Edwin Pell 2013-09-14 02:04
We have a nice strong 1.5 degree (min to max) oscillation in cell A external and a 0.4 degree (min to max) oscillation in cell A ambient.

I can't tell who is the leader and who is the follower.

This differential method is not working. The reference side is not stable.

0 #33 Ecco 2013-09-13 21:28
US Cell 1.3B (active) is more sensitive to ambient temperature variations than US Cell 1.3A. I don't think this is due to a LENR effect:

0 #32 charlie tapp 2013-09-11 19:17
@robert greenyer on the subject of magnetic alignment, whan welding sometimes it is required after the weld is done to ping the metal to realign the ions of the metal for a stronger weld. only works with hot metal. throw a bass speaker in front of it for some vibration. just some dumb ideas i have.
0 #31 Robert Greenyer 2013-09-11 05:46
@Charlie tapp

It is to do with magnetic alignment in the body of the wire and how that relates to the Hydrogen. It is not about inducing electric current in the wire.
0 #30 charlie tapp 2013-09-10 22:19
@mfmp i am very interested why the magnet effects things it only should if physicaly moved. the magnetic field is stationary. how about for a quick test you guys swing a piece of metal on a string back and fourth over it causing turbulance in the magnetic field and see what that does to the data. not so close that it actualy moves the magnet. water being dripped next to it would be neat to see also kind of like a kelvins thunderstorm apparatus induceing a charge on the wire.
0 #29 Ake malhammar 2013-09-10 18:50
I am very sorry for my linguistic mistake in the comment below. It should be “assures that the air flow enters horizontally”.
0 #28 Ecco 2013-09-10 11:24
@Ake Malhammar: I'm glad I'm not the only one who thinks so. I've been saying since last year that differences in wire emissivity could account for much of the excess heat calculated. Most of the rest might be due to random differences in air flow/currents around the cells, since they're not in a tightly controlled testing environment.

However instead of painting (did you actually mean that?) the wires dark, MFMP could more simply paint the glass tubes opaque, or using (metal) tubes instead of glass ones, which would be a better solution but would also leave potential room for contamination. Foil wrapped on the glass tubes could also be used but might bring a different set of problems.

From the latest MFMP video it's very apparent that the NiCr wires in Cell A are shiny metal gray colored, while the active ones in Cell B are dark brown:

+1 #27 Ake Malhammar 2013-09-10 09:19
I do not new if this is news for you but I will pass it on anyway.

The wire is cooled by convection and radiation. Some of that radiation escapes through the tube glass wall without impacting its external temperature. A metal shiny wire radiates less than a dark wire. As a result the control cell must dissipate more heat form its external surface and thus be warmer.

This could explain the differences for low input powers in the curves shown in the video. So it might be that the actual difference in heat dissipation is larger than the difference of the temperature curves. A remedy could be to colour the passive wire dark.
0 #26 Edwin Pell 2013-09-09 01:27
At this point I agree with Ecco excess heat is not correlated with flux. It seems to be correlated with time of day. It seems to peak at midnight in Greenwich, England. I guess about 7pm in the lab. Is there any automatic air handling turning off for the night around 6pm?

0 #25 Ecco 2013-09-08 23:52
I've already posted this on ECW in their latest MFMP blogpost, but thinking a bit more about it, I guess it's perfectly fine here.

@MFMP: What about putting both cells in two separate open tubes (made of plastic? That would have some advantages. Or of metal with an external insulating jacket) with a temperature controlled constant input air flow, and then comparing the output air flow temperature of both cells? You could set up the cells in a vertical orientation which would help not only keeping the flow controlled and uniform along the cell length, but also saving space.

This would an improvement over the current setup, I think. It will ensure that both cells are receiving about the same air flow. and therefore that they're heating up in the same way, in addition of protecting them a bit from random air currents from lab activity.

EDIT: I'm aware (actually, I just remembered) that you already had something similar in place when you were running this experiment with the v2.0 protocol, but it wasn't quite the same thing as what I am suggesting here.

EDIT2: I couldn't resist making a diagram of what I have in mind. However I can see there might be problems in setting up a good, undisturbed air flow and choosing carefully a good place for measuring output air temperature.


Maintenance (ie disassembling/r eassembling the thing) might also get tricky.
0 #24 Edwin Pell 2013-09-07 17:42
At this time delta T is rising and flux is flat and low. Could this be due to increasing loading?

Please do not change applied power. Let's see how far this will go.
[add] not much flux but rising delta T.
[add] plot update 8:40pm east, 5 degrees with little flux? 34 continuous hour of excess heat Letts' Magnets rule!
[add] >0.2MJ

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