Do you have a chart of heat recovery rate at steady state for a wide range of powers?
My earlier variable power supply was too unstable to do sufficient quality measurements so I just have 200 W input with a dedicated 50 V DC power supply for now. I will add more power options at some point once other issues are squared away, mainly airspeed sensing/logging capability.
It would have to, since you reduced the inlet size. The question is: did the airflow velocity measurement change enough to make the balance zero? It wouldn't be exact, but it should be close. This is a good test of your instruments. But in real experiment you should not change a parameter such as the inlet size.
The heater core temperature dropped about 5 C when the inlet was restricted, so there was some element of actual increased heat exchange going on.
Test run MR2.4 has started. The live video is available at https://www.youtube.com/watch?v=nJNRhdIUY_8
For this test the cell was baked out and 370 Pa of pure Deuterium was added from a new lecture bottle. The same treated Ni+Pd mesh from previous runs was left in place.
On a related note, this is the calibration curve I am using:
| Input power | Temperature |
| 0 | 18.44 |
| 20 | 88.34 |
| 30 | 119.81 |
| 40 | 149.18 |
| 60 | 202.15 |
| 80 | 248.46 |
| 100 | 289.29 |
| 120 | 325.83 |
| 140 | 359.29 |
| 160 | 390.86 |
Run so far for the 20W input power step
It appears that temperatures are somewhat higher than calibration but higher power levels should be better calibrated.
The situation up to the 30W input power step:
The entire run with data manually sampled every 5 minutes from the live dashboard:
Using my calibration curve I get a roughly constant excess of a couple watts until about 60W. However my calibration values differ slightly from those of magicsound. Data is from the final (settled) Tcell 2 values.
| Input power (W) | Tcell2 (°C) | Output power (W) | Excess (W) | Efficiency |
| 19.737 | 94.83 | 22.01 | 2.27 | 111.50% |
| 29.955 | 127.90 | 32.69 | 2.73 | 109.11% |
| 39.938 | 155.90 | 42.39 | 2.45 | 106.14% |
| 49.705 | 182.20 | 52.14 | 2.44 | 104.90% |
| 60.082 | 206.60 | 61.81 | 1.73 | 102.88% |
Since the apparent excess is not increasing with temperature I think this could be a calibration artifact.
Thanks can for your prompt and excellent graphing work. I agree with your conclusion that there was not excess heat above the measurement accuracy of the experiment. I attribute the measurement offset primarily to the long thermal settling time of the cell, apparently more than two hours at the low power levels used last night. I will delay the second half of the test ( the higher power steps) to consider the implications going forward. The complete data files are available at:
One other potentially important observation is the elevated Neutron detection seen during the first hour of the test. Here is a plot of the 5-minute forward average showing the trend line. The duration of the elevated level is too long to be caused by environmental disturbance like dust in the air, but it could be solar or cosmic in origin. For that reason, I think it's worth repeating MR2.4. I'm running an extended 20 Watt step this morning, starting with the cell cold and the pressure at 352 Pa.
I can't access the datafile page due to permissions required.
Neutron count was already high in the beginning when temperatures were low and PdD was starting to decompose from the increasing temperatures. Was it elevated all along, and only started to decrease after applying power? If it's the case it could perhaps be an idea to perform room-temperature testing and monitoring the neutron count at varying D2 pressure levels, before/after applying a vacuum, etc. Since heating will be disabled it should be safer to run unattended over many hours.
So last night I warmed up the calorimeter box to steady state, then fiddled with the inlet restriction again...
I measured the outlet and inlet air velocity with the hot wire anemometer.
The outlet air was 32 C, Ambient air temperature was 18 C.
The anemometer probe tip temperature was adjusted by inserting into outlet air for the required time to heat up, then cooling towards ambient, and repeated as necessary to maintain the anemometer at the desired probe temperature during each measurement.
The hot wire anemometer probe temperature is the temperature reported below.
Inlet velocity with 1/2 of calorimeter inlet hole covered by tape:
2.29 m/s @ 20.1 C
2.20 m/s @ 23.0 C
Outlet velocity with 1/2 of calorimeter inlet hole covered by tape:
1.67 m/s @ 20.1 C
2.21 m/s @ 25.0 C
2.69 m/s @ 28.4 C
2.88 m/s @ 30.1 C
// .........
Inlet velocity with normal inlet opening:
1.28 m/s @ 20.0 C
1.18 m/s @ 21.1 C
Outlet velocity with normal inlet opening:
1.91 m/s @ 21.0 C
2.31 m/s @ 25.1 C
2.50 m/s @ 28.0 C
2.93 m/s @ 30.0 C
// ........
3.47 m/s @ 14.0 C both outlet air and probe temperature.
3.57 m/s @ 20.0 C both outlet air and probe temperature.
I can't access the datafile page due to permissions required
Neutron count was already high in the beginning when temperatures were low and PdD was starting to decompose from the increasing temperatures. Was it elevated all along, and only started to decrease after applying power? If it's the case it could perhaps be an idea to perform room-temperature testing and monitoring the neutron count at varying D2 pressure levels, before/after applying a vacuum, etc. Since heating will be disabled it should be safer to run unattended over many hours.
Permissions are updated, and the posted link should now work.
I agree regarding the re-test of Neutron count. I'll leave the heat off for several hours while collecting date, to see if it's time-of-day related. Then a second test for PdD loading or decomposition later today.
Here's a graph of the run using the full data.
I can see that neutron readings were the highest at the beginning with power off, but there's also an odd periodic signal seen in pressure and temperature though several input power steps. This feature could be a good sign as Mizuno observed something along these lines as well, although with larger magnitude and longer period.
https://www.lenr-canr.org/acrobat/MizunoTincreasedea.pdf (slides 22 and beyond)
Here are two sections in detail where the periodic variation is more visible:
Pressure changes first, then Tcell1 follows almost simultaneously, and Tcell2 comes after that but by briefly increasing in temperature.
I ran a background check of the Neutron count yesterday, over the same time interval as the 2.4 active run. The cell content was unchanged.
No heater power was used and the range of temperature was 19-23°C. Another test seems justified, after bake out and adding fresh D2.
Experiment details hae been added to the live document at https://tinyurl.com/vudbmro
Do you realize you have an accuracy issue of you actually get excess heat? Try plotting your cal curve out to 400W or so...
I now have the entire 9 hours of the 120 W calibration. I have uploaded the spreadsheet here:
Jed, is this 120 W calibration spreadsheet still available?
Do you realize you have an accuracy issue of you actually get excess heat? Try plotting your cal curve out to 400W or so...
Kirk, you may be referring to the modified calibration chart previously posted by can. The polynomial resulting from that chart will clearly have an inflection point above the displayed range, to which I suspect your comment alluded.
That said, the safe operating region for the heater in this apparatus is limited to 170 watts, corresponding to ~400°C at the outside surface of the cell. The equivalent temperature of the heater thermo-well is probably 600°C or more. Therefore accurate calibration at higher temperatures cannot be experimentally tested without risking damage to the apparatus in its current form.
Below is the master calibration chart for the system, extracted from a series of 10 calibration sequences . I have added an approximate logarithmic curve to estimate the temperature up to 400 watts. While I claim no particular accuracy for such an estimate absent experimental confirmation, it does suggest that 400 watts would yield an exterior surface temperature of around 550°C.
My point was that it isn't safe to extrapolate a statistical fit very far. The cubic fit gives the inflection point as you noted which I think your system isn't going to show. Therefore trying to use the cal curve at any temperature in excess of say 10% over your maximum value in the cal data set introduces a 'modeling error' of unknown magnitude. The only way to know what that is is to calibrate at higher input powers, which you unfortunately can't do due to equipment limitations. But this error means that in general one should run the experiment to keep the final temp in the calibrated range. IOW, a COP of 2 due to some 'excess heat' would mean the input should be no higher than half of your calibration range.
I have to wonder how many people doing the Mizuno (and other) experiment(s) do that...
The previous test ("MR2.4") had input power in the 20–60W range, while the calibration curve was up to 160W. The polyfit might have not been perfect, but no extrapolation above the calibration input power range was performed.
On Monday I baked out the cell for 12 hours, at up to 130°C The final vacuum was 2E-5 Torr. Tuesday I added 363 Pa of D2 and saw loading (pressure decline) almost immediately.
That continued for about 8 hours, reaching 215 Pa. The pressure remained at that level overnight.
MR2.5 active run is now live at https://www.youtube.com/watch?v=Vy90oJPZxxo
Of particular interest is the neutron count during initial heating with 20 watts.
The 20W and 30W steps so far. No particular feature from partial 5 min data that I can see.
On a loosely related note, I use a spreadsheet like this to create these graphs from manually inserted data.
