FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

A Slow Climb

Written by Ryan Hunt on .

The temperatures and power out had a nice run up today for almost the last 24 hours.

And the temps all rose, albeit, slightly.  About 12:20, after struggling to determine a valid comparison, we upped the power to 48.8 watts.  Then we dropped it back to 48.2 a little later to be closer to our target and other calibration points.

Our first suspect was that the Ambient might be changing.  It was nice and steady as long as the rise was.  When it rose at the end, it made the P_xs get all choppy and appear to drop.  We're only talking about 0.4C range, though.  And notice that little 0.25 C spike at the beginning?  I have absolutely no idea what might have caused that.  It happens suspiciously close to the beginning of this whole rise, though.

This long rise started after we dropped the power by switching from heating with both wires to heating with only the Celani wire.  After the temperatures fell, they jumped back up and started rising.  Just before that is the mysterious T_Ambient spike.  The impedance of the Celani wire also started rising.

And the impedance has continued to rise smoothly ever since except for when we changed the power level.

The next suspect is the pressure dropping, which it did all day.  As the pressure drops, the gas effectively insulates better.  

So we dug into this to find out how the current conditions compared to other runs.  Since we are using the first Helium run with the Celani wire as our current baseline for measurements, we new that the cell was at a lower temperature than those runs.  The graph below contains reference points from ALL the calibration runs.  The little black triangle is the high point from today.  We are above any of the calibrations done before the Celani wire was installed, but still below either of the runs done in Helium with the same wire presumably before it was loaded with Hydrogen and active.

By the way, if there is a software guy, or Excel wizard out there that wants to help, I would be thrilled to get a script that gets the latest data from data.hugnetlab.com and plots that point on a graph with those reference lines behind it.  That would be so much faster and easier to tell how the cell is performing

But back to the pressure.  To figure out how the cell is performing as the pressure is dropping, Malachi the wonder-engineer graphed the rise of the cell temperatures above ambient versus pressure for the calibration runs with this same gas mix.  He extracted points at 48 watt power levels from each calibration run.  The circles are from the first run attempt on Nov 12th.  The squares are from this morning.  And the triangles are from this afternoon after it had adjusted to a new 48W power level.  The general trend is that as the glass temperatures vary little with pressure while the T_mica and T_well rise as the temperatures fall.  We are computing our excess heat based on the T_GlassOut for that reason.  

So, pressure drop may account for the rise in T_mica, but may not account well for the rise in the glass temperatures.

None the less, if we are getting excess energy we are clearly not getting enough to be indisputable.  In Celani's presentation, he claims it took him over 4 days to reach 10 watts.  Since we have a questionable wire that we have attempted to "repair", who knows what to expect.  

I think I will let it simmer over the holiday weekend and see if it continues to rise.

There have been some questions about the gamma detector.  It is set up and running.  The NaI detector is outside the shield, or about 15+cm from the cell.  Every day or two I save the spectrum and start a new one.  I have not taken time to dig into them.  Anybody want to check them out and see if anything stands out?  

Add comment


Security code
Refresh

Comments   

 
0 #120 123star 2012-12-06 00:26
[please delete, posted in the wrong place]
Quote
 
 
0 #119 Sanjeev 2012-11-26 20:53
@Ecco/Ged,

If I recall correctly, when the indirect heating was on, the conditions were very different, like pressure/ambien t or the loading were different.

Now we have 1 W excess and a different situation. If when the power switches to inactive wire and the P_Xs drops to 0 or less, we can conclude that the 1 W is due to Celani wire.

It may remain at 1 W or it may rise a bit because there is indirect heating. These possibilities are also there.
Quote
 
 
0 #118 Ron B 2012-11-26 18:12
When I look at the data from 8:30-9:00 PST I see at 8:47 there was a jump up in the ambient temp and at the same time there a jump in power in. At that same time T_well shot up high while glass_in dropped and then shortly after that T_mica dropped.
Very curious chain of events
Quote
 
 
0 #117 Ged 2012-11-26 17:44
@Ecco,

Indeed, last try it was all negatives, but that was when the wire was potentially "dead" and before the fix. Now that it is "repaired", perhaps that experiment will yield new results. I think Sanjeev's idea is a good one, and definitely worth trying. Should help tease out some of the variables.

@Ryan

We're back to hitting over 1 W, but there is definitely something that goes on in that room or affects the cell that strips heat from the outer glass every now and then; or, the LENR reaction can "flicker". I don't see anything in the data that could tell us exactly what causes those transient drops.

I think it might be a really good idea to set up an airflow sensor near the reactor. That should illuminate a lot of information for us.
Quote
 
 
0 #116 Ron B 2012-11-26 17:40
It would be nice to have the display able to show unscaled values for selected channels. Just fit the data samples to the window so that we can see small changes in each signal at the same time on the same screen.
As it is now I have to switch between channels and try to remember exactly where one channel lines up with another. If there's already a way to do this I'm not sure how.
Quote
 
 
0 #115 Ecco 2012-11-26 17:21
@Ryan Hunt: excluding anomalous LENR effects, it is always possible that something in the reactor room due to failure, chance or something else (a change in air pressure due to somebody opening doors or windows, etc), suddenly moved causing short lived air movement near the reactor which temporarily stripped some heat from the glass tube causing a drop in P_Out. A webcam (even just for internal use) might be useful to check out things like this too.

@Sanjeev: I agree too about testing by applying the same power to the heater wire. There's the chance that, if the active wire is currently producing very slight excess power (~1 W), using the other wire instead might help it increase more, as documented by Celani in his ICCF17 presentation.

This test has already been made some time back, but at that time, P_xs was constantly below the zero line if I remember correctly, and personally I assumed it wasn't producing any excess heat at all.
Quote
 
 
0 #114 Ryan Hunt 2012-11-26 17:03
The drop at 10:47 local time was not something we can account for from our actions, changes in power in, or changes in ambient. Can anybody else?
Quote
 
 
0 #113 Sanjeev 2012-11-26 16:43
The P_in for inactive wire should be about 20- 24 W, to ensure that the Celani wire is not heated indirectly and remains totally inactive, for this check I suggested.
Quote
 
 
0 #112 Sanjeev 2012-11-26 16:36
Is it possible to turn off the Celani wire and heat it only using the inactive wire at 48 W for some hours to see if it shows 0 P_Xs ?

It looks like the P_Xs returns to 1W after disturbances in input, so the experiment can continue without problems after this little check.
Quote
 
 
0 #111 Ged 2012-11-26 16:36
Pumping up the juice didn't seem to help it any. But pressure drops may be the culprit now. This wire sure likes to tease, but not make our lives easy. Lots of interesting data though.
Quote
 
 
0 #110 Red 2012-11-26 12:01
Well that's disappointing :-/
Quote
 
 
0 #109 Ged 2012-11-26 01:06
According to the graphs, it seems the hydrogen pressure may be falling low enough to now be negatively impacting heat transfer from the mica to the glass; which will lower P_xs.

I wonder if there's any other way to ratchet up the cell temperature while keeping the wire at this amiable power in.
Quote
 
 
0 #108 Paul Hunt 2012-11-26 00:53
It appears that the glitch we caused in the software yesterday left some bad offset calibration data in the endpoint that reads voltage and current. Today we realized that and reset the board at about 16:05. The voltage, current, impedance and power were all off a little bit during that 22 hours. They should be more accurate now.
Quote
 
 
0 #107 Ecco 2012-11-25 22:56
@Ryan Hunt: by the way, I was thinking that another, quicker way to test for possible thermal influence on the reactor by radiation from the surrounding causing apparent excess power, would be increasing input power by roughly the average calculated excess power produced, which so far today has been around 1W. Input power has already been increased by 0.25W about 45 minutes ago, so it would have to be increased by an additional 0.75W, for a total of 48.9W as of now.

If the excess power is real, but it's also not some sort of gain (rather an absolute value), the cell should resume producing those 1.0-1.2W that in relatively short time, like yesterday after the short unexpected shutdown. If it yet again takes days or otherwise an unreasonably long amount of time to resume producing it, then there's a high chance we've not been witnessing excess heat.
Quote
 
 
+1 #106 Ecco 2012-11-25 22:08
I don't know if it was intentional, but the high power "tick" preceding the increase of input power back to 48W caused the red wire impedance to drop by 0.05 Ohm. I would repeat that again to see if it's possible to induce further decreases.

I wonder if it actually changed or if this is only a measurement artifact.

@Rats: actually P_xs has remained fairly constant over the past 24 hours. I wonder if it leveled out or if it still has potential to increase.

As for your question it's hard to tell. With an input power of 48W, I would personally say at least 10 Watt of excess power to confidently rule out measurement and calculation errors. That still would not be conclusive for many people, however, at least not with the current experimental set-up.
Quote
 
 
0 #105 Rats 2012-11-25 21:56
Interesting ... still climbing after several days. I have a question please. At what level of P_Xs could we confidently say it is above noise/error levels and the cell is definitely producing excess energy?
Quote
 
 
0 #104 Ecco 2012-11-25 18:25
A request: could P_xs color on the real-time chart be changed to something easier to read like bright orange?
Quote
 
 
0 #103 clovis ray 2012-11-25 15:01
hi, guys,
wow, 1.1904 and climbing.
Quote
 
 
0 #102 Jack Cole 2012-11-25 04:21
Pxs is negatively correlated with red channel power, blue channel power, and with pressure (psi). But note that since impedance of both channels is positively correlated with Pxs, dropping the power will drop the impedance possibly leading to no significant change in Pxs.

If the relationship is causal, the only change that would lead to increased Pxs is reduced pressure, perhaps leading also to increased impedance (negative correlation). It may not actually be causal. At a minimum, a controlled drop in pressure would help to demonstrate the presence or absence of a direct effect of pressure on Pxs.
Quote
 
 
0 #101 Jack Cole 2012-11-25 04:20
I ran some statistical analyses to see if I could discover associations that might help improve the Pxs. I would suggest trying the following in graduated steps (perhaps separated by a couple of hours to see the individual effect of changing each variable):

1) Lower pressure, 2) Lower red channel power, 3) Lower blue channel power.

Correlational data and other data here:
lenr-coldfusion.com/.../...
Continued in next post.
Quote
 

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