Two Curious Events

Geschrieben von Ryan Hunt am 16. November 2012.

Two interesting little anomalies in last nights data that I am having a tough time explaining.

The Ambient fell suddenly just after midnight

The Excess power reacted as expected. The P_xs continued to look positive for the next 8 hours, when I had expected that it would equilibrate back to where it started after 45 minutes, though. Perhaps I am wrong.

But why did the Internal cell temp go up??

The Celani Wire Impedance just continued to decrease during this time.

Curious that it didn't restabilize and that the temp inside went opposite the temp outside.

Then, later on, after we had gone up in power, we replaced the room heater and the T_ambient rose again sharply.

And, the P_Excess went down again, as we mathematically expect when the air changes faster than the cell. And it appeared to mostly recover.

Now this is the unusual part. When the Ambient temp rose 1 degree, the internal cell temperature dropped 2.5!

And then it recovered again! Meanwhile, the impedance dropped VERY slightly (probably a reflection of the temperature dropping)

The dropping in the impedance resulted in the Power in going up 15 or 20 mW. And still, the temperature dropped significantly and recovered.

This seems to imply that either something endothermic suddenly happened, or something exothermic ended briefly.

Curious that the temp inside went opposite the temp outside. Any ideas?

< Zurück Weiter >

Add comment

Comments

0 #24 drew 2012-11-17 16:54

Celani in his paper did say that somehow variations in the ambient room temperature seemed to help the reaction get going!

Quote

0 #23 JOB001 2012-11-17 07:07

It appears the cell is responding correctly assuming constant voltage. With constant input voltage power increases rather than be controlled when impedance changes. Input power P=VI=RI^2 so a 10% impedance change creates a 1.1^2 = 1.21 or 21% power change.
The rest is time delay effects from heating and cooling. Someday maybe a constant power supply control or a constant impedance control will be tested. For now it appears correct, good work!

Quote

0 #22 Ecco 2012-11-16 23:32

I see you have started a new run with an initial pressure of around 0.5 bar (something more). Is this one with hydrogen?

Quote

0 #21 Wookie 2012-11-16 22:18

@Robert, @Ryan, @ David , @everyone involved

great job, keep it up!
I*d like to express my joy and admiration, watching these wonderful scientiffic research LIVE with all the great comments from people actively helping to improve the experiment.

Thank you all.

and make it watertight :)

Quote

0 #20 Jack Cole 2012-11-16 22:15

Re-reading the article, I may have misunderstood what they were talking about with the temp ranges, but maybe somebody can make some sense of it.

Quote

0 #19 Jack Cole 2012-11-16 22:10

Here is a paper that within discusses the optimal hydrogen loading temperature for material prepared similar to Celani's method. It suggests that the ideal temps for the loading may occur between 225 and 255 C. Go much above or below that and not much happens.

144.206.159.178/.../...

Quote

0 #18 Robert Greenyer 2012-11-16 21:38

I concur with Ryan - some great suggestions

@David

You do not need to apologise for posting regularly when you are doing such a great job!

Thanks to all for your thinking

Quote

0 #17 Ryan Hunt 2012-11-16 21:28

Lots and lots of good comemnts here.
We are using DC, so in this case Impedance just means the same thing as resistance. We labeled that column generically in the power meter and haven't customized it for this experiment.

I absolutely love the idea of adding annotations as we do anything, change anything, or observe anything. That feature has been on my list for a long time, but it all takes time to accomplish. We will be open sourcing the software and publishing the API for others to use next week. Hopefully we can get some good assistance.

I will blow warm air on the instrumentation boards yet this afternoon when it settles from a tiny change I just made in the power.

In the not too distant future we will have a temperature stabilized air calorimeter.

Thanks for all the brainstorming. If I missed any good suggestion, I will try to pick them up over the next day.

Quote

0 #16 AB 2012-11-16 21:19

Quoting Ged:

@ AB,

Well, resistance and impedance both basically boil down to R=V/I. But only resistance can "burn off energy" in the form of heat. Impedance includes the resistance but also the reactance of the wire, which is how it changes in response to frequency. AC is definitely a little more complex than a DC current.

Let me rephrase the question:

Is AC used in these experiments or DC? Because Celani was definitely using DC.

Quote

0 #15 Ged 2012-11-16 21:18

@ AB,

Well, resistance and impedance both basically boil down to R=V/I. But only resistance can "burn off energy" in the form of heat. Impedance includes the resistance but also the reactance of the wire, which is how it changes in response to frequency. AC is definitely a little more complex than a DC current.

@ David Jones,

That's another good idea. That doesn't change the P_xs consideration, other than keeping convection down would allow us to properly detect any excess heat better, as it isn't being lost before we can monitor it. Could explain the rise, but all this of course depends on what happened at 12 am (a heater turning down since people aren't supposed to be around makes sense though), and how well isolated from air flow their cell is.

Quote

0 #14 AB 2012-11-16 21:05

This may be a stupid post as I'm not technically qualified to comment, but it could be important.

Impedance implies alternating current, right? Celani uses direct current though, or at least he did during the NIweek demo.

Quote

0 #13 David Jones 2012-11-16 21:05

Sorry, but another point - What I have said below could be all wrong. It could simply be that at the first event when the room temperature dropped (presumably because some heating turned off) that less air is circulating around the cell - that is the air is more still and hence less heat is drawn away from the cell - this affect should also therefore show up as an increase in the glass wall temperature too. The second event results in more air circulating.

Now, I will shut up

Quote

0 #12 Lu 2012-11-16 21:04

I think the rise in the T_Board temperature is telling. I assume it's probe on the board somewhere. Not sure where the T_Ambient probe is.

Quote

0 #11 Ged 2012-11-16 21:02

That's a good point, David. I hope they give that a look, and at least it should be easy to test that. Also like Alain's suggestion a lot.

So many mysteries to solve; science is a constantly unfolding detective's novel.

Quote

0 #10 David Jones 2012-11-16 20:52

Final note - then I will shut up

Your temperature probe boxes may have the facility to do a differential reading from two probe wires - by-passing the temperature reference control circuit.

Quote

0 #9 David Jones 2012-11-16 20:47

My point about differential readings is as follows. Temperature probes connected to a box use an internal temperature compensation circuit to act as a temperature reference point for the subtraction. If this is not working correctly then cooling the room (ambient) and hence the probe box may give a false indication that the cell temperature has risen. Connecting the ambient and cell probes together removes this problem as they directly read the differential temperature. You simple read off the temperature from the calibration table provided for the particular wires you are using.

Of course this might not be the problem at all...

Quote

0 #8 Alain Coetmeur 2012-11-16 20:45

Just an idea: add similar small periodic signal of ambient temp, an make analysis of the spectral response... it is used to detect strange correlations in noise...

now whether it is artifact or fact... but the causality will be clearer.

Quote

+1 #7 Matt 2012-11-16 20:38

I really think there should be annotations associated by time with the data stream. Information such as when the power blew out, what gas is in the chamber at the time. Any time relevant data should be viewed in the data view.

I'm also interested in adding click and drag to zoom capacities. I think it would really improve the ease of use in terms of exploring the data.

I think these two things combined would aid interpretation. Also perhaps clearer labels and information regarding the experiment in the data viewer itself.

Quote

0 #6 David Jones 2012-11-16 20:31

My guess would still be that your temperature probe boxes or some monitoring components are not stable/immune with respect to changes in their temperature. Simple test is to heat components gently with a low flow heat gun and see if this affects readings.

Quote

0 #5 Ged 2012-11-16 20:19

Fascinating data.

What does this all imply? T_ambient fluctuations should not be expected to affect the temperature of the cell in an inverse way. I would also say that while T_ambient is part of the excess power calculations, the affect it's having on the excess power trace is much larger than it should. Or put another way, that the changes in excess power calculation are too big to be simply from T_ambient values alone. Easy way to check that would be to dissect the equation and run the numbers.

But, it seems like the cell, at these internal temperatures, is very sensitive to heat flux. The colder room allowed a better flux, and it may be the -flux- that affects the reaction events and readings as well as (or more than) absolute.

Quote

Refresh comments list
RSS feed for comments to this post

JComments