Sweep with Both Wire Results

Scritto da Ryan Hunt on 19 Novembre 2012.

Here are some interesting results from last night's run where we scanned a range of temperatures using both wires so the current in the Celani wire was in a similar range to what he demonstrated. The graph below gives you the timeframe and a view of the temperatures achieved at each step.

Next we have the power steps for the NiCr wire (Blue channel) and the Celani Wire (Red Channel).

The Celani wire ran more current because it has a lower resistance and they are both run off the same voltage. Although, I am seriously considering adding another power supply specifically for the NiCr wire. It would be interesting to hold the cell temp constant while changing power through the active wire.

The impedance data was interesting, again. We saw a small range of steps in which the impedance rose with the temperature and then started to decline. The range in this experiment was at much hotter temperatures than the sweep test with just one wire. We still don't know what this means, exactly, but I think it is a hint. So, while I am writing this up, we stepped the power back two steps to the highlighted point.

The Excess Power data looked uneventful to a casual observer. What we saw last night:

What I "think" I see in it may be more intersting. This 5 minutes averaged data makes it much more readable.

Go ahead and tell me if you think I'm grasping at straws. I know the downward spikes should grow with the size of the power increase, which increases with the square of the voltage, so I am largely at a loss to explain this shape. Or the downward trend of the top of the first 6 steps.

The temperatures in the first few higher points are rather close to the range in which we saw decreasing impedance over time in the Mini Sweep done a few days ago. See: Sweep of Loading Temps Write Up

This is totally making me wonder, though, since it is consistently beneath the power levels seen in the first runs with the Celani Wire in Helium. Were we seeing excess energy from the "unloaded" wire and we aren't anymore? Or is the difference between Helium and Hydrogen? We'll have to plot a few more temps on calibration curves and get another perspective.

< Prec Succ >

Add comment

Comments

0 #20 Pierre 2012-11-20 00:19

I wonder if there is a variation in the internal resistance of your psu at certain current levels. It could be verified by using an ordinary copper-wire / resistor.

0 #19 Ryan Hunt 2012-11-19 22:13

We are working on the control loop for the power supply currently. It may be up to 3 weeks, but when we get it working we will be able to set the voltage, current, power, or a temperature.
Unfortunately the data collector computer is bogged down with something for the moment and not uploading the data. We are trying to figure it out.
It was suggested that running at low pressures of Hydrogen like we are may slow down the recombination of H2 and improve the loading. I was getting antsy to up the pressure back to 3.5 bar again. I am not too concerned about oxygen or nitrogen because I have need no hint of the pressure rising in the cell over time.

0 #18 JOB001 2012-11-19 22:06

Why run at less than 1 bar? This creates the potential for N2 and O2 contamination and criticism due to leakage. Isn't the wire is sensitive to O2?

0 #17 Ecco 2012-11-19 22:04

@Ryan Hunt: thanks for the change. It appears that other channels were affected by this issue as well, not just P_Out.

By the way, the data feed doesn't seem to be updating at the moment.

0 #16 Michael Kussmann 2012-11-19 21:54

Could you modify the control of the PSU so that you get the Pout controlled in a closed loop? This would make P in to the wires control easier . (My lab PSU has that functionality and it is sometimes really helpful). As Hydrogen has a significantly higher specific heat and a higher thermal condictivity than helium, this could explain the "negative excess" energy.

0 #15 Ryan Hunt 2012-11-19 21:36

@ Ecco - the change you suggested is made and will be installed shortly. Unfortunately, we can't make it retroactive.

0 #14 Ryan Hunt 2012-11-19 21:35

At Mathieu's suggestion, I flipped the power off on the wires for about 3 seconds and then back on, again, just to perturb the system. Also, the ambient is rising because the main building is rising because today is unseasonably warm and sunny. It stays fairly constant in here, but not within a tenth of a degree.

0 #13 Ged 2012-11-19 20:54

On another note, it really does seem like P_xs has a slight upwards trend to it; as it's getting consistently more on the positive side of the line the longer this run goes. At least so it seems to my watching. This is getting more and more tantalizing. Strange it's so sensitive to the presence of people.. could it be from air displacement?

0 #12 Ecco 2012-11-19 20:53

By the way, gaps in the data feed make P_Out drop to zero causing chart plotting problems. Can this be fixed, for example by simply leaving a gap and not assuming the missing data as zero?

0 #11 Ged 2012-11-19 20:50

@Ryan

Apparently it's got performance anxiety :D.

0 #10 Ryan Hunt 2012-11-19 20:38

At 14:20 and 14:30 I took a bunch of photos of the wire, trying to get enough detail to see the texture and color of the wire. Both times it correlated to a P_xs drop. Who knew it wouldn't like getting it's picture taken?

0 #9 Jack Cole 2012-11-19 20:14

It doesn't really seem to be that closely correlated with the ambient T changes. If you look at the graph, ambient goes up and down and back up again. Pxs just goes up to the new level and has stayed fairly consistent.

Was anything new turned on in the room upon entering (lights, etc...)?

-1 #8 Ecco 2012-11-19 19:25

@Ryan Hunt: ok I see, thanks for answering, surely the reactor at the moment seems behaving interestingly, worth to keep an eye on as it is.

Another question: is it physically possible to remotely (= without entering the reactor room) decrease/increa se input voltage by 0.05 - 0.10 V or so just to give the reactor a little "nudge" and see how it reacts in these conditions (with an apparently anomalous behavior) to a very small input, but not microscopic, power change?

Also, regarding the power supply: is it designed for quick and small input voltage changes (up to 2-3 times per second and more) ? I was thinking that one could try inputting a sort of "square wave" and see if the reactor "likes" it more than constant input.

[just brainstorming. But if sudden small changes on external conditions can cause significant reactor changes, that might be something worth to explore]

0 #7 Ryan Hunt 2012-11-19 19:20

There is a gamma detector nearby. When I went to save the spectrum file and accidentally deleted it for the last day :c No undo button.
I was about to try to up the power so we could explore the 300+ range. Perhaps that should wait and let it sit.

0 #6 Ged 2012-11-19 19:10

@Ryan,

Fascinating, hard to reason why such a small ambient change would change the other values so disproportionat ely and in a sustained way after ambient stabilized again.

Maybe we simply need to wait? Once we are at the favorable conditions, perhaps triggering is a slow progressive event with this design? I would say maybe we should run it all day at the sweet spot to see if time is a factor here.

0 #5 Ged 2012-11-19 19:08

I will say, the P_xs is definitely above the other runs we've seen starting with the helium conditions. Not sure what this implies, since it's basically averaging around 0 now, instead of -2 W as previously in the same conditions but lower temps.

It is... interesting. Maybe we just need more ways to measure and cross check power out, like IR imagery.

0 #4 Ecco 2012-11-19 19:07

@Jack Cole: It looks that started right when ambient temperature also increased markedly. However, shouldn't a rise in ambient temperature cause the calculated excess power decrease?
Internal reactor temperatures appear to be slightly hotter, and heater wire impedance appears to reflect this too.
Hydrogen pressure also appears to have behaved a bit irregularly over the past 3 hours.

Unfortunately this rise is still within measurement and calculation error margins, so if there actually is excess heat, it is inconclusive at best.

Is there a beta/gamma radiation detector near the reactor, by the way?

0 #3 Ryan Hunt 2012-11-19 19:06

Yes, Jack, it does look interesting. We went into that room about 12:00. We were disturbing it by trying to do a study of the air flow around the cell with a neutral buoyancy smoke pencil. We noted before we went in that the P_xs was at -5 to -7, new record lows. We were trying to explain it with air flow around the glass, maybe, although nothing had changed in that room all morning. While we were in there working around the cell, we apparently caused the room temp to warm up by 0.5C. Usually when the Ambient rises quickly, the cell takes a while to adjust and the result is a lower apparent excess energy value. This went the other way. The ambient seems to have settled now. The External glass temperature has settled 6 or so degrees higher than the low point before noon. A very noticeable change and somewhat inexplicable.

0 #2 Jack Cole 2012-11-19 18:51

Starting after 12pm today, the excess power graph is looking more interesting. It above the baseline that it had previously settled in on and showing above 0 part of the time.

0 #1 JOB001 2012-11-19 18:17

The impedance rise per step is increasing from 0.02 per step at the start to 0.04 for the last step. More significantly, at 290C you may be18w * 2^-60C/5.59 = 0.01w, or way below assumed 18w output at 350C, so 0.01w is below experimental error and higher temperature is required.

Refresh comments list
RSS feed for comments to this post

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