FacebookTwitterDiggStumbleuponGoogle BookmarksRedditTechnoratiLinkedin

RF Experiment Calibrations soon Underway [UPDATE #3 - Interesting Data Already?]

Scritto da Wes Baish on .

I toyed with preliminary application of waveforms across the powder yesterday, just to get a feel for the behavior of the cell and gain fluidity in my use of the controls for both the generator and oscilloscope. Of course, we ran into a few hangups that are not a detriment to the experiment, but do require slight tweaking in the protocol. I will highlight those tweaks along with a few additions to the test that merit mentioning before we get serious (this week, tentatively). 

First, our resistance circuit wiring requires adjustments. Instead of a 2.5V source we're now using a 50V source across the powder with a 10KOhm resistor. The input is now a high voltage source and must be put on a high voltage HUGnet pin; it differs from the low voltage pins in that it has a 100KOhm input resistance and 1KOhm resistance to C ground, yielding an effective Voltage of ~25V across the powder. This concept shifted through several stages yesterday, finally settling on the wiring diagram added to the protocol doc. We arrived here because of the excellent resolution numbers; 0.15 Ohm resolution from 1-100Ohms P_resistance, 0.6Ohms at 10KOhms, and 2,000Ohms at 1MegaOhm. Not bad - considering that's 0.5% of the total value. We chose to have better resolution on the low end, where the powder is loaded and the "magic" happens. Otherwise the high resistance values (in the MegaOhms) aren't much use to us. Powdered solids are fickle creatures and MOhms are just astronomical to deal with. 

The protocol doc leads me to my second point. There is now an info link available on the test page. It appears on the homepage of the data stream just behind the test name (there's one on the CTC air jacket test). I will post the RF info doc there as another convenient place to find it. It's growing to be quite a useful page!

Lastly, there's a clear link between AC frequency and signal noise on the entire HUGnet board for R3. The AC pulses perturb the resting voltage in the common ground plane that all the sensors are tied into. The slower the frequency, the more data reads per second pick up a higher common voltage during the pulse's relaxation time (~15ms). Due to this issue we won't dabble in frequencies below 2KHz. R3's power parameters (like all reef power parameters) are noiseless with the waveform addition because those measurements are done on a different HUGnet board. 

That's all for now. I'll update this post with the official announcement for when I'll characterize unloaded powder with different AC frequencies. Remember to watch the powder experiment log for the times I change the frequency. The log will closely match the data timestamps in UTC. Fingers crossed we can get rolling tomorrow. 

 


UPDATE#1 - AC Calibrations Today

Though "calibrations" isn't the preferred term, I will be applying the first round of AC characterizations this afternoon (1800 UTC). Again, there are no statistical analyses planned in the protocol or any power steps in this experiment to date. The characterizations are just a background check on unloaded powder at room temperature to see which frequency looks the most promising. Without good calorimetry for the device, actual power steps and statistics don't seem a very practical effort.

On to step 2! 

 


UPDATE#2 - Results of AC Calibrations

The RF protocol results from yesterday are now posted in the google spreadsheet, and quite the bombshell: there resulted no significant data.  P_resistance is too high to feel such a small poke from the AC generator, and this outcome was far from unexpected. We simply had to check for the sake of checking.

Notice that cell temperatures drop as time goes on, which is also expected. Earlier that morning I was searching for the H_Power that got T_Powder to ~108°C, shutting it off in due time before waveform applications. The thick insulation around the cell takes a good while to cool off. The slightly-higher than ambient temps were an artifact of the morning's power levels. 

So with step 2 (and 3, subsequently) out of the way we move on to step 4 of the protocol - H2 loading. Step 4 of the protocol doc is worded a little funky and I will clarify this small point to avoid any confusion down the road. We keep our H2 at room temperature, but we want the cell at operating temp during loading. We do not load with the cell at RT. 

One last thing: the applied frequency is still open to speculation. We won't be doing any frequency sweeps but instead want to hold it at one frequency for extended periods of time. I'd prefer the largest possible combination of square waves, amplitudes, and frequency, which happens to be at about 300KHz. As with everything, this is subject to change (with sound warrant). If nothing exciting happens at 300KHz, we'll try another frequency, probably the highest possible by the generator, and move down the protocol from there. 

Loading is the fun part. Enjoy!


UPDATE#3 - Interesting Data Already?

I loaded R3 this afternoon, the graph of the resulting data below. I have several remarks for this figure, but more so I have questions. . .

Here we see the first pressure charge in R3 for the RF test. Though the gold pressure line should be a straight, one-time shot to 150 psig, I happen to be a little rusty and didn't get it right the first couple times. However this slight blunder has offered us an interesting opportunity to watch the resistance flicker from "infinite" to 14 Ohms. The first charge at 63 psig forced a plummeting resistance value, only to skyrocket back up to the same infinite range s few short minutes later. In that point in time I was fussing with the cell's bellows valve to charge the cell again. I missed a closed valve along the H2 line to the reef which is why the pressure doesn't increase (oops), but the perturbation from my shaky hands handling the cell disturbed the ever fragile electrical path through the powder.

Brushing off the rust, charging a second and third time went much smoother and the resistance drop stayed permanently. While this well illustrates my lamentations of powder resistance, it also adds a slight bit of assurance that our data is trustworthy and not misguided by extraneous factors. It's uncommon to see such severe drops in resistance; however it is key to remember that this cell was never formally baked off and thus may still have some H2 loading present from its paces last year, leaving this oddity in the grey realm of possibility. Are there any alternative explanations to this resistance skyscraper? We sort of take an attitude of "guilty until proven innocent" in weird-looking data such as this case. You have to in this field! 

I would also like to call attention to the current rate of loading, shown above. We're steadily dropping ~1psi/20mins. With partially loaded powder, lower loading temperatures, and larger Ni powder (2 micron), this slope remains entirely possible - if not expected! 

The really interesting feature is in the temperatures. . .

T_Powder (in red) very quickly coupled to T_Shell's temperature with H2 charging, just as expected (beautiful, right?). The addition of room temperature H2 slightly decreased the cell temperature, again as expected. As of now (2107 UTC) though, those temperatures haven't risen again and still retain the slight negative slope in league with the declining pressure and P_Resistance. We've never loaded at this low of a temperature, is H2 dissociation that endothermic and drawing heat? Does it take this long to equilibrate? Is H2 absorbing more heat than in a vacuum? 

Though there is already puzzling data to behold, I can't complain - only if the data is genuinely interesting and not bogus. I don't even wanna think about leaks right now. . .

Add comment


Security code
Refresh

Comments   

 
0 #13 Wes Baish 2013-07-02 15:43
We should also consider the electrical path through the powder and the size of its cross section. Is there just 1 path of least resistance or does the current seek out multiple paths to induce such low P_Resistance values? How big are these paths?

It could additionally be a problem in the unit that supplies the powder's voltage. This is a simple, inexpensive supply that doesn't have our control modifications installed on it. Once it's plugged in it simply outputs 48V. 4mV jumps could be some slight change that starts at that source and trickles down 4 orders of magnitude.

We certainly know the powder is physically unstable and can shift very easily, but this is just micron-sized powder. It makes me excited to get more nanopowder cells running live.

The difficult part is that these small anomalies leave us with so many "could-be's" and few certainties.
Quote
 
 
0 #12 Wes Baish 2013-07-02 13:45
@AlanG It's not a WFG artifact because I haven't turned it on yet. Though it's a pretty small blip (4mV), I think it might lend some merit to odd behavior in the powder. On Sunday I was in the lab figuring out why R3's data turned off and shortly after fixing the problem P_Resistance did another small but noticeable jump.
docs.google.com/.../edit
Again this is small 4mV jump, but the instantaneous peak is puzzling. I brushed off Sunday's event, chalking it up to some effect of the data coming back on or some sort of tiny vibration from my lumbering around the lab shifting the powder just slightly. However this new anomaly can't be dismissed so easily as the motion sensor of the US1.3 tests indicate no one was in the lab at that time.

We may find it telling that these two separate leaps are the same size and could be a result of the endpoint's sense limits. One of its architects might know more.
Quote
 
 
0 #11 AlanG 2013-07-01 22:33
Quoting Wes Baish:
The powder should certainly see an effect; whether or not it's a positive one I can't say. . .


P_voltage jumped up at 22:21 utc along with P_resistance. Is this an artifact of turning on the waveform generator? There's a tiny increase in the pressure as well, but it's very close to the noise level.
Quote
 
 
+1 #10 Jamie Sibley 2013-07-01 22:08
I am putting this forward as a proposed experimental apparatus.



With the spring loading and constant vibration from the piezo transducer, we should be able to get much more stable resistance readings.
Quote
 
 
0 #9 Wes Baish 2013-07-01 19:06
So pressure has flat-lined over the weekend. Additional power to the heater didn't seem to help. Are there any objections to turning on the WF generator? The powder should certainly see an effect; whether or not it's a positive one I can't say. . .
Quote
 
 
0 #8 AlanG 2013-06-29 00:13
One idea is that the nano powder is packed so tightly that the hydrogen can't fully penetrate at lower pressures. With enough pressure it is forced into the voids and eventually permeates the whole powder volume. This creates a bulk thermal conduction path to the T_powder TC. The inter-particle voids represent a very small fraction of the powder volume, so only a small drop in pressure right at the transition would be expected.
The hydrogen somehow also creates a bulk electric conduction path to the core, and how that works is the real question. If the powder had been exposed to air at some point, maybe the process is simply reduction of the non-conductive oxide layer in the bulk powder. It could also be something like electron migration from H+ (proton) absorption into the nano nickel. That doesn't seem likely at this temperature and voltage.
Quote
 
 
+2 #7 AlanG 2013-06-28 17:40
Regarding the RF stimulus voltage, we discussed earlier the probable need for much higher voltages before any effect is seen. One easy way to achieve this with the current apparatus would be to use a matching transformer. Even something as crude as a plate transformer salvaged from an old AM radio would give several hundred volts of AC waveform from the function generator, up to 10 kHz anyway. The goal of this is to fluidize the powder by piezoelectric agitation rather than to supply direct emf stimulation of the Ni.
Quote
 
 
0 #6 Wes Baish 2013-06-26 15:09
@all Ryan and I had a discussion this morning concerning calibrations for powder cells. We haven't planned any calibrations in powder protocols - much like this test (protocol is located on the info link next to the test name as well as a number of other places scattered in powder docs), where the "calibrations" we mention are just one-time frequency sweeps with and without heat. Our determination of excess heat in powder has simply been if powder temperature exceeded shell temperature.
However this method of determination is not set in stone! We have considered doing an average of each temperature rise at equilibrium which, right now, seems a good option. There is also potential for He atmosphere calibrations. The thermal coupling in a vacuum compared to a H2 atmosphere is cause for concern, and H2 loading happens almost immediately following charging (adding the H2 to the cell) which adds complications.
Are there any other supposed options for more legitimate calibration?
Quote
 
 
0 #5 Robert Greenyer 2013-06-26 07:40
The Celani V2.0 protocol experiment comparative test data is not useful at the moment as the cells are loading. When they are in active run - the calculation for the excess heat based on calibrations is here

quantumheat.org/.../...

on the first spreadsheet.

the Powder cell will likely have something similar when it has finished calibration.
Quote
 
 
0 #4 Robert Ellefson 2013-06-26 04:19
Quoting h veeder:
What is the formula? Is it displayed with the data?


Good question! In theory, you should be able to find it in the experiment's protocol description, located under "Follow"->"Expe riment Overview"->"Pow der Test Cell #1"->"Info"

However, it seems that this document does not yet describe the analytic models applied for derived results. Although this experiment is not reporting P_xs, it is reporting power and resistance values, which are derived from raw sensor readings.

I'm currently working up an experiment definition document protocol that will contain both human-readable and machine-readabl e sections. There will be a checklist of information that needs to be included in all experiment descriptions, including raw sensor data descriptions that have references showing where in the apparatus the measurements are taken, as well as derived data descriptions with formulas. I'll be posting the live document soon.
Quote
 
 
0 #3 h veeder 2013-06-26 03:14
Quoting Robert Ellefson:

Excess power is not a data point that is measured specifically; rather, it is an analytic result that comes from applying measured data to a formula intended to model the behavior of our apparatus.


What is the formula? Is it displayed with the data?
(The formula should be easy to find. )
Quote
 
 
+1 #2 Robert Ellefson 2013-06-25 22:13
Quoting frasp:
Can I now trust the excess power readings that are shown here and there on the live data pages?


Excess power is not a data point that is measured specifically; rather, it is an analytic result that comes from applying measured data to a formula intended to model the behavior of our apparatus. Based on the results, a number is produced, which we are calling excess power. It is your responsibility, in fact it is your contribution, to determine how trustworthy this analytical result is. If you have reason to believe it is incorrect, by all means, please share!
Quote
 
 
0 #1 frasp 2013-06-25 21:39
Couple of retard questions again...

Can I now trust the excess power readings that are shown here and there on the live data pages?

How much active material is there on one meter of Celani wire?
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