Here's another cool iceberg photo for the desktop -->
http://socivilized.files.wordpress.com/2009/02/berg.jpg
| ||||||
  Loading ... |  Click here to read | Be certain to do a simple RC time constant decay test on your capacitor to measure it's parallel resistance before connecting it to the LED. I just did some measurements on various LED's with my electrometer, and was getting ~ 7mV DC with the green LED, a red LED, and another red LED. I thought it was odd that all three of these LED's were producing the same DC voltage, so I tested the capacitor's parallel resistance, a 0.1uF metal film. It's junk! Metal films are great capacitors, and in fact, I recently learned that the 353mV record was a 1.0uF metal film. Anyhow, the 0.1uF tested somewhat less than 1Gohm, which is terrible for such capacitors. Note, this was a short term test, as in minutes. I swapped the 0.1uF for a different 0.1uF metal film capacitor, and the DC voltage went up to 48mV DC. The new 0.1uF cap tested at 5Gohms, which is about what I'd expect from such a caps short term parallel resistance test. So you're probably wonder how the 1.uF metal film was charged up to 353mV. The answer is that the parallel resistance of such capacitor increases as the frequency decreases. For example, if you charge the cap from 0mV to 50mV, for at least a few minutes it's most likely going to be less than 10Gohms parallel resistance. Although, if you keep the 50mV voltage source on the cap for say an hour, then remove the voltage source and test parallel resistance you should see a considerably higher parallel resistance. There's a EE on the usenets that did very long term tests on a 0.1uF cap who was getting resistance values around 7.2E+22 ohms (72000000000000000000000 ohms) --> So a good capacitors parallel resistance should increase over time so long as the voltage source remains DC. Lets say my green LED was producing 10pA during the 353mV record. That puts the 1.uF metal film capacitor at 35.3Gohms. Measuring your capacitors parallel resistance: http://www.cvs1.uklinux.net/cgi-bin/calculators/time_const.cgi Leave the resistance field blank. Enter your capacitance and select "µ". Enter 1200 for the time and select "l". Percentage is calculated by taking the final DC voltage divided by 1.5V. If the final voltage was 1.44 volts, then 100% - (1.44/1.5 * 100%) = 4 percent. Click "calculate now." For this example the parallel resistance would be 5.87918 Gohms. A good capacitor should have at least 5Gohms parallel resistance-- short term testing. So my recommendation is to test a bunch of capacitors to find the one with the highest parallel resistance. You could even try various types of capacitors such as mylars and metal films. I loved this photo so much that I made it my computers desktop --> http://aemcomp.googlepages.com/glacierice.jpg or in my Google photos --> Just click Download Lets hope our great great grandchildren will live the day to see such natural beauty!!! It's nice that some people are now replicating the simple LED & cap experiments, or at least they will be within a week. Unfortunetely people still don't get it. Perhaps they're not reading all of the blogs. In short, if you want to see the LED's notable DC voltage, then make sure they LED has ***not*** been producing any notable DC current for some time, perferablly for at least one month. So when you get the LED, just solder the mylar cap to it, try to solder quickly so as not to heat the diode junction by that much, then quickly place it inside it's shielded dark contaimment for 2 to 10 hours. The cap will slowly charge during this period. Then measure the DC voltage that should be on the cap. There's *no* need to desolder the cap from the LED before taking the measurement. After you take the voltage measurement, then remove the cap from the LED to allow the LED time to recover. Perhaps I'm beginning to see a TED trend. The past weeks LED/photodiode experiments will be release when the data is more conclusive, but for now I'll say that a pattern appears to be forming. It appears (still inconclusive) that diodes with higher zero bias resistance tend to be less stable. This would match data on Seebeck coefficients with respect to dopant densities in that semiconductors with less dopant densities (higher resistance) tend to have higher Seebeck coefficients. So far my green LED has been the least stable of all my tested diodes. The NIR (Near Infrared - closer to visible red light) seems to be the more stable then the green LED. The Judson MIR (less resistance than the NIR LED) seems to be more stable than the NIR LED. The SMS7630 diodes appear to be more stable than the Judson MIR LED. Although this data is difficult to say for certain given the instability of the diodes. It just occurred to me that the TED effect could some cause problems with *large* capacitor tests. Every LED/diode would be different, where some LED's less sensitive than others. My green LED was connected to a 1.0uF metal-film capacitor purchased from radio shack, while my infrared LED has a 4.7uF mylar. I'm theorizing that the higher the zero bias resistance would tend to result in less stable diode. If true, then red LED's would tend to be more stable than green LED's, but green LED's could theoretically produce more DC voltage and power. Following the same pattern, infrared LED's would tend to be more stable than red LED's. Regarding the LED experiments, I would recommend the following experiments: Experiment #1: Experiment #2: What I do *not* recommend, unless investigating the TED effect, is conduct too many experiments on the same LED within a relatively short period of time. I don't have an equation to predict how long to wait, yet, but my SMS7630 diode array required sometimes as long as two months to recover from a lot of measurements. I believe the problem could be in the amount of DC current the diodes produce; i.e., they're only good for a certain amount of current per given period of time. I will go out on a limb and guesstimate that one LED is good for one good experiment per one to two months. The capacitance of the capacitor depends on your voltage meter. For 2Mohm meters, ~ 50uF. For 10Mohm meters, ~ 5uF. For 100Mohm meters, ~ 0.5uF. For 15Gohm meters, ~ 1uF is low enough. The green LED I used as mostly likely a 100mA high intensity, but I can't be that certain. Another thought just after posting the Diode caution again blog is that it's possible the 10pA DC current comes from the rectification of *high frequencies* ambient thermal AC noise. The SMS7630 diodes are microwave diodes that appreciably rectify up to 100GHz. LED's and photodiodes obviously react to high frequencies, infrared to visible and higher frequencies. Perhaps this will lead to a new and more accurate diode equation. IMO, the 1N914 diodes were simply too disturbed, as for years they were stored in a non-shielded plastic box. Such diodes produce dozens of milli volts while unshielded. Given their 4pF capacitance, there will always be significant AC current flowing through such diodes while unshielded. The topic of diode instability cannot be stressed enough, the TED effect. That is not say you will not measure a DC voltage, as it is very easy to obtain such proof. So far, all of my LED's, photodiodes, and microwave diodes have succeeded in detecting a definite DC voltage. Note that I do not include common pn junction diodes because yesterday I was unsuccessful in detecting a DC voltage produced by 5 in-series 1N914 diodes, but I used the AM240 meter, which can only detect down to 0.1mV. So I have no idea how much DC voltage the 1N914 5 diodes in-series were producing, except that it's less than 100uV. The predicted DC voltage is ~ 0.45mV. One important note is that all of my SMS7630 microwave diodes, that have always produced a DC voltage, were not tested immediately. In every case, the SMS7630 diode array was placed inside the metal shielded containment for at least 3 weeks. My new 1N914 diode array was immediately tested after making it. It's also possible that only high frequency diodes (such as the SMS7630 microwave diodes), LED's, and photodiodes produce the ~ 10pA of DC current. Perhaps simple pn indirect bandgap diodes such as the 1N914 produce significantly less DC current. Lots of unanswered questions. I'm nearly certain the 353mV record was achieved with a metal-film capacitor purchased at radio shack --> This green LED is very old, and has not been used that much. Not sure if that makes any difference. It has a typical transparent round plastic lens. The 1.0uF is ~ a year old, and also has not been used much. I don't think there's anything special with the green LED or 1.0uF capacitor, except perhaps that it was left undisturbed for ~~ a week (possibly less) before the 353mV test. The green LED test was contained in compete darkness, but not inside a metal box because I have one metal shielding system that is presently occupied with the infrared photodiode experiment, which BTW obtained up to 128mV DC across a 4.7uF mylar capacitor. People should understand they do not have to purchase an AM240 DMM. The AM240 was used to immediately measure the DC voltage across the green LED, which of course shunted the green LED since the AM240 is only 15Gohms in 400mV setting. If you voltage meter is low, 10Mohm, then get a 4.7uF good capacitor. Note that a lot of cheap voltage meters are even lower, 2Mohm, which means you'll need ~ 25uF. Please do not use electrolytic capacitors since they contain electrolytic material and have low parallel resistance. Before conducting the LED experiments, you should do some quick and easy capacitor measurements to see how fast your voltage meter discharges the capacitor. Charge the capacitor to 1.5 volts with a battery, *remove the battery*, then place your voltage meter on the capacitor. It would also be nice to do a 0.5 volt test, but not required. Please see my previous blog post on "old TED." I cannot stress the utter importance of knowing about passive diode instability. TED (Thermal Equilibrium Diode) effect is an effect I discovered long ago regarding the instability of diodes near thermal equilibrium. Data is beginning to offer further details of diode instability. Such details is regarding some old facts on a topic I discussed a year or so ago in my email list, Thermoelectric effects. It is a fact that when current flows through a junction, that a temperature gradient begins to form-- the Peltier effect. It is a fact this temperature gradient thus causes an opposing voltage on the current-- Seebeck effect. In my old email list I went into details about the effects on such thermoelectric effects. What I could not determine was how significant this effect is on diodes near thermal equilibrium because diode manufacturing companies do not provide the needed near zero bias thermoelectric coefficients. It now seems possible this could be a main cause for the TED effect. How "possible" remains to be seen. Semiconductors have the highest thermoelectric coefficients of all materials. Consider the following example. The diode rectifies ambient thermal energy, thus producing a DC current. A temperature gradient begins to form at the junction due to the Peltier effect. The rate at which the temperature gradient rises depends on the amount of mass (diode, the casing, wires, etc.) and the thermal conductivity of such materials. As the temperature gradient increases, so does the opposing voltage produced by Seebeck effect. Now under normal conditions, the temperature gradient would rise to its peak within minutes, but not at power levels ranging from pico to zepto watts. The amount of time required to create a microscopic temperature gradient across the junction on a common diode that's producing atto watts is staggering. In short, *if* such thermoelectric effects play a significant role in passive diode arrays, then the DC current would slowly come to a near halt. Such thermoelectric effects will *not* prevent the diode from violating the 2nd Law of Thermodynamics! By allowing the diode array to rest (open circuit), the temperature gradients slowly fade away. Lately I've been extraordinarily interested in cold fusion. There's a youtube video showing a convention of a lot of major Universities putting a great amount of effort into cold fusion. The evidence appears overwhelming. Everyone USA citizen knows of the popular American investigative television show 60 Minutes. The recent 60 Minutes section on Cold fusion was amazing. Here's a video clip of this Cold Fusion episode on YouTube. I would highly recommend everyone watch this youtube clip of the 60 Minutes Cold Fusion episode. If I was not flat broke, I'd spend part of my time working on cold fusion research. There is a video show casing a lot of top Universities around world that are now working on Cold Fusion. Cold Fusion is alive & well. The problem cold fusion scientists are having is instability. They believe it is due to the process of making the material. So for instance imagine having a dozen rods made, and a certain percentage of the rods produced far more energy than what was put into the experiment, while the others did not. That is the present problem. I would encourage scientists to work on Cold Fusion research. If you consider yourself to have a sharp mind, or you're a scientists (even a EE) by profession, and you're interested in doing some Cold Fusion research, then please contact me and I'll email you a link to a forum where a lot of good scientists hang out who are working on Cold Fusion. It should not take you long to get the basics down. Sure, it would help if you specialize in Quantum Physics, but it's not required. Just learn the basics, and begin experimenting. Using logic, your experiments could make a major discovery. It should be said that special electrical shielding caution should be take for high performance diodes, such as schottky diodes, which is why I've spent the past 1.5 years taking as much precaution with my SMS7630 diode array. Most LED's and photodiodes are usually low performance, so often you can get away without any metal shielding, but they need to be in a dark room. Testing high performance diodes such as schottky diodes requires some patience. The SMS7630 diode arrays that I built require two to three weeks while contained within proper metal shielding to recover before they began producing their normal DC voltage. I'd imagine the same would apply to high performance LED's and photodiodes. At least out here in Los Angeles, CA, such diodes experience appreciable current while outside of metal shielding containment due to external RF signals, which from my experience has shown to always disturb the diode arrays thereby significantly decreasing the DC voltage they produce. Therefore, when testing high performance diodes such as schottky diodes, it's highly recommended that you allow the diode array to sit inside a *good* metal shielding with the charging capacitors, preferably unloaded, for at least two weeks. Yesterday I realized that my AM-240 voltage meter by Amprobe would not be able to measure the DC voltage on the green LED experiment today because the DC voltage is most likely far above 400mV. My AM-240 while in 400mV setting was measured at 15Gohms, but in the higher voltage settings it uses an entirely different input circuit that is 10 to 20Mohms, which is roughly 1000 times less resistance. So the AM-240 will drain the 1.0uF cap too fast. So I'm uncertain what to do about the green LED experiment for now. Perhaps take the time to get my electrometer circuit going again-- the batteries are drained, and I would have to change the gain resistor since it's geared to measure nano and microvolts. Or if I had ~ 1Gohm and 100Mohm resistor I could build a voltage divider, and place the AM-240 across the 100Mohm resistor while in 400mV mode. Here's more data from another diode, a photodiode, experiment that appears to match the predicted voltage --> Photodiode - Part # is unknown. Appears to be an infrared photodiode: The above data was added to Diode breakthrough.
Predicted DC voltage produced by a diode with proper load: Predicted voltage across a capacitor where capacitor is still charging and capacitor voltage is significantly less than predicted Vout: The following data was taken from yesterdays measurement on the green LED --> Cheap green LED - part # is unknown: The above adds to the list of diode measurements that are close to the predicted values. In fact, the predicted values are far closer than would be expected. The following data was posted in a previous blog post. The above data was added to Diode breakthrough. Given if the diodes are truly converting ambient thermal energy into usable energy, then a conservative power figure per diode is 1nW. Such diode array fabrication does *not* require 100% flawless productions since a broken diode merely cuts out that entire line of diodes, which might be 100 diodes in-series. Semiconductor chip fabrication does *not* create each component at a time. The *entire* semiconductor chip is created at the same time, or more specifically each layer is created simultaneously. A diode array chip may require less than 10 layers. So an entire diode array chip 1 meter by 1 meter would be created at once. Given present technology, each diode could easily be 50nm by 50nm. A one square meter chip with 100nm from diode to diode would contain 10 million by 10 million diodes for a total of 100 trillion diodes. At 1nW per diode, the total power comes to 100000 watts or 100KW! With large production, the retail market price per 100KW chip could be a few hundred US dollars, coming to $0.002 or 0.2 cents per watt. There is far far far lower than solar cells, no comparison. Furthermore, there's no comparison in that solar cells rely upon good weather, no clouds, day time, will not work at night, will not work inside, must be pointing toward sun. Whereas diode array chips would work at any hospitable location on Earth, even inside a cave at night. | ||||
Disclaimer: The owner of this website assumes no liability for any incidental, consequential or other liability from the use of this information. All risks and damages, incidental or otherwise, arising from the use or misuse of the information contained herein are entirely the responsibility of the user. Copyright © 2010 Global Free Energy Blog - All Rights Reserved
Powered by WordPress & the Atahualpa Theme by BytesForAll. Discuss on our WP Forum | ||||||
