User talk:DieSwartzPunkt

From Wikipedia, the free encyclopedia
Jump to: navigation, search

North America blackouts[edit]

See my explanation at Northeast America blackout of 1965. Simply south ...... time, department skies for just 8 years 21:20, 26 June 2014 (UTC)

@Simply south:Your explanation is invalid. The article is not about blackouts throughout the world but about a specific blackout that occured in Northeast America. Also as stated there is no such recognised compass point as 'Northeast North'. DieSwartzPunkt (talk) 10:20, 27 June 2014 (UTC)
I never said it was, I said it was located in North America, as in the continent. Simply south ...... time, department skies for just 8 years 10:55, 27 June 2014 (UTC)

Was it your intention to simply undo the move? Or was it your intention to move the articles to a completely new title as you in fact did. If so then you were as guilty of moving without consensus as User:Simply south was. Any way, the move has now been undone and both articles move protected. I B Wright (talk) 16:36, 27 June 2014 (UTC)

Oooops! My intention was to move the article back to its starting place. Sorry anout that. DieSwartzPunkt (talk) 10:54, 28 June 2014 (UTC)

I have set up an RM instead. Simply south ...... time, department skies for just 8 years 19:53, 28 June 2014 (UTC)

In RM it has every relevance as it show the links of what is and what could be. Simply south ...... time, department skies for just 8 years 20:12, 7 July 2014 (UTC)

Memory effect on lithium-ion cells[edit]

Why did you remove the content mentioning the memory effect on lithium-ion cells? Although it is true that Li-ion cells don't have a pronounced memory effect like NiCd batteries do, some researchers have identified a slight memory effect in some cells. The text that you deleted even included a reference, and not just any reference, but a reference to a published scientific article: Sasaki, Tsuyoshi; Ukyo, Yoshio; Novák, Petr (14 April 2013). "Memory effect in a lithium-ion battery". Nature Materials. doi:10.1038/NMAT3623.  ---- (talk) 04:26, 11 July 2014 (UTC)

@ The memory effect in the majority of nickel-cadmium batteries is a myth. This is discussed at length on the talk page. The memory effect exists solely in a certain construction of sintered cells made specifically for aerospace use. Even then it only manifests itself if the batery is discharged to exactly the same level before being recharged. This occured when the batteries were used in satellites when the discharge time was the fixed time that the satellite was on the dark side of the earth. The Underwriter's Laboratory (and several other organisations) have attempted to reproduce the memory effect in general purpose cells and have been unable to do so. The effect was exploited by marketing men trying to sell the then new nickel metal-hydride batteries which (at the time) offered no appreciable advantage at typically five times the price. The myth has persisted ever since even though no one had heard of it or experienced it before Ni-MH batteries appeared.
References used on Wikipedia must meet the requirements of being a reliable source. The reference provided falls far short of that requirement. First, the reference as accesible is only an abstract. The full paper is only accesible by shelling out large quantities of cash (any reference is required to be available without payment (per WP:ELNO)). There are exceptions but this isn't one of them. The paper is clearly self published which is wholly unacceptable (per WP:SELFPUB). A reference of this type must have the backing of a recognised third party authority on the subject being discussed or, at the very least, be published in a recognised peer reviewed scientific journal. is just a web based magazine that acepts articles from anyone (including you or I if either of us cared to write one) - effectively a blog for reference purposes. The paper is written by employees of Toyota, who although are battery users are not authorities in the technology itself (WP:NOTPROMOTION). I have attempted to find references to memory effect in lithium ion batteries from other sources but I have drawn a blank. The battery manufacturers explicitly state that there is no such effect (see last para.). Without any recognised support, we cannot tell if Messrs Sasaki, Ukyo and Novák are publishing in good faith or have an agenda. Wikipedia's assume good faith applies to its editors and not to references. One relevent question is: why did they publish this where they did and apparently nowhere else?
On the other hand the statement, "Lithium ion batteries do not suffer from memory effect" is easily cited to Wikipedia standards because a recognised authority and one of the larger battery manufacturers, Panasonic, says so [1]. DieSwartzPunkt (talk) 12:26, 11 July 2014 (UTC)
Please bear with me because I am absolutely speechless on the way you have handled this. I will go through some points.
  • First and foremost, the reference cited is reliable.
  • The style guide you mentioned WP:ELNO is for "external links", not for citations. In the lede it reads, "This guideline does not apply to citations to sources supporting article content."
  • The correct guideline is WP:RS. In this guideline there is nothing saying that the source should be gratis (free of charge). "When available, academic and peer-reviewed publications, scholarly monographs, and textbooks are usually the most reliable sources." Most of these resources have a price tag. Of course, not everybody is able to get a subscription to Nature and other leading scientific journals, but that does not mean the resource is not legitimate. There are many people employed worldwide in universities, research centres, laboratories and private companies with access to such literature, and any of them can verify this citation. I can verify it. If you are not able to, perhaps you should not be editing this article, or at least not this part.
  • "The paper is clearly self-published..." I'm sorry but, what? Where did you get this notion from? Do the authors own the Nature Materials journal?
  • The paper is published in the peer reviewed journal Nature Materials ISSN 1476-1122, not in a "" website. Of course, like many other scientific publications, it makes its content available online, but that does not mean the website is some sort of blog where anybody can publish anything. Yes, both you and I, and any person can publish there, that's a good thing, not a bad thing. What do we need to publish? Solid, reproducible, and well-developed scientific results that can stand the questions of the scientific community, just as these researchers have done. I'm afraid that you only saw the .com address and were quick to draw up conclusions on that. The Nature Materials journal has an impact factor of 35, that's rather high! Its full-time editorial board is composed of scientists. There can be almost no better source to make a reference from.
  • The authors are affiliated to Toyota, this is correct. You say "... who although are battery users are not authorities in the technology itself". Well, my question to you is, how do you even know? You may not be aware of this, but car manufacturers actually invest hundreds of millions of dollars, pounds, and euros in expanding their research and development facilities to investigate new energy storage methods. Even if their main business as seen from the general public is just cars, I can assure you that behind doors they have thousands of full-time scientists doing research on new technologies you are not even aware of. For them it is an investment to do research in house in order to develop new products and subsequently commercialise them. Believe me when I tell you that many of these car manufacturers have transcended that moniker, they don't just sell cars, they are technology-innovation powerhouses, looking into many things like electrochemistry, batteries, fuel cells, carbon composites, multiphysics simulation software, heat exchangers, refrigeration systems, and others. The authors are specifically with "Toyota Central R&D Labs., Battery Laboratory, 41-1, Yokomichi, Nagakute, Aichi 480-1192, Japan" and with "Electrochemistry Laboratory, Paul Scherrer Institut, CH-5232 Villigen PSI, Switzerland". You would do well in at least giving them some credit.
  • Did you even read the paper? Do you agree with the results? Do you think they did the measurements wrong? Do you think they omitted discussions on entropy, on Peukert's law, on the activation energy? How can you question its scientific exposition if you have not even read the paper? You are just making blind assumptions. Do you have a question for the authors? By all means contact them. Most scientists would welcome criticism to their papers, and that's precisely why they publish, to expose themselves to the broader community and defend their own ideas, methods and theories.
  • One of the authors is none other than internationally recognised electrochemistry expert Prof. Petr Novák, by the way. He is with the Paul Scherrer Institute, and a professor with ETH Zürich, the leading technical university in Switzerland.
  • "Why did they publish this where they did and apparently nowhere else?" In scientific publishing, the general rule is you only publish original content, meaning that a paper that is published in one journal cannot be published exactly the same in another journal (it would not be original anymore). They could have published their results in the Journal of the Electrochemical Society, the Journal of Power Sources, Electrochimica Acta, or others. Why did they decide on this? I am not sure. Perhaps it just so happened that Nature Materials is a very high-quality journal, with a very high visibility that publishing there was better. Again, they published in a scientific journal, not in a blog.
  • "I have attempted to find references to memory effect in lithium ion batteries from other sources but I have drawn a blank." In this I must agree with you. The "memory effect" described by Novák et al. was certainly surprising to me when it was published a couple of years ago. The repercussions of this have not been felt so far. There hasn't been follow up research on this, thus why I liked the wording "no significant memory effect". It gives the idea that there is none, but if there is, it is only marginaly detected, as the paper authors suggest. This is a significant point because people in the Li-ion battery community are not aware of this and don't seem to give it importance. In my opinion, it does not hurt to mention it.
  • Using the Panasonic reference to state that there is no memory effect is acceptable. However, as per WP:RS#Some types of sources, "However, some scholarly material may be outdated, in competition with alternative theories, or controversial within the relevant field." It is also a primary source, not a peer-reviewed publication. I'm not saying that they did, but Panasonic could have basically written whatever they wanted in that document without having to face scrutiny from other people. So, please don't be closed to the idea that some memory effect does exist, especially when the claim is made by recognised scientists that have much more experience in electrochemistry than you do.
I've seen your edits and I think you are WP:BOLD, however, I also feel that sometimes you act in a reckless way. I do hope you make edits in good faith and not just trying to impose your view and tone on articles. If there is something that I don't like in Wikipedia is spending some time editing and collecting references, and then just see my work undone in the blink of an eye by a less-than-expert editor that is only concerned with getting his version of the story.---- (talk) 00:08, 15 July 2014 (UTC)
@ It is to your credit that at least you have not attempted to edit war on the subject and have followed policy on this. I can think of more than a few editors who would have edit warred and refused to discuss. I still believe that the reference falls below the bar of reliability. I am not interested in my version, your version or any other editor's version. What I am interested in is getting the correct and suportable version, and on that we should be on the same page. However, I suggest you follow WP:BRD and propose your change on the article talk page with your reference and let's see what others think of it. I will not contribute to the discussion and we can both then abide by the consensus of others. DieSwartzPunkt (talk) 12:17, 15 July 2014 (UTC)
I think opening the topic for discussion in the talk page is okay, however, for this small issue, just making a single change to the article, I think it's going overboard. I would be satisfied if we can just edit that sentence at the beginning with a suitable wording like "... no significant memory effect<ref>Although it has been generally accepted that there is no memory effect in Li-ion cells, in 2012 researchers from Toyota and the Paul Scherrer Institute described a kind of memory effect in cells with the LiFePO4 chemistry... {{cite journal| ...}}</ref>.---- (talk) 23:18, 15 July 2014 (UTC)
I see you have put it on the talk page, so let's see what happens. DieSwartzPunkt (talk) 15:49, 16 July 2014 (UTC)

CRT horizontal resolution constraints[edit]

So we agree that the triad density constrains the horizontal resolution, but Jeh's point about the H and V size/position controls still deserves a response. On my own CRTs, which are high end GDM trinitrons, when you make these adjustments using the on screen display, it is only in discrete steps. But I wonder if older CRTs had truly analogue controls for H and V adjustments. If so, then this presents something of a paradox, agreed? Spacediver (talk) 17:44, 13 July 2014 (UTC)

@Spacediver: As far as the CRT itself is concerned, apart from the upper limit set by the triad density, there is no constraint whatsoever over the resolution that can be displayed both horizontally or vertically. Indeed, monochrome tubes without these triads have no set upper limit. The maximum resolution that may be achieved in practice is limited by the bandwidth of the video amplifying chain coupled with any parasitic capacitance between the tube gontrol grids or cathodes and the other electrodes. You are perhaps getting confused because you are describing a computer monitor which is resolving an inherently digital source signal onto an inherently analogue display. Such monitors are only capable of resolving the fixed signal resolutions and timings that are set up in their firmware. But this was not always true of computer displays because in the olden days, computers output their displays as an analogue signal, often via a modulator that allowed the signal to appear on a standard television set. Such a signal starts life in the computer in digital form, usually set by the size of the video memory. Very early systems were of fixed resolution, but later systems started to offer several video modes that may or may not have had differing resolutions.
While the connection from the computer to the monitor was analogue, the monitor merely had to be capable of synchronising to whatever television standard the computer output (usually 312 line 50 frame progressive or 262 line 60 frame progressive - these worked with standard 625 line and 525 line TV systems respectively but without the interlacing). The actual horizontal resolution of the video data in the computer was immaterial. As monitors improved, they acquired digital circuitry and became able to convert (what was by now) an RGB based analogue signal into its digital counterpart keeping the 'pixels' that were descernable in the analogue signal in 1:1 correspondence with the real pixels in the digital signal. The monitor's firmware limited the resolutions and frame rates that were displayable as noted above. DieSwartzPunkt (talk) 12:57, 14 July 2014 (UTC)
Thanks for the reply. When I mentioned older CRTs that had truly analogue controls for H and V adjustment, I was referring to color CRTs. Whether we're talking about a slot mask or an aperture grille, I don't see how one can "slide" the entire raster in any direction with arbitrary precision, and still maintain a resolved image. The only way I can see H and V adjustments working, in a color CRT, is if the controls operate on discrete multiples of the triad width. I should say that I have absolutely no idea if such units exist (i.e. color CRTs with arbitrary precision of H and V adjustment), but if they do, then an explanation of how this is achieved is required, unless I'm missing something fundamental here.Spacediver (talk) 16:55, 14 July 2014 (UTC)
{replyto|Spacediver}} Nope. The image can be displayed absolutley anywhere along the horizontal or vertical axes. It is because the colour triads are small compared to the effective resolution that this is so. Some people get confused over resolution in the digital domain and the analogue domain. Although they are not the same they can be (very) broadly equated. Resolution in the digital domain is in discrete pixels. The highest resolution is obtained when each alternating pixel is the same colour (let's assume black and white here). So across the screen, you hava a black pixel followed by a white; then black and then white and so on. In the analogue domain you display a sine wave of the highest frequency able to be displayed. This will display as a black/white pattern similar to the digital domain except that there will be grey in between the white and black. In the digital domain, we express the resolution as so many pixels per screen width (or height). In the analogue domain, we express it as so many line pairs per screen width (or height) but it can also be expressed in Megahertz. Broadly similar resolutions are obtained when the analogue line pairs is half the number of pixels (think pixel pairs here).
Although we just attempted to equate one to the other we cannot really do that. If we shift the analogue sine wave by a quarter of a cycle, the black/white pattern simply moves by half a line across the screen but remains visible exactly the way as before. This is because the display is not comprised of discrete elements (the colour triads do not count here because they are much smaller than the resolved line pairs). If we attempt to move the digital pixel pattern by the same amount (half a pixel in this case), the pattern completely dissappears because each pixel can be considered to be attempting to resolve half a pixel of white and half a pixel of black - which it cannot do as a pixel is a discrete element. DieSwartzPunkt (talk) 17:16, 14 July 2014 (UTC)
Ok here is what my mental model of the situation is. In this image, the maximum horizontal resolution is being utilized. The dots represent the peaks of the sine waves. Now if we shift the peaks of these dots a bit to the right, how on earth will they land on the correct parts of the triad? Wouldn't the beam be landing on the wire of the grille itself, during the peaks of the waveform? See this image (middle). I do, however, appreciate that at lower horizontal resolutions, the wavelengths encompass several triads, and sliding the raster is not an issue here. I'm also a bit confused about your discussion of digital vs. analogue domain. I appreciate that modern video cards have a DAC that converts a digital representation into an analogue one, and that in the past this was unnecessary, but what do you mean when you say "attempt to move the digital pixel pattern...". Are you still talking about a CRT here? Although I do see your point about how a digital representation is all or none, and there is no sinusoidal falloff from peak to trough, and sliding wouldn't be accommodated if the signal was a "discrete square wave" Spacediver (talk) 23:36, 14 July 2014 (UTC)

────────────────────────────────────────────────────────────────────────────────────────────────────Maybe the illustration that you are looking at is the problem. The illustration appears to show the electron beam from each gun landing on only one colour phosphor dot or line. This does not happen in practice. The size of the spot to which you can focus an electron beam has a certain minimum size. This is a function of the density and speed of the electrons in the beam and also the tube size but comes about because the electrons are mutually repelling each other. One way of reducing the size of the spot is to increase the accelerating voltage. This means that the electrons have less time to repel each other but it does require a larger deflection current (because the beam spends less time under the influence of the deflecting field). The colour triad size is always arranged to be much smaller than this minimum spot size. This means that the electron beam will never illuminate just one colour dot or stripe as the illustration shows but a number of them (usually at least 3 or 4 - more for higher end displays). Shifting the sine wave pattern described above will just move the peaks and troughs to a new part of the triad pattern and the ones in between will now display the shades of grey. Also the minimum spot size should be smaller than best resolution that the tube is capable of (but this was not always the case in lower end television tubes).

Thanks, this makes a lot of sense. Just to clarify, the aperture grille would still prevent the beam from striking the inappropriate parts of the triad, even though the point spread function and spot size is such that multiple triads will be bombarded for each "pixel signal", right? I find it fascinating that the geometry of of the guns and the wires are such that they are so good at masking the beam appropriately. It seems like an impossible feat of geometry! I've recently ordered a high quality loupe and am looking forward to examining the screen more closely (and potentially sharing any images I'm able to acquire). I really do appreciate your patience here, and your generosity with information. At the risk of pushing my luck, is there any way I can contact you directly for a couple more CRT related questions? I'm in the process of gathering more information so I can ask the questions more intelligently, but one of them deals with thermionic emission and electron gun life span, and the other deals with the relationship between cutoff voltages and the burden on the amplifier circuitry. I can always post the questions on this page, or link to a forum post where I post the questions. Alternatively email could work. Mine is (talk) 17:50, 15 July 2014 (UTC)
@Spacediver: The geometry of making the aperture grill and the dot pattern match doesn't come into it. It's very simple how it is achieved. The aperture grill (shadow mask or whatever) is paired with the screen before any phosphor is applied. They spend the rest of their lives together. The screen is coated with one phosphor mixture, and is then illuminated with ultra violet light from a spot source located at the ADC (apparent deflection centre) for the relevant gun. This 'fixes' the phosphor so illuminated and the rest is washed away. Repeat and rinse for the other two colours and you have a screen with phosphors accurately matched for the apperture grill. It follows that any other aperture grill will not work with the screen so produced.
Ask your questions here and I will answer them if I can. Don't worry about filling up the talk page because I regularly purge it when discussions become stale. DieSwartzPunkt (talk) 15:56, 16 July 2014 (UTC)
Thanks, I appreciate it. I'm still having trouble understanding how this is possible. I did see a video a while back describing the fixing of the phosphors, and it made sense qualitatively, but the geometry of the situation is confusing to me. Here are two possible scenarios that I'm imagining, both constructed from a bird's eye view of the situation. In the top image, the beam is shown in two raster positions. In each position, the beam magically splits up when it hits the grille such that only the relevant phosphors are hit. Clearly this is not realistic, which led me to the bottom part. But in the bottom part, I don't see how the grille can selectively allow the beam from one gun to hit only one of the stripes, while allowing the other guns to selectively hit their respective stripes. In this situation, a substantial portion of the entire phosphor coating will never be struck by any of the three beams. UPDATE: This image is another attempt, and I imagine this is closer to the truth?Spacediver (talk) 03:02, 17 July 2014 (UTC)
Your second image is closer to reality. If you were to make the beam wider than the size of the hole in the mask, it would be a reasonable accurate portrayal. In reality, the beam is a truncated cone with the apex (ideally) just past the phosphor covered screen (in much the same way as focusing light to a point would be a similar cone). The focused 'point' of electrons is a finite size because it represents a projection of the effective diameter of the electron gun just as focusing a light source that has finite diameter will always be to a circle of light and not an infintesimal sized point. The mask is exactly that: it prevents the electrons from striking the wrong phosphor because they are in the shadow of the mask (hence 'shadow mask' for the delta gun configuration). The other two guns are positioned such that they strike the correct phosphor and the 'wrong' phosphors are in the shadow.
I have tried to illustrate it here. In this case the electron guns have been defocused; the slots and phosphor stripes are larger than reality, the geometry doesn't quite fit and I'm not going to win any prizes for artistic merit but it should illustrate the point. One of the big failings of colour tubes is that less than 33% of the electron beam from each gun lands on its intended phosphor. The delta gun tube is the worst in this respect as the shadow mask has considerably more mask than hole area. This meant that these tubes had to have high beam currents to provide any reasonable brightness. It was this high current that contributed to their relatively short lives. These tubes were also the most complicated to set up for colour purity and convergence (at which I was very proficient having set up hundreds of them in varying aircraft and ground installations). Their one big advantage was that they could have finest pattern of colour phosphors on their screen. The trinitron tube had the highest ratio of hole to mask of any tube which meant that it gave good bright and well focused image (the lack of top and bottom borders to the apertures contributed in no small part). The trinitron came fairly close to the delta gun for fineness of the phosphors on the screen, but the fact that the aperture grill consisted of just vertical wires made it very suseptible to vibration. This limited the fineness of the grill which often had one or more horizontal support wires which annoyed some users. The slot aperture tube lay between the two, but offered the big advantage that (in its final incarnation) that the magnetic fields required for convergence was substantially the same between different tubes of the same type. These magnetic fields could be 'printed' into the deflection yoke assembly, the only field adjustments required were for colour purity. This appealed to TV manufacturers who were always wanted to simplify the manufacturing process. This type of tube was rarely used for high definition (i.e. computer type displays) as the compromises in design meant that high resolution displays could not be adequately converged - there being no dynamic adjustment. DieSwartzPunkt (talk) 12:51, 17 July 2014 (UTC)
Thanks for the illustration! Let me test my understanding: So the focus grid will refocus the beam (which starts by fanning out), so that it's fanning inwards by the time it reaches the grille mask. Now if you were to get rid of the horizontal and vertical deflection yokes and/or their circuits, and only fire the "red" gun, you might have a single red spot in the center of the screen. This spot would be the result of a focused beam that goes through multiple slots in the grille (perhaps 3 or 4, depending). This spot would have a luminance profile along two dimensions - the point spread function. It wouldn't be a continuous profile along the horizontal dimension, as the electron beam has been split up at the grille so only the red stripes are struck by electrons. So two questions follow: 1:In normal operation (with active deflection yokes etc.) does the spot size (and its profile) remain constant across all resolutions? 2:Does the horizontal luminance profile of a vertical line depend on the luminance profile of the spot itself? Or does it depend on the temporal profile of the signal (i.e. the sinusoidal modulation of voltage)? Or does it depend on both? I just browsed through a chapter in Winn L. Rosch's Hardware Bible, and found some good information there. I'm gonna try hunt down some books/papers that go into more detail, but not sure if this is possible due to the proprietary nature of some of this tech. Found some interesting descriptions in the patent realm, but nothing comprehensive. If you have any recommendations, I'd love to know :) Spacediver (talk) 00:05, 18 July 2014 (UTC)

edit point[edit]

────────────────────────────────────────────────────────────────────────────────────────────────────Your first half has it absolutely spot on (the point about the vertical spread function is valid for a trinitron tube but not the other two type as the mask is segmented vertically as well as horizontally.

Q1. The spot size remains fixed for all horizontal and vertical resolutions. This ignores that focus is difficult to maintain when the spot is away from the centre of the screen as the distance travelled is greater and some defocusing is also inevitable as the spot becomes an ellipse. Just remember the CRT is a strictly analoge device (and apart from the arbitrary triad size), the resolution is constrained at an upper limit by spot size. Probably the best way to visualise this is to imagine a screen with an infinite number of colour triads and a mask with a correspondingly infinite number of slots or holes. There is no lower limit or any constraints to discrete resolutions, apart from the physical size of the screen.

Q2. The vertical resolution characteristics are identical to the horizontal (ignoring asymetrical spot geometry). The vertical resolution for a television or computer display is constrained only by the number of scanning lines. This is an entirely arbitrary choice, and for any given CRT, the actual number has no constraint other than the maximum resolution determined by the tube characteristics. It is theoretically possible, if pointless, to have a television display with only two active lines! This constraint disappears when vector scan is used rather than raster (The arcade game 'Asteroids' was a prime example of a vector scan system). Vector scan systems are not generally popular because the horizontal deflection circuitry, though technologically simpler, is actually more complex and power hungry than for raster scan systems. DieSwartzPunkt (talk) 10:13, 18 July 2014 (UTC)

I'll try to rephrase Q2. Look at this image. Top left shows the luminance profile of the spot. Top right shows the voltage signal for a plain white (or red, green, blue) field with a resolution of 6 x 6 (ignore the lack of square pixels!). The voltage changes sinusoidally across time as it scans left to right with each successive row. Blanking retrace lines are not depicted here. Now there are two functions at play here: One: the luminance profile of the spot itself, and another: the change in voltage (and hence beam intensity) over time, and -- because the beam is sweeping across space -- over space as well. Bottom left shows what the signal for a vertical line is. My question is this: If you were to measure the horizontal luminance profile of this line, it would, in my mind, reflect both the spot profile, and the voltage "profile". Would it be a convolution of these functions? A multiplication? And, following up to this, consider the horizontal line in bottom right. In my mind, the vertical luminance profile of this line should only reflect the spot profile, since the voltage profile is only modulated horizontally across space. Have I got this correct? Spacediver (talk) 19:32, 18 July 2014 (UTC)
I think you may be trying to over complicate matters in your mind. Your spot luminance profile is good, though the profile may, in reality, have a flat top to it depending on the precise geometry. For the 6x6 pattern sine waves, it is easy to ignore the square pixels because, as already stated, there are none to ignore. A CRT used for any form of display, be it an oscilloscope; television display; radar display or whatever, is only ever capable of displaying a single spot at any instant in time. So let's stick wih this for a moment. With the grid at the same potential as the cathode (in practice a little negative) the spot is at its brightest. Your top left picture is more or less, what is seen by the viewer and you have got the general idea as to what the luminance profile looks like from one side of the spot to the other. If the control grid is made more negative relative to the cathode (or usually in practice, the cathode is made more positive relative to the grid), the spot becomes dimmer. The luminance profile is exactly the same but it just has a smaller amplitude. The overall luminance of the spot, in theory, is inversely proportional to the delta voltage applied to the grid or cathode. I say, 'in theory' because it is not in practice. CRTs have a distinct 'triode' characteristic and the voltage to luminance is very non linear. This actually turns out to be an advantage because the eye's perception to brightness is also conveniently non linear (more or less logarithmic).
The luminance profile of a displayed line (produced with a sawtooth deflection signal) would be difficult to quantify because a line is never displayed - at least not all at the same time. The perception to an observer, if the frequency were high enough, would be of a line with a uniform luminance along the length of the line, falling off at the edges and ends with the slopes of the spot profile.
If the beam were to be modulated with a sine wave as you suggest, the displayed spot would be luminance modulated as outlined above. The perception to an observer is a sine wave modulated line with varying grey between the peak white and the peak black (I am obviously assuming a white phosphor here). Beyond that: I am not sure what you are trying to achieve. I do not know of anyone who has attempted to describe what is observed mathematically. I have never seen it described in any text book. I have worked all my life with display systems (and trained others) without ever needing to. DieSwartzPunkt (talk) 11:08, 19 July 2014 (UTC)
I'll try to explain the motivation behind this question. There has always been something that didn't quite click with me when it came to understanding the way "pixels" are generated on a scanning CRT, and then you made this comment that made me realize exactly what didn't click: "In the analogue domain you display a sine wave of the highest frequency able to be displayed. This will display as a black/white pattern similar to the digital domain except that there will be grey in between the white and black.". My reading of this comment is that the gradual, rather than sudden, transition from black to white in a rendered pixel is due to the sinusoidal rise and fall of voltage. But your later comments (and, indeed, those by others, including Charles Poynton) emphasize the spot profile of the beam as being responsible for this fall off.
Let me try to get at the same issue in another way. Suppose you send a signal representing a single white pixel in the center of the screen surrounded by black. And suppose that the shape of the scanning spot is a perfect circle with a gaussian luminance profile along both dimensions. Now, the only way I can see how the rendered pixel will be a perfect circle is if the rise and fall of voltage is an impulse function. If the voltage rises for any given length of time before it falls off, the rendered pixel will, in my mind, appear as a horizontally extended rectangle with rounded edges. The luminance profile of this "pixel" would be influenced by the spot profile, the voltage function, and the temporal rise and decay function of the phosphor (somewhat similar to how the BOLD response of a voxel in fMRI is the convolution of the neural activity in that voxel and the temporal hemodynamic response function). In fact, I just noticed that in the last figure of this document, they do indeed use a 10 mm wide moving filter to obtain the underlying spot profile, presumably because the voltage rises for a "length" of 10 mm.
p.s. I wasn't thinking clearly in my earlier response when I made reference to the lack of square pixels - thanks for catching that. Also, the fact that the triode transfer function of the gun results in a perceptually uniform relationship between voltage and luminance is really damn cool (and under appreciated by many who bemoan the emulation of CRT gamma characteristics in modern video decoding standards). Spacediver (talk) 18:45, 19 July 2014 (UTC)
I am going to have to duck out of this dicussion. It is getting way out of my depth. This may be of interest to a designer and/or manufacturer of tubes, but as a mere user I have not had to get into those sorts of details. I had to study electron dynamics in my apprenticeship but that was as sophisticated as it got (and I have probably forgotten more than I remembered). DieSwartzPunkt (talk) 08:12, 20 July 2014 (UTC)
Completely understand. I'm extremely grateful for your patience and insights - this discussion has been very enlightening for me. I hope I can bug you again in the near future with those other questions (which are actually important beyond academic considerations). Spacediver (talk) 22:31, 20 July 2014 (UTC)
If I can, I will. Best regards. DieSwartzPunkt (talk) 11:17, 21 July 2014 (UTC)


I've replied to your comment posted on my Talk page. Clevelander96 (talk) 13:41, 21 July 2014 (UTC)

CRT triode characteristics[edit]

Ok I think I'm ready for the next set of questions! I'm working on a color calibration guide for the sony trinitron line of CRTs using the service adjustment mode. It involves a piece of software called WinDAS (the DAS stands for digital alignment system), and a special cable that connects a pc/laptop to the service port of the monitor. The WinDAS software instructs you to load up certain test patterns, and then allows you to adjust sliders until various luminance and chromaticity targets (as read with a colorimeter or spectroradiometer) are met . Early on in the procedure, you are asked to load a grayscale pattern, and adjust the G2 voltage until the second bar is barely visible against the pedestal bar. I think I have a pretty good understanding of G2 voltage, and its relationship to the G1 cutoff voltage, thanks to Kenneth Compton's text. From what I understand, the G2 voltage is the first real accelerator in the system: due to its positive charge, the electrons experience a pulling force along the tube in the direction of the viewscreen. The G1 grid, which is negatively charged relative to the cathode, acts as a valve of sorts. Lowering the (+ve) G2 voltage means that the (-ve) G1 voltage doesn't need to be as low in order to prevent electrons from being accelerated through the tube. In other words, lowering G2 means the cutoff voltage at G1 doesn't need to be as low, which means cathode loading is reduced (too low a cutoff voltage means that electrons are pulled from a smaller area of the cathode). However, I'm not as confident in my understanding of some of the following steps in the calibration procedure, and their relationship with each other. Right after the G2 adjustment, you are asked to keep the grayscale pattern on the screen, and adjust "G_Bkg_Brt_Cent" until the second bar is barely visible against the pedestal bar. During this part of the procedure, the screen has a green cast. I think the register stands for green blanking brightness Center, and I think the purpose of this step may be to set the voltage necessary for blanking the green gun when brightness is in the center of its range. Perhaps this is something like a Cutoff voltage for the green gun? Later, you are asked to make adjustments for Cutoff Max (with a black pattern), Cutoff Min (with 30 IRE), Drive (100 IRE), etc., but you're never asked to adjust the B_Bkg_Brt_Cent or R_Bkg_Brt_Cent in the same way as the Green one. You do adjust the B_Bkg_B_Cent and the red one during the Cutoff Min adjustment, but here you are just setting a chromaticity target with a 30 IRE pattern. I'm guessing the reason that only the green version of this register is done using the grayscale pattern has something to do with the fact that changes in "green" wavelengths have the largest effect on luminance.

So the questions:

1) Am I on the right track with these interpetations of the registers? I have a feeling I might be missing something, and I'm not sure how the RGB Bias and Gain parameters fit into my above story.

2) Everything I've read indicates that setting the G2 voltage too high is what is "unhealthy" for the tube. The only side effect of setting it low enough for really deep blacks is that the resulting gamma of the display is very high (image is crushed). However, this is easily remedied with a LUT adjustment after the WinDAS procedure. Are there any other considerations when setting the G2 voltage? Should I be worried about setting it too low?

3) I've heard that you can influence the gamma of the display by choosing different luminance targets at the different stages of the WinDAS procedure (instead of the targets WinDAS instructs). Do you have any guidance about tube health considerations here? Clearly, I want to keep peak luminance below 100 cd/m2 - my peak luminance ends up around 85-90 cd/m2 after the procedure. But what about during the Cutoff Max (0IRE) and Contrast_Min (30 IRE) adjustments. I'm willing to experiment with different luminance targets to see how it affects the gamma, but it would be nice to know that I can adjust these within a moderate range without risking tube health.

I recognize that these questions may be completely idiosyncratic to Sony's way of doing things, and that it's asking a lot, but I don't have anywhere else to turn to for answers at this moment, and a number of people including myself have been trying to figure this out for a while now. If you're unable to engage me here, for whatever reason, absolutely no offense taken - you've already given a great deal of time and energy here. Spacediver (talk) 07:13, 25 July 2014 (UTC)

@Spacediver: I am aware of software driven adjustment systems on modern monitors and TV sets (and also video cameras). However, I have never used them myself. I have been around long enough that every monitor; television or video camera that I was required to set up, I had to do the old fashioned way. This would mean that you needed a video pattern generator and possibly a test card generator. For a camera, an illuminated test pattern was required.
As you surmised, the Grid 2 voltage (often called Anode 1 by European manufacturers) in a CRT is the one that sets up the gun cut off point (usually a positive voltage of 100-400ish volts relative to the cathode). As you note it is frequently adjusted such that the darkest (primary) colour bar is just visible. Repeated for all three guns, once they are the same, you are in a position to get a good white balance with the same hue over the full contrast range. In the good old days, as the three cathodes in the guns wore, this colour balance would change and a colour cast to a picture was not uncommon after two or three years of use. More modern CRT systems are equipped with a circuit that automatically restores the three guns' cut off points to that at which it was set during adjustment. If you look carefully, you can see it working as the tube warms up.
As far as tube longevity is concerned, the luminance from the screen versus the beam current per unit area of the cathode is the factor that limits tube life. The shadow mask tube was the worst culprit here as the beam current was the highest for any of the three systems. The beam current was so high that it was not unknown for the cathode to develop a 'skin' over the oxide coating that had a lower emission than the original oxide. The silvery highlights on the picture was the easily recogniseable symptom of a worn out tube (often not that long outside the five year warranty). The quick fix was to connect all the electrodes of all three guns (except the final anode and cathode) together and apply the 240 volt mains supply via a 15 watt bulb between the cathode and these connected electrodes (having first powered up the heaters of course). One counted slowly to 10 ignoring all the sparking and flashing inside the tube. The result was a tube that once again displayed a good picture without the silvery highlights. The best one could expect was an extra year or so's life out of it though.
The aperture grill in Sony's trinitron tube has a much larger transmission than the shadow mask and consequently the beam currents are much lower for a significantly brighter picture. So much so that I seldom had to scrap a television or monitor with a trinitron tube with completely worn out cathodes. Provided that you do not exceed the published limiting factors for your tube, I would say that you are unlikely to affect its longevity. Good data sheets should give the voltage range over which G2 should be adjusted, but some just give the maximum and others give the average, so interpret with care. DieSwartzPunkt (talk) 12:23, 25 July 2014 (UTC)
Thanks for the response. Your method of restoring the cathode sounds like really old school rejuvenation! One last quick question. I've read that the gun can be driven either by directly changing the cathode voltage, or by driving G1. Do you know which method was used in modern trinitrons? Spacediver (talk) 22:41, 25 July 2014 (UTC)
The only method possible for a trinitron is to drive the cathode. Trinitrons only have a single G1 for all three electron beams. DieSwartzPunkt (talk) 11:17, 26 July 2014 (UTC)
Makes sense, thanks. Spacediver (talk) 21:00, 26 July 2014 (UTC)

Capitalisation of London Underground lines[edit]

Hi, I've just reversed your revision of my edits at King's Cross fire. Transport for London has an editorial style guide which states specifically that when referring to tube lines the "line" should be lowercase. The guide can be found here (under TfL corporate design standards) and the specific page of the guide is here.--DavidCane (talk) 21:30, 28 July 2014 (UTC)