Editing Cathode-ray tube (section)

====Convergence and purity in color CRTs====
Due to limitations in the dimensional precision with which CRTs can be manufactured economically, it has not been practically possible to build color CRTs in which three electron beams could be aligned to hit phosphors of respective color in acceptable coordination, solely on the basis of the geometric configuration of the electron gun axes and gun aperture positions, shadow mask apertures, etc. The shadow mask ensures that one beam will only hit spots of certain colors of phosphors, but minute variations in physical alignment of the internal parts among individual CRTs will cause variations in the exact alignment of the beams through the shadow mask, allowing some electrons from, for example, the red beam to hit, say, blue phosphors, unless some individual compensation is made for the variance among individual tubes.

Color convergence and color purity are two aspects of this single problem. Firstly, for correct color rendering it is necessary that regardless of where the beams are deflected on the screen, all three hit the same spot (and nominally pass through the same hole or slot) on the shadow mask.{{clarify|reason=See Talk page for apparent contradiction|date=August 2019}} This is called convergence.<ref>{{cite web |url=http://www.ultimateavmag.com/howto/305picture/|archive-url=https://web.archive.org/web/20091126054428/http://www.ultimateavmag.com/howto/305picture/|archive-date=26 November 2009|title=Picture This|date=March 2005|author=Norton, Thomas J. |work= UltimateAVmag.com}}</ref> More specifically, the convergence at the center of the screen (with no deflection field applied by the yoke) is called static convergence, and the convergence over the rest of the screen area (specially at the edges and corners) is called dynamic convergence.<ref name="auto48"/> The beams may converge at the center of the screen and yet stray from each other as they are deflected toward the edges; such a CRT would be said to have good static convergence but poor dynamic convergence. Secondly, each beam must only strike the phosphors of the color it is intended to strike and no others. This is called purity. Like convergence, there is static purity and dynamic purity, with the same meanings of "static" and "dynamic" as for convergence. Convergence and purity are distinct parameters; a CRT could have good purity but poor convergence, or vice versa. Poor convergence causes color "shadows" or "ghosts" along displayed edges and contours, as if the image on the screen were [[Intaglio (printmaking)|intaglio printed]] with poor registration. Poor purity causes objects on the screen to appear off-color while their edges remain sharp. Purity and convergence problems can occur at the same time, in the same or different areas of the screen or both over the whole screen, and either uniformly or to greater or lesser degrees over different parts of the screen.

[[File:Magnet on TV.webm|thumb|A magnet used on a CRT TV. Note the distortion of the image.]]
The solution to the static convergence and purity problems is a set of color alignment ring magnets installed around the neck of the CRT.<ref>{{Cite web|url=https://www.repairfaq.org/samnew/tvfaq/tvcrtpura.htm|title=SER FAQ: TVFAQ: CRT purity adjustment|website=www.repairfaq.org}}</ref> These movable weak permanent magnets are usually mounted on the back end of the deflection yoke assembly and are set at the factory to compensate for any static purity and convergence errors that are intrinsic to the unadjusted tube. Typically there are two or three pairs of two magnets in the form of rings made of plastic impregnated with a magnetic material, with their [[magnetic fields]] parallel to the planes of the magnets, which are perpendicular to the electron gun axes. Often, one pair of rings has 2 poles, another has 4, and the remaining ring has 6 poles.<ref>{{cite web |url= http://www.arcaderepairtips.com/wp-content/uploads/2016/01/guide_setup_adjust_arcade_monitors_v1.2.0.pdf |archive-url=https://web.archive.org/web/20170612054932/http://www.arcaderepairtips.com/wp-content/uploads/2016/01/guide_setup_adjust_arcade_monitors_v1.2.0.pdf |archive-date=2017-06-12 |url-status=live|title=A Guide to Setup and Adjust the CRT of an Arcade Color Monitor |last= Karlsson|first=Ingvar |website= www.arcaderepairtips.com|access-date=2020-12-11}}</ref> Each pair of magnetic rings forms a single effective magnet whose field [[vector field|vector]] can be fully and freely adjusted (in both direction and magnitude). By rotating a pair of magnets relative to each other, their relative field alignment can be varied, adjusting the effective field strength of the pair. (As they rotate relative to each other, each magnet's field can be considered to have two opposing components at right angles, and these four components [two each for two magnets] form two pairs, one pair reinforcing each other and the other pair opposing and canceling each other. Rotating away from alignment, the magnets' mutually reinforcing field components decrease as they are traded for increasing opposed, mutually cancelling components.) By rotating a pair of magnets together, preserving the relative angle between them, the direction of their collective magnetic field can be varied. Overall, adjusting all of the convergence/purity magnets allows a finely tuned slight electron beam deflection or lateral offset to be applied, which compensates for minor static convergence and purity errors intrinsic to the uncalibrated tube. Once set, these magnets are usually glued in place, but normally they can be freed and readjusted in the field (e.g. by a TV repair shop) if necessary.

On some CRTs, additional fixed adjustable magnets are added for dynamic convergence or dynamic purity at specific points on the screen, typically near the corners or edges. Further adjustment of dynamic convergence and purity typically cannot be done passively, but requires active compensation circuits, one to correct convergence horizontally and another to correct it vertically. In this case the deflection yoke contains convergence coils, a set of two per color, wound on the same core, to which the convergence signals are applied. That means 6 convergence coils in groups of 3, with 2 coils per group, with one coil for horizontal convergence correction and another for vertical convergence correction, with each group sharing a core. The groups are separated 120° from one another. Dynamic convergence is necessary because the front of the CRT and the shadow mask are not spherical, compensating for electron beam defocusing and astigmatism. The fact that the CRT screen is not spherical<ref>{{cite web |url=https://wiki.arcadeotaku.com/images/d/dc/Nanaoms2932.pdf |archive-url=https://web.archive.org/web/20170605081955/http://wiki.arcadeotaku.com/images/d/dc/Nanaoms2932.pdf |archive-date=2017-06-05 |url-status=live |title= General notice|website=wiki.arcadeotaku.com |access-date=2020-12-11}}</ref> leads to geometry problems which may be corrected using a circuit.<ref>{{Cite web|url=https://patents.google.com/patent/US3825796A/en|title=Crt geometry correction network|accessdate=18 December 2022}}</ref> The signals used for convergence are parabolic waveforms derived from three signals coming from a vertical output circuit. The parabolic signal is fed into the convergence coils, while the other two are sawtooth signals that, when mixed with the parabolic signals, create the necessary signal for convergence. A resistor and diode are used to lock the convergence signal to the center of the screen to prevent it from being affected by the static convergence. The horizontal and vertical convergence circuits are similar. Each circuit has two resonators, one usually tuned to 15,625&nbsp;Hz and the other to 31,250&nbsp;Hz, which set the frequency of the signal sent to the convergence coils.<ref>{{Cite book|url=https://books.google.com/books?id=53nnX4fnnNIC&q=crt+convergence&pg=PA110|title=Colour Television: Theory and Practice|date=March 1994|publisher=McGraw-Hill Education |isbn=9780074600245}}</ref> Dynamic convergence may be accomplished using electrostatic quadrupole fields in the electron gun.<ref>{{Cite web|url=https://patents.google.com/patent/EP0739028A2/en|title=Method and apparatus for controlling dynamic convergence of a plurality of electron beams of a color cathode ray tube|accessdate=18 December 2022}}</ref> Dynamic convergence means that the electron beam does not travel in a perfectly straight line between the deflection coils and the screen, since the convergence coils cause it to become curved to conform to the screen.

The convergence signal may instead be a sawtooth signal with a slight sine wave appearance, the sine wave part is created using a capacitor in series with each deflection coil. In this case, the convergence signal is used to drive the deflection coils. The sine wave part of the signal causes the electron beam to move more slowly near the edges of the screen. The capacitors used to create the convergence signal are known as the s-capacitors. This type of convergence is necessary due to the high deflection angles and flat screens of many CRT computer monitors. The value of the s-capacitors must be chosen based on the scan rate of the CRT, so multi-syncing monitors must have different sets of s-capacitors, one for each refresh rate.<ref name="auto4"/>

Dynamic convergence may instead be accomplished in some CRTs using only the ring magnets, magnets glued to the CRT, and by varying the position of the deflection yoke, whose position may be maintained using set screws, a clamp and rubber wedges.<ref name="auto48"/><ref>{{Cite web|url=https://patents.google.com/patent/CA1267682A/en|title=Deflection yoke for adhesive assembly and mounting|accessdate=18 December 2022}}</ref> 90° deflection angle CRTs may use "self-convergence" without dynamic convergence, which together with the in-line triad arrangement, eliminates the need for separate convergence coils and related circuitry, reducing costs. complexity and CRT depth by 10 millimeters. Self-convergence works by means of "nonuniform" magnetic fields. Dynamic convergence is necessary in 110° deflection angle CRTs, and quadrupole windings on the deflection yoke at a certain frequency may also be used for dynamic convergence.<ref>{{cite web |url=http://sbe.org/handbook/fundamentals/Video/Video-Electron_Optics.pdf |title= Video electron optics|website=sbe.org/handbook |access-date=2020-12-11}}</ref>

Dynamic color convergence and purity are one of the main reasons why until late in their history, CRTs were long-necked (deep) and had biaxially curved faces; these geometric design characteristics are necessary for intrinsic passive dynamic color convergence and purity. Only starting around the 1990s did sophisticated active dynamic convergence compensation circuits become available that made short-necked and flat-faced CRTs workable. These active compensation circuits use the deflection yoke to finely adjust beam deflection according to the beam target location. The same techniques (and major circuit components) also make possible the adjustment of display image rotation, skew, and other complex [[raster graphics|raster]] geometry parameters through electronics under user control.<ref name="auto4"/>

Alternatively, the guns can be aligned with one another (converged) using convergence rings placed right outside the neck; with one ring per gun. The rings can have north and south poles. There can be 4 sets of rings, one to adjust RGB convergence, a second to adjust Red and Blue convergence, a third to adjust vertical raster shift, and a fourth to adjust purity. The vertical raster shift adjusts the straightness of the scan line. CRTs may also employ dynamic convergence circuits, which ensure correct convergence at the edges of the CRT. Permalloy magnets may also be used to correct the convergence at the edges. Convergence is carried out with the help of a crosshatch (grid) pattern.<ref name=":98">{{cite web |url=http://educypedia.karadimov.info/library/DX2P.pdf |archive-url=https://web.archive.org/web/20160428122828/http://educypedia.karadimov.info/library/DX2P.pdf |archive-date=2016-04-28 |url-status=live |title=1DX2P Geometry, Convergence, and Purity Adjustments |website=educypedia.karadimov.info |access-date=2020-12-11}}</ref><ref name=":99">{{Cite web|url=https://www.repairfaq.org/samnew/tvfaq/tvcrtcona.htm|title=SER FAQ: TVFAQ: CRT convergence adjustment |website=www.repairfaq.org}}</ref> Other CRTs may instead use magnets that are pushed in and out instead of rings.<ref name="auto96"/> In early color CRTs, the holes in the shadow mask became progressively smaller as they extended outwards from the center of the screen, to aid in convergence.<ref name="auto55"/>