Mod a CRT to increase its TVL

The place for all discussion on gaming hardware
Post Reply
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Mod a CRT to increase its TVL

Post by LukeEvansSimon »

In order to help newcomers to this thread, I have edited the original post to make it easy to jump into the latest status. Expand the spoiler sections below to get the details. Many people tend to have a strong negative reaction to this mod at first, because many people have an incorrect understanding of how CRTs work, and so I'd encourage every one to first read the summarization of their contents in the original post in the spoilers section below, which contains citations for all controversial claims. The cited technical publications describe in detail how the electron gun works and how electron beam spot size is a key limiting factor for CRT resolution and these 20 year old publications also explain how electron beam width (aka "spot size") can be electronically controlled using cathode and anode voltage differentials within the "triode" section of the electron gun. If you aren't going to bother to read these citations, then consider avoiding posting any negative replies, as your negative reaction is likely due to your lack of understanding of CRT technology.

High-level Explanation Of How This Mod Works
Spoiler
Cathode - G1 voltage acts as an electromagnetic iris. Increasing cathode - G1 voltage shrinks the diameter of the aperture of the iris, which more tightly packs the electrons into a thinner, sharper cathode ray, which creates a smaller illuminated spot size on the screen. However, this also changes the black levels, so G2 - G1 voltage needs to be increased to calibrate black levels, and the G2 - G1 voltage needs to be increased by more than the increase in cathode - G1 voltage. In addition, a higher G2 - G1 voltage also serves to increase the acceleration of the electrons in the cathode ray, which further improves spot size because the electrons have less time to spread out. A much more detailed explanation is in the next spoiler section.
Original Post With Detailed Explanation (August 17 2020)
Spoiler
It is well known that many consumer CRT TVs can be modded to make them support RGB input, but apparently nobody in the CRT gaming scene has looked into what it would take to modify a consumer CRT TV to make it high TVL. This post provides an overview of the major factors that influence CRT TVL, and calls out one of the factors that can be modified in any CRT: spot size. The main factors that influence the effective TVL of a CRT display are:

Input signal quality
Composite video bottlenecks the effective TVL to around 450, s-video, component, and RGB allow for 800 to 1000 TVL. This bottleneck is well known in the CRT gaming community and RGB mods exist to remove the bottleneck from consumer CRTs.

Supported input resolutions aka "Addressability"
In this post, I am limiting my discussion to 15khz CRTs, which allow for 240p/480i resolution. Obviously this limits the effective resolution of the display, but for retrogaming, the 240p aesthetic with "thick scanlines" is the goal we are after. CRT resolution technically considers "addressability" to be distinct from the resolution of the CRT itself, because addressability is a limitation of the circuitry that drives the CRT, not the CRT itself. Detailed discussions of CRT resolution versus addressability are in (this reference).

Spot size
"Resolution is primarily a function of CRT spot size" (see reference). When the electron gun beam hits the phosphor of the CRT screen, it creates a spot of light. The spot size refers to the diameter of the spot. A CRT moves this spot very quickly, left to right, across the screen for each illuminated scanline of video. The smaller the spot size, the thinner the illuminated scanlines and the thicker the black, unilluminated scanlines. A larger spot size causes the illuminated scanlines to bleed into each other, causing the unilluminated scanlines to be very thin to the point where two illuminated scanlines can bleed into each other, decreasing vertical resolution. Similarly, a larger spot size decreases horizontal resolution.

For 240p gaming, addressability does not increase or decrease when the TV has a high TVL or low TVL. What makes the games look better is the high resolution of the CRT.
To stress the difference between "resolution" and "addressability" for CRTs, which many people seem to get hung-up on, here is a screenshot of the most important section in this reference:

Image



Mask pitch aka Dot pitch
Dot mask, slot mask, and aperture grille mask the phosphors on a color CRT. A smaller pitch allows for a higher TVL. This is clearly not moddable, as the mask is bonded to the inside of the CRT. However, consumer CRTs 24-inches and larger have a spot size that is at least twice the size of the mask pitch. A deeper discussion of the interplay between mask pitch (sometimes called dot pitch) and spot size is in this reference in section "A Discussion of Issues Relating to Monitor and CRT Resolution". This means that the effective TVL of a CRT can be doubled before the mask pitch becomes the bottleneck. In fact, the aforementioned reference indicates that spot size can be smaller than the dot pitch and yield a pleasing sharp look.

Large Spot Size
Image

Small Spot Size
Image

Cathode Ray Tube Theory
Before I get to how the spot size can be modified in a consumer CRT, let's first overview the theory of how CRT works, as it will set the context for the description of the mod. I am going to skip lots of details and start with what happens in the CRT's neckboard printed circuit board. The pre-amplified RGB video signals run through high voltage RGB amplifiers that dramatically increase the voltage of the RGB video signals so as to drive the electron guns in the CRT. There is a separate cathode K for each color R, G, and B. For simplicity of this explanation, the picture below is a diagram of a single color. Convergence magnet rings around the CRT's electron gun ensure that all 3 colors converge at the same spot.
Image

The amplified voltage waveform that drives the cathode K, determines how much that color component of the electron beam will vary as the beam scans the screen of the CRT. The G2 anode also known as the "screen" or "accelerating anode" is a hollow metal cylinder that the neckboard electrifies with high positive voltage. Electrons are negatively charged, so the positive voltage of the G2 anode attracts the electrons from the cathode K. The G1 anode, also called the "control grid", has a voltage that is always lower than the voltage of the electron emitting cathode K. The G1 serves two purposes: it sets a voltage floor that allows for "cut off" voltage for the cathode, so that no electrons from the cathode are allowed to pass the G1 anode and enter the electron beam that reaches the screen. This is needed for true black levels. The G1 also starts shaping the electron beam because it has a small hole causing electrons that pass through the G1 to "squeeze" through the hole. The G3 for "focus" anode allows for further focusing the electron beam so that it is in focus at all positions on the screen. There is a forth anode, G4, which has an extremely high voltage of around 30,000 volts. This causes the electrons to rapidly accelerate to the from of the CRT. The dangerous suction cup shaped plug that people discharge on a CRT, as a safety measure, is the G4. The G1, G2, G3, and G4 are often called "grids".

More detailed accounts of how a CRT works are at the following links:
Link 1
Link 2

Modifying Spot Size
There is an excellent 3-part slide deck on SlideShare that provides all of the details. Part 2 here has the most relevant parts. The most important slides are inlined in this post.

Image
Image
Image

The electrons from the cathode K are negative charged and repelled by the negative G1 voltage. However, the G1 anode is a cylinder with a hole at the end that the electrons can pass through. As the G1 voltage is made more negative, the electromagnetic aperture that electrons can pass through is made smaller, causing the electrons that pass through to be more focused into a smaller spot size. The same technique is used by cameras with a physical aperture.
Image

Consumer CRTs have the G1 voltage tied to ground, that is, a voltage of zero. Why do they do this? It is inexpensive. The electron gun only needs the G1 to be a voltage that is lower than the voltage of the cathode. Sure it limits TVL, but since many consumers used their CRT TVs with RF or composite video... low costs matter more than anything else. Consumer CRTs are full of parts that create positive voltages, but additional parts are needed to create a negative voltage and that costs money. It doesn't get any cheaper to just wire G1 to ground.

The modification to decrease spot size is 2 parts: Tie the G1 to a negative voltage power supply so that the G1 is set to a constant value between 0 volts and -100 volts. The spot size decreases as the voltage is more and more negative. Since the electromagnet aperture created by the G1 anode is half the size at -100 volts, the electron beam that can pass through the G1 is more focused, but it also means a higher positive G2 screen voltage is required to increase the pull of electrons through the smaller electromagnetic aperture. The G2 screen voltage can be tuned as if you are setting the black level of the CRT by tuning the “screen” knob on the flyback.
Before & After Photo of Prototype (September 21 2020)
These before and after photos were taken on a hand held iphone. So the exact angle and position of the photos is a bit off because my hand moved between shots. Also, I did the prototyping work in my garage, so the CRT screen is dirty and there is sun glare on the glass in the photos. Sorry, I am not a photographer. The prototype in these photos is an old used 27-inch curved slot mask consumer Panasonic TV, model number CT-27L8G. This TV's un-modded TVL is roughly 450TVL.

2-chip SNES 240p Test Suite Displaying a Solid Green Screen Over Component Video (left before, right after):
Spoiler
Image
PS2 Castlevania SoTN (left before, right after):
Image

SNES Super Mario World (left before, right after) (credit goes to ElBartoME):
Image
Click here for high resolution: https://i.imgur.com/LTCwG3e.jpg

SNES Super Metroid (left before, right after) (credit goes to ElBartoME):
Image
Click here for high resolution: https://i.imgur.com/W2oilYX.jpg

SNES Super Metroid (left before, right after):
Image

Minimum CRT Requirements For Mod To Achieve 1000 TVL
Spoiler
Even if your CRT does not meet these minimum requirements, the mod will still increase your CRT's TVL, but it will not reach 1000 TVL unless it meets these minimum requirements. Looking at my consumer CRT collection, most of the TVs 27-inch and above meet these requirements.
R=Attribute&Value&Explanation R=RGB Drive Amp Bandwidth&~25MHz&RGB cathodes need to be varied approximately 1000 times per the approximately 39.45uS of time per 3/4 of a horizontal scanning line R=Phosphor triad + mask width&1/1333th width of screen&if every other triad is illuminated, this is 1333 alternating white and black vertical lines and 1000 alternating white and black lines over the first 3/4 of the width of the screen


When measured in units of TVL, a 15khz CRT’s resolution (as opposed to addressability), can be calculated using the following formula:

Code: Select all

CRTResolutionTVL = minimum( CathodeTVL, PhosphorTVL, SpotTVL )

...where...

CathodeTVL = (RGB drive amp bandwidth in megahertz) * 39.45

PhosphorTVL = (3/4) * (CRT width) / (phosphor triad width)

SpotTVL = (3/4) * (CRT width) / (cathode ray spot width)
Step-by-step modding instructions
Spoiler
The changes are all made on the CRT neckboard, and these steps assume you already have a 250v power supply. The next spoiler section describes how to make such a supply for about $1:

1. Cut trace between ground and the trace to the G1 anode. My TV had a jumper between ground and G1, so desoldering that jumper was all I needed to change. Another option is to pull the CRT neck socket pin for G1.

2. Connect an 1uF and approximately 350v film capacitor between ground and the trace to the G1 anode. This keeps the G1 voltage from jittering. It is optional if the power supply you use in the next step has a good smoothing capacitor built in.

3. Connect the positive rail of an approximately 250v power supply to the ground leg of the capacitor from step 2. A higher voltage will make the spot size even smaller, but if you go too high, you will get arcing between the G1 and cathodes.

4. Connect the negative rail of the 250v power supply to the G1 leg of the capacitor from step 2.

5. Using the 240p test patterns (or equivalent), turn clockwise, the screen potentiometer on the flyback to increase the screen voltage until proper black levels are restored.
Last edited by LukeEvansSimon on Sun May 16, 2021 8:33 am, edited 44 times in total.
User avatar
Josh128
Posts: 2123
Joined: Thu Jan 16, 2014 9:01 am

Re: Mod a consumer CRT television to make it high TVL

Post by Josh128 »

Sounds like a fun project. Good luck!
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

In theory, decreasing the size of the electromagnetic aperture at the G1 anode will also make it easier to tune the spot to be in focus simultaneously at both the center of the screen as well as the corners of the screen as the depth of field for focus will be larger, meaning the spot remains in focus for a larger variation in distance from the screen. This will address the other main shortcoming of many consumer CRT TVs: inability to bring both corners and center of the screen into sharp focus at the same time. This issue occurs because of the small depth of field in which the spot stays in focus when G1 anode is zero volts. The distance the electron beam travels changes as the beam scans the edges of the screen versus the center of the screen.

I am going to build a dual power supply by wiring two Maida high voltage regulated power supplies in series with the center 0VDC rail tied to the chassis ground, so that I have an adjustable 0VDC to +220VDC rail for driving the RGB cathodes and a 0VDC to -220VDC rail for setting the G1 voltage and in doing so, adjusting the spot size. The benefit of using a regulated voltage for the RGB drive will be better contrast and less blooming and breathing during bright scenes.
User avatar
maxtherabbit
Posts: 1763
Joined: Mon Mar 05, 2018 4:03 pm

Re: Mod a consumer CRT television to make it high TVL

Post by maxtherabbit »

tag, very interesting
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

A deeper discussion of the interplay between mask pitch (sometimes called dot pitch) and spot size is in this reference in section "A Discussion of Issues Relating to Monitor and CRT Resolution". This means that the effective TVL of a CRT can be doubled before the mask pitch becomes the bottleneck. In fact, the aforementioned reference indicates that spot size can be smaller than the dot pitch and yield a pleasing sharp look.

The main goal of this project is to allow consumer CRT TVs to have a high TVL, RGB look, but with that huge fat boy 27-inch (or hell, 36-inch) consumer CRT goodness. Since the G1 voltage will be adjustable between 0 volts down to -220 volts, the spot size will be adjustable. For some people, the 1000 TVL look is too extreme, so having a dial will let people hit the sweet spot that works for them. I prefer the look of around 700 TVL.
SamIAm
Posts: 475
Joined: Thu Mar 03, 2005 1:09 am

Re: Mod a consumer CRT television to make it high TVL

Post by SamIAm »

This doesn't make a great deal of sense to me.

No matter what kind of mask they're behind, the red, green and blue phosphors in a color CRT are arranged in clusters. Suppose we have a group of six clusters in a horizontal row like this:

R-G-B R-G-B R-G-B R-G-B R-G-B R-G-B

The maximum number of alternating white and black vertical lines that this group will be able to resolve - which is how TVL is defined - is six. Altering the way the beam is focusing on any particular phosphor element won't change the fact that each cluster as a whole can only be one color.

Yes, there are vagueries with CRTs and not everything works in neat little pixel-units. Nonetheless, I think that at best, all you're going to be able to do is make a low TVL set have a superficially high TVL set look, i.e. thin scanlines and fat black spaces. You're not going to make a 450 TVL set produce 700 distinct alternating white and black vertical lines (over a horizontal span equal to the height of the picture).
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

SamIAm wrote:This doesn't make a great deal of sense to me.

No matter what kind of mask they're behind, the red, green and blue phosphors in a color CRT are arranged in clusters. Suppose we have a group of six clusters in a horizontal row like this:

R-G-B R-G-B R-G-B R-G-B R-G-B R-G-B

The maximum number of alternating white and black vertical lines that this group will be able to resolve - which is how TVL is defined - is six. Altering the way the beam is focusing on any particular phosphor element won't change the fact that each cluster as a whole can only be one color.

Yes, there are vagueries with CRTs and not everything works in neat little pixel-units. Nonetheless, I think that at best, all you're going to be able to do is make a low TVL set have a superficially high TVL set look, i.e. thin scanlines and fat black spaces. You're not going to make a 450 TVL set produce 700 distinct alternating white and black vertical lines (over a horizontal span equal to the height of the picture).
Read the references that I linked, which make it clear that spot size is typically 2 times the dot pitch, and a higher resolution is possible by decreasing spot size. I checked my CRTs and they clearly have at least 2 RGB triplets per illuminated scanline.

In your example a vertical line uses two triplets because spot size illuminates 2 RGB triplets. With a spot size of half the diameter, only one triplet is illuminated for a single horizontal line. This effectively doubles the TVL.
User avatar
mikejmoffitt
Posts: 629
Joined: Fri Jan 08, 2016 7:26 am
Location: Tokyo, Japan

Re: Mod a consumer CRT television to make it high TVL

Post by mikejmoffitt »

Improving beam focus will never overcome the phosphor pitch's limitations.

TVL has nothing to do with how "thick" or sharp the scanlines are. That correlation is simply a coincidence of higher-end monitors having superior tubes, regulation, condition, etc in general.

The evidence for this is empirical. Just get up close to the tube, and... look at it. Look at the phosphors! For medical imaging, they may be discussing monochrome CRTs, where this is a different matter.

If playing with G1 gives you any results, I'd be interested to see what you get, but I think the best you can expect is an improvement in beam focus. There are many other things you can do to make that better before you consider anything higher effort (adjust the focus knob, calibrate cutoff and gain levels correctly, replace capacitors to improve geometry, ensure the yoke is centered and mounted firmly, perform static convergence adjustment using the rings, fix dynamic convergence with yoke adjustments and/or strips...)
Image
User avatar
maxtherabbit
Posts: 1763
Joined: Mon Mar 05, 2018 4:03 pm

Re: Mod a consumer CRT television to make it high TVL

Post by maxtherabbit »

ok we've got 2, who's the next poster that's going to say the same thing without reading the OP?
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

maxtherabbit wrote:ok we've got 2, who's the next poster that's going to say the same thing without reading the OP?
Every point they raised is addressed in the OP, and references are provided, which go into explanations of addressibility versus resolution, and that resolution is determined by spot size and dot pitch, with spot size being the bottleneck in most monitors because most monitors use a spot size of 2 RGB triples, with some using a smaller spot size.

The comments about phosphors shows they don’t understand what contributes to resolution too, and focus? The G3 anode for focus functions more like an electromagnetic lens, whereas the G1 control grid anode functions like an aperture.

My worry is convergence will be much harder as there are 3 spot sizes that get smaller: R, G, and B. A slight misconvergence of each beam will be much more noticeable with a smaller spot size.
User avatar
Syntax
Posts: 1774
Joined: Wed Aug 09, 2017 12:10 am
Location: Australia

Re: Mod a consumer CRT television to make it high TVL

Post by Syntax »

Ill bite.

All you are doing is sharpening the image.
Reducing dot size reduces bleeding and increases the dark areas around each pixel. It reduces brightness somewhat though.

Would make a really poor set sharper sure, but it's not increasing tvl at all.
SamIAm
Posts: 475
Joined: Thu Mar 03, 2005 1:09 am

Re: Mod a consumer CRT television to make it high TVL

Post by SamIAm »

LukeEvansSimon wrote:
maxtherabbit wrote:ok we've got 2, who's the next poster that's going to say the same thing without reading the OP?
Every point they raised is addressed in the OP, and references are provided, which go into explanations of addressibility versus resolution, and that resolution is determined by spot size and dot pitch, with spot size being the bottleneck in most monitors because most monitors use a spot size of 2 RGB triples, with some using a smaller spot size.
If you modify a TV so that its smallest possible "spot" goes from using at least two RGB triplets across the horizontal axis to only one, that's great, but your title is misleading. If your tube is, under any circumstances, able to resolve 700 alternating white and black vertical lines over a horizontal span equal to the display height, then by definition it is at least a 700 TVL tube, and that's probably the spec the manufacturer would list.

One wonders why a manufacturer would bother putting a 700 TVL tube into a set where the surrounding electronics prevented it from resolving anywhere near that, but perhaps they had their reasons. I'd still be shocked if you managed to get 700 alternating white and black vertical lines to display on a TV where the spec sheet says it's 450 TVL.
Last edited by SamIAm on Wed Aug 19, 2020 4:47 am, edited 4 times in total.
User avatar
Guspaz
Posts: 3136
Joined: Tue Oct 06, 2015 7:37 pm
Location: Montréal, Canada

Re: Mod a consumer CRT television to make it high TVL

Post by Guspaz »

I don't think a smaller dot would reduce brightness, not like you'd expect anyway: you're getting the same amount of energy in a smaller space, so the phosphors will glow brighter. The average brightness of the screen is the same.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

SamIAm wrote:
LukeEvansSimon wrote:
maxtherabbit wrote:ok we've got 2, who's the next poster that's going to say the same thing without reading the OP?
Every point they raised is addressed in the OP, and references are provided, which go into explanations of addressibility versus resolution, and that resolution is determined by spot size and dot pitch, with spot size being the bottleneck in most monitors because most monitors use a spot size of 2 RGB triples, with some using a smaller spot size.
If you modify a TV so that its smallest possible "spot" goes from using at least two RGB triplets across the horizontal axis to only one, that's great, but your title is misleading. If your tube is, under any circumstances, able to resolve 700 alternating white and black vertical lines over a horizontal span equal to the display height, then by definition it is at least a 700 TVL tube, and that's probably the spec the manufacturer would list.

One wonders why a manufacturer would bother putting a 700 TVL tube into a set where the surrounding electronics prevented it from resolving anywhere near that, but perhaps they had their reasons. I'd still be shocked if you managed to get 700 alternating white and black vertical lines to display on a TV where the spec sheet says it's 450 TVL.
“TVL” is not a specification of a CRT. It is a specification for a display that is a combination of a chassis and a CRT. How the chassis drives the CRT determines TVL. I mentioned why consumer CRTs did not bias the G1 anode to -100 volts: it would increase costs because a negative power supply is required.

But hey, don’t read any of the references where biasing the G1 is described as a technique for creating a smaller spot size and a higher resolution. The slides I linked are from a pro CRT company that specialized in high resolution CRT monitors. I guess you think their detailed plots, schematics, explanations for developing high resolution CRT displays is made up?

Finally, you are confusing addressability with resolution. TVL measures resolution, not addressability. A 15khz, 240p display can only address 480 vertical lines that span the width equal to the display height on a 4:3 CRT. I spoke to this in the OP. When you play 240p content on a 700 TVL, 15khz display, you get a sharper image due to two factors: smaller spot size and smaller dot pitch. Your game is not actually able to alternate 700 white and black vertical lines in a horizontal width equal to display height. If you read my references, you’d know this.

I am not claiming that shrinking spot size will increase addressability. The CRT TV will still be a 15khz display. It can’t display 1080p content with a smaller spot size, but high TVL 15khz displays can’t display 1080p content either. Addressability is not the same thing as resolution. Read the references if you still do not understand.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

Syntax wrote:Ill bite.

All you are doing is sharpening the image.
Reducing dot size reduces bleeding and increases the dark areas around each pixel. It reduces brightness somewhat though.

Would make a really poor set sharper sure, but it's not increasing tvl at all.
A smaller spot size doesn’t increase resolution? Please inform Texas Instruments and Display Labs Inc that their published research on CRT technology is wrong. See, I provided authoritative references for my statements, and you just made a statement without any references to back it up.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

Guspaz wrote:I don't think a smaller dot would reduce brightness, not like you'd expect anyway: you're getting the same amount of energy in a smaller space, so the phosphors will glow brighter. The average brightness of the screen is the same.
The paper from TI and the slides from Display Labs Inc mentions that screen voltage does need to be increased to set black levels. The beam’s energy is more concentrated so the spot is smaller but brighter.
SamIAm
Posts: 475
Joined: Thu Mar 03, 2005 1:09 am

Re: Mod a consumer CRT television to make it high TVL

Post by SamIAm »

There is a lot that I'd like to respond to, but this has to come first:
LukeEvansSimon wrote:A 15khz, 240p display can only address 480 vertical lines that span the width equal to the display height on a 4:3 CRT. I spoke to this in the OP.  When you play 240p content on a 700 TVL, 15khz display, you get a sharper image due to two factors: smaller spot size and smaller dot pitch.  Your game is not actually able to alternate 700 white and black vertical lines in a horizontal width equal to display height.  If you read my references, you’d know this.
I've read this a few times, and each time I'm left wondering if you really said what you seem to be saying.

You're not telling me that I could connect a pattern generator with an adjustable dot-clock to a pro monitor spec'd at 700 TVL and not be able to see, with my eyeballs, above 480 alternating vertical lines (in the relevant horizontal space) just because the monitor's horizontal refresh rate is 15khz, are you?

Because that would be a very extraordinary claim.
User avatar
Syntax
Posts: 1774
Joined: Wed Aug 09, 2017 12:10 am
Location: Australia

Re: Mod a consumer CRT television to make it high TVL

Post by Syntax »

LukeEvansSimon wrote:
Syntax wrote:Ill bite.

All you are doing is sharpening the image.
Reducing dot size reduces bleeding and increases the dark areas around each pixel. It reduces brightness somewhat though.

Would make a really poor set sharper sure, but it's not increasing tvl at all.
A smaller spot size doesn’t increase resolution? Please inform Texas Instruments and Display Labs Inc that their published research on CRT technology is wrong. See, I provided authoritative references for my statements, and you just made a statement without any references to back it up.

You will only see a resolution increase if you move the now smaller dots closer together, how will you do that?

All you've suggested is to make them smaller and increase the dead space around them.

Why would I need to give reference for something as simple as that?
User avatar
Xer Xian
Posts: 881
Joined: Sun Feb 06, 2005 3:23 pm
Location: Italy

Re: Mod a consumer CRT television to make it high TVL

Post by Xer Xian »

@SamIAm - I think what he wanted to say there was that 15kHz consoles had too low a dot clock to be able to render high resolutions.

I might be wrong, but it seems to me that all the objections raised here up to now are missing the mark. The OP already acknowledged that the mask pitch is a limiting factor, but provided some reference indicating that it may not be the actual bottleneck for some CRTs.

I think the thread is interesting and I'm eager to see the findings to come out of this. I would be very surprised if run-of-the-mill consumer CRTs turn out to be able to resolve 700 TVL with such a basic mod. At least in the case of very cheap sets, I would expect the quality of the electronic components and of the optics * to be other relevant factors in limiting the TVL count, so that reducing the spot size will surely achieve thinner scanlines (thicker blanklines) but not increase the TVL count substantially.


* this is brought up on the first article linked in the OP:
Spoiler
Video bandwidth is related directly to the horizontal addressability of a display, as it deter- mines how many individual points may be con- trolled along each horizontal scan line. The video amplifier must be "flat" (have equal output at all frequencies throughout the bandwidth range) in order to produce the correct shading for all-sized objects on the screen. It must not "ring" or have excessive "overshoot" which can cause problems in displaying sharply defined objects.
This small excerpt is also worth a read for its clarity (taken from 'Image Performance in CRT Displays', page 47):
Spoiler
Image
An article that goes into more detail about optic-related factors than the one cited above can be found here.

Edit: incidentally, but mostly unrelated to the thread, I've just bumped into this article on reducing the spot size by acting on the yoke rather than the anodes (unfortunately, it is in japanese)
User avatar
Josh128
Posts: 2123
Joined: Thu Jan 16, 2014 9:01 am

Re: Mod a consumer CRT television to make it high TVL

Post by Josh128 »

Syntax wrote:
LukeEvansSimon wrote:
Syntax wrote:Ill bite.

All you are doing is sharpening the image.
Reducing dot size reduces bleeding and increases the dark areas around each pixel. It reduces brightness somewhat though.

Would make a really poor set sharper sure, but it's not increasing tvl at all.
A smaller spot size doesn’t increase resolution? Please inform Texas Instruments and Display Labs Inc that their published research on CRT technology is wrong. See, I provided authoritative references for my statements, and you just made a statement without any references to back it up.

You will only see a resolution increase if you move the now smaller dots closer together, how will you do that?

All you've suggested is to make them smaller and increase the dead space around them.

Why would I need to give reference for something as simple as that?
This is an interesting question, and perhaps either OP or you could shed some light on what makes the difference between the granularity and spacing of horizontal dots or pixels on a line between say, a consumer Trinitron and a BVM Trinitron. Is it not simply the pitch of the aperture grille and its distance from the phosphor coating of the screen (for focus)? Or is it the phosphor coating itself? Its certainly not the input horizontal resolution because feeding each of the sets with say a standard SNES 256x224 yields vastly different results in terms of horizontal granularity, which appears to be different altogether from focus.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

SamIAm wrote:There is a lot that I'd like to respond to, but this has to come first:
LukeEvansSimon wrote:A 15khz, 240p display can only address 480 vertical lines that span the width equal to the display height on a 4:3 CRT. I spoke to this in the OP.  When you play 240p content on a 700 TVL, 15khz display, you get a sharper image due to two factors: smaller spot size and smaller dot pitch.  Your game is not actually able to alternate 700 white and black vertical lines in a horizontal width equal to display height.  If you read my references, you’d know this.
I've read this a few times, and each time I'm left wondering if you really said what you seem to be saying.

You're not telling me that I could connect a pattern generator with an adjustable dot-clock to a pro monitor spec'd at 700 TVL and not be able to see, with my eyeballs, above 480 alternating vertical lines (in the relevant horizontal space) just because the monitor's horizontal refresh rate is 15khz, are you?

Because that would be a very extraordinary claim.
Is your pattern generator running on a 240p/480i gaming console/arcade? Again, you are confusing addressability with resolution. The OP said that addressability is not increased, and since the OP is only scoping to 240p and 480i gaming content, addressability is off topic. None of the 15khz gaming hardware in consideration can address 700 TV lines. The high TVL look for 240p content is influenced by spot focus and dot pitch.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

Syntax wrote:
LukeEvansSimon wrote:
Syntax wrote:Ill bite.

All you are doing is sharpening the image.
Reducing dot size reduces bleeding and increases the dark areas around each pixel. It reduces brightness somewhat though.

Would make a really poor set sharper sure, but it's not increasing tvl at all.
A smaller spot size doesn’t increase resolution? Please inform Texas Instruments and Display Labs Inc that their published research on CRT technology is wrong. See, I provided authoritative references for my statements, and you just made a statement without any references to back it up.

You will only see a resolution increase if you move the now smaller dots closer together, how will you do that?

All you've suggested is to make them smaller and increase the dead space around them.

Why would I need to give reference for something as simple as that?
Being able to move the TV lines closer together is determined by addressability, not resolution. The term”resolution” has a specific definition in CRT literature and it is not the same thing as the addressable resolution or input resolution. None of the retrogaming 15khz retrogaming consoles are capable of addressing more than 480 TV lines.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

The research paper on using yoke voltages to decrease spot size is interesting, but the entire paper is in Japanese. Can somebody get an English translation? I am trying to compile and read as much CRT literature as possible. It is a shame if we lose all of this tech AND the knowledge that goes with it, because that means we will never manufacture this stuff again if the knowledge is lost. Also, there are going to be cool mods and projects that come from better community understanding of the tech.

I own a BK 490b CRT analyzer, and I have used it to analyze and rejuvenate old CRTs. It has a setting that allows for testing the CRT with different G1 voltages applied. In vacuum tube literature, this is often referred to as the "bias" of the tube, and the bias is typically ground (zero volts) or it is negative.
Sencore made the highest end CRT analyzers and rejuvenators, and their literature here provides more explanation of how the G1 anode's negative voltage is used as an iris.
Spoiler
Image
This article, on the CR7000, remarks that high resolution CRTs use a negative bias (that is a negative G1 voltage):
Spoiler
Image
User avatar
Xer Xian
Posts: 881
Joined: Sun Feb 06, 2005 3:23 pm
Location: Italy

Re: Mod a consumer CRT television to make it high TVL

Post by Xer Xian »

LukeEvansSimon wrote: Being able to move the TV lines closer together is determined by addressability, not resolution.
I'm no expert but I would say it's the other way around - TV lines measure resolution. Unless you're trying to say something else here?

Also, it somehow slipped my mind but upon rereading I realized that my post pretty much reiterated what SamIAm said, which makes me think some of the confusion expressed here was warranted. What you're trying to do should be clear by now, but the thread title should probably be ".. to make it high TVL-looking".
User avatar
maxtherabbit
Posts: 1763
Joined: Mon Mar 05, 2018 4:03 pm

Re: Mod a consumer CRT television to make it high TVL

Post by maxtherabbit »

if you visualise a beam sweeping continuously across each horizontal line at a fixed speed, and then reduce the diameter of the beam, you should easily be able to understand how that would accommodate more horizontal detail if the source dot clock is sufficiently high and the mask pitch is sufficiently small
SamIAm
Posts: 475
Joined: Thu Mar 03, 2005 1:09 am

Re: Mod a consumer CRT television to make it high TVL

Post by SamIAm »

Well, I think I get what you want to do, and I don't think it would be productive to sit here and argue about wording, so best of luck to you. It would indeed be neat if what you're theorizing turns out to be doable.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

SamIAm wrote:Well, I think I get what you want to do, and I don't think it would be productive to sit here and argue about wording, so best of luck to you. It would indeed be neat if what you're theorizing turns out to be doable.
There is value in reading and discussing all of the published research on CRT technology and then designing mods. This is how the RGB mods came about. This ideation process is how other mods will come into existence. Also, the CRT gaming community would benefit from using the standard terminology from the published research literature when talking about CRTs. It is clear that many in the community have no idea that their high TVL CRT's addressability is not what gives the pleasing high TVL aesthetic for 240p and 480i gaming. Your Genesis and SNES cannot address more than 480 TV lines. So spot size and dot pitch are what gives the high TVL aesthetic.

The OP mod can be applied to professional broadcast CRTs too to increase TVL for those that want a little more sharpness and thicker unilluminated scanlines. I checked my Panasonic BT-H1350Y's schematic and the G1 anode is tied to chassis ground (zero volts). I then hooked its M34JEF037X picture tube to my BK 490b and tested the cutt-off emissions of each electron gun for various voltages between -100 volts and 0 volts. The most catastrophic thing that could happen if the G1 anode is made too negative, such as -500 volts, would be electrical arcing between the cathode K and the G1 or arcing from the G1 to G2. But a quick test with my BK 490b shows that these tubes can easily handle -100 volts.

I have kids and a job, so getting the time to put together the mod is restricted to a few hours block on Saturdays and Sundays. I've placed a mouser order for the parts: transformer, rectifier, filter capacitor, and high voltage regulator. I will try it with and without a voltage regulator. The benefit of the regulator version will be a very smooth voltage to the G1. The unregulated version will be a cheaper mod, but may cause the spot size to jitter, which will look horrible.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

I finally bought a bench power supply. I should have done this a long time ago. Would speed up experimenting with so many projects. The power supply has adjustable voltage up to 120 volts and current limiting up to 3 amps. I will use a 1N5402 diode to tie the positive output of the supply to the chassis ground that the entire CRT TV's circuit that sits on secondary side of the TV's isolation transformer. I will then experiment with low current and low absolute value negative voltages, slowly dialing them up until spot focus no longer decreases... or I blow up the electron gun :)

The voltage range I should experiment with is clear to me, but the current range is not clear. My understanding is that the max current should be very low, less than 10mA because the G1 anode is not connected to the cathode or the other anodes. Anybody have a better understanding of the electrical characteristics of vacuum tubes? What is the best way to determine the max current that I should allow to run through the G1 anode?
User avatar
buttersoft
Posts: 383
Joined: Sun Jul 24, 2016 7:49 am

Re: Mod a consumer CRT television to make it high TVL

Post by buttersoft »

Current ranges for later, very high-end rejuvenators that had a super low "re-activate" setting (super-low rejuv current for up to 30s) were still only using about 1mA. Not sure if that helps.

I'm not sure i grasp how you're going to reduce spot size. My understanding is that lowering the screen grid voltages will lower the number of electrons and reduce the brightness. This reduces spot size, everything else being equal, because the focus grid is now capable of delivering more energy per electron vs time, keeping the beam tighter. Simply adjusting the brightness or even the Screen control, the latter normally on the flyback of a set, has this effect (and it implies such in that CRT's for medical imaging paper you linked). If in a given set you bypass the HV regulation and simply make beam focus tighter, but keep everything else the same, you'd be lasering x-rays through the leadglass - or having that effect. Note that i said, in a a given set. Different sets are designed with different parameters.

Addressability would simply seem to be a way of measuring the dot-pitch against the chassis bandwidth with spot-size thrown in, i'm not sure i see the big deal there. You might be able to put slightly faster, slightly more expensive transistors onto the neckboard and improve things very slightly, but i don't get how else you'd attack the issue. My feel is that to make any real gains in a particular set using a fixed tube, you'd need to redesign the entire chassis. There's a reason old NANAO/Sega MC-2000S arcade monitors from the early 80's were like 8 times the price of their nearest rivals, eve ntough they were only slightly more reliable and giving a slightly nicer picture, albeit noticeably so. And the same goes for BVMs vs PVMs vs Prosumer vs consumer CRT's.

Don't mistake me, i'd love to be proved wrong, and if you could find a way to make this happen it would be completely awesome.

EDIT: I think the lion's share of what i'm really trying to say here is that surely every consumer set is designed to run up against its limits in terms of HV regulation and x-ray production. If you dial up the brightness you increase the grid voltages, if you increase the grid voltages you increase the beam current, and if you increase the beam current the focus voltages drop to spread the beam wider and reduce x-ray production. Though now i read what i've written, i want someone to help me examine that last part. Does spreading the electrons out reduce x-ray production? Perhaps it's simply that the focus winding in the flyback is fixed (at any given time) and as you increase beam current it's just not able to deliver as much energy? Any thoughts on that? Surely it can't just be phosphor halation? No matter how you dial the knob around, a very bright image is always in poorer focus than a dim one. But why not simply add more energy? There'd be a reason why, and that reason is usually to do with regulation. Along with which there must physical limitations to focus imposed by the quality of the neck assembly in CRT tubes. And how exactly do the bias and drive voltages figure into this? Either way you look at it, changing the focus alone, or the G1 or G2 alone, or any other important part of the regulated circuitry, is likely to have secondary effects.

I should add that i'm using an unusual chassis in one of my arcade machines, and slightly overdriving a consumer tube to about 105% brightness. The resulting x-ray production isn't geometric/much increased, but going any higher, to say 110% brightness, spikes x-ray production to about 200%. So it might be worth measuring for any side effects like that, if you do achieve a noticeable gain in one parameter.
LukeEvansSimon
Posts: 128
Joined: Mon Aug 17, 2020 7:28 pm

Re: Mod a consumer CRT television to make it high TVL

Post by LukeEvansSimon »

Each grid in a CRT serves a different purpose. G2 is a particle accelerator that also determines how many particles get pulled off of the cathode. G4 is a particle accelerator. G1 acts like an iris, and G3 acts like a lens. So while an iris does increase and decrease the amount of particles that pass through it, it also filters the particles down to those that are more uniform in their motion along the beam. This causes the spot size to be smaller AND it causes the depth of field for spot focus to be larger. G2 does not have this iris effect because of both the shape of the anode and its positive charge.

Reading published literature for high resolution CRTs, it looks like it was common to have G1 set to a constant -70 up to around -100 volts so as to keep spot size small.
Post Reply