Improving CRT geometry by upgrading capacitors

[momaka]

Re: Improving CRT geometry by upgrading capacitors

But what strategy would I need to get video out of something with DVI or HDMI to an analog monitor?

That depends on your video card.
Analog output is no longer built into current-gen video cards anymore. In fact, the last (and best) nVidia video card to have analog output (through DVI-I) is the GTX 980 Ti. And for AMD, it was the R7/R9 series, I believe... though I can tell you for a fact that not all 290's come with DVI-I (like on my ASUS DCuII, where the designers simply chose not to have a DVI-I connector, despite the card GPU chip actually having a RAMDAC.)

Thus if you want analog output from a video card without having to use an _active_ adapter/converter, then you will need either a GTX 900-series or older video card or Radeon R7/R9 series or older. Anything newer doesn't have a RAMDAC and hence no analog output.

...and another Hitachi ir similar with the R, G, B, H, V BNC connector cable.

You can still get BCN-VGA [D-SUB] cables on sites that sell used stuff, like eBay. In fact, I got one last year, since I have one at my uncle's house, but not one at my house, where I still have and use old CRT monitors. For $8 shipped, I got a genuine good quality used BNC-VGA cable. Beware of the cheap stuff from China. I suspect they may have cheaper shielding, which could cause ghosting or poor image at high resolutions. Or maybe they are fine, I don't know, I haven't tested them. But I am always weary of cheap stuff from China.

Maybe I should bust these out and try some modern games on them.

Honestly, I don't think you would be disappointed.
You may not be able to run the games at such high resolutions. But then again, I don't really recommend going for the highest resolution anyways. Rather, try to balance things between resolution and refresh rate. For the latter, 75 Hz minimum and 85 Hz preferable. If your rig can run the games well in excess of 85 Hz, you can even try V-sync for extra smooth experience. In general, it's almost impossible to see image tearing on a CRT. But with V-sync on, the image will look even smoother. Or you can jump up to 100 or 120 Hz for super-smooth gameplay.

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

Another picture:


800x600 with a $20 VGA card.
For people who think that CRT is outdated technology.


There is room for higher resolution up to 1280x1024, maybe 1600x1200 with a better video card.

Well, the quality maxes out at 1024x768 in terms of sharpness which is even more defined than 800x600. Capacitors which passes video signals where upgraded with very low E.S.R. capacitors and decoupling for video amplifier was "improved".

The most difficult task was guessing parts which needed better components and places were avoiding low E.S.R. capacitors was a better choice (e.g. quality went up when upgrading VDDA-VSSA coupling capacitor on the digital deflection IC).

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

I rely on a PCIe VGA card with direct VGA output.

By the way, the CRT is still working perfectly. Planning to find a bigger monitor and tune everything till the geometry is almost perfect.


The picture looks real, incredible. Not bad for a 2002 monitor.

[stj]

Re: Improving CRT geometry by upgrading capacitors

check your DVI, there are different pin patterns,
some output VGA aswell as digital, others are digital only and have some pins missing.

if your lucky, you can just use a DVI to BNC cable (or to a 13W3 for the SGI monitor)

[Logistics]

Re: Improving CRT geometry by upgrading capacitors

How are you guys connecting to said CRT's? Are these TV's with SVGA-in's? I understand soandso was using a Trinitron monitor. But I still have a big SGI out in the container, and another Hitachi ir similar with the R, G, B, H, V BNC connector cable. Maybe I should bust these out and try some modern games on them. But what strategy would I need to get video out of something with DVI or HDMI to an analog monitor?

[clearchris]

Re: Improving CRT geometry by upgrading capacitors

The yellowing can be revesed to a large extent, see http://retr0bright.com/

The plastics turning brittle, I'm not sure anything can be done, but it, like the yellowing might not be a one-way process.

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

I agree.

Too bad that time is passing and the plastics, even with very good monitors, are both yellowing and degrading. The age of the plastic material goes the point that you have to be careful even removing screws (when removing e.g. the main board) or mounting back everything will be impossible. Faston are to be swapped to get rid of corrosion, you have to refresh most of the solder points and so on.

And metal boxes with shielding are oftenly rusted, so restoring takes days and days and is an extremely challenging task where a single error makes everything useless.

[momaka]

Re: Improving CRT geometry by upgrading capacitors

Because the quality of the picture, its contrast, the blacks, the colours and the sharpness beats hands down the quality of any LCD I've encountered so far.

Exactly!

Well, I don't know about the sharpness of a CRT beating any LCD. In fact, CRTs are typically a lot less sharp than LCD monitors (due to pixels in CRTs not being exactly square, unlike LCDs.) But you are absolutely right about the colors, contrast, and deep blacks. Also, don't forget that CRTs do not show any tearing on a moving image, regardless if the image is v-synced or not. This is why video/moving image on a CRT running even at 60 Hz still looks more fluid than an LCD. And also why analog video looks better on a CRT than LCD... well, it's that, and the fact that CRTs have a more neutral mid-tone colors. LCDs tend to "exaggerate" colors near mid-brightness levels, and this makes compression artifacts on video (especially low-bit rate and non-HD) show a lot more.

Apart from that, picture is outstanding and even movements on both movies and games have perfect responsiveness already at 85Hz.

Indeed.
This is why I still actually use CRTs for gaming. Many people ask me why I keep those old things until they try them out. The difference really is noticeable.

Unless you have some really top of the line CRT, running any modern software will be quite painful on a low-res crt.

Not really.

Most web pages still render fine on 1024x768, and that's the resolution I still use quite frequently myself. For other software, bigger resolution may be needed to fit more stuff on the screen. Personally, though, I don't like that, because then text becomes to small to read. I like to sit fairly far from my monitors, so very larger resolutions start to bother me a lot if the text becomes too small to read. I don't like hunching over the screen to see it.

That said, I only switch to higher resolutions when gaming... and by higher, I mean 1152x864 or 1280x960. On a CRT (especially standard 17" and 19"), those resolutions don't look grainy at all, unlike on an LCD. Very rarely, I may do 1600x1200 on my 21" CRTs, but it's rare. I rather sacrifice resolution (especially on older PCs) to gain framerate and higher refresh rate. 85 Hz on a CRT looks amazing for gaming.

In terms of image color quality, I wouldn't say that a large "top of the line" CRT necessarily looks better. Most CRTs actually have a trade-off between contrast and screen size. For a given tube anode voltage and current, bigger screen means light output from the tube is spread over a bigger area, and that means lowered contrast. I find this to be especially true for Sony CRTs. Both of my 21" Sony's have significantly worse contrast than the 17" and 19" models from the same line. So if I want better "cinematic experience", I go for the smaller screens. On that note, I find 19" CRTs to be the best compromise between contrast and screen size.

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

So what's the use case for keeping a CRT running? I'm sure anyone who hangs here can find a broken LCD and make it work.

Because the quality of the picture, its contrast, the blacks, the colours and the sharpness beats hands down the quality of any LCD I've encountered so far.

Obtaining a perfect geometry is howewer almost impossible.
You have to either:
1) Restrict the picture horizontally, obtaining an almost acceptable geometry, leaving black borders on both left and right.
2) Enlarge the picture, which exposes unlinearities on the edges of the tube, but no black borders.

So, you have to accept a compromise.

Apart from that, picture is outstanding and even movements on both movies and games have perfect responsiveness already at 85Hz.

Are you running old school nintendo and want to play "duck hunt" with the orignal zapper?

No, this is a PC CRT with Triniton tube. On CRT TVs coming with an acceptable geometry is much easier.

[clearchris]

Re: Improving CRT geometry by upgrading capacitors

I totally agree, I fix everything I can. Unless you have some really top of the line CRT, running any modern software will be quite painful on a low-res crt.

[eccerr0r]

Re: Improving CRT geometry by upgrading capacitors

It runs... save the environment: use till it breaks, then fix it, and use it some more until the cathode is contaminated and the heater burns out...

[clearchris]

Re: Improving CRT geometry by upgrading capacitors

So what's the use case for keeping a CRT running? I'm sure anyone who hangs here can find a broken LCD and make it work.

Are you running old school nintendo and want to play "duck hunt" with the orignal zapper?

[R_J]

Re: Improving CRT geometry by upgrading capacitors

The B+ that supplies the horizontal output transistor is usually around 130vdc, but the transistor is usually rated for 1500v or more and disc capacitors in the horizontal circuit can be rated 2~3KV it has is due to the high frequency ac peak voltage not the dc voltage.

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

I suspect it as well. My multi-meter is quite cheap though so I take its reading with a grain of salt and derate. Not that I would not derate nevertheless when it comes to delicate circuitry with almost-irreplaceable transistors.

[R_J]

Re: Improving CRT geometry by upgrading capacitors

Other informations for the readers and the members of the forum.

Found a very old CRT monitor which effectively uses MKP capacitors for this exact section. It parallels four MKPs: 820nF+820nF+1uF+1uF. They are 250V parts despite the voltage being no more than 60V (quickly measured with a cheap multi-meter in DC).

You are forgetting about the high frequency pulse that is on that cap, I suspect it is greater than 60v peak

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

Other informations for the readers and the members of the forum.

Found a very old CRT monitor which effectively uses MKP capacitors for this exact section. It parallels four MKPs: 820nF+820nF+1uF+1uF. They are 250V parts despite the voltage being no more than 60V (quickly measured with a cheap multi-meter in DC).

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

Other useful informations for the members of the forum and whoever arrives here with the same question:
1) Higher capacitance made no differences to geometry. Tried 6.8uF, 10uF, 12uF, 22uF, 22uF+12uF with zero changes to geometry. No overheating of the transistor by using higher capacitance but I cannot say if it causes problems on the long run. Not advised going higher than 15uF has I've seen monitors with such capacitance but never more.
2) Any electro capacitor, even a beefy crossover capacitor rated at 100V and able to sutain 900mA of alternate current, which is quite high, failed in a matter of minutes. The test was made while keeping a 3.3uF MKT capacitor in circuit to prevent failures of the horizontal transistor so we can assume that the electro capacitor was even not sustaing the whole current (!).
3) Installing 6.8uF capacitor (MKP/MKT) made the monitor more silent at 100Hz. Same obviously happened with higher capacitances.
4) If the original capacitor in this area measure much more than its rating (when I opened the monitor for the first time checking for faults the capacitor was a 3.3uF measuring 3.8uF with other parameters perfect but the screen was somewhat noisy) then you have to replace it. Producers back in the day used very cheap capacitors despite knowing that this component is vital for the working of the entire monitor.

Even if the replacement is a MKP capacitor, use at least two identical capacitors in parallel. Better if you use three in case of cheap/no-name MKP capacitors.

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

I've used a 0.1uF/250V MKP capacitor and called it a day.

[Hitto]

Re: Improving CRT geometry by upgrading capacitors

I've used a 3.3uF/50V electrolytic capacitor (made for horizontal deflection) and paralleled it with a small 0.1uF/50V MKP/Film capacitor (it should help with very high frequencies which the electrolytic is fisically unable to smooth).

It works okay.

Do you guys think that any MKP capacitor will be good, considering that it will only filter a small percentage of the total current, or should I replace it with something beefer?

I ask this because a lot of datasheets warn about using Film/MKP/MKT capacitors "with other types of waveforms".. and horizontal deflection on CRTs is a sawtooth.

Should I derate the MKP/Film capacitor in order to avoid problems?

[stj]

Re: Improving CRT geometry by upgrading capacitors

you may want to think about how low the current handling of the old electrolytics was.
use the 3uf