SCART to YUV box Adjustment Guide

The place for all discussion on gaming hardware
Post Reply
User avatar
James-F
Posts: 87
Joined: Fri Mar 23, 2018 11:01 am

SCART to YUV box Adjustment Guide

Post by James-F »

Everyone got this Scart (RGB) to YUV (Component, YPbPr) box from ebay/Amazon but can't get proper colors out of it...
Here is a guide how to get proper colors, you'll need an oscilloscope and 240p test suite able console (or a pattern generator).

Image
Image

1.
Plug everything and make sure the output of this scart-to-yuv box is terminated to 75ohm (like the OSSC or a TV).
Make sure you use a well balanced console that has equal RGB values when terminated to 75 ohm and can run the 240p test suite.
Start will all trimmer pots at middle position, towards the middle leg.

2.
Display SMPTE Color Bars 100% from the 240p test suite and scope the Y (Luma) black to white (no sync) line from the scart-to-yuv box (terminated of course), write down this number.
EDIT: The Y output amplitude is fixed, but adding a trimpot can change that, see fourth post of this thread.
Image

3.
Scope Pr and adjust Pr_Amp till it has the same amplitude peak-to-peak as the Y line.
Image

4.
Scope Pb and adjust Pb_Amp till it has same amplitude as Y, this trimmer doesn't have enough range so I set it to Max (fully clockwise).

Now the important color fixing procedures. :)

5.
Scope Pr and adjust Pr_Step (Pr Offset) till the is no visible step (see image with marked colors where the step is).
Play with the trimmer first to see the right side of the step go higher and lower than the left side of the step.
You can change the voltage deviation in your scope to a lower value to show the step more clearly resulting in finer adjustment.
Image

6.
Scope Pb and adjust Pb_Step (Pb Offset) till there is no more visible step, exactly the same as you did with Pr in point #5.
Image

7.
Continue to scope Pb and adjust Pb_Ratio till the amplitude difference between the first-to-second and second-to-third peak is equal (see image).
Play with the trimmer first to check how it affects the middle peak.

*Note the voltage delta in Pb between Cyan-Magenta-Blue in the chart at the bottom of this post.
Magenta-Cyan = 231.7-118.3 = 113.4
Blue-Magenta = 350-231.7 = 118.3
The result is roughly the same, so just make the voltage delta on both equal.

Image



Now start over and check that everything did not change, fine tune, etc, etc...
Note that the trimmer pots are different in tolerance (around 20% for these Chinese trimmers), so just copying my trimpot positions from the image will not do, might even get things worse. :shock:
With proper adjustment I get very close to RGB quality colors and grey balance when the YPbPr signal is decoded with the OSSC, this box is truly useful now.


Here's a useful chart:
Image

Here is a very useful document about Component (YPbPr) video signal:
https://cdn.rohde-schwarz.com/pws/dl_do ... 107_0E.pdf

PS.
Better to use a 7.5v or 9v power supply with this box since the LM7805 inside gets too hot with 12v.

Cheers. :D
Last edited by James-F on Thu Nov 22, 2018 5:23 am, edited 14 times in total.
User avatar
maxtherabbit
Posts: 1763
Joined: Mon Mar 05, 2018 4:03 pm

Re: SCART to YUV box Adjustment Guide

Post by maxtherabbit »

hell yeah, nice writeup

gotta ask - what's the middle unlabeled pot do?
User avatar
James-F
Posts: 87
Joined: Fri Mar 23, 2018 11:01 am

Re: SCART to YUV box Adjustment Guide

Post by James-F »

Well... nothing obviously. :mrgreen:
I see no difference in measurement or color on screen when turning this trimpot in any direction, keep it centered.
User avatar
James-F
Posts: 87
Joined: Fri Mar 23, 2018 11:01 am

Re: SCART to YUV box Adjustment Guide

Post by James-F »

The Y (Luma) output of this box is fixed and hotter than input making some digitizer devices (OSSC, RetroTink2x, Framemeister) clip the highest white shades, can reach 800mV output from 700mV input.
To fix this replace the 470 ohm feedback resistor in the final Y (Luma) amplifier with a 1k multi-turn trimpot (shown at the bottom of the image).
Scope Y and make white to black voltage at output the same as input with grey steps test pattern, then redo the calibration in the first post.
As a result of this mod the Pb channel can now be calibrated to to the same amplitude which was slightly lower output than Y and Pr.

Image


Another matter is Lowpass Filtering of the Y and Pr channels of this box which may (are) reduce the sharpness of these channels.
The 330pF cap (bottom of image) and 75ohm resistor form a 6.43 MHz lowpass filter on the Y channel which reduces sharpness, just remove it since all devices down the chain already have much better LPF.
The 2.5nF and 100 ohm form a 637 kHz LPF on the Pr channel which is very low and smears the red colors.
The 2.4nF cap for Pb LPF is also huge and smears the Pb channel.

EDIT:
After some testing I just removed all LPF caps and the resulting image is as sharp as the original signal.
It is important though that the device down the chain has a SDTV (6.5MHz - 9MHz) lowpass filtering since this conversion box uses a lot of amplifiers that create a tiny shift between the channels which result in very high frequency (very thin) color haloing, this ringing will cause havoc with low sampling rate digitizing.

EDIT2:
I have reverted the changes and added back 2.2nF caps for Pb and Pr, also the 330pF cap for Y.
Apparently without the chroma caps the bleeding/haloing is out of proportion with some TV that have higher than SDTV lowpass filtering, which is most modern TVs that designed to accept higher resolutions through Component inputs,, or with low sampling rates which may sample the ringing and make it much thicker than it should be.
All in all, just keep these caps stock, yes, there will be some minor chroma bleeding and loss of sharpness but it is much better than rainbow haloing/ringing.
Note that on a typical consumer CRT this bleeding will be barely visible, it is more visible with line doubler like the OSSC in higher multiplication factors.
Don't worry about it, it is still a great device after calibration.

Image

Plenty of people buy this converter with crappy colors out of the box, it needs some love before it is compliant. :P
someitalian123
Posts: 10
Joined: Mon Mar 06, 2017 1:32 am

Re: SCART to YUV box Adjustment Guide

Post by someitalian123 »

Can the pots be replaced with simple resistors, I do not have an oscilloscope to test exact values.
fernan1234
Posts: 2175
Joined: Mon Aug 14, 2017 8:34 pm

Re: SCART to YUV box Adjustment Guide

Post by fernan1234 »

This is very good info. For those able to do this, it's a cheaper solution than using something like the RGB2COMP.
Taiyaki
Posts: 1050
Joined: Fri Apr 04, 2014 11:31 pm

Re: SCART to YUV box Adjustment Guide

Post by Taiyaki »

Fantastic guide.
User avatar
kitty666cats
Posts: 1270
Joined: Tue Nov 05, 2019 2:03 am
Location: Massachusetts, USA

Re: SCART to YUV box Adjustment Guide

Post by kitty666cats »

I'm apparently one lucky dude, because the colors on mine looked fine right outta the box. Are the screwed-up colors truly as common as people make it out to be?

I've long since moved on to the RetroTink transcoder, and I have one of Jam's (linuxbot3000) coming in the mail that I bought for kicks/to support him because he's a cool-ass dude :) I will still always have love for the good ol' CSY-2100 clone though, haha
User avatar
Shelcoof
Posts: 1520
Joined: Mon Nov 03, 2008 9:36 pm
Location: Canada

Re: SCART to YUV box Adjustment Guide

Post by Shelcoof »

Wow where was this guide when I bought my Scart to YUV years ago.

Luckily for me I was able to eye out the colors while adjusting the pots and I'm happy with the results.
But man of man it took a lot of fiddling around to get the colors just right for my liking/
Post Reply