Geometry calibration - red and white grid

The place for all discussion on gaming hardware
Post Reply
Cavia
Posts: 10
Joined: Thu May 07, 2015 5:58 pm

Geometry calibration - red and white grid

Post by Cavia »

Hello,

given that many are using the '240p test suite' as a test/calibration tool for the geometry of a CRT, I was wondering if this very common (I've seen it in a lot of CRT including PVMs) problem can be fixed.
I'm referring to the case where the white part of the grid in the 240p test suite, is not correctly aligned with the red grid.
For example in the image below, you can see that the white grid starts correctly on the bottom, but it ends on the opposite side when reaching the top corner.
https://pasteboard.co/JJtIRdq.jpg
This picture was taken from a PVM2042QM (source 240p TS from a Nintendo Wii).

Is that fixable? The reason for my question is that, even after achieving a close to excellent geometry like this one
https://pasteboard.co/JJtK9A8.jpg
as soon as I change the source (say a RGB-pi or others), the lines on screen (especially the white ones from menus) are usually deformed instead of straight/aligned (as they are following in fact the white grid).
This is very annoying.

Any help is welcomed.
Thanks.
User avatar
Josh128
Posts: 2124
Joined: Thu Jan 16, 2014 9:01 am

Re: Geometry calibration - red and white grid

Post by Josh128 »

Whats happening in this pic is that your horizontal convergence is "keystoning" from the bottom of the screen to the top, also called C-cross if it only occurs at the sides or Y-cross if it happens in the center. If you look close you will notice that the blue is misconverged to the right side of the white at the bottom, where at the top you can see it is misconverged to the left of the white (while the red does the exact opposite). Very easy to see whats happening if you look at the first white vertical line just to the right of the red/white joined line. I have some similar issues on my 36" with what is called X-cross, or vertical convergence keystoning from side to side. I was able to get it a bit better by adjusting the XCV reactor on my set.

Adjustments for this are as follows:

1.) Use the YBWU and YBWL bus service mode registers on units that have these registers. Many units do not have these registers.

2.)Tilt the deflection yoke up and down

3.)Adjust the YCH Reactor (should be one of the inductive reactors on either the yoke or neckboard.

4.) You can also try to barely move the H-stat adjustment. You might be able to get the top and bottom looking better, but it will also affect the center so you need to find the best balance.

**Edit It also looks like you have some vertical misconvergence at the top left as well -I can see the blue on top of the horizontal white line and red on the bottom of that line. Does it cross to the opposite way at the other side of the screen? If so, you also have some X-cross, just as I did. Be warned that trying to get the "perfect" convergence can be a VERY time consuming and painstaking process. Adjusting pots and reactors is easy enough, just make sure to count the turns and directions of the adjustments you make. Its very easy to fix a problem on one part of the screen only to later realize your adjustment caused a new problem on another part of the screen. You likely may be able to improve what you have, but at some point you may have to call it "good enough".

I would start with a very fine tweak of the H-STAT pot and be sure to scan all areas of the screen after each adjustment. Once you start moving that fucking deflection yoke shit gets real. Thats usually my last resort and I almost always avoid it if possible, as its rare for it to move from the factory adjustment, which is usually close to as good as it gets....usually.
Cavia
Posts: 10
Joined: Thu May 07, 2015 5:58 pm

Re: Geometry calibration - red and white grid

Post by Cavia »

Josh128 wrote:Whats happening in this pic is that your horizontal convergence is "keystoning" from the bottom of the screen to the top, also called C-cross if it only occurs at the sides or Y-cross if it happens in the center. If you look close you will notice that the blue is misconverged to the right side of the white at the bottom, where at the top you can see it is misconverged to the left of the white (while the red does the exact opposite). Very easy to see whats happening if you look at the first white vertical line just to the right of the red/white joined line. I have some similar issues on my 36" with what is called X-cross, or vertical convergence keystoning from side to side. I was able to get it a bit better by adjusting the XCV reactor on my set.

Adjustments for this are as follows:

1.) Use the YBWU and YBWL bus service mode registers on units that have these registers. Many units do not have these registers.

2.)Tilt the deflection yoke up and down

3.)Adjust the YCH Reactor (should be one of the inductive reactors on either the yoke or neckboard.

4.) You can also try to barely move the H-stat adjustment. You might be able to get the top and bottom looking better, but it will also affect the center so you need to find the best balance.

**Edit It also looks like you have some vertical misconvergence at the top left as well -I can see the blue on top of the horizontal white line and red on the bottom of that line. Does it cross to the opposite way at the other side of the screen? If so, you also have some X-cross, just as I did. Be warned that trying to get the "perfect" convergence can be a VERY time consuming and painstaking process. Adjusting pots and reactors is easy enough, just make sure to count the turns and directions of the adjustments you make. Its very easy to fix a problem on one part of the screen only to later realize your adjustment caused a new problem on another part of the screen. You likely may be able to improve what you have, but at some point you may have to call it "good enough".

I would start with a very fine tweak of the H-STAT pot and be sure to scan all areas of the screen after each adjustment. Once you start moving that fucking deflection yoke shit gets real. Thats usually my last resort and I almost always avoid it if possible, as its rare for it to move from the factory adjustment, which is usually close to as good as it gets....usually.
Thanks a lot for your explanation Josh. That was very helpful.
Adjusting the static convergence (H-STAT or V-STAT) will not do anything major at the borders without screwing the center (since the static convergence targets explicitly the center of the screen). For border convergence there is only some slight yoke tilt + magnets (external or strips). But I'm not worried a lot about borders convergence. What I'm really trying to solve is the C/Y-cross problem you mentioned.
I'll check if the PVM2042QM has the reactors of the registers you suggested, cause I'm also trying to keep the yoke as a last resort.

Thanks.
User avatar
Josh128
Posts: 2124
Joined: Thu Jan 16, 2014 9:01 am

Re: Geometry calibration - red and white grid

Post by Josh128 »

The only reason I mentioned the H-Stat pot (and put it at the bottom of the list) is because its so easy to try. Adjusting it definitely does have some effect on the sides as well , but yeah its not going to affect the Y-cross as far as fixing it in the purest sense. Just something to try, you may find you can help minimize the visible cross at the edges at the expense of some very slight misconvergence in the center. I would only make very slight adjustments and you will quickly see if it can help or just make things worse.

Good luck and post back, with comparison pics if possible!
Cavia
Posts: 10
Joined: Thu May 07, 2015 5:58 pm

Re: Geometry calibration - red and white grid

Post by Cavia »

Josh128 wrote:The only reason I mentioned the H-Stat pot (and put it at the bottom of the list) is because its so easy to try. Adjusting it definitely does have some effect on the sides as well , but yeah its not going to affect the Y-cross as far as fixing it in the purest sense. Just something to try, you may find you can help minimize the visible cross at the edges at the expense of some very slight misconvergence in the center. I would only make very slight adjustments and you will quickly see if it can help or just make things worse.

Good luck and post back, with comparison pics if possible!
Totally agree.
Unfortunately, the PVM 2042QM doesn't have any advanced convergence controls (no coils around the yoke's neck, or even in the service manual), or at least I wasn't able to find them. Probably because it's one of the earliest models. Apparently, another manual mentions that the converge adjustments on YCH XCV TLV TLH YBH, are in fact to compensate the imperfections derived from the yoke's adjustments.
https://ia600600.us.archive.org/1/items ... Manual.pdf Page 22.
So I'm not going to mess with the yoke which is anchored with a ton of apoxie anyway.
User avatar
Josh128
Posts: 2124
Joined: Thu Jan 16, 2014 9:01 am

Re: Geometry calibration - red and white grid

Post by Josh128 »

OP, any luck so far?
Cavia
Posts: 10
Joined: Thu May 07, 2015 5:58 pm

Re: Geometry calibration - red and white grid

Post by Cavia »

Josh128 wrote:OP, any luck so far?
As you can see from the reply above, there is not much I can do on that monitor.
User avatar
Josh128
Posts: 2124
Joined: Thu Jan 16, 2014 9:01 am

Re: Geometry calibration - red and white grid

Post by Josh128 »

Ah, thats a shame. But your pics really dont look that bad anyway IMO. Just enjoy the set. I think you will find very, very few CRTs capable of 15KHz 240p that can attain near perfection in regards to convergence and purity. Even those that seem to come close, there is usually always some small issue. Nature of the tech. If each of us had our brand new childhood CRTs that we played 8 bit and 16 bit era games on, today, and checked all this stuff on them, we would probably be shocked at how off they were. Back then, we didnt have a clue, just enjoyed the games. Simpler times, lol.
Last edited by Josh128 on Tue Jan 19, 2021 9:52 pm, edited 1 time in total.
Cavia
Posts: 10
Joined: Thu May 07, 2015 5:58 pm

Re: Geometry calibration - red and white grid

Post by Cavia »

Josh128 wrote:Ah, thats a shame. But your pics really dont that bad anyway IMO. Just enjoy the set. I think you will find very, very few CRTs capable of 15KHz 240p that can attain near perfection in regards to convergence and purity. Even those that seem to come close, there is usually always some small issue. Nature of the tech. If each of us had our brand new childhood CRTs that we played 8 bit and 16 bit era games on, today, and checked all this stuff on them, we would probably be shocked at how off they were. Back then, we didnt have a clue, just enjoyed the games. Simpler times, lol.
Amen.
Post Reply