Undesired Jail bars on French NES and checkerboard pattern

Discuss technical or other issues relating to programming the Nintendo Entertainment System, Famicom, or compatible systems. See the NESdev wiki for more information.

Moderator: Moderators

alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

The BC327-25 has a typical gain of 250; in contrast the original 2sa937 ... oh, it's also binned. The one in my NES says "A937E L" but the datasheet I found doesn't mention "E" or "L" as a binning label (only P Q R or S)... eh, that's probably fine.
Lol I hope I did not damage anything! I will put back the other one.

The jailbars are due to the digital activity that the PPU is doing (fetching values from its own memory at specific times). But ... because this is a PAL set, the checkerboard pattern should be consistent from vsync to vsync, so I don't know why it'd drift ... unless it's actually noise from the crystal inside the metal box?
I also replaced that crystal with a brand new one, but saw no improvement. Maybe I need to adjust the trimmer?

You could just try swapping the metal boxes between the two NESes.
Yes, but I think this would be more difficult; it is such a pain to desolder those boxes. Besides, I had to repair the other NES because when I desoldered this metal box for replacing the capacitors inside it, I accidentally broke some of the connections between the box and the main board (it was a mess), but since I repaired it and is working fine now, I think it would be too much of a hassle to desolder the box from the other nes!
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by lidnariq »

alfredocalza wrote: Mon Dec 16, 2019 12:15 am Lol I hope I did not damage anything [by swapping the output transistor]! I will put back the other one.
It certainly won't lead to damage. I'd guess it might possibly get darker, or get worse color rendition, but the output structure of the PPU just isn't a shape that would be damaged by drawing too much power out of it.
I also replaced that crystal with a brand new one, but saw no improvement. Maybe I need to adjust the trimmer?
I don't think so, but it can't hurt.
I accidentally broke some of the connections between the box and the main board (it was a mess),
... this is probably a bad time to point out the advice is to desolder the pins inside the metal box instead of off the NES mainboard...
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

Ok, so I recently got the CXA1621S ICs that I order. They seemed to be new old stock since they were in excellent condition. I also got a couple of V7021 which have clearly been used (no wonder why the CXA1621S were twice as expensive as the V7021).

I replaced the old V7021 with a new CXA1621 and after adjusting the crystal a little bit, the picture quality was noticeably better than with the old V7021. However, the jail bars and checkerboard pattern were still there.

I have not tried this:
"The two resistors together serve to move the average voltage around. The V7021 has specific requirements on how the video enters it: namely, it must
1- be roughly 2.7V DC
2- be 0.36Vpp

If those aren't complied with, video may be off, but I wouldn't anticipate jailbars as a plausible symptom. It'll be hard to check whether the latter is right without an oscilloscope, but you could measure the voltage on V7021 pin 27 ("VIDEO IN") and see if it's roughly 2.7V."

I will try to test the voltage here to see if there is an issue.

On the other hand, I've recently installed the NESRGB module on my other console and experienced a similar checkerboard pattern when using the composite SYNC signal instead of the CSYNC signal.

Going back to this console, I wonder If by adding a SYNC stripper to the composite signal and extracting the CSYNC, I can get rid of the checkerboard pattern and maybe the jail bars?

If so, which SYNC stripper should I use? LM1881? Any ideas?
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by lidnariq »

alfredocalza wrote: Sun Jan 05, 2020 7:22 am If so, which SYNC stripper should I use? LM1881? Any ideas?
The LM1881 is just the archetype of the family. There are others, too, that may be less expensive. Quickly checking: BA7046, EL1881, LMH1980

I haven't any recent experience with any of them.
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

I just tried measuring the voltage on pin 27 of the CXA1621S ic. The voltage was 2,711 volts DC according to my multimeter, so I guess that is not the issue.

Regarding the sync separator, I think I will just buy the LM1881 which is like 3 Euros only and use the opportunity to buy other components to have around, and some soldering points (I already broke 2 of them!)

I have a question, where exactly should I pick up the video signal to feed the LM1881? Should I use the signal that comes out of the PNP transistor? or should I use the unamplified signal from the PPU? or should I use the video in signal that goes into pin 27 of the CXA1621S? or should I use the SYNC signal that comes out of pin #2 of the CXA1621S? I would like to pick up the signal as early as possible to minimize the possibility of any interference that other components on the board and other lines might be causing on the video signal line.

btw, thank you lidnariq for sharing all your knowledge, It has been really helpful!
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by lidnariq »

alfredocalza wrote: Tue Jan 07, 2020 1:09 pm I have a question, where exactly should I pick up the video signal to feed the LM1881? Should I use the signal that comes out of the PNP transistor?
Probably?
or should I use the unamplified signal from the PPU?
Probably not... Input impedance of the LM1881 is "more than 10kΩ" but the output impedance of the 2C07 is already quite high.
or should I use the video in signal that goes into pin 27 of the CXA1621S? or should I use the SYNC signal that comes out of pin #2 of the CXA1621S?
Trying those should be fine too.
I would like to pick up the signal as early as possible to minimize the possibility of any interference that other components on the board and other lines might be causing on the video signal line.
I don't think you'll have any problems because this is just sync separation. There's a lot less opportunity for analog wonkiness as a result.
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

Ok, so I got the LM1881 ICs and some other stuff. I'm trying to plan exactly what I am going to do before doing anything over the weekend. I looked at krzysiobal French NES schematics for the modulator box and noticed that the sync out signal coming out of the V7021 IC uses a transistor (my guess is that this transistor amplifies the signal), an inductor (I don't know what this does to the signal) and two resistors (to weaken the signal?).

My question is, if I use the composite video line that enters the V7021 to feed the LM1881, should I then use a transistor to amplify the sync signal that comes out of the V7021? Should I do anything else with the sync signal before connecting it to the multi-out connector?

I also noted that the R, G and B signals coming out of the V7021 are connected to the multi out via transistors. Are these transistors used to amplify the RGB signals? What if I use a THS7314 RGB amp to amplify those signals instead? Could the jail bars or checkerboard pattern issues be hidden in here? Would using an rgb amplifier instead of the transistors make the video look any better, or fix the issue?

Thanks!
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by lidnariq »

alfredocalza wrote: Wed Jan 15, 2020 4:39 pm sync out signal coming out of the V7021 IC uses a transistor (my guess is that this transistor amplifies the signal)
That is a common collector amplifier. It is a voltage buffer, or current amplifier.
an inductor (I don't know what this does to the signal) and two resistors (to weaken the signal?).
The resistors are there to reduce the magnitude of what comes out to be whatever SCART expects on its sync pin. The inductor reduces RF crosstalk.
My question is, if I use the composite video line that enters the V7021 to feed the LM1881, should I then use a transistor to amplify the sync signal that comes out of the V7021? Should I do anything else with the sync signal before connecting it to the multi-out connector?
If the LM1881 is separating out the sync signal from whatever source, why does the V7021's sync output matter any more?

If you mistyped, and meant "should you use a transistor to amplify the output of the LM1881" - I've never needed one before. Bare sync inputs are usually high impedance. But I don't know if your TV is.
I also noted that the R, G and B signals coming out of the V7021 are connected to the multi out via transistors. Are these transistors used to amplify the RGB signals?
They're the same common collector amplifier you've seen elsewhere.
What if I use a THS7314 RGB amp to amplify those signals instead?
That would work.
Could the jail bars or checkerboard pattern issues be hidden in here? Would using an rgb amplifier instead of the transistors make the video look any better, or fix the issue?
Very unlikely.


Given the number of variables here, I'd strongly recommend setting up your LM1881 on a breadboard first and playing around with values until you get something acceptable. No point in soldering ... and resoldering ... and resoldering ... just to find out what works.

I'd also strongly recommend, if you have a TV anywhere that accepts baseband PAL, seeing if the video artifacts are also present on the baseband PAL signal out of the 2C07 before it gets to the V7021.
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

Ok, thank you!. I will use a breadboard to play a little with the LM1881. I'm not going to add the RGB amp since it doesn't seem to be necessary. Unfortunately, I do not own an old TV :( so I can't test the composite signal..
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

I was reading through the LM1881 data sheet and it says it is optional to install a lowpass filter in the composite video input line before it enters the LM1881. For what I could understand, low pass filters filter out any high frequency signals. Would this help eliminate any of the undesired artifacts?

According to the data sheet, the low pass filter consists in a 620ohm resistor with a 510pf capacitor. I do not have a 620 ohm resistor nor a 510pf capacitor, however I could use 2 resistors that I do have to get to that value (500 + 120), and use two capacitors in parallel (220pf + 330pf, I know this is not quite 510pf).

Also, if I install this low pass filter, would I need to amplify the SYNC signal after it comes out of the LM1881? It makes sense to me that the 620ohm resistor would make the video input signal smaller?
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by lidnariq »

alfredocalza wrote: Sat Jan 18, 2020 3:35 am I was reading through the LM1881 data sheet and it says it is optional to install a lowpass filter in the composite video input line before it enters the LM1881. For what I could understand, low pass filters filter out any high frequency signals. Would this help eliminate any of the undesired artifacts?
Yes, and all the rest of the video too :P

In a frequency sense, a spectrogram of video will look something like this:

Code: Select all

|\
| \    |                  |
|  ----|-----\            |
|             ------------|\   F
|  A   B    C     D       E -------
+----------------------------------------> increasing frequency
the things on the graph are
A: Brightness changes vertically
B: Hsync pulse
C: Wider brightness changes horizontally
D: a mix of wider brightness changes and color information
E: Color carrier
F: a mix of narrow brightness changes and color information.

Increasing the lowpass filter knocks out F, E, D, C in order, until you get just the hsync pulses desired by the LM1881.

But it should be clear why this won't help with visual artifacts if you want to keep the video.
According to the data sheet, the low pass filter consists in a 620ohm resistor with a 510pf capacitor. I do not have a 620 ohm resistor nor a 510pf capacitor, however I could use 2 resistors that I do have to get to that value (500 + 120), and use two capacitors in parallel (220pf + 330pf, I know this is not quite 510pf).
The important part is 620Ω · 510pF = 320ns. You should be able to pick other choices of R and C that are comparable.
Also, if I install this low pass filter, would I need to amplify the SYNC signal after it comes out of the LM1881? It makes sense to me that the 620ohm resistor would make the video input signal smaller?
Except that the purpose of the LM1881 is to take the analog input and generate a digital output. It'll be fine.
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

Hi Lindariq, this is very interesting and helpful information, thank you so much for sharing!. Anyway, yesterday I did not have enough time to build the low pass filter, so I just built the sync stripper on a breadboard without it (it is not necessary anyway). At first, I sourced the video in signal from the same line that goes into the CXA1621S. This did not work; I got no image and no sound when I turned on the NES. I tried multiple times after cleaning the cartridge and re-inserting it, but it just did not work. I don't know if the video in signal was too weak?

The second time I tried using the composite line that goes into the rf modulator; right before the large metal box and right after the 330 pf capacitor. The first time I tried this, I got some horrible green jail bars:
Green Jail bars
Green Jail bars
At this moment I thought I had fucked up my NES, but I took out the cartridge, cleaned it again and re-insterted it, and got the same old checkerboard pattern and same old jail bars. I think that the sync stripper inside the CXA1621 was definitely not the issue.
LM1881 SYNC stripper - French NES
LM1881 SYNC stripper - French NES
I think it is time to rule out any components inside the rf modulator box as the source of the video artifacts. Going back to the PPU, is there anything else I could try? Also, Is there a way of knowing if the composite video line that comes out of the PPU is picking up any noise from anywhere else?
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by lidnariq »

alfredocalza wrote: Mon Jan 20, 2020 12:46 am Also, Is there a way of knowing if the composite video line that comes out of the PPU is picking up any noise from anywhere else?
Not without better diagnostic tools than you have. A TV that accepts baseband PAL, an oscilloscope, ... maybe an AM or SW radio, but I'm not certain about that last one.


Did you say you'd already swapped the metal boxes between the two NESes?
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

lidnariq wrote: Mon Jan 20, 2020 12:55 am
alfredocalza wrote: Mon Jan 20, 2020 12:46 am Also, Is there a way of knowing if the composite video line that comes out of the PPU is picking up any noise from anywhere else?
Not without better diagnostic tools than you have. A TV that accepts baseband PAL, an oscilloscope, ... maybe an AM or SW radio, but I'm not certain about that last one.


Did you say you'd already swapped the metal boxes between the two NESes?
No, I have not done that. My other NES has the RGB mod and I would hate to take it apart since I have taken good care of it, except for when I took the modulator box out of that NES to replace the caps inside it, I broke some of the lines that go from the main board into the box. I had to repair it by soldering some cables on the lower side of the board, and would not like to take that apart either.

Unfortunately, I do not have an oscilloscope. Are they too expensive?

I found this guy online who managed to solve the jailbar issue on the famicom by relocating the video line that goes from the PPU into the transistor and adding some 1,5uf ceramic caps instead of the standard 0,01uf caps on the board.

http://vaot.mydns.jp/fc/noise5.htm

http://vaot.mydns.jp/fc/noise4.htm

Do you think that doing something like this on my french NES is worth a shot? I already tried the tantalum caps in the range of 10 to 45 uf on the 5v input of the PPU and that did nothing on the jailbars.
alfredocalza
Posts: 111
Joined: Sat Jun 22, 2019 1:06 pm
Location: Madrid, Spain

Re: Undesired Jail bars on French NES and checkerboard pattern

Post by alfredocalza »

Also, I do not have any 1.5uf ceramic caps; I only have 1.0uf caps. Maybe I can use two of these?
Post Reply