"Jail bars" in RGB output of Philips NMS 8250

By triskalguilo

Resident (40)

triskalguilo's picture

17-02-2017, 01:03

Hello all, 'tis I again! Smile

I have a Philips NMS 8250, and the RGB video output of it has "jail bars" in it: vertical lines that are particularly visible in areas of solid color. The lines are typically slightly brighter than the colors around them. For example, in MSX-BASIC, there are light blue vertical lines across the screen. Their position across the screen is very regular -- every few pixels -- not at all random.

My Sony HB-F700D, when connected to exactly the same setup, does not have this problem -- so I don't think my setup is the problem.

Anyone know what causes this, and how to fix it? Maybe I need to replace some capacitors or resistors?

Login or register to post comments

By Wolverine_nl

Paladin (699)

Wolverine_nl's picture

17-02-2017, 10:12

Hello triskalguilo,
I have both machines too, but I do not have this problem. It could be also a monitor thing.
So you use for both the same monitor input?

By hamlet

Paragon (1227)

hamlet's picture

17-02-2017, 11:54

Could it be a bad RGB cable?

By Wolverine_nl

Paladin (699)

Wolverine_nl's picture

17-02-2017, 12:30

I know there is a slight signal difference between the HB-F700 and the NMS 8250, but my monitor corrects it. Cannot find it here, but I know there is. But it should not give bars in the screen, it flickrs the screen slighty

By Wolverine_nl

Paladin (699)

Wolverine_nl's picture

17-02-2017, 12:31

the RGB video out of the HB-F700 has the following: 0-0.7V, 75 ohms
That of the NMS 8250, I cannot find that easily.

By RetroTechie

Paragon (1414)

RetroTechie's picture

17-02-2017, 14:08

Looks like a power supply problem... If original electrolytic capacitors are still in there: replace the ones on the power supply board with same-value (uF) and same or higher voltage rating (V). Do the same on mainboard with capacitors near the point of entry of the power cable (see schematic & board itself for which ones). Probably it wouldn't hurt to replace a few caps elsewhere on mainboard or analog board, but often you can leave those alone.

A quick look at schematic...

C1050, C1051, C1052, C1053, C1054 (power supply board)
C101, C102, C103 (near keyboard connector)
C106 (near the VDP)
C115 (near main RAM / VRAM chips)
C148 (next to Z80)
C171 (next to cartridge connector)
On analog board: C339, C341 (next to power connector), C325 (near MC1377 PAL encoder), C355 (next to RF modulator)

If possible: measure supply voltages on +5V, +12V and -12V lines, and write those down. After replacing those caps, repeat. Please report back with results when done? Also have a look at the clockchip battery (long yellow thing) for signs of leakage. Or remove/replace anyway.

FYI: the vertical bars would be due to ripple voltage on the supply lines getting worse because filtering caps don't do their job anymore. Regular distance is due to some variation in the machine's power consumption occurring at a regular frequency. For example by ROM IC's getting enabled/disabled for Z80 instruction fetches. Or whatever.. just occurring at a regular rate most of the time.

By Meits

Scribe (4650)

Meits's picture

17-02-2017, 14:17

In all the years this machine has existed I never saw a 8250 (and up) that didn't have this issue. It's visible on my Philips RGB screens and on my Samsung flatpanel. It's visible from an original 8250 in its original case and its own PSU but also from my 8250 in PC case with a brand new 450 watt PSU which has been tested on all the voltages.
I'm not a hardware guru and may not stand in the shadow of RetroTechie. I only conclude from personal experience that it's a typical Philips 8250 (and up) thing.

By RetroTechie

Paragon (1414)

RetroTechie's picture

17-02-2017, 14:41

Meits wrote:

In all the years this machine has existed I never saw a 8250 (and up) that didn't have this issue. It's visible on my Philips RGB screens and on my Samsung flatpanel. It's visible from an original 8250 in its original case and its own PSU but also from my 8250 in PC case with a brand new 450 watt PSU which has been tested on all the voltages.

Very possible! In that case PC psu would deliver nice smooth voltages on its output. And they get on the mainboard... and travel over a (relatively) long distance of pcb track... and then (for example) a set of DRAM chips suddenly increases its power consumption on the +5V line... causing a voltage drop on that +5V line on that section of the board. To go right back after those DRAM's are disabled again. And such quick voltage drops & bumps may be directly visible on screen.

A good power supply at point of entry can help, but not prevent the above issue. That why there are locally placed bypass capacitors spread all over the board, to counter this problem locally for each section of the board. The often used 100 nF caps next to logic IC's are the most important for that. But these age well & rarely have problems. Electrolytic caps on the other hand... Evil

I'm not even sure problem as described is caused by bad caps. But symptoms are like what you'd expect if those caps are bad. So... still original -> re-cap before investigating further. (and often that'll do the job Smile2 )

By Wolverine_nl

Paladin (699)

Wolverine_nl's picture

17-02-2017, 15:03

Strange, my 8250 was a virgin, no strange things on my flatscreen monitor (with scart), the monitor is also Philips. My Sony HB-F700 gives a small flickering screen, when I change the channel and then go back to the SCART channel, it is gone. That is the only difference I see.

By triskalguilo

Resident (40)

triskalguilo's picture

18-02-2017, 18:56

Wolverine_nl wrote:

Hello triskalguilo,
I have both machines too, but I do not have this problem. It could be also a monitor thing.
So you use for both the same monitor input?

I do indeed. It could indeed be the DVDO Edge, since although it is broadly compatible, it probably wasn't designed for this particular case Smile

hamlet wrote:

Could it be a bad RGB cable?

I don't think so, because I'm just using an ordinary male-to-male SCART cable, so there's not a lot that can go wrong with it -- just some connectors and some wire Smile And the connectors at the end don't unscrew and open, so I doubt that there's anything weird in them, like extra capacitors or resistors.

----------

Thanks for the wealth of helpful information from everyone else Big smile If I have the time and expenses, I will certainly try re-capping things. But in light of what Meits said, I'm not sure it's worth it...

I did connect the computer to another flat panel monitor, and the jailbars were visible there too, albeit not quite as pronounced. So it could be that that monitor is better at filtering out the jailbars, as Wolverine_nl mentioned regarding his/her monitors. I only have one other monitor that is 15khz capable -- my Amiga 1080 -- so I may try connecting it to that, and see if (and to what degree) I still get jailbars.

That reminds me, though -- I did do all of the audio mods as recommended (i.e. replacing the incorrect resistors with capacitors, inverting the one capacitor, etc.), and I do not remember if I got the jailbars before that -- surely those changes couldn't have caused the jailbars, could they?

My MSX profile