Vertical Lines on Screen in 2 Different MSX2 HB-F9s model!

Página 1/3
| 2 | 3

Por Dhampird

Hero (589)

Imagen del Dhampird

13-10-2010, 03:34

Hello, in another post earlier I had commented on about a strange effect that I saw on the monitor screen with a MSX2 HB-F9S,which I made some updates on it including VDP9958 update, this vertical lines were not very visible but if i adjust the monitor brightness and contrast this vertical lines accentuated more, first of all i blamed the RGB cable that follows the same RGB - SCART cable pattern showed in http://www.faq.msxnet.org/connector.html

img547.imageshack.us/img547/3814/rgbs.jpg

In my RGB cable the only difference is that when i opened the SCART case i saw three 10V 100uF capacitors conected to R,G,B lines in the SCART conector, i removed that capacitors but the vertical lines still appear in screen.

Well, i get another HB-F9s without scratches in the case and best state than the older, this new HB-F9s has no mods, all components are original, i noticed that the same vertical lines appear like in the older HB-F9s with mods and V9958, the new have V9938 on it. These are the Vertical Lines Effect:

Vertical Lines in HB-F9s with V9958 and another mods:

img222.imageshack.us/img222/8648/verticallines.jpg

Vertical Lines in HB-F9s with V9938 and original without mods:

img63.imageshack.us/img63/3778/verticallines2.jpg

To get that images i had to play with monitor controls brightness and contrast and after that i touch a little in PC to that v.lines be more visible here, without this adjustments or using the MSX monitor normally configured the vertical lines pass quite unnoticed to the eye.

The strange thing is that in two same MSX2 HB-F9s models the same effect happens, the vertical lines shape are a little different in each MSX2 model. I tried the same DIN-RGB cable in Turbo-R A1-GT and the image is perfect, without any vertical lines, then i suppose that the cable is OK (all test without the 3 capacitors in SCART RGB lines), anyways i´m wondering if that vertical effect could be corrected with any Resistor conected to the RGB cable, i saw another MSX RGB cables diagrams that uses some Resistors, i post this with the hope if anyone with a MSX2 HB-F9s or similar have the same effect problem, will be great to know if some MSX2 HB-F9s not have the problem and how is wired their RGB cables, you know, will be great to avoid the problem modifing only the RGB cable.

Anyways, like a expert user of the forum told me, the problem could be due to:

"Power-supply related, cabling. This is one of the things a filter on VDP's (DAC) power might fix, or cause when missing. Also IMHO cables should be wires only, if extra components are needed they should be in equipment on either side, not in the cable..."

That makes me wonder if a 100nF Capacitor (or higher) between VDP PIN 20 (GND/DAC) and VDP PIN 21 (VDD/DAC) could be a possible solution.

If anyone have a HB-F9s please comment if you see this vertical lines or not.

Thanks a lot for read.
Regards.

Login sesión o register para postear comentarios

Por RetroTechie

Paragon (1563)

Imagen del RetroTechie

13-10-2010, 12:52

Can't say I ever noticed similar disturbances with my HB-F9P.

Including resistor(s) in RGB cable won't do much good (btw has long is that cable?). If Turbo-R has good image with same cable, and both HB-F9S show same pattern, then maybe it's result from something they have in common? Schematic says VDP's power supply is filtered through L6 (22 uH) and C77 (470 uF). So you might try: short L6 (sometimes a coil in power line causes distortion rather than prevent it), or replace C77, or put 100 nF ceramic across C77. If that doesn't make a difference: I see that plain +5V (which might include digital switching noise) is used as supply in transistor output stages, possibly that causes this issue? Unsure about easy way to filter that supply, it would depend on circuit board layout & how much current those output transistors are using (read: trial & error, difficult to figure out) ...

Maybe you just have a noisy mains supply (230V) & there isn't much you can do about it. Eek!

Por pitpan

Prophet (3152)

Imagen del pitpan

13-10-2010, 13:32

I'm not much of a harware geek, but you didn't point out the following:

- Are all the wires properly shielded?
- Are both connectors properly grounded?

I know they are very basic ideas, but I've experienced a similar problem with a different cable type and it was related to interference patterns that got solved by following these two tips.

Por RetroTechie

Paragon (1563)

Imagen del RetroTechie

13-10-2010, 13:47

Read: make sure that cable shielding (you are using a cable with shielding, are you?) is connected to GND point on at least 1 end of the cable (=so that it may act as shielding properly). Or on both ends, in which case it would also act as return wire for signals (which may or may not be a good thing).

Higher-frequency/long cables like for VGA might use individual shielded wires for R/G/B, but that's really not needed to get good image quality in our case.

Por flyguille

Prophet (3028)

Imagen del flyguille

13-10-2010, 20:09

those vertical lines are not from MSX -> TV connector, they are from a bad filtered PSU unit at MSX.

if you looks, its frequency is of a period of like 12 dots wide? and in a horizontal tracing period (+-16k (to be round)) fits 22 cycles of that interference.

so, its source is running at 16k*22 = 352khz....... and that is a sub of some a digital circuit working.... all digital noice is feeding by PSU to VDP analogues outputs.

if you activated some heavy work, like FDD acces, you will see the interference grown in its intensity.

all that, is because

msx PSU are most, standard regulators PSU (no switching PSU in msx)
filters on its PSU are getting older and are not LSR type like PCs has.
anyway power lines in msx boards are not the best on most models
worst, analogue circuits power supply is not separated from digital PSU lines.
desing desing desing tips...

wait, I remember the same interference coming from my msx2 system, a DAEWO like msx....

so it is common.

Por Dhampird

Hero (589)

Imagen del Dhampird

14-10-2010, 02:15

Hello and thanks for replies...About the cable:

Including resistor(s) in RGB cable won't do much good (btw has long is that cable?)

The DIN-SCART cable is 1 meter and 20 cm long. I have here another cable from a SATURN console, is 3 meter long, the cables are more fat, also ground is more fat, no espiral mesh around it, and include circular magnet at each end. I´ll build this to compare result with the other, but 3m is long i think, probably need resistors or capacitors, i´ll try first without it and after i´ll comment results.

- Are all the wires properly shielded?
...make sure that cable shielding (you are using a cable with shielding, are you?)

Yes, the wires seem good shielded, the Ground is nude along the cable and also have a espiral mesh around it (like antena cable type but more thin), perfect soldered points and insulation for each point with heat shrink tube, the only thing that i dont like is that the cables are too thin, i think, and another point is that at each end there is NO the typical magnet like in another RGB types.

- Are both connectors properly grounded?
...make sure that cable shielding is connected to GND point on at least 1 end of the cable (=so that it may act as shielding properly). Or on both ends, in which case it would also act as return wire for signals (which may or may not be a good thing).

Yes, grounded in DIN conector to the circular metalic case and also grounded in the RGB SCART metallic case, about if the problem will be a possible return i could try disconect one GROUND point in the cable, but in which point will be better do that? in the SCART metallic case? and leave conected the ground into the DIN metallic case?

About internal fix in HB-F9s:

So you might try: short L6 (sometimes a coil in power line causes distortion rather than prevent it)

I removed the L6 and rejoin its points to let the 5V pass through where the L6 is placed, no changes in vertical lines... same effect again.

or replace C77, or put 100 nF ceramic across C77

I placed again the L6 and i dont have here anyone C77 (470 uF) but i tried install a 100 nF ELECTROLITIC across C77 and the vertical lines effect changes a lot, i can say that the vertical lines have been reduced about a half, i have to get a ceramic 100 nF and try again, also i´ll change the C77 (470 uF) with another new, i have to buy one, i have here a new LT1084CP-5 i´m wonder if will be a good idea change it for the original or will help with the v.lines effect.
Well, then the effect is reduce now a half, Big smile , then probably this could be a hint to complete fix the problem... Anyways i´m wondering the effect to install 100nF Capacitor (or higher) between VDP PIN 20 (GND/DAC) and VDP PIN 21 (VDD/DAC)

Here is where i placed the 100 nF ELECTROLITIC capacitor with correct polarity (placed in red):

img252.imageshack.us/img252/2438/vdpview.jpg

Thanks a lot for the fully help, again.

I´ll try with the ceramic capacitor and another cable....

Por AuroraMSX

Paragon (1902)

Imagen del AuroraMSX

14-10-2010, 22:39

Your monitor isn't a Trinitron one by any chance?

Por Dhampird

Hero (589)

Imagen del Dhampird

15-10-2010, 05:24

Hello, thanks "flyguille" for info about PSU, anyways i could reduce (only half but now the lines are not much visible but i could see it because i know that there are) the v.lines effect like Retrotechie saids, today i change the electrolitic capacitor for one ceramic, also i changed the C77 (470 uF)- original 6.3V, i changed with one 16V 470 uF, the v. lines effect have not improved since the change, but at least now the effect has been reduced a lot. Ok, i´m wonder if i could reduce more the effect if i install 100nF Capacitor (or higher) between VDP PIN 20 (GND/DAC) and VDP PIN 21 (VDD/DAC) or try another thing...

Your monitor isn't a Trinitron one by any chance?

No, my monitor is a PHILIPS CM 8802, but i´m planning to get a Sony (Triniton) KX-14CP1 nearly...why you asked about Sony Triniton monitor? Trinitron monitor is more prone to these failures?, better?worse?

Thanks a lot for info and help.

Por Dhampird

Hero (589)

Imagen del Dhampird

14-02-2013, 03:37

This is an old theme, but the other day i get a HB-F700S, the image on screen appear very clean, no shadow effects even after install a new VDP9958, perfect image. Then, i remembered that maybe the HB-F9S ghost effect on screen could due to a bad filtered PSU or mobo design, then the idea now will be try use the HB-F700S PSU signals: +5V, +12V, 0V, -12V instead the HB-F9S original voltages (all this without removing any original HB-F9S PSU parts or components, only obviosly without connecting the original source to 220V because i´ll trying use external voltage signals) only for test purpose. The idea then will be mantain 2 PSU´s, HB-F9S original and new one external PSU (using HB-F700S PSU for now for test)

Follow HB-F9S datasheet i located voltage points, then i ´ll need to know if i place HB-F700 voltage points like diagram shows, and if the original HB-F9S PSU is not conected, if this mod will work or not, or if will need cut some points on board or another extra modifications, here the diagram i made to hook up new external HB-F700 PSU signals:

Por Daemos

Paragon (1978)

Imagen del Daemos

14-02-2013, 14:14

352khz. Is the frequency that causes the problem so thats what you could try to filter out of the PSU.

Try to build a filter as close to the the 5 V powersupply of the VDP as possible. Your cutoff frequency is very high so you should use 5nF together with 100 ohms to ground. Try ceramic or if you have money silver mica.

You propably need to add more filters in series if your problem doesn' t go away 100%

Por Jipe

Paragon (1552)

Imagen del Jipe

14-02-2013, 15:04

for using HB700 PSU on the HBF9 you must remove
IC1 SI-3502V REG
IC3 UPC78M12H REG
IC2 NJM79L12A REG

connect
+5V to pin 2 of IC1
+12V to pin 1 of IC3
-12V to pin 3 of IC2
0V ground to pin 1 of IC1

Página 1/3
| 2 | 3