Forums

Hello everybody. I know there are some knowledgeable people here and I hope you guys can help, since I’ve been looking at service manuals and technical info for a few days now and I can’t get my head around it.

Ok, so I have a potential fault with an SMS II that to be honest, I’m not entirely sure is a fault. Let me explain: For most of my games, the console plays them just fine. Some games that don’t seem affected by this problem at all include Golden Axe, Fantasy Zone, Space Harrier and Outrun.

But there are other games in my collection that exhibit strange behaviour when drawing sprites to the screen. It’s almost like a ghosting effect. As if the console is drawing parts of the sprite to the wrong location of the screen. It doesn’t break up the sprite entirely, just makes chunks of the sprite appear where they shouldn’t. Like it’s not mapping the sprites correctly. As if it's not drawing the screen quickly enough...

The problem seems to occur most in games that push a lot moving sprites (particularly if they appear on the same horizontal plane). Bubble Bobble is the worst case and borderline unplayable, but it’s a distraction in R-Type, Sagaia, Fantasy Zone II, Paper Boy and a few others.

It’s not my TV, because I tested the console on a small B/W TV I own and the problem remains. It’s not the cap in the cart because I replaced one in the hope it might fix the problem and it didn’t. I’ve opened my machine up (UK model) and can see nothing visually wrong (no bulging caps, and I did some re-flow work on any iffy soldering). I’m not technical, but could it be a problem with the VDP? The RAM chips? Bad Z80? Could the voltage regulator be dying and feeding too much power to the chips, causing the graphics to go screwy? (I plan to order a new one of these anyway, because it looks like it’s had a hard life!).

Really, I’m at a loss. I have decent bodging/soldering skills but no multimeter or oscilloscope, so I’m just hoping somebody else knows exactly what this problem is and how to fix it (I fixed a ZX Spectrum with a RAM issue because fault finding documentation out there for that computer is amazing, but the official SMS service manual seems pretty vague to me, at least to a lay person).

My Master System has had this display problem for as long as I can remember, so it’s not a new issue (though it may be getting worse). Anyway, I’d really like to fix this issue if possible. It’s not a big deal in most cases, just a minor irritation, but it would be nice to get the Master System I’ve owned since 92 finally working the way it should. I’m hoping from my description that someone out there can easily diagnose this problem, but I have a bad feeling I’ll at least need to buy a multimeter and do some more thorough testing…

Also, I did note that using a power supply that only outputs 7 volts made the problem a little worse, which leads me to think it might be a power issue after all?

Any suggestions would be appreciated.

Quote

Also, I did note that using a power supply that only outputs 7 volts made the problem a little worse, which leads me to think it might be a power issue after all?

The input voltage might be a bit too low for SMSII to function correctly and/or the voltage after the 5V regulator is being pulled down by too much load, making the ram/vdp glitches appear.

Having a good power supply is a good start, replacing the 5V regulator is another

That's about all I can say. I've not had that problem, so I'm not sure what else it could be.

You need to post a photo of the screen. "Ghosting" is usually an analogue artifact caused by things like the cable between the SMS2 and the RF switch box.

I was just trying the 7 volt power supply to experiment and see if it made a difference. It did (with more sprite break up being the result)

But I typically use a couple of 1602-05 Sega power supplies, which seem to work just fine with my Game Gear and Model 1 Mega Drive. It also worked fine with my Mega CD before that died (though come to think about it, maybe a faulty power supply did for that, too?)

Anyway, it's rated for 10V and 1.2A, which is a little higher than the original supply (9V 500ma) but I felt it would be safe to use. Perhaps I was wrong? As far as I understand it, the console should only use what amps it requires.

Sorry I wasn't clear enough in describing the glitching as a kind of "ghosting" effect. It's nothing like the typical RF ghosting you get on old consoles like these. I've attached some images (sorry for the poor quality, my phone camera is not the best). I've circled the areas I'm talking about (in one of the images, I think part of a bubble appears near bub's head but I haven't circled it).

It doesn't look too bad as a static image, but try to imagine this kind of thing happening for every frame and you can see how it might be a bit distracting!

Edit: Oh, I should add that I've tested Bubble Bobble on my Game Gear with the converter and it plays fine. No glitches at all.

glitches.jpg (372.29 KB)

glitches2.jpg (532.5 KB)

OK, I see, that's not ghosting at all. That could be a faulty VDP, faulty VRAM, VRAM that's slightly out of timing spec, VRAM bus, the power supply...

Checking the power to each chip by measuring the voltage is always an early troubleshooting step. Here in Australia, you can buy a cheap multimeter from most electronics stores for a measly $10. Do you have the same in the UK?

Thanks for your suggestions on where to start some fault finding. Sadly, we don't have any specialist electronics shops since Maplins went bust recently, but they are easy enough to find online.

I'm a little bit apprehensive about using a multimeter, to be honest. But I guess I have no choice but to learn how to use one now. Hope I don't fry something...

BTW, If it does happen to be the VDP, that's a proprietary Sega component right? There are no readily available replacement parts or a fan developed FPGA clone to replace it with? Not that I'd be up for replacing it anyway. That thing looks like a nightmare to unsolder...

Replacing the VRAM doesn't look too tricky, though. Is it still easy to find replacement VRAM that I can just drop in?

Still, I do wonder why this issue affects some games but not others...

It seems like a VRAM problem. There's probably an issue writing to VRAM, it could be just slightly out of timing spec, who knows. Every game has it's own way of conducting when it writes to vram some are strict on when it happens while others may load objects into vram whenever available, so it would make sense why it's fine for some games but not for others. You should Still be able to find RAM chips within their spec, but you should investigate traces between the system bus and the VDP as well as traces between the VRAM and VDP first.

Quote

I'm a little bit apprehensive about using a multimeter, to be honest. But I guess I have no choice but to learn how to use one now. Hope I don't fry something...

Don't worry, it's easy. You won't fry anything. I'll tell you how to do it.

Quote

BTW, If it does happen to be the VDP, that's a proprietary Sega component right? There are no readily available replacement parts or a fan developed FPGA clone to replace it with? Not that I'd be up for replacing it anyway. That thing looks like a nightmare to unsolder...

The VRAMs are also a nightmare to unsolder without the right equipment. But before we contemplate replacing chips, there's a few simpler possible problems to rule out and these are not too hard to fix.

Quote

The VRAMs are also a nightmare to unsolder without the right equipment.

Tell me about it! fixing the RAM on a ZX Spectrum was a pretty traumatic experience. The solder pads on that thing would lift if you just looked at it funny. It has a really, really horrible quality PCB. So I think swapping VRAM on a Master System would be a cinch by comparison.

Anyway, I think I'm a little bit closer to figuring out what's wrong.

Ok, so I had a hunch that it might be a problem with the oscillator and I read somewhere on the Internet that it's possible to temporarily "fix" a bad crystal by heating it up with a hair dryer.

So that's what I did and incredibly, it worked! The glitches cleared up completely! Of course, they returned again after the crystal cooled. So this leads me to make an (un)educated guess that something's gone wrong in the Oscillator circuit. I can't discount the possibility that there might be an issue with power being supplied to the circuit, or that it's bad caps causing the issue, not until I get a multimeter.

But if it is the crystal, then it's bad news. There seems to be no easy way to source a replacement. I've looked everywhere I can think of and they don't manufacture this crystal anymore. So how to proceed? The easiest way is to gut another Master System, but I'd hate to do this, especially if it's still working. And there's no guarantee that crystal won't also go bad soon (even the newest Master System is over 20 years old now).

Any ideas?

For oscillators, you can find them. It's just, do you want to gamble on it being the correct item from aliexpress. Or you can get an adjustable oscillator and preset it, but the cheap ones aren't too reliable on staying that frequency. Another option is to source one running at the NTSC compatible frequency and just have an NTSC system.

Hope you don't have to go that far and that it's really just an easier fix. Good luck.

I'm surprised the crystal is far enough out of spec to impact the VRAM access but not far enough to impact the TV screen or colours.

I would for sure try replacing the crystal even if you have to purchase it off aliexpress. It's a cheap and easy fix if it works. Good luck!

Just out of interest, how did you manage to heat up the crystal oscillator with a hair dryer without heating up anything else like the PCB or chips?? Because depending what else you heated up in the process may mean the fault is somewhere else and you sort of found it.

I, Uh, didn't... Ha ha. Everything got pretty hot (but the oscillator chip heated up the most).

At this point, all I know for sure is that it's something on the bottom right of the motherboard. I guess I could heat everything up and try spraying individual components with a quick freeze spray to see if I could diagnose the problem that way (bit crude, but it might work). That stuff's expensive, though.

Anyway, I re-flowed all the components on that side of the board, just in case it was a faulty connection being reconnected by the board flexing with the heat but no joy.

I also found a cheap multimeter and did some continuity checks and tested the voltage regulator (pretty close to 5V, so it can't be that). I just can't see what else it could be...

And I'm not sure how to go about doing further testing of voltage around the rest of the board. I'll need to do some more research.

At this point, it might be sensible just to buy a new Master System and put my old one out to pasture. Well, at least until I can find a cheap oscillator or a faulty SMS II to gut (I know you guys said they were out there but really, they're not. Not anywhere I've looked... Closest I found in stock was 50mhz)

Anyway, thanks for the help. On the plus side, I managed to test my Mega CD with the multimeter to determine that the fuse had blown, so it's not all bad!

Oscillators @aliexpress

Like I said, they're out there supposedly its just if you want to gamble aliexpress and hope you get the correct part.

IllusionOfMana, your link is breaking the forum layout.

thanks for letting me know, I fixed it.

Quote

I, Uh, didn't... Ha ha. Everything got pretty hot (but the oscillator chip heated up the most).

Haha, that's what I thought when I went away and thought just how would that work with a hair drier.

When you do get around fixing it some time in the future, we'll do what we can to help you fix it.

I had a thought about what you said about the hair dryer and tried heating the board pretty much anywhere and sure enough, that also "fixed" my problem.

My best guess now is that there's a bad solder joint (maybe somewhere on one of the VRAM chips) that expands and gets reconnected as the board heats up. But try as I might, I just can't seem to find it. All pins seem to test fine for continuity.

I'm going to try to re-flow the joints on the top of the board and see if that works. (which is a bit tricky with all the components in the way).

It's so frustrating to feel so close, yet still so far from fixing the problem. But I've not given up yet.

I'll let you guys know if it works.

Author	Message
aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Help diagnosing a PAL UK SMS II video issue... Posted: Mon Nov 12, 2018 11:30 pm
aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Hello everybody. I know there are some knowledgeable people here and I hope you guys can help, since I’ve been looking at service manuals and technical info for a few days now and I can’t get my head around it. Ok, so I have a potential fault with an SMS II that to be honest, I’m not entirely sure is a fault. Let me explain: For most of my games, the console plays them just fine. Some games that don’t seem affected by this problem at all include Golden Axe, Fantasy Zone, Space Harrier and Outrun. But there are other games in my collection that exhibit strange behaviour when drawing sprites to the screen. It’s almost like a ghosting effect. As if the console is drawing parts of the sprite to the wrong location of the screen. It doesn’t break up the sprite entirely, just makes chunks of the sprite appear where they shouldn’t. Like it’s not mapping the sprites correctly. As if it's not drawing the screen quickly enough... The problem seems to occur most in games that push a lot moving sprites (particularly if they appear on the same horizontal plane). Bubble Bobble is the worst case and borderline unplayable, but it’s a distraction in R-Type, Sagaia, Fantasy Zone II, Paper Boy and a few others. It’s not my TV, because I tested the console on a small B/W TV I own and the problem remains. It’s not the cap in the cart because I replaced one in the hope it might fix the problem and it didn’t. I’ve opened my machine up (UK model) and can see nothing visually wrong (no bulging caps, and I did some re-flow work on any iffy soldering). I’m not technical, but could it be a problem with the VDP? The RAM chips? Bad Z80? Could the voltage regulator be dying and feeding too much power to the chips, causing the graphics to go screwy? (I plan to order a new one of these anyway, because it looks like it’s had a hard life!). Really, I’m at a loss. I have decent bodging/soldering skills but no multimeter or oscilloscope, so I’m just hoping somebody else knows exactly what this problem is and how to fix it (I fixed a ZX Spectrum with a RAM issue because fault finding documentation out there for that computer is amazing, but the official SMS service manual seems pretty vague to me, at least to a lay person). My Master System has had this display problem for as long as I can remember, so it’s not a new issue (though it may be getting worse). Anyway, I’d really like to fix this issue if possible. It’s not a big deal in most cases, just a minor irritation, but it would be nice to get the Master System I’ve owned since 92 finally working the way it should. I’m hoping from my description that someone out there can easily diagnose this problem, but I have a bad feeling I’ll at least need to buy a multimeter and do some more thorough testing… Also, I did note that using a power supply that only outputs 7 volts made the problem a little worse, which leads me to think it might be a power issue after all? Any suggestions would be appreciated.

pyroesp Joined: 14 Mar 2013 Posts: 62 Location: Belgium	Posted: Tue Nov 13, 2018 11:10 am
pyroesp Joined: 14 Mar 2013 Posts: 62 Location: Belgium	Quote Also, I did note that using a power supply that only outputs 7 volts made the problem a little worse, which leads me to think it might be a power issue after all? The input voltage might be a bit too low for SMSII to function correctly and/or the voltage after the 5V regulator is being pulled down by too much load, making the ram/vdp glitches appear. Having a good power supply is a good start, replacing the 5V regulator is another That's about all I can say. I've not had that problem, so I'm not sure what else it could be.

asynchronous Joined: 14 Aug 2000 Posts: 742 Location: Adelaide, Australia	Help diagnosing a PAL UK SMS II video issue... Posted: Tue Nov 13, 2018 1:01 pm
	You need to post a photo of the screen. "Ghosting" is usually an analogue artifact caused by things like the cable between the SMS2 and the RF switch box.

aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Posted: Tue Nov 13, 2018 8:38 pm
aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	I was just trying the 7 volt power supply to experiment and see if it made a difference. It did (with more sprite break up being the result) But I typically use a couple of 1602-05 Sega power supplies, which seem to work just fine with my Game Gear and Model 1 Mega Drive. It also worked fine with my Mega CD before that died (though come to think about it, maybe a faulty power supply did for that, too?) Anyway, it's rated for 10V and 1.2A, which is a little higher than the original supply (9V 500ma) but I felt it would be safe to use. Perhaps I was wrong? As far as I understand it, the console should only use what amps it requires. Sorry I wasn't clear enough in describing the glitching as a kind of "ghosting" effect. It's nothing like the typical RF ghosting you get on old consoles like these. I've attached some images (sorry for the poor quality, my phone camera is not the best). I've circled the areas I'm talking about (in one of the images, I think part of a bubble appears near bub's head but I haven't circled it). It doesn't look too bad as a static image, but try to imagine this kind of thing happening for every frame and you can see how it might be a bit distracting! Edit: Oh, I should add that I've tested Bubble Bobble on my Game Gear with the converter and it plays fine. No glitches at all. glitches.jpg (372.29 KB) glitches2.jpg (532.5 KB)

asynchronous Joined: 14 Aug 2000 Posts: 742 Location: Adelaide, Australia	Help diagnosing a PAL UK SMS II video issue... Posted: Wed Nov 14, 2018 2:35 am
	OK, I see, that's not ghosting at all. That could be a faulty VDP, faulty VRAM, VRAM that's slightly out of timing spec, VRAM bus, the power supply... Checking the power to each chip by measuring the voltage is always an early troubleshooting step. Here in Australia, you can buy a cheap multimeter from most electronics stores for a measly $10. Do you have the same in the UK?

aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Posted: Wed Nov 14, 2018 8:47 pm
aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Thanks for your suggestions on where to start some fault finding. Sadly, we don't have any specialist electronics shops since Maplins went bust recently, but they are easy enough to find online. I'm a little bit apprehensive about using a multimeter, to be honest. But I guess I have no choice but to learn how to use one now. Hope I don't fry something... BTW, If it does happen to be the VDP, that's a proprietary Sega component right? There are no readily available replacement parts or a fan developed FPGA clone to replace it with? Not that I'd be up for replacing it anyway. That thing looks like a nightmare to unsolder... Replacing the VRAM doesn't look too tricky, though. Is it still easy to find replacement VRAM that I can just drop in? Still, I do wonder why this issue affects some games but not others...

IllusionOfMana Joined: 08 Sep 2018 Posts: 270	Posted: Wed Nov 14, 2018 9:29 pm
IllusionOfMana Joined: 08 Sep 2018 Posts: 270	It seems like a VRAM problem. There's probably an issue writing to VRAM, it could be just slightly out of timing spec, who knows. Every game has it's own way of conducting when it writes to vram some are strict on when it happens while others may load objects into vram whenever available, so it would make sense why it's fine for some games but not for others. You should Still be able to find RAM chips within their spec, but you should investigate traces between the system bus and the VDP as well as traces between the VRAM and VDP first.

asynchronous Joined: 14 Aug 2000 Posts: 742 Location: Adelaide, Australia	Help diagnosing a PAL UK SMS II video issue... Posted: Thu Nov 15, 2018 1:08 am
	Quote I'm a little bit apprehensive about using a multimeter, to be honest. But I guess I have no choice but to learn how to use one now. Hope I don't fry something... Don't worry, it's easy. You won't fry anything. I'll tell you how to do it. Quote BTW, If it does happen to be the VDP, that's a proprietary Sega component right? There are no readily available replacement parts or a fan developed FPGA clone to replace it with? Not that I'd be up for replacing it anyway. That thing looks like a nightmare to unsolder... The VRAMs are also a nightmare to unsolder without the right equipment. But before we contemplate replacing chips, there's a few simpler possible problems to rule out and these are not too hard to fix.

aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Posted: Thu Nov 15, 2018 9:43 pm
aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Quote The VRAMs are also a nightmare to unsolder without the right equipment. Tell me about it! fixing the RAM on a ZX Spectrum was a pretty traumatic experience. The solder pads on that thing would lift if you just looked at it funny. It has a really, really horrible quality PCB. So I think swapping VRAM on a Master System would be a cinch by comparison. Anyway, I think I'm a little bit closer to figuring out what's wrong. Ok, so I had a hunch that it might be a problem with the oscillator and I read somewhere on the Internet that it's possible to temporarily "fix" a bad crystal by heating it up with a hair dryer. So that's what I did and incredibly, it worked! The glitches cleared up completely! Of course, they returned again after the crystal cooled. So this leads me to make an (un)educated guess that something's gone wrong in the Oscillator circuit. I can't discount the possibility that there might be an issue with power being supplied to the circuit, or that it's bad caps causing the issue, not until I get a multimeter. But if it is the crystal, then it's bad news. There seems to be no easy way to source a replacement. I've looked everywhere I can think of and they don't manufacture this crystal anymore. So how to proceed? The easiest way is to gut another Master System, but I'd hate to do this, especially if it's still working. And there's no guarantee that crystal won't also go bad soon (even the newest Master System is over 20 years old now). Any ideas?

IllusionOfMana Joined: 08 Sep 2018 Posts: 270	Posted: Fri Nov 16, 2018 12:00 am
IllusionOfMana Joined: 08 Sep 2018 Posts: 270	For oscillators, you can find them. It's just, do you want to gamble on it being the correct item from aliexpress. Or you can get an adjustable oscillator and preset it, but the cheap ones aren't too reliable on staying that frequency. Another option is to source one running at the NTSC compatible frequency and just have an NTSC system. Hope you don't have to go that far and that it's really just an easier fix. Good luck.

asynchronous Joined: 14 Aug 2000 Posts: 742 Location: Adelaide, Australia	Help diagnosing a PAL UK SMS II video issue... Posted: Fri Nov 16, 2018 1:30 am
	I'm surprised the crystal is far enough out of spec to impact the VRAM access but not far enough to impact the TV screen or colours. I would for sure try replacing the crystal even if you have to purchase it off aliexpress. It's a cheap and easy fix if it works. Good luck!

asynchronous Joined: 14 Aug 2000 Posts: 742 Location: Adelaide, Australia	Help diagnosing a PAL UK SMS II video issue... Posted: Fri Nov 16, 2018 4:37 am
	Just out of interest, how did you manage to heat up the crystal oscillator with a hair dryer without heating up anything else like the PCB or chips?? Because depending what else you heated up in the process may mean the fault is somewhere else and you sort of found it.

aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Posted: Sat Nov 17, 2018 12:33 am
aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	I, Uh, didn't... Ha ha. Everything got pretty hot (but the oscillator chip heated up the most). At this point, all I know for sure is that it's something on the bottom right of the motherboard. I guess I could heat everything up and try spraying individual components with a quick freeze spray to see if I could diagnose the problem that way (bit crude, but it might work). That stuff's expensive, though. Anyway, I re-flowed all the components on that side of the board, just in case it was a faulty connection being reconnected by the board flexing with the heat but no joy. I also found a cheap multimeter and did some continuity checks and tested the voltage regulator (pretty close to 5V, so it can't be that). I just can't see what else it could be... And I'm not sure how to go about doing further testing of voltage around the rest of the board. I'll need to do some more research. At this point, it might be sensible just to buy a new Master System and put my old one out to pasture. Well, at least until I can find a cheap oscillator or a faulty SMS II to gut (I know you guys said they were out there but really, they're not. Not anywhere I've looked... Closest I found in stock was 50mhz) Anyway, thanks for the help. On the plus side, I managed to test my Mega CD with the multimeter to determine that the fuse had blown, so it's not all bad!

IllusionOfMana Joined: 08 Sep 2018 Posts: 270	Posted: Sat Nov 17, 2018 2:01 am Last edited by IllusionOfMana on Sat Nov 17, 2018 4:47 am; edited 1 time in total
IllusionOfMana Joined: 08 Sep 2018 Posts: 270	Oscillators @aliexpress Like I said, they're out there supposedly its just if you want to gamble aliexpress and hope you get the correct part.

pyroesp Joined: 14 Mar 2013 Posts: 62 Location: Belgium	Posted: Sat Nov 17, 2018 4:15 am
pyroesp Joined: 14 Mar 2013 Posts: 62 Location: Belgium	IllusionOfMana, your link is breaking the forum layout.

IllusionOfMana Joined: 08 Sep 2018 Posts: 270	Posted: Sat Nov 17, 2018 4:48 am
IllusionOfMana Joined: 08 Sep 2018 Posts: 270	thanks for letting me know, I fixed it.

asynchronous Joined: 14 Aug 2000 Posts: 742 Location: Adelaide, Australia	Help diagnosing a PAL UK SMS II video issue... Posted: Sat Nov 17, 2018 1:22 pm
	Quote I, Uh, didn't... Ha ha. Everything got pretty hot (but the oscillator chip heated up the most). Haha, that's what I thought when I went away and thought just how would that work with a hair drier. When you do get around fixing it some time in the future, we'll do what we can to help you fix it.

aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	Posted: Sun Nov 18, 2018 6:18 pm
aliensyndrome Joined: 12 Nov 2018 Posts: 6 Location: UK	I had a thought about what you said about the hair dryer and tried heating the board pretty much anywhere and sure enough, that also "fixed" my problem. My best guess now is that there's a bad solder joint (maybe somewhere on one of the VRAM chips) that expands and gets reconnected as the board heats up. But try as I might, I just can't seem to find it. All pins seem to test fine for continuity. I'm going to try to re-flow the joints on the top of the board and see if that works. (which is a bit tricky with all the components in the way). It's so frustrating to feel so close, yet still so far from fixing the problem. But I've not given up yet. I'll let you guys know if it works.

Forums

View topic - Help diagnosing a PAL UK SMS II video issue...