Forums

Sega Master System / Mark III / Game Gear
SG-1000 / SC-3000 / SF-7000 / OMV
Home - Forums - Games - Scans - Maps - Cheats - Credits
Music - Videos - Development - Hacks - Translations - Homebrew

View topic - Game Gear power board default behaviour

Reply to topic
Author Message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Game Gear power board default behaviour
Post Posted: Sat Oct 10, 2015 6:19 pm
Greetings,

I'm debugging a GG power supply here. When not connected to the mainboard, does the power supply shutdown after some miliseconds or does it supply the typical 34V and 5V full time?
  View user's profile Send private message Visit poster's website
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Sat Oct 10, 2015 11:24 pm
(Self answering)

Repaired one of the power supply boards I have here.

A working power supply generates +34V, +5V and a +1.28V voltage reference without any sort of feedback from the mainboard.
  View user's profile Send private message Visit poster's website
  • Joined: 29 Jan 2014
  • Posts: 92
  • Location: Brasilia, Brazil
Reply with quote
Post Posted: Mon Oct 12, 2015 3:52 pm
I did some testing with a working 171-6483B power board taken from a 1-chip GG unit (IC-BD GG MAIN VA4 837-9537) and I determined that it requires a load to operate properly.

Without a load it will drift outside of the normal operating frequency and will over-voltage. (peaks at about 7v on 5v line) That will cause it to shut down (similar as how it would without the caps working properly).
And that is the behavior you were asking about it shutting down while disconnected from the main board

As a piece of advice, it's very dangerous to test switching power supplies without a load (most switching power supplies have dummy loads built in to prevent them from drifting out of normal operational range so my advice might sound inane but trust me it's not...). So to avoid risk damaging the main board I would suggest you use resistors as load while testing these power boards.

Switching power supplies running from mains voltages can even explode if powered on without a load (3DO FZ-10 is a good example where that can happen if you power the power supply without it being connected to the mainboard).
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Mon Oct 12, 2015 6:53 pm
Thanks for your advice.

The problems I had were caused by minimal solder paste, together with flux and some ethanol remaining on the pcb. After soldering I always like to have a clean pcb but it seems I didn't remove all the "liquids".

I believe you, but I have been monitoring the circuit and (with the 2 fps voltage update of my multimeter) the +5V line never goes above 5V, usually around 4.95V.

Another problem was caused by the replacement of the Vcc decoupling capacitors nearby the logic ICs with higher capacity ones, for instance a 1000 uF to decouple the large ASIC in the 1 ASIC board. This triggers the MB3775's timer latch's short-circuit protection, which is configured by the CPE capacitor connected to MB3775 pin 15 (see datasheet).

As you can see, the above problems were the main reason I created this topic.
  View user's profile Send private message Visit poster's website
  • Joined: 29 Jan 2014
  • Posts: 92
  • Location: Brasilia, Brazil
Reply with quote
Post Posted: Mon Oct 12, 2015 7:34 pm
What is the reference code for your power board?
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Tue Oct 13, 2015 7:59 pm
As any other I have (4 total), 837-9131 (front) 171-6391b (back).
  View user's profile Send private message Visit poster's website
  • Joined: 29 Jan 2014
  • Posts: 92
  • Location: Brasilia, Brazil
Reply with quote
Post Posted: Tue Oct 13, 2015 8:09 pm
All of them are identical?

There are variations... I ask because I have two GGs and this one I used as reference is the "cheapened" one with 1 LSI and different LCD (this one has TMSS and can't use the TV tuner). This one powers on and gives sound but the LCD has 1/3 part of it's display area dead vertically.
So I sure expected the power board to be different.

I have a TMSS-less unit but I didn't disassemble it as it's fully working and I didn't feel like disturbing it much.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Tue Oct 13, 2015 9:31 pm
Very identical. The only difference is the appearance of the big inductor L2, possibly from a different manufacturer. The position of the diode D1 also changes from parallel to inductor JG to perpendicular.

Tomorrow, I'll take a couple of photos with a better illumination.

From what I am able to see, all the board versions are in accordance with the official Sega schematics.

By the way, to prevent shutdown problems as I had previously (because of the larger decoupling capacitors), I increased the CPE (C1) from 0.68uF to 4.7uF. This way, the time constant for the short circuit protection (t = C1*0.6) increased from the default 0.4s to 2.82s :D

This is the only way I can have a couple of 1000uF capacitors inside the GG to minimize ripple. I should add some 100nF ceramic capacitors too.

I have to say, the machine feels (and appears) as brand new. The only artifacts in the screen are some vertical trails extending from certain lengthy sprites. This somewhat bothers me, sometimes.
  View user's profile Send private message Visit poster's website
  • Joined: 29 Jan 2014
  • Posts: 92
  • Location: Brasilia, Brazil
Reply with quote
Post Posted: Tue Oct 13, 2015 9:48 pm
That artifact is likely this:



It's called LCD Ghosting and the LCD on the GG is pretty slow so very susceptible to it...

If the LCD was not supposed to suffer that kind of issue, wrong values on parts could be cause of such a issue. But I am sure I remember as a kid that the GG had pretty horrible ghosting even when it was new.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Thu Oct 22, 2015 12:50 pm
Sorry for the delay.

No, it's not temporal ghosting. It's a general problem in all the game gears I have seen. Perhaps someone has solved it by now:



Note the vertical extensions from the building.

And here you have the two types of supply boards I have here:

  View user's profile Send private message Visit poster's website
  • Joined: 29 Jan 2014
  • Posts: 92
  • Location: Brasilia, Brazil
Reply with quote
Post Posted: Thu Oct 22, 2015 1:06 pm
The stripes are certainly related to power supply circuitry. Maybe the voltage is out of range or the capacitors are not doing their job properly.

As you noticed it's best to stick to the original values due to how the thing is designed.

Have you replaced all the electrolytic capacitors on the main board?
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Sat Oct 24, 2015 4:07 pm
It's the same with all the GGs I have seen, whether the capacitors are equivalent or not.

It is very likely to be related to the LCD's external circuitry or even its internal workings. As I don't have an oscilloscope, I can't go far in further research.
  View user's profile Send private message Visit poster's website
  • Joined: 29 Jan 2014
  • Posts: 92
  • Location: Brasilia, Brazil
Reply with quote
Post Posted: Sat Oct 24, 2015 4:16 pm
It's so frustrating see these things aging and not be able to do much to conserve them... D:

My working GG has lots of stuck pixels and one of the three regions of the screen is visibly dimmer than the other two.

At least it doesn't have that horrible banding. Have you tried to wash the circuit board with soap and water? (to remove capacitor residue)

I did that on mine (I did protect the LCD assembly with plastic before doing it.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Sat Oct 24, 2015 5:03 pm
Yep. I'm perfectionist with my surgical operations. I clean the pads with a precision knife, ethanol, then I solder the components with flux and then after, clean again with ethanol. I like to have the solder joints shiny ;)

Naturally, whatever I can clean in the pcb, I clean. Especially those yellow/green almost carbonized solder joints around leaking capacitors. If I have to, I reflow them.

Now with the LCD, it's a whole different question. These are hard to replace (but I have a method), neither improved (much)...
  View user's profile Send private message Visit poster's website
  • Joined: 28 Oct 2020
  • Posts: 5
Reply with quote
Post Posted: Thu Oct 29, 2020 2:29 pm
Hello C3R14L.K1L4,

I'm very much interested in your project with reducing ripple. I sent you a pm about it but I'm not sure if you check pms.

If you could tell me your purpose for adding the 1000 uf caps specifically. I know you mentioned reducing ripple. Does this also mean reducing the electronic noise and static that a game gear has?

I'd like to experiment and try myself.

Where do you add the 1000 uf capacitors precisely.

And I'm wondering where the cpe c1 capacitor is that you mentioned.

Thanks a lot!

P.S.- sorry to other members for bumping such an old post. I'm really hoping to find information on this topic.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Thu Oct 29, 2020 3:56 pm
The ripple on the gg is negligible. And with the original power supply adding large capacitors will trigger the MB3775's short circuit protection and you won't be able to use the system unless you change that protection. You can always add ceramic capacitors (eg 100nF) nearby each IC to reduce noise. The logic system is fairly resistant to noise/ripple.
I haven't added anything to the mainboard but I did replace the power supply in one of my GGs with step-up/down converters (not yet finished but working, nothing special).
Why are you interested in reducing ripple (or similar changes)?
  View user's profile Send private message Visit poster's website
  • Joined: 28 Oct 2020
  • Posts: 5
Reply with quote
Post Posted: Thu Oct 29, 2020 10:48 pm
Thanks for your reply.

So basically I'm trying to reduce the buzz and humm of a game gear. I'd like to make it as quiet as possible. That's my main goal. It's seems that when you add a new LCD it adds more electronic noise and static too.

So I was told by someone who repairs these that he always uses a 460 uf or 680 uf decoupling capacitor to the power board to help with noise.

He said he puts it on the 5v and ground pins on the underside of the power board. Just attaching it via wires so it can be on top of the board.

You mentioned chips. So I looked at the chip diagram on the power board and there is a vcc and ground. I'm wondering if that would accomplish the same thing.

Are you saying that adding 100 uf capacitors to chips on the main board can help with noise?

How many chips on the main board can I do that too? And where would the locations be?

Could that be done to the amp chip on the sound board too?

I'm also wondering if there is anyway to reduce static coming out of the headphone jack.

Thanks a lot. I haven't found any real information on this topic.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Fri Oct 30, 2020 9:20 pm
Quote
So I was told by someone who repairs these that he always uses a 460 uf or 680 uf decoupling capacitor to the power board to help with noise.

Quote
Are you saying that adding 100 uf capacitors to chips on the main board can help with noise?


I'm pretty sure those sizes will trigger the power supply's short circuit protection (SCP). And large electrolytic capacitors won't help with high frequency noise (they have large impedance/losses at high frequencies), it's better to use small ceramic capacitors or inductors. Large capacitors are better for slower (> 0.01 s) voltage variations.

Quote
So basically I'm trying to reduce the buzz and humm of a game gear. I'd like to make it as quiet as possible. That's my main goal. It's seems that when you add a new LCD it adds more electronic noise and static too.


You want to reduce noise where? In sound? If so can you analyse the background noise with an oscilloscope / frequency analyser? Otherwise you could use your pc's sound card line-in (using a small 10uF capacitor in series). When my phd lets me I may try to check the audio noise of one of the working gg I have.

If you want to reduce noise in the lcd that's a different beast...

Quote
You mentioned chips. So I looked at the chip diagram on the power board and there is a vcc and ground. I'm wondering if that would accomplish the same thing.


Yeah, adding/replacing the power supply decoupling capacitors to each IC. The standard approach is one "larger" electrolytic capacitor (10 uF or lower because of SCP) and a smaller ceramic capacitor (<470 nF) in parallel with the power rails of each IC. Adding inductors will be problematic as they have to be in series with the voltage supply of each IC...

Quote
How many chips on the main board can I do that too? And where would the locations be?

All of them ;) Ideally it should also include oscillators, there is one for the transformer driving the CCFL which feeds from the 5V supply (and thus feedbacks back to the system) and to another I think in the LCD's contrast (I may be wrong in this), so I would also filter the voltage sources (5V and 34V) in this section.

Quote
Could that be done to the amp chip on the sound board too?

Quote
I'm also wondering if there is anyway to reduce static coming out of the headphone jack.

In my opinion this may be where noise may be most disturbing, as I said above. That will require first knowing the source of the noise. As before, you can always filter the TDA2822's positive voltage rail. If the noise comes from the sound generator inside the ASIC your task will be more complicated...

Quote
Thanks a lot. I haven't found any real information on this topic.

This approaches a level of perfectionism I don't see anyone else bothering with. You have my support ;)
  View user's profile Send private message Visit poster's website
  • Joined: 28 Oct 2020
  • Posts: 5
Reply with quote
Post Posted: Sat Oct 31, 2020 3:27 pm
Thank you very much for the assistance and feedback! I tend to tinker and tinker until I break something. I start over and continue tinkering.

I'm not sure exactly what a completely optimized game gear would be, but I'd like to try and find out.

The sound board does seem like a larger part of noise.

I'm wondering if there is a better mode line filter that could be changed out.

But yeah, it's worth finding out what will happen if you cap that sound board amp chip...

I'm also thinking of trying audio grade electrolytic capacitors.


On a power board related question. I've put a 1000 uf cap in replace of the 820 uf cap. Do you know anything about the location of that cap and what it does for the circuit? The 1000 caps are easier to source for me in smaller voltages. Just wondering if it's a better change, neutral, or shouldn't be done.

Thanks!
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Thu Nov 12, 2020 7:48 pm
Quote
I'm wondering if there is a better mode line filter that could be changed out.

Replacing the common mode choke in the sound board could reduce the noise but perhaps not by much. There will be noise coming in from the audio source, the supply rails and high frequency electromagnetic interference all around the game gear. I have the feeling that the ccfl and the high voltage transformer are problematic in that sense. Hence a possible nightmare to solve...
Quote
On a power board related question...

The 820uF is the bigger capacitor on the power board. Be careful when replacing with larger capacitors (because of the short circuit protection) as well as capacitors with lower ESR (the mb3775 datasheet dedicates a couple of pages to that issue). If you note problems with the system turning off after power-on or some strange phenomena (instabilities) it may be because of this.
  View user's profile Send private message Visit poster's website
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Fri Nov 20, 2020 10:39 am
I'm also interested in finding the culprit causing this vertical smearing issue. I think the LCD would be much clearer if it's solved, if at all solvable. Although you nowadays have better solutions like McWill, I very much like modifying as little as possible.

I also don't have an oscilloscope, but I'm planning to buy a second hand analogue one. When looking at the schematic in the VA1 service manual, I also see some decoupling capacitors on the 3 buses going to the LCD, i.e. {DS, P[1..4]}, {CL2, DO[1..4]}, and {DW, CLA[1..3], CLB[1..3]}. Caps C18-31 are all 47pF, and C17 is 1200pF. Do you think it might be worth replacing these with larger capacitance ceramic ones?

I'm also wondering, if you do a video out mod, the signal you get on the TV does not show the same issue. If the power decoupling capacitors were the culprit, wouldn't you also see the same effect on TV? If we take Tim Worthington's mod as an example, you see that he is taking the signal from DO[1..4], so these lines should be fine. HSYNC, I think, is taken directly from IC1, which is not connected in the Game Gear. Time says, 'The only problem is that the RGB data is sent serially over four digital lines, one colour at a time.', so the way the signals are separated on the Game Gear must not be very good, with signal somehow leaking over multiple lines. Either that, or it's the fault of the LCD and that's the end of it; from what I gather, the lines DO[1..4], go directly into the LCD, so the signal is somehow processed into appropriate electrical current on the LCD itself.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Mon Nov 23, 2020 1:25 am
Quote
I'm also interested in finding the culprit causing this vertical smearing issue. I think the LCD would be much clearer if it's solved, if at all solvable. Although you nowadays have better solutions like McWill, I very much like modifying as little as possible.


I am assuming you are referring to this wonder:



Quote
I also don't have an oscilloscope, but I'm planning to buy a second hand analogue one.

One of the best tools you could have. If the purpose is to analyse GG's signals you should get at least one with 100 MHz of bandwidth. But still, a slower one is better than none.

Quote
Do you think it might be worth replacing these with larger capacitance ceramic ones?

I'm pretty sure interesting results may be obtained with smaller capacitors or even removing them. By having large values you'll be filtering the pixel's rapid changes and thus left with slower variations in the image. I'll try to get a look onto those signals this week (but I do not guarantee, I'm pretty overloaded with work).

Quote
If the power decoupling capacitors were the culprit, wouldn't you also see the same effect on TV?

Very likely.
  View user's profile Send private message Visit poster's website
Reply to topic



Back to the top of this page

Back to SMS Power!