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 losing power immediately after being turned on

Reply to topic
Author Message
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Game Gear losing power immediately after being turned on
Post Posted: Mon Nov 02, 2020 4:51 pm
Nearly 9 years ago, I tried fixing a Game Gear I had gotten from a friend, that would turn immediately off. Several days before I got the Game Gear, my friend had checked that it was working. The day I was going to get it he turned it on again and it displayed the issue I describe. It sounded like a typical leaky cap issue. Back then I recapped the main board only without any success. 4 years later I tried troubleshooting again, I even did an LED backlight mode to no avail. The problem persists, backlight and screen go on for a fraction of a second, and the power led does not turn on.

After having seen many repair Youtube videos and having done a few repairs myself, I thought I'd give this sucker another go. Here are the things I tried:

1) Checked power board voltages.
2) Tried using a known working power board from another game gear.
3) Checked for continuity and investigated thoroughly with a loop for damaged traces, blown fuses, or transistors.
4) Checked all replaced caps to see if they were replaced with the correct capacity/voltage and correct polarity.

Something seems to make the power board cut off power as the output voltages indeed drop to zero very quickly on the power board side.

The only ideas I have to continue further is to remove all capacitors and check them with the multimeter (I don't know if the caps are of the best quality, they are made by Lelon). I don't want to admit this Game Gear has beat me, but I'm afraid the only way forward is to start replacing commonly failing parts! Do you have any suggestions for further debugging this issue? I can add pictures if it helps.
  View user's profile Send private message
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Mon Nov 09, 2020 11:53 pm
I tried removing all replaced capacitors and measured all of them. They were all within 20% from the rated capacitance. I also checked the variable resistor and it seems to work perfectly.

I then resoldered all caps back and added a 22uF capacitor in parallel with C1 of the powerboard to increase the under-voltage protection time. The only thing I was able to find out is that the rail that is being under-voltaged, is the +34V one, which comes up as +26V. I double checked everything that was receiving the +34V (see last page of the service manual) and nothing looks out of the ordinary. This part of the circuit goes into the last few pins of the LCD. I also checked LCD pins for continuity etc.

The board actually looks quite clean. I don't think the caps leaked at all before I replaced them, 9 years ago. So either some caps are bad even though the capacitance measures fine, or some IC got fried.
  View user's profile Send private message
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Tue Nov 10, 2020 9:39 am
Last edited by Grieverheart on Tue Nov 10, 2020 4:29 pm; edited 5 times in total


Here is a video of what the LCD is doing. Note that here I fiddle with the power connector cables, but in reality it only does so when first turned on. I confirmed there's nothing wrong with the power cable by soldering the power board output directly to the appropriate pads on the main board.

I'm wondering if IC3 is fried as, from my understanding, is the one responsible for the LED turning on as well as for driving the LCD?
  View user's profile Send private message
  • Joined: 18 Jul 2020
  • Posts: 21
Reply with quote
Post Posted: Tue Nov 10, 2020 12:43 pm
Maybe adding some pics of the PCBs would be a good idea to get some more eyes on it.
  View user's profile Send private message
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Tue Nov 10, 2020 2:56 pm
Here are some pictures of the main pcb.
IMG_4176.JPG (2.73 MB)
Main board without caps - Right
IMG_4176.JPG
IMG_4177.JPG (2.9 MB)
Main board without caps - Left
IMG_4177.JPG
IMG_4178.JPG (3.01 MB)
Main board without caps - Right
IMG_4178.JPG
IMG_4200.jpg (2.87 MB)
Main board with caps
IMG_4200.jpg
IMG_4201.jpg (3.3 MB)
Main board with caps - Left
IMG_4201.jpg
IMG_4202.jpg (2.69 MB)
Main board with caps - Right
IMG_4202.jpg
IMG_4203.jpg (2.66 MB)
Main board front
IMG_4203.jpg

  View user's profile Send private message
  • Joined: 18 Jul 2020
  • Posts: 21
Reply with quote
Post Posted: Tue Nov 10, 2020 4:13 pm
I'm no expert, I just so happen to just do a recap and a mcwill install recently. I'm sure there are far more experienced people here. But this is what I observed.

We have similar models. From the looks of it, your C37 has a 47uf / 10v cap. From the pic I took of mine before replacing the caps, it was originally 68uf and 6.3v. It seems the cap uf is lower than the original. I've read it should be equal or higher than the original. C45's positive connection looks sus to me, maybe re-adjust it.
  View user's profile Send private message
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Tue Nov 10, 2020 4:23 pm
Thank for you reply. Indeed I have a 47uF there. At first I had a 47uF together with a 22uF in parallel so that I could get 69uF, which is close to the 68uF that is actually installed. When I took all caps out a couple of days ago, the 47uF actually measured around 60uF, so I removed the 22uF one just in case that was the issue. I'm actually wondering how close the capacities should be to the original values as I'd think that the Game Gear was built with at least a 20% tolerance in mind?
  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 7:40 pm
I'm pretty sure the game gear turning off immediately after powered on is due to short circuit protection in the MB3775 chip, itself on the power board.

Disconnect the mainboard from the power board and check for a circuit or low resistance between +5V and ground in the mainboard. Or perhaps you replaced capacitors with larger ones which "short circuit" the supply during the first instants it is powered on.
  View user's profile Send private message Visit poster's website
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Thu Nov 12, 2020 10:58 pm
Quote
I'm pretty sure the game gear turning off immediately after powered on is due to short circuit protection in the MB3775 chip, itself on the power board. Disconnect the mainboard from the power board and check for a circuit or low resistance between +5V and ground in the mainboard.


That's what I also thought at first, but I cannot find any shorts at all, that's why I bypassed the short (under-voltage) protection by adding a higher capacitance to C1. Here are the resistance readings for each rail:

+5V -> 324 Ω
+9V -> 12.6 kΩ
+1.28V -> 7.25 MΩ
+34V -> 70.7 kΩ

I guess the +5V one is so low because of the led mod I've done. Maybe someone can measure these resistances on their twin ASIC game gear to pinpoint on which rail the problem is.

Quote
Or perhaps you replaced capacitors with larger ones which "short circuit" the supply during the first instants it is powered on.


If that was indeed the case, which it isn't as I've checked all caps multiple times, bypassing the protection, would make the game gear work as the capacitors would get charged.

I took some further steps in investigating any potential problems. My current theory is that IC3 is not working properly, but I'm not sure how I could prove that non-destructively. I measured all voltages going into IC3, but got nothing strange. When I check voltage across e.g. R11 which is on the output of the LED driver, I don't get anything (although I'm not sure the multimeter would register a short voltage spike). Also when I measure the crystal, X1, I get absolutely nothing.

Now, why would the +34V rail be undervoltaged? Perhaps because IC3 is not turning on the LCD display? Does my theory make sense?
  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 11:25 pm
What's IC3? MB3775?

You only need +5V to test the logic board, if the system boots with a game it should play some sound.
I suggest you disconnect the internal power connector and use a pair of wires to supply both +5V and ground to the mainboard.
You can't (easily) test the resistance (load) of the system as it is composed of dynamic and non linear devices (semiconductors, sequential logic, etc.). One simply way is to put an ammeter in series with the rails and measure the flowing current, then use ohm's law.

I woulnd't be surprised if MB3775 is damaged (or one of the nearby transistors). Following the leaky capacitors, that chip is typically the next to meet its creator. The undervolted +34V rail may be a sign of it, of damaged transistors or a schottky diode but also of an overwhelmed +5V rail (the +34V is derived from it, check the powerboard schematic).

Check the system with an external +5V supply (preferably with short circuit protection). A 7805 would do the job and you can check its temperature to see if there is a short :D
  View user's profile Send private message Visit poster's website
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Fri Nov 13, 2020 8:25 am
Quote
What's IC3? MB3775?


IC3 is the IC you see in the second from last page in the service manual. The IC is called SCA, I'm not sure what its exact function is.

Quote
You only need +5V to test the logic board, if the system boots with a game it should play some sound.
I suggest you disconnect the internal power connector and use a pair of wires to supply both +5V and ground to the mainboard.


Thanks, I'll try that out. Are the other voltages not required? What's the 9V and 1.28V actually doing, then? They are split into VREF, VRES, and VONF of the IC, but I cannot find anywhere what the function of these pins is.

Quote
You can't (easily) test the resistance (load) of the system as it is composed of dynamic and non linear devices (semiconductors, sequential logic, etc.). One simply way is to put an ammeter in series with the rails and measure the flowing current, then use ohm's law.


I thought measuring resistance to ground in logic boards and comparing them to reference values of working boards is a common debugging practice. See, e.g. https://logi.wiki/index.php/Resistance_to_Ground_List. More often, diode mode with the red lead on ground is used instead for faster readings.

Quote
I woulnd't be surprised if MB3775 is damaged (or one of the nearby transistors). Following the leaky capacitors, that chip is typically the next to meet its creator. The undervolted +34V rail may be a sign of it, of damaged transistors or a schottky diode but also of an overwhelmed +5V rail (the +34V is derived from it, check the powerboard schematic).


As I mentioned, I have another power board that I know is functioning correctly and I see the exact same behaviour. I indeed had MB3775 get fried on another Game Gear I repaired. The power boards correctly give +34V, but as soon as it's connected to the main board, the voltage drops, which is triggering the MB3775 circuit protection. I think, that the voltage might be dropping because the LCD is not turned on (by IC3?).

Quote
Check the system with an external +5V supply (preferably with short circuit protection). A 7805 would do the job and you can check its temperature to see if there is a short :D


I was thinking of buying a bench power supply as a last resort to further debug the issue. If there is a short, you'd expect the component that's shorting to heat up. I don't have a lot of experience with components like the 7805. I wish there was a list with useful components to have, like the 7805 or a 555 timer, etc. I just ordered caps from digikey, and I had to fill my cart with more caps to reach the limit of 50 euro they impose.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Fri Nov 13, 2020 7:23 pm
Quote
The IC is called SCA, I'm not sure what its exact function is.

That's an ASIC and depending on the mainboard it could one or two physical chips. It's what basically glues all the other chips together with logic gates, logical blocks, etc.

Quote
What's the 9V and 1.28V actually doing, then?

Where do you see +9V? The 1.28v is used internally, I am not sure what for, see
https://www.smspower.org/forums/18144-PurposeOfGameGears128Vref

You should only need +5 and +1.28 to boot the system and have sound.

Quote
I thought measuring resistance to ground in logic boards and comparing them to reference values of working boards is a common debugging practice

One thing is measuring resistance/impedance, another one is power/current consumption. The resistance is measured with a ohmmeter by using a known voltage/current so it is dependent on those parameters. For instance, the forward current of a diode depends non-linearly on the forward voltage, so you would have a resistance curve and not a fixed one: https://cdn.sparkfun.com/assets/4/4/a/5/b/5175b518ce395f2d49000000.png
By measuring directly current you could get an average consumption depending on the circuit dynamics better than by measuring resistance.

Quote
I think, that the voltage might be dropping because the LCD is not turned on (by IC3?).

The +34V goes directly to the LCD's contrast driver and other stuff inside the LCD (ie lateral chip) I don't know much about. But it shouldn't affect the backlight/ccfl. Disconnect +34V and check if the backlight turns on as it feeds on the 5V rail.

Quote
there was a list with useful components to have

A 7805 or equivalent is useful to have but you can also order from ebay a couple of step-down/step-up converters. Search for MP1584, MP2307 or LM2596 on ebay. Except the last one, the others all small enough to fit inside the GG.

Check also this post to help fill your drawers:
https://www.eevblog.com/forum/chat/good-ics-to-have-lying-around/
  View user's profile Send private message Visit poster's website
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Sat Nov 14, 2020 8:48 am
Thank you very much for your reply. There is a lot of useful information for me.

+9V is named VBAT in the service manual and is generated by the power supply even if no batteries are connected. It's the 3rd pin from the +34V, between the one that's not connected and VREF.

I just found something out which, if it turns out to be the problem is a bit embarrassing. I found out on the functioning Game Gear I got recently, that if I don't connect the sound board, it does not start. Basically it shows the same behaviour as the one I'm trying to fix. Now, I connected the corresponding sound board to the non-functioning Game Gear and it's still not working, but I have not recapped the sound board so it could very well be that it's bad. I'm waiting for the caps to arrive, hopefully that will fix it, but in any case, if it doesn't, I will have to redo all debugging with the sound board connected.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Sat Nov 14, 2020 1:49 pm
ah... 9V is from the battery :D I didn't tough of that one. I always saw it as the external power source. It doesn't have to be 9V, could be higher if for instance coming from an external transformer.

Leaky capacitors on the sound board? I assumed you replaced all of them.

Glad to be of help, cheers.
  View user's profile Send private message Visit poster's website
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Sun Nov 15, 2020 7:41 pm
So, I don't know what went wrong last time I checked, but I tried turning on my good Game Gear without the sound board connected and it run just fine. So the sound board does not need to be connected for the Game Gear to function properly.

Now, I soldered all leads from the power board to the main board, except the +34V one, and I also removed the additional cap I added to the powerboard to bypass the undervoltage protection. When turning on, the backlight turns on like usual and it stays on, but the LED still doesn't turn on. I also don't hear anything if I connect the sound board from the working Game Gear. Note that the working one is a single ASIC while the bad one is a twin ASIC, and although the sound boards are the identical, the connector is slightly different so I had to solder directly.

I have really no clue what's wrong. I checked voltages on IC3 when everything's connected and I do see voltages present. Everything's pointing to the +34V rail being the culprit but other than that I don't know how to proceed.
  View user's profile Send private message
  • Joined: 11 Aug 2007
  • Posts: 87
  • Location: Coimbra, Portugal
Reply with quote
Post Posted: Sun Nov 15, 2020 10:38 pm
The +34V rail is not the culprit if it is disconnected and the red led fails to turn on. The led is connected to the +5V rail and driven by the ASIC, ie switched to ground internally if everything is ok. Check the right corner of IC1:

https://d3nevzfk7ii3be.cloudfront.net/igi/JEbeSiNTVS14BN6Y.huge

If the capacitors have been replaced, the power supply is ok and yet the red led does not turn on I am afraid the issue is complicated, probably inside the ASIC.
  View user's profile Send private message Visit poster's website
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Mon Nov 16, 2020 7:36 am
Yes, that was thus my theory. I have a twin ASIC one, and the corresponding pin on IC3 does not seem to short to ground. I don't know, though, what the complete logic behind the condition for the LED to turn on is, i.e. maybe the IC checks for the correct functioning of other components first. So it could even be that the LCD is faulty, or either some IC, causing the LCD to not power on.

As a last resort, and as a practice for my SMD soldering skills, I'm considering replacing IC3 with one from another faulty board I have, although I'm not 100% sure the ICs there are good either.
  View user's profile Send private message
  • Joined: 23 Dec 2011
  • Posts: 20
Reply with quote
Post Posted: Sun Nov 22, 2020 8:26 am
I managed to desolder IC3 on both faulty Game Gears (315-5378A on one board and 315-5378B on the other), and solder one on the other. The LED on the Game Gear I was trying to fix was not turning on, while on the other faulty Game Gear it was. When I swapped the ICs, the LED was turning on. It took me a long time, but i finally found out that IC3 must be bad on both boards -- it might be possible that when I tried fixing the boards 10 years ago, as I was inexperienced, that I shorted something causing the ICs to go bad.

Unfortunately, I cannot find any spares online and I doubt I will. Let me know if you see it being sold anywhere.
  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 1:14 am
Well that's sad mainly because of the wasted work. The easiest way is to get them (the chips) from a donor GG...

I assume you cleaned the mainboard of electrolyte before testing the GGs as that stuff could conduct current between the logic pins making the system unstable.
  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!