Saturn switching PSU repair

[nick3092]

I have an internal switching power supply from a Sega Saturn that is acting up.  It's creating noise in the video in the form of lines that roll through the picture, and occasionally I can hear what is most likely some coil whine.  As I understand it, most of the after market PSU's designed for Saturns are not engineered very well.  So I'd like to fix mine if possible rather than go a different route.

Sega used several different PSU's over the years.  And I don't have a schematic for it.  So I started by doing a recap using some 105c Panasonics, but it didn't seem to help much, if at all.  Someone suggested it may be the rectifier bridge, or possibly the transistor going bad on it.  So I figured I'd try replacing those two whenever I place another parts order.  But I'm not that good at finding comparable parts, so I was hoping to get some advice on what to replace them with.

The bridge rectifier is marked "S1NB" and "60 58" below it.  I believe this is the data sheet for it:
https://www.mouser.com/datasheet/2/366/J534S1NB-1203035.pdf

Looking at mouser, I think this rectifier (DF06-G) would be a similar (if not direct) replacement, based off the data sheet values.  Its wider, but it looks like on the PCB, they drilled holes and extended the traces to be able to use a rectangular bridge in place of the square one originally used.
https://www.mouser.com/datasheet/2/80/DF005-G_Thru323050._DF10-G_RevD-1481243.pdf

One question I do have on the rectifier is that according to the data sheet, it's rated for 1A average/30A surge.  But the 5v rail is silk screened on the board as being rated for up to 3A.  So would I be better off maybe moving up to a 2A average bridge?  Or just go with whats there?  I will most likely be using some power hungry accessories that may draw more than the stock Saturn was expecting, but should still be within the 3A rating of the board.

The transistor is marked "C4507" and "63" below that.  I think its a 2SC4507 NPN power transistor.  It has the same logo as an ebay auction for a pulled Fujitsu of the same part number.  I didn't come across a Fujitsu data sheet, but did find this one:
https://www.alldatasheet.com/datasheet-pdf/pdf/271933/SAVANTIC/2SC4507.html

I don't know much about transistors and what values are critical to match, and which ones can go higher.  Looking at the data sheet below (part number FJPF5021OTU), this one seems to be a good match in the sense that the max collector current matches, and all the maximum voltages (CBO, CEO, EBO) all meet or exceed the original.  And the collector dissipation is the same.  The only thing I'm not sure on is the old part had collector and emitter cutoffs of 100uA, but the new one is only 10uA.  I'm not sure if that's a value that can change.  The DC current gain also looks to be a a little lower on the new part.
https://www.mouser.com/datasheet/2/308/FJPF5021-D-1809405.pdf

Thanks in advance for any advice on this.  Of course I understand this could be a futile effort and that some other part could be bad. Just figured I'd start with something hopefully easy to replace.

-Nick

[BrokenYugo]

I wouldn't be so sure the problem is in the power supply, could be bad caps or something else in the system itself leading to internally generated noise. I'm unfamiliar with the system but that's right around the time of those early SMD electrolytic caps that were put in lots of stuff and leaking/failing now. If throwing parts at a SMPS fixes it the output filter caps and other smaller electrolytics usually does the job.

What do you have for test equipment? A simple diode test on a multimeter will check the bridge and to some extent the power transistor, but that stuff usually works or fails shorted and doesn't work at all anyhow.

The bridge is rated for less than the current on the 5 volt rail because it's 1 amp at mains voltage vs 5V, a 3 amp 5 volt switching power supply should draw well under an amp, think power (watts), not current.

[james_s]

I really don't think the problem is the rectifier, what makes you suspect that? The only failure mode I've ever seen with rectifiers is they fail shorted.

[m k]

How those lines are rolling, 60Hz something?

[nick3092]

I'm sure the issue is the PSU and not the main board of the Saturn. The main board was recapped as well, with no change in the issue. Also, I tried a different power supply that utilizes an external power brick, and the issue went away. But I'd prefer to use the internal supply if possible instead of having an extra power brick kicking around.

The thought about it possibly being the rectifier or the transistor was just an idea someone mentioned to me that knows more about SMPS than I do. My electrical theory is pretty weak beyond the basics. But I'm good at doing rework on through hole parts. As for troubleshooting tools I have a multimeter and a Rigol 1054 (I think) scope. But I'm not great with using the scope.

Most of the other components in the board are film caps and resistors, which as I under stand it generally have very low failure rates. Outside of that, I see two inductors/coils (one large dual unit near the AC end of the board, one small one on the 5v output), two small transistors (I can't make out the part numbers without removing them), 3 or 4 discrete diodes, a transformer, what I believe is a dual Schottky diode in a TO220 package, and a smaller part I can't identify. It's marked sec/pri like a transformer and marked PI1, which I am not familiar with. The board was cracked and one trace broken before I had it. But the broken trace has been reinforced with a solid bar or solder. I can post a couple pics if it would help.

I'm not sure how to confirm if the rolling lines are 60hz or not. Is there any easy way to tell?

[BrokenYugo]

In lieu of a schematic pictures of both side of the board would be helpful. I struggle to wrap my head around it to a useful level, but with periodic video noise you can get an idea of the frequency, and thus source, based on the spacing, shape, etc. Pictures and/or video of the interference could be useful.

As to not being good with the scope, past knowing what NOT to do (see: https://www.youtube.com/watch?v=xaELqAo4kkQ, long story short the ground clip is grounded and will short whatever it's connected to straight to earth ground) you get good by playing around with it. You could learn a lot about the problem by probing the output rails (stay away from the mains side though) and seeing what noise is there, again, pictures are good.

[james_s]

This kind of problem is often something like a missing shield, improperly routed cable or bad ground connection somewhere.

Be very, very careful working on switchmode power supplies. They are somewhat dangerous and quite unforgiving of mistakes. Dave's video on how to not blow up your scope is highly recommended, and in general just stay well away from the primary side, I would advise keeping anything connected to the scope out of the power supply entirely until you feel competent with the scope.

[nick3092]

Thanks for the tips on the scope and avoiding the mains side.  I work on old tube radios as a hobby, so I'm at least familiar with avoiding high voltage and mains voltage with my fingers, and I know not to  put a scope on mains.  I do have an isolation transformer that I use for "hot" radio chassis that I can power this with as well if needed.

As for missing shielding or a miss routed wiring, there really isn't anything missing or miss routed.  The main board is completely shielded and gorunded, and the mains wiring is pretty much only routed in one spot where the power switch connects.  The rest is all contained on the board.  The power connector slides over 4 prongs coming up off the main board.  Even though there are 4 connectors, it's just 5v and ground.  The two 5v connections are tied together on the PSU PCB, as are the two ground connections. And the issue comes and goes, and when the video issue is present, I can hear a high pitched whine.  My guess is that it's coil whine coming off the mains inductor, but not sure.  It may also be tied to load, as in the high the load on the PSU, the more likely the issue is to occur.  So I'm fairly certain it's not any of those issues.  Unless the elctrolytics are bad, but they were brand new Panasonics, bought from a reputable reseller.

I've attached pics form the top, both sides, and the bottom.  As I mentioned earlier, someone broke the PCB when they were removing or inserting it in to the case.  The broken trace was replaced with a solid line of solder which seems to have reinforced the break pretty well.  That trace happens to be the positive line coming off the rectifier bridge, if it matters.

The middle TO220 part is the NPN power transistor, the one with the heatsink closer to the DC output connector is the dual diode.  I was able to read the part numbers off the small transistor next to the DC output.  Looks like a NEC C1093, which is a shunt regulator.  I can't read the one closer to the mains end without removing it, but can do so if needed.  The unknown (to me) part marked PI1 on the PCB is the little tan part near the transformer.  Its marked "6D (or 0, not sure)  P521  G"  Which now that I google that, appears to be an optocoupler/photo transistor?

[BrokenYugo]

Good idea on the isolation transformer, never a bad idea when working on stuff with exposed mains referenced circuits. Also makes probing about with the scope somewhat safer, at least for the scope (clipping the ground in won't hurt the scope, but it will reference the circuit back to earth, reintroducing shock hazards).

PT1 appears to be an optocoupler (feeds back the output voltage to the mains side for voltage regulation without making any electrical connection), the SEC/PRI dividing line there is a strong clue. PT probably refers to the phototransistor side of it.