Sony HB-F1XDJ PSU swap: new SMPS outputting +5V and +12 with inverter IC for -12V?

By SamIAm

Rookie (21)

SamIAm's picture

02-01-2019, 07:12

I recently pulled an SMPS out of a broken device and was interested to see that it produces +12V and +5V. It just so happens that I have a Sony HB-F1XDJ that exhibits odd behavior which I believe is connected to the power supply, so this seems like a great match for a replacement.

Obviously, it's missing -12V. I was imagining that I might get around this, though, by setting up an inverter IC on the +12V line. Whether this would work depends on how much current is normally drawn from the -12V rail, because the actual voltage output from these inverters drops sharply in proportion to the current. Does anyone have any idea? I wouldn't be connecting any external devices; the only thing this particular system will ever run going forward is a MegaFlashRom. Is the -12V only used as a reference?

Thank you!

Login or register to post comments

By sd_snatcher

Prophet (2914)

sd_snatcher's picture

02-01-2019, 09:10

What odd behaviour does your XDJ exhibits?

By SamIAm

Rookie (21)

SamIAm's picture

02-01-2019, 15:12

It's a long story, but basically, the Megaflashrom begins to fail to write ROMs after the system has been on for roughly fifteen minutes. It's extremely consistent about this. Flashing aside, there are no problems.

I have not been able to test the cart on another MSX. Discussions about this in the Megaflashrom megathread have not been fruitful, however, I have heard that the cart is very picky about voltage when flashing. One thing I do have is an oscilloscope, and I see that the power rails have lots of high-frequency (>1MHz) noise on them, at roughly 90mVp-p. The Megaflashrom contains a 3.3V linear regulator, and I wonder if some of the noise is getting through it since they are not great at ripple rejection at high frequencies and there are sadly only electrolytic capacitors in the cart doing decoupling, which high frequencies sail right past.

Whether the noise changes after the unit has been on long enough, I need to test some more to confirm. The character of the noise changes depending on what is on screen, which is interesting.

The 5V rail measures an even 5V via DMM. The 12V measures 12.6V and the -12V measures -11.6V.

Perhaps the cause of the problem lies elsewhere, but I'd like to be able to say with certainty that issue is not in the power supply.

By Meits

Scribe (5298)

Meits's picture

02-01-2019, 15:35

Have you visually inspected the board of your XDJ? They tend to destruct themself quite severely. Especially the HIC board below the diskdrive. Other caps break easily as well. They all go one day and lately they do by the masses.
Recapping the whole machine would do it well. If it's not the cause of your present problems, at least you've got that one handled.

By SamIAm

Rookie (21)

SamIAm's picture

02-01-2019, 15:47

I've already completely recapped it. Sorry, I should have mentioned that!

By SamIAm

Rookie (21)

SamIAm's picture

03-01-2019, 15:10

I'm sorry if I'm bumping this needlessly, but I observed something interesting today.

Previously, whenever the Megaflashrom began to fail to flash, I could put my hand on the case above the power supply and feel that the unit was giving off very noticeable heat. Today, it being a cold day outside, I put on my jacket and opened the windows in my game room to prevent that heat from building. Whereas in previous tests in warmer air, the Megaflashrom would fail to flash after 15-20 minutes of on-time, today it never failed once even after two hours.

Thus, it appears to be a thermal issue. Maybe reflowing some solder joints would help?

My MSX profile