MBFM PSU trouble

[offe]

Hi,

I'm building the Midibox FM and have had some problems getting the +-12v psu to work properly. I've built the psu using this http://www.midibox.org/users/talion/symetric_psu_12V.zip board mentioned on the OPL3 module page. Everything seems fine, but when I connected a 2x12v transformer to the circuit, it ended up with "releasing the magic blue smoke" and the primary side of the transfomer seems to overload (the primary side coil seems to be burned). This is the second transformer (have already rebuilt the circuit once) that behaves like this. Oh, and by the way, I did put a 1A fuse on the primary side the second time to prevent this from happening. Am I doing somthing seriously wrong here? The circuit seems to work fine for a minute or two (though it produces a low humming sound) before the transfomer dies. Anyone got a suggestion?

/offe

[smashtv]

This sounds a bit like two power supplies fighting each other....Without a deep technical explanation let's just say polarity -can- be an issue when dealing with AC-AC transformers.

I have a powered firewire hub connected to my powerbook.  The mac feeds the usual fw power down the cable also, but you don't really want to stack cable powered fw devices up on a rather expensive (in a laptop) fw chipset.  Both the power adapter on the mac and the wal wart on the hub are 2 prong, no chassis ground.  If I flip the mains plug on one of the two but not both, the hub will start to meltdown....It becomes a new and not engineered ground path (or neutral return path) for the rather beefy power supply on the mac.  The hub enclosure is still a bit deformed from this lesson, and I'm really lucky I did not have to replace the mac motherboard, or burn my house down. :)

My bet is your supply will run all day long powering only the FM module, but as soon as you connect other modules with separate supplies or other equipment (PC gameport cable? other MIDI devices with +5 on the MIDI connectors?) share a ground rail all hell breaks loose.

I would try firing up all modules/devices involved without interconnecting them, and test for a large voltage differential between the ground rails.  If it's there swap your primary lines to only one of the transformers and see if the readings change.

This kind of scenario will leave a fuse on the primary "hot" wire intact even if the transformer is going balistic.  Personally I would fuse all primary lines including chassis ground temporarily until the cause is hammered out, a look at which fuse blows next time will be a big hint to cause.

I'm at a loss for a good way to explain exactly what is happening in these out of phase/sympathetic power scenarios, but it's way past time for a good wiki entry explaining what that ground lift trace is for on my Core boards.

Best

Smash

[offe]

Thanks, smash!

I think your explanation is quite close to what's happening, at least this was the case with the first transformer. The second one, however, was not even connected to the opl3 board when it blew. That was really not what I had expected. I'm going to get a couple of new transformers and go with your suggestion to fuse all lines to try to isolate the problem. I'd really like to use this board but if the next batch of transformers goes up in smoke, I'll take it as a sign to power the mbfm the wallwart way.  :)

/offe