A photo of exactly what I'm looking at is on the Syntaur website:
http://www.syntaur.com/images/3200-Lg.jpg
My power supply board is exactly identical.

The diodes are also clearly marked on the PCB with D1 and D2 being the spherical shaped diodes at the left next to the large 6800 uF cap. I guess if I were to look for a suitable subtitute, I'd need something that's pretty close in forward and reverse voltage, current and speed. Without having specs on what I'm taking out, I can either coming with a way to try to figure out what the specs are, or take a wild guess and substitute whatever I think will work and just test it (hopefully not the smoke-type of test).

The caps look brand new and I went ahead and pulled the 4 larger caps and went ahead and connected them up an ESR meter. No problems there. All are reading well within limits.

I'm not entirely understanding of your 'brown-out' characterization but basically the display indicates the random character problem occasionally and randomly at power up. The only predictable behavior I've noticed is that it usually powers on correctly on the initial power up (after the keyboard has been powered off for an extended period of time). Then after having been left on for a several minutes or so, I can cycle the power switch on and off and get the display problem again with random characters.

So to answer your last two questions, if the keyboard is left on (with a normal display at power-on) it seems to continue to work properly as it should. If I cycle the power off and then back on (after having been left on for a while) I can usually get the display to indicate faulty with the random characters.

As far as measuring the voltages at the parts, I haven't take a try at it. The schematic for the Matrix 6 is uploaded in the files section. I can't read most of the text so it's not entirely useable. F.I. is an abbreviation for...

--- In oberheim@yahoogroups.com , Martin Ator <cyllall@...> wrote:
>
> Have you tested the diodes You could upload some pictures, just in case what you think are diodes, may not be diodes. Depending on what diodes they are, modern 4148, 1n4001, 1n4004 can do the job of most oldies.The last 2 are most likely in a PS.
>
> What about capacitors    Electrolytics can smell a bit cat pissy if they have problems.Are you able to measure the correct voltages now
>
>
> There may be a faulty decoupling ceramic capacitor near to a logic chip,or an intermittent logic chip involved in the display control. I'm guessing you may be having a 'brown out' on power up if it's the power supply.
> Power regulators don't usually have problems unless something somewhere else has made them work too hard.
>
>
> Sometimes it's easier to measure power at the components themselves.If a schematic says 15 volts F.I, then you can expect to find a little bit more or less on testing. But I would suspect a problem if I found 13.5v instead F.I.
>
>
> If you switch your Matrix on and the display is OK, does it go funny after a while
> If you let it warm up and then do a power cycle, does the display then go funny (try to answer these last two)
>
>
>
> ________________________________
> From: wesleyknick <wesknick@...>
> To: oberheim@yahoogroups.com
> Sent: Sunday, 5 August 2012, 3:54
> Subject: [oberheim] Re: Oberheim Matrix 6, Dead
>
>
> Â  
> Assuming you were to replace the diodes on the power supply, is there a suitable substitute I'm having some trouble identifying the 2 different diode types visually. 6 of the diodes on the power supply are marked with what appears to be a 'T85' followed in smaller font by a '5D' or '5C'. The two separate spherical diode types are marked with what appears to be a 'T3B5H'. A Google search of any combination of the markings brings up very little in the way of any reference to diodes. None of the manuals uploaded on various websites have any additional information and the only place the diodes are depicted is the block diagram. Anybody have a parts list out there they can upload Most of the Korg and Roland service manuals have a parts list...
>
> --- In oberheim@yahoogroups.com , jim finch <jahg88@> wrote:
> >
> > You didn't mention how you were reading the powersupply voltages-but I didn't read the whole thread.If you use a O-scope you can see what is actually there,not just the D.C component.  Most likely there is garbage riding on the +5V rail making the digital components see it as logic level shifts.
> >
>