Audio Research SP11 power supply missing…can It be DIYed

I have an Audio Reasearch SP11 I got for a good deal in pristine condition the catch is it is missing the SP-11PS the power supply. Is there away to build a power supply for it? Thank you would like to here your suggestions..
 
I am in the process (long delayed) of "repairing" my SP-11 power supply. One of the filament windings of the transformer failed. (I have heard that others have had issues with the power supply.) It's been an "off-and-on" project for years.

Their engineers relied upon the MOSFET to dissipate quite a bit of heat. Today you can chassis mount a 15W Ohmite 300 Ohm resistor for R112 to drop 39V ahead of the MOSFET. (Speaking of which, Tomchr (www.neurochrome.com) has a very nice HV regulator.) It's performance should far exceed the original AR design and it is quite compact. The more heat you can dissipate with the resistor, the less cumbersome the MOSFET heat-sinking becomes.

You may need two transformers -- AR ran the filaments of V3,4,5,6 in series (6DJ6) @12.7V. V2 is elevated filament, V1 run at 6.5V. Also need DC aux voltages for the soft-start.
 

Attachments

  • SP11 Power Supply.png
    SP11 Power Supply.png
    479.5 KB · Views: 181
I am in the process (long delayed) of "repairing" my SP-11 power supply. One of the filament windings of the transformer failed. (I have heard that others have had issues with the power supply.) It's been an "off-and-on" project for years.

Their engineers relied upon the MOSFET to dissipate quite a bit of heat. Today you can chassis mount a 15W Ohmite 300 Ohm resistor for R112 to drop 39V ahead of the MOSFET. (Speaking of which, Tomchr (www.neurochrome.com) has a very nice HV regulator.) It's performance should far exceed the original AR design and it is quite compact. The more heat you can dissipate with the resistor, the less cumbersome the MOSFET heat-sinking becomes.

You may need two transformers -- AR ran the filaments of V3,4,5,6 in series (6DJ6) @12.7V. V2 is elevated filament, V1 run at 6.5V. Also need DC aux voltages for the soft-start.

I looked at the power supply schematic briefly. I can't find a data sheet for the MTM2N45 but apparently it came in a TO3 package which are pretty scarce these days. A re-engineered version would probably end up with a TO-220 which is even worse for heat sinking. But even with a big heat sink what would worry me in that design would be the SOA at start up. With 400 V on the input and 0 V on the output, even the relatively small output caps - I could only find 2 2 uF caps - would give you a pretty hairy trip through the SOA. (And forget about short circuit survival.)
Maybe you could use your idea of increasing R112 but instead of using one 300 ohm and a single big a$$ FET, divide the current up with say 4 75 ohm resistor/FET paths in parallel. That would give you power sharing but also pretty close to true SOA sharing.