Krell KSA 100mkII Clone

[PWatts]

Here're two very preliminary 1kHz and 10kHz squarewave tests of a proto board. This is minus output stage, 0 bias and at only 35V rails so not at all proper operating conditions, but it gives an idea. Transistors were 2SA970/2SC2240 input, 3310 mosfers, 2SA968/2SC2238 predrivers and MJE15032/MJE15033 drivers. R and C values are the same as in the BOM.

[PWatts]

10kHz..

[Flodstroem]

Originally posted by loek
Hi Flodstroem, it would be nice to see a full body picture here of your stuffed board maybe?
Greetings, Loek

Ok loek here we go.........

Originally posted by PWatts
Here're two very preliminary 1kHz and 10kHz squarewave tests of a proto board. This is minus output stage, 0 bias and at only 35V rails so not at all proper operating conditions, but it gives an idea. Transistors were 2SA970/2SC2240 input, 3310 mosfers, 2SA968/2SC2238 predrivers and MJE15032/MJE15033 drivers. R and C values are the same as in the BOM.

very nice indeed PWatts. Yes it gives us an idea and thats not a bad one.

[Flodstroem]

This is my solution for the drivers heat sink: on the rear side of the main board (solder side). Board will be clamped by the drivers heat sink to the ventilated bottom of the chassis, something like this:

[Flodstroem]

another couples of views:

[Flodstroem]

last one:

[PWatts]

One thing I can't remember if I've mentioned: the boards were designed exactly according to the originals. The driver transistors' screws were not insulated with a plastic bush i.e. the screws made contact with the collectors. At the bottom of the board, they tightened it with tooth washers, so in effect the collectors made good contact with the power planes directly via the screws. However, this means that the heatsink must be well-insulated. What Krell did was to drill the holes quite a bit larger, and then fit 3mm inside diameter bushes inside the holes to ensure there will be no electrical contact between the screws and heatsink. The easiest though is to simply use Nylon screws of course.

[Flodstroem]

Originally posted by PWatts
One thing I can't remember if I've mentioned: the boards were designed exactly according to the originals. The driver transistors' screws were not insulated with a plastic bush i.e. the screws made contact with the collectors. At the bottom of the board, they tightened it with tooth washers, so in effect the collectors made good contact with the power planes directly via the screws. However, this means that the heatsink must be well-insulated. What Krell did was to drill the holes quite a bit larger, and then fit 3mm inside diameter bushes inside the holes to ensure there will be no electrical contact between the screws and heatsink. The easiest though is to simply use Nylon screws of course.

I dont think Nylon screws are the best solution for this application though its seems to be the easiest. Nylon screws do change their strange when mounted in a hot environment, they could even melt down if to hot. Best solution as I can see it would be to use the original "Krell method" As insulator bushes you can use a flexible urethan/glass fiber reinforced tubing. Those type of insulating tubes could withstand several hundreds of °C.
And also could be easily cut with the help av a scissor or a wire hand cutter.

As you maybe could observe at my boards, the pre-drivers are fully insulated types and for those transistors there is no need for insulating hard wares.

The heat sink for the drivers (on rear side of main board) will be fully grounded by the mounting screws from the bottom chassis. Because of that the driver transistors had to be mounted with all necessary insulating hard wares.

Regards

[thomasfw]

Can I still chance to get 2 pcs of the PCB, thanks !

Thomas

[analogair]

I'm interested in buying a set of the Krell KSA 100mkII Clone PCB, like the ones Flodstroem uses

By the way, Nice Heatsinks Flodstroem