Revisiting some "old" ideas from 1970's - IPS, OPS

Hey!

The front-end has been built and tested a few times with excellent results.
Talking about the full amp - Lichtstark-X - it has not been built as yet, as far as I know.

Cheers,
Valery

Huh ) I was actually wrong - Jeff has prototyped it and published his initial test results in "Sons of VHex" thread.
I'm also going to prototype and test it in the nearest future.
 
Close to finish

I feel that now it's the time for final procedure!:eek:
 

Attachments

  • DSC09608.jpg
    DSC09608.jpg
    271.3 KB · Views: 324
  • DSC09610.jpg
    DSC09610.jpg
    222.9 KB · Views: 332
Now you're into the fun part!If your transformer will fit the amp would work out nicely with the transformer attached to the power connector end of the chassis with the supplies stacked back to back next to it. That would keep all the noisy parts away from the input end and everything in one box. Cooling would be improved if you flipped the whole thing upside down putting the output devices below the center of the heat sink.
 

Attachments

  • DSC09612.jpg
    DSC09612.jpg
    610.7 KB · Views: 271
Now you're into the fun part!If your transformer will fit the amp would work out nicely with the transformer attached to the power connector end of the chassis with the supplies stacked back to back next to it. That would keep all the noisy parts away from the input end and everything in one box. Cooling would be improved if you flipped the whole thing upside down putting the output devices below the center of the heat sink.
Aha i see Jeff, you prefer that solution....all in one box.
1)I dont think that transformer will fit.
2)I want to use the power supply as passepartout!;)
As for the cooling...my plan is to drill the bottom place.The upper will be logo, cnc machined ,cooling helper logo.
Any disadvantage putting the transformer,rectifier out of the box?
 
It sounds like a great looking chassis is in the works! Having the power supply in a separate box works but you need to increase the conductor size due to the extra length and you'll have a lot more area producing noise so you would need to be careful with interconnect wiring.

The optimal location for placing the output devices is slightly below center of the heat sink. Wakefield recommends 40% up from the bottom spread out the entire length of the heat sink. This will only become an issue if you are running above 60V rails though.
 
It sounds like a great looking chassis is in the works! Having the power supply in a separate box works but you need to increase the conductor size due to the extra length and you'll have a lot more area producing noise so you would need to be careful with interconnect wiring.

The optimal location for placing the output devices is slightly below center of the heat sink. Wakefield recommends 40% up from the bottom spread out the entire length of the heat sink. This will only become an issue if you are running above 60V rails though.
Does it mean that would more efficient turning this upside down??:confused:
 
Does it mean that would more efficient turning this upside down??:confused:

If you flip your heat sink upside down it will put the row of output transistors right in the optimum location for cooling. When I did the layout I chose the position of the output transistors to position them correctly on a 6" tall heat sink with the transistors mounted below the board. With the heatsink mounted as you have it they are about 1.5" above the optimum position so the heat won't transfer to the bottom of the heat sink, reducing cooling efficiency.
 
Tribute 3000

I’m back at the Tribute 3000 and am having a problem. I have the driver boards up and running but I have an issue when I wire it all into the chassis. With a bulb in series with the mains the bulb goes bright then dims then pulsates bright, dim about once a second. I haven’t left it powered on more than a few seconds but even without the bulb in series I can see my bench lights pulsate a little. Looking at the schematic I remembered I don’t have the thermistor from the original output transistors connected. Could this be the problem? Without that, part of the protection circuit is incomplete.
Thank you for any ideas
 

Attachments

  • sansui protection.png
    sansui protection.png
    212.7 KB · Views: 250
  • sansui output.png
    sansui output.png
    156.5 KB · Views: 244
  • IMG_4102.jpg
    IMG_4102.jpg
    727.6 KB · Views: 244
Last edited: