Tubelab SE inverted assembly

[Nikon1975]

Dear George,

I am assembling the TubelabSE with the big components on the lower side of the board. I assembled the capacitors, but I am not sure about the semiconductors. How hot do they get ? I would not like to cook the board putting them down.

By the way I took some pictures of the boards, and I will be happy to send them to you if you want to use them for documentation.

I already assembled the Simple SE, and everything is working fine. A couple of questions: what happen if I switch on the diodes rectification by mistake when the rectifier tube is installed ?

What id behind the choice to lift the heater ?

Best Regards,

Davide

[rknize]

Originally Posted by Nikon1975

I am assembling the TubelabSE with the big components on the lower side of the board. I assembled the capacitors, but I am not sure about the semiconductors. How hot do they get ? I would not like to cook the board putting them down.

They get fairly hot, especially the MOSFETs. I originally built my TSE this way. Note that when you mount them on the bottom, you have to turn them around such that the heat sink legs don't line up with the PCB. If you turn them around and space them away from the board carefully to clear some of the resistor leads that are right there, it can be done. The 5-lead regulator cannot be done this way and the only way to do it is to dnagle it off the PCB line I did here:

(YATSE) Yet Another Tubelab SE (sorry...long)

I already assembled the Simple SE, and everything is working fine. A couple of questions: what happen if I switch on the diodes rectification by mistake when the rectifier tube is installed ?

Nothing. The silicon diodes have lower forward drop, so they will take the load away from the tube. You will just be burning power running the tube heater for no reason. I have a lot of capacitance in the supply on my SSE and it takes several seconds for the B+ to settle when switch from silicon to tube rectification.

[Nikon1975]

I am thinking of putting them up, I don't like to have them down, they will heat the electrolitic caps.

So I am thinking of using something with the shape of a cilinder between the chassis and the tube socket. My question now is: all the 300B have the same diameter for the base ? if the socket is installed lower than the plate of the chassis, how big should the hole be ?

Is there any way to electrically isolate the heatsink from the semiconductor, so that I don't have B+ on them ?

Thanks,

Davide

[rknize]

Originally Posted by Nikon1975

My question now is: all the 300B have the same diameter for the base ? if the socket is installed lower than the plate of the chassis, how big should the hole be ?

They are roughly 1-3/8", so you would want to allow for at least 1.5". You'd want to go considerably more if you are hoping for some convection.

Originally Posted by Nikon1975

Is there any way to electrically isolate the heatsink from the semiconductor, so that I don't have B+ on them ?

Yes. There is a classic mica wafer and white grease approach or you can go with the fancy thermal pads that aren't as messy and often perform better. Get the thickest ones, if that is the case.

[SGregory]

If the heatsink is anywhere exposed that would allow someone to touch accidentally, then I would isolate with mica/grease or a silpad and then bond the sink to ground.

For mine, to pull the heat away from close proximity to the caps, I made a heat sink plate out of 1/8" aluminum plate. The chassis stil gets warm, but the intense heat is further away. Also be wary of the 5W, 10k (R6) resistor, it gets extremely hot as well. Keep it as far away from the board/components as possible.

[SGregory]

For R6, I was looking at installing a Caddok MP820-10.0kf-ND or MP925-10.0kf-nd on a heatsink.

[BillEpstein]

I'm planning on running wires from the tube socket holes on the board to chassis mount sockets. Being able to insert the stoppers at the sockets should negate the downside of socket extensions?

Have I failed to anticipate any other catastrophic or annoying issues?

[rknize]

Are you referring to the input/driver tubes? I put a carbon comp on the PCB and have not had problems with oscillations even with a "socket-saver" as an extension. But yeah, moving the resistor closer to the socket won't hurt.

[BillEpstein]

Getting cold feet on wiring the tube sockets and have decided to simply put them on the revrse side. If I add a nice long piece of 1/8" aluminum bar to the doubled up heat sinks at the edge of the board would that keep things cool enough?

[rknize]

It will help. All I have done to mine so far as the chassis comes together is put this little 1" angle stock on. It's enough to prevent thermal shutdown with 300Bs:



They key is airflow. Adding a larger sink to an enclosed chassis will delay the inevitable. If the bar attaches to the chassis somehow, it should be fine.

I hope you aren't planning to mount the PCB upside down. The tube sockets will have their pins swapped. It won't matter for the heater, but having the grid and plate pins swapped will be bad.