Building amps is fun. Getting the case/enclosure work is not.

Building amps is fun. Getting the case/enclosure work is not.

 

I really enjoy the case and design aspect.

Ditto but sometimes it is just a pain. 

I have never used tubes before, so I am a complete novice when it comes to filaments.

Thank you Kevin and Victor for putting up with my dense lack of knowledge...

Here are the transformers and wiring I am using for my build.

The transformers are all Triad Magnetics and their product # is in the PDF file.

Feel free to berate me if I screwed up

https://www.dropbox.com/s/75av9kajl6zin6u/Megatron%20tube%20transformers.pdf

Edited May 17, 201412 yr by eggil

That looks good.  I still don't know why we made the front end filaments 6.3V though.  They really should have been 12.6V or even 25V.  Hell a LM7824 with a diode on the ground pin would be perfect for that. 

That looks good. I still don't know why we made the front end filaments 6.3V though. They really should have been 12.6V or even 25V. Hell a LM7824 with a diode on the ground pin would be perfect for that.

What makes using 12 or 25 volts better than the current 6? And would it be possible to use the current boards at those filament supply voltages?

Higher voltage means less current so easier to power using regulated DC.  No possible to do on the current boards without cutting some traces. 

I was thinking on using these Amphenol connectors for the umbilical cord:

 

MS3106F28-12S

http://www.mouser.com/Search/ProductDetail.aspx?R=MS3106F28-12Svirtualkey65400000virtualkey654-MS3106F28-12S

and

MS3102A28-12P

http://www.mouser.com/Search/ProductDetail.aspx?R=MS3102A28-12Pvirtualkey65400000virtualkey654-MS3102A28-12P

Edited June 6, 201412 yr by eggil

There are some vintage? military connectors listed on ebay that would fit the bill, I can't remember in which thread they appeared. Definitely less than 74 a pop

you only need 6 pins which are a lot cheaper

You mean you are installing all the filament transformers in the amplifier, not in the power supply box?

The filament transformer for the CCS tubes have to be close to the amplifier board. 

actually you need 8 pins. the ccs transformers absolutely have to be inside the amp box, and you

might as well put the other one there too.

I would also run the raw mains on a separate umbilical.  For safety and all that something like the Neutrik Powerkon would be perfect. 

OK, then:

1- the Neutrik Powerkon for the raw mains going to the filament transformers.

2- the 8 pin Amphenol for the main transformer.

Edited June 7, 201412 yr by eggil

I'll be doing something like that.  I'll probably rewire the front end filaments to run on 25VDC too. 

Here is first power up. All LED's turned on. The 2 closest to the tubes not as bright as the other 2.

Tubes got warm but did not see any glow. Where is the glow??

Got cold feet after a couple of minutes and turned it off.

Thumping heart rate @ 140 per minute.  

Nothing exploded. 

 

Edited June 24, 201412 yr by eggil

Nice going. 

Oh my goodness! They are so beautiful!

I can now see why people like tubes!

My wife is in Finland with my camera. iPhone pics.

Edited June 28, 201411 yr by eggil

Noice!

Congratulations on the successful power up. I am far behind you. Just got my casings and my amp board isn't even populated yet.

Almost to the point I can hear music. Perhaps today.

The casing though.... that will take time

DC Offset is staying around 4V for both channels.

I am concerned however that the heat sinks for the 8N80C's are a toasty 80 degrees Celsius.

I know there is more current than the KGSSHV because the 3904's are not installed, but isn't this a bit too much?

My KGSSHV's were in the 60 degrees range.

And another Megatron comes to life!

Bach sounds glorious

Many thanks to Kevin for this wonderful design, to Birgir for preventing a major disaster with the PSU, and to RiStar and Victor for their much needed assistance.

Edited June 28, 201411 yr by eggil

I'm considering tricking mine out with another hv supply for the 300v line, anyone know of appropriate low power (ie reasonably sized) transformer? how does the srd generate bias in that small package?