Tube DIY Asylum

Posted by John Swenson (A) on May 07, 2008 at 03:25:43

Its time for an update on the BDT preamp saga. For those that don't remember the original posts this is a preamp made with Beam Deflection Tubes (BDT) which were originally made as color subcarier demodulators for color TVs. BUT they just happen to be one of the most linear amplifiers ever made and behave like differential amplifiers with balanced in and out (they make great phase splitters too), with one really neat feature: the gain can be adjusted with a DC voltage on the control grid.

Anyway the original was cap coupled from one output, then I added a Hammond 124B transformer to the output to combine both outputs, this sounded quite good but the tubes were microphonic and quite susceptible to hum pickup. I used it for awhile then tried a bunch of other topologies. Recently I brought it back out and added a Lundahl 1660 output, this was way better than the cheap Hammond, but I kept on thinking I could do better.

Next I talked to Dave Slagle and he designed a custom transformer for this design, this was significantly better than the 1660 (but more expensive). This was getting really good, but was still plagued by hum pickup.

At this point I decided I needed a little more gain but couldn't get it. The effective mu from the deflectors is quite low. So I decided to try cascoding the outputs. I put together a circuit using a CCS into a resistor as a voltage reference and fed that to MOSFETS used as cascode elements between the plates and transformers. This worked great. The gain went up and the distortion went down. (I know its not tubes, but its way easier to build than with tubes). But the hum was still there, thus started the hum eradication efforts.

First to go was the AC heaters, DC helped a lot. Then I determined it was picking up hum from the power supply iron. The chokes disappeared to be replaced by single low C filter followed by active regulators. I know this flies in the face of fashion, but the chokes just give off too much field. Next came the power transformers. I tried several things and wound up with an Antec toroidal, with the primary wired as 240 but actually ran at 120. The fields from this were REALLY low.

I thought this would do it, but no, I still had hum, less than before but still annoying. I eventually traced it to the power supply for the big 813 power amp. I wasn't going to rebuild THAT! So that left shielding. I decided to try thick aluminum. I had heard that thick aluminum makes a good magnetic shield due to eddy currents induced in the metal, but it had to be pretty thick. First I tried Front Panel Express, but it was going to cost almost $600 to have an enclosure done through them with thick aluminum. After seeing that figure I went DIY. I found a guy on ebay selling 1/4" aluminum plates at very low rates so I bought a bunch. I then spent three weeks cutting, drilling, tapping and countersinking aluminum to make the box. This is split into two compartments so there is a 1/4" plate between the PS and the "signal" section.

I put the tubes on their own separate 1/4" thick plate which was isolated from the rest of the box to try and cut down on microphonics. After getting it all built the hum was gone when I listened to it (YAY) but I could still see some hum on the scope when I turned the gain all the way up. I did some more testing and determined it was direct magnetic field pickup by the tubes. The aluminum box did a good job of shielding, the field inside the box was much less than outside, but it wasn't perfect.

I decided to try other forms of magnetic shielding, so I ordered some metglas from lessemf. I made a little cardboard "chimney" wrapped in metglas to put around the tubes. With this the output was clean, no hum visible on the scope at all.

Then I listened to it. Jaw on floor etc. etc. This was magic. The only change was the shielding around the tubes, no wiring change at all. It went from sounding wonderful to magic. Voices in particular are so real sounding its scary. Its hard to describe the difference with and without the shield other than to say the music sounds so much more life like. Its real, its alive. Its detailed without being harsh or edgy or throwing detail in your face; its smooth but not syrupy or slow.

This sounds way better than the best "passive preamp" I have on hand, don't ask, I don't know why this is so. I even tried running the DAC direct to the power amp and it does not sound as good.

I'll post schematics in a couple days if I can pry myself away from the listening room.

John S.