Furniture Update: Studio Desk

Here's an update on the Foray into Furniture:

The first photo is an overall shot, including the CPU, Digi 003 rack interface, monitors and wall-mounted display. All the wood is cherry; the top is 3/4" plywood with veneer banding on the edges. The sides and back are 1/4" plywood. The CPU shelf is 1/4" solid cherry. The rest is solid cherry 1×2s. I used a tongue & groove router bit to make the simple rail, stile & panel construction. The slot cut for the panel also serves as a mortise for the rails. Some dowel joints were also used for the frame under the CPU and to attach the top. The finish is Tried & True oil finish, which is a polymerized linseed oil and beeswax product.

The drawer is made with birch plywood prefab drawer sides from Rockler (they come precut to width, with a routed groove for the bottom panel). I used an inexpensive Wolfcraft dovetail jig to make the drawer box. Side mount drawer glides attach to the upper rails. The only screws used are on the drawer glides and to attach the drawer front to the box.

A rectangular hole was required for the cables to pass through the top. The DVI connector for the video monitor is the longest and the IEC power connector was the widest. I dressed the opening with a custom rectangular grommet (somewhat visible in the photo) made from a scrap of cherry.

Mostly, everything works as planned, except that our low-noise CPU enclosure has a front door to cover the CD/DVD drives. The CPU tower has to be pushed forward so that the hinge of this door doesn't interfere with the leg of the desk.

This last photo is unrelated to the studio desk. This was a mini-project from the summer. I bought the clock movement with the intent of making some kind of desk clock. An ECC832 triode I ordered for an amplifier project arrived DOA and the dealer didn't ask for the defective one's return. I fashioned the clock loosely after a vintage radio and epoxied in the triode as a decoration.

Amplifier Tweaks: Final Lama Kazu 8SE

The tweaks to a simple single-ended amp could be endless. I think this will be the last adjustment to the Lama Kazu 8SE, however. (Note: It wasn't! Some edits to this post were made in October.)


I decided to try a 12BZ7 in the second gain stage/cathode follower. Of course the bias had to be adjusted for that. Not knowing how exactly the 12BZ7 would act at clipping, I decided to add a small (3.9kΩ) grid stopper on the socket.

After using the amp for a couple months with the 12BZ7 installed, I ended up switching back to a 12AX7, but kept the hotter bias caused by the 1.5kΩ cathode resistor I put in for the 12BZ7.

I added a 560kΩ resistor in parallel with the drive potentiometer. This serves to cut the bass response as the drive is increased, and really reduces the muddiness at high gains.

I also decided to go over the whole thing and check my original calculations; it appears I chose the wrong tap on the output transformer when I built the thing. I changed it over from the 2.5kΩ to the 5kΩ tap. It's now actually capable of producing the 8-or-so watts of power I originally expected!

Now with the 'BZ7 there's not a "normal" tube in the whole setup!

Coming soon: Sound samples of everything...

Amplifier Tweaks: Lama Kazu 8SE

The Lama Kazu 8SE was revised last fall; I never got around to mentioning it here until now. The revised schematic is at right. Most of the changes were to reduce the gain, smooth out the overdrive character and clear up some blocking distortion.

• V2 was changed to an ECC832, (JJ's equivalent to 12Dw7) with the high-μ triode at the input and low-μ triode at the driver stage.
• V1 was changed to an ECC81/12AT7.
• The volume potentiometer was changed to 1MΩ and the attenuator resistor was removed.
  
• Coupling capacitors into the tone stack and power amp were reduced to 22nF to drop some of the "mega bass" muddiness.
  
• Cathode bias resistors were changed to 4.7kΩ at the second stage and 3.9kΩ at the driver stage.
• Load resistor on the driver stage was reduced to 39kΩ. This, combined with coupling and valve changes really reduces the output impedance so that grid clipping at the power stage is more graceful.
  
• The 68kΩ dropping resistor in the power supply was reduced to 39kΩ, since the second preamp stage is using more current with the ECC81.

Sound files will be posted eventually. This amp works really well with my Rangemaster clone, known as the "Buffalo Box" and based on an old B175 PNP Germanium transistor I found in my junk bin.