|
46 Single-Ended Amplifier46 output tubes in single ended operation. Ultrapath connection, direct coupled, reactance drive. 6AN4 driver tubes. I built this over a couple of months with help from John, Gary, and Jack. Thanks! I designed all the major components to sit inside the amp chassis. The tubes sit on top. The power supply is external. Currently the amp is in a wine crate. I am having some nice amp chassis built up from quilted maple MDF. I'll take pictures when these are in. The 46 vacuum tube is actually a dual grid directly heated tube, which in this amplifier is connected in triode mode. The 46 was designed for multiple use, principally for single ended operation (class A) as a driver tube, and for push-pull operation (class B) as an output tube (often in the same circuit). This circuit was designed by Jack Elliano of Electraprint. He designed it for a 45 output tube, then I changed the operating points to accommodate a 46. The circuit with voltages is shown below: 
This circuit is not optimized. Currently we have the 6AN4 running a plate voltage of 188V and a current of 9.4 mA. Tube data for this tube lists 200 V and 13 mA. Originally I had only 4.5 mA running through this tube because I mistakenly put in a 270 ohm cathode resistor. Jack said that it needed decent current (10 mA or more) otherwise the highs will be rolled off. Now we have 135 ohms, the data sheets call for 100 ohms. I'll change it if I get time. The 46 is designed to run at 250 V max and 22 mA in class A (5.5 Watts dissipation). However, the plate structure appears capable of dissipating much more power than this. We decided that 10 W was pretty safe. The tube now runs at 272 V and 38 mA. This is a plate dissipation of 10.5 W. The grid bias is only -23 V, where the tube data calls for -33 V. I don't know what effect this has. Warning: this current is way over the listed specification for the 46 tube, you run this operating point at your own risk. I have now found out that a few members of the Joelist are using similar operating points for their 46s. I achieved approximately 2 Watts power output at 5% distortion, according to brief measurements at John's house. Components20 Henry choke load (also known as "Direct Reactance Drive" by Jack Elliano, or the "monkey" amp on Joenet). from Electraprint. BE5KB single ended output transformer from Electraprint. I believe that the quality of these transformers play a large part in the fantastic sound from this amplifier. Their size and weight is pretty large for their stated purpose (45, 50 tubes and 10W maximum), I am sure there is hardly any core saturation. The 5.3 K primary is good for power output from the 46, but results in higher distortion. A 7 K or even 10 K primary would result in much less harmonic distortion, at the expense of power output. 50 uF 440VAC oil capacitor (very big, 9 inches x 4 inches x 2 inches) for the Ultrapath connection. I will try some smaller size capacitors soon. I think the "slow" sound from this amp might be due to the large capacitors. Hum pots are used, and two more wirewound pots used for adjusting the circuit voltages (at the 269 ohm resistor). Other resistors are various wirewounds from Dale Vishay etc. and a combination metal film and carbon comp at the 135 ohm driver tube (two 270 ohms in parallel). A Black Gate capacitor is used for the 6AN4 cathode bypass. I bought most of the parts from Handmade Electronics and some from Mouser. The power supply is a Lambda regulated bench supply capable of 525 VDC. The filaments for the 46 and 6AN4 are supplied with separate transformers in a filament supply box (grey box above the Lambda supply). It takes three switches to turn on the amplifier. Total weight for the amplifier and power supply is 92 lbs, or a Dunker factor (weight to power) of 46 lbs/Watt. On the workbench before the big 50 uF ultrapath caps are fitted (10" x 17" chassis plate): 
Completed (in temporary box !): 
On stand with power supply: 
Close-up view: 
Testing at John's with Fostex 208 Sigma in TQWT: 
And the sound… ?Very very nice, very detailed and very sweet. Not too jazzy "in your face", but smooth. Good bass definition, beautiful mids. I'm not really an expert, but John L. who has heard a lot of tube amps liked it. They play very well and loud on my Hammer Dynamics Super 12s. I put on some Bob Marley, the track Exodus on vinyl and was awed with the detailed presentation. What stood out was the backing vocalists, whose voices are now much clearer. It is just a lot of fun to listen to. Update 10 July 2002:My long term impressions? Well, I had not directly compared this amp to anything other than Johns 845SE, so I just used it at home and enjoyed it. Good points:Still a very smooth sound. Non-fatiguing, you can listen for long periods. Mellow, fairly detailed. A little fruity and warm. Low noise, hum about 5 mV. Relaxing sound. Not bright or jazzy or forward. Questionable traits:Slow, languid sound. A bit like listening through treacle. The hard consonants (k, p, t, etc.) and notes with a sharp attack seem slowed or rounded. Also the highs seem missing. Not quite bright or sparkling enough. I just heard a Joseph Esmilla 71A amp right after listening to my own 46 on Sherwin's system. I also heard the Larry Moore 10 Squared amp, and Wavelength Cardinal 300Bs. All in a couple of hours. While my 46 amp was extremely listenable and not disgraced by these superb amps, I would personally like some more pace and attack, and better highs would be nice. Sherwin thought these traits (slow sound) are often due to oil capacitors. My amp has one oil cap in the signal path, the 50 uF ultrapath "return" cap, and also oil caps in the power supply (possible second return signal path). Forward signal is direct coupled from tube to tube. So possible fixes include changing the ultrapath cap, or adding a film cap to the ultrapath connection. Also building a new power supply may help. Currently the supply is tube rectified, tube regulated, and lots of large oil caps. If changing the ultrapath oil cap and/or power supply doesn't help, I would attribute the slow sound to the 46 tube. Sherwin thought other 46 amps he has heard did sound rolled off in the highs, but not necessarily "slow". Another reason the highs may be missing is if the 6AN4 is not getting enough current. They have approximately 9 mA now, specification calls for about 13 mA (I think). I'll try some modifications when I have time. Links
Now playing: "Rachmaninoff Concerto No. 3" Vladimir Horowitz, recorded live at Carnegie Hall, 1978 (on LP). Was playing: "Eusa Kills" (on LP) by The Dead C (on the Flying Nun record label). From New Zealand. Top
|
