My first reaction was to pull out a 6SL7. Ahhh I thought; a mu of 70 ought to help. But examining the curves I see that almost no matter WHAT I do with this tube (within the voltage contraints I am using) it stays pretty darn linear. A good thing, no? Well ... yes. Except that part of the "charm" of the Darling is what might be a bit of anti-1626 curve built into the driver. With the 6SL7 I get the gain, very low distortion and I will definitely hear the 1626 ... for whatever it sounds like.

So started a path of tube searching.
Objective 1. Don't BUY any more tubes! I got a closet full already! I examined the 12AX7 (the universal gain cure), but found no happiness there. I rarely find happiness in a 12AX7. I looked at a lot of my favorites and they are all below u40. Sure a 6DX might sound fine here. But then I might as well use another GE 5965 and it'll sound better. This I already know.

So find a tube with an anti-curve at about 100-150V plate with high mu and lots of current. Hmm ... back to the 5965 except the mu. The 12AV7 looks interesting. Mu is too low [sigh].

Ojective 2. Keep the 1626 input grid resistor at about 220k. Sure, with the self bias one might venture into the 500k range and I might try it. But for now, I'll keep it at 220k. Don't want to raise the grid resistance too much on what is essentially an RF amplifier tube.

Then I tried a 12AT7. A "cousin" they say to the 5965 with a mu of 60. Well, it must be a distant cousin. They don't even sound alike. But that's where I am. A shoot out between the 6SL7 and the 12AT7 driving the 1626. Which will be "better"? The linear driving 6SL7 or the very unlinear 12AT7? That's where we're headed this week.

Later ...
OK next. I like driving a tube from as low an impedance as I can. I try to keep it no worse than 1/3 of the input Z. I could operate the 6SL7 using a 68k plate resistor. But then I'd throw away all the gain. I layed out a curve for 120k and calculated the AC curve. Sure it tilts up as it must, but the tube stays in the linear region well past my needs here and I get to keep some gain. So we have a firing solution for the 6SL7.

Now to the 12AT7. Here I have a lot more leeway except I WANT the tube close to the dirt. My B+ is only 240V. Let's do 47k. Calculating the AC curve tells me quickly that I'm still in the same basic shape with no risk of hitting the bottom before the output tube is already past its ability. OK then. Can't put it TOO high on the voltage side or I will hit the rock. So we'll have to settle on about 115V plate.

Still much later ...
Well I did the dirt, but frankly the 1626 sounds just fine on its own merits. I guess that should not come as a surprise. So the final circuit is a 6SL7. The AT7 worked fine and sounded nearly as good as the 5965 I built earlier. I wanted something different.

So the 6SL7 ended up at a 120k plate load driving 220k g1's at the 1626. After calculations and listening, this is working out well. I've plenty of gain so I don't use the input stage bypass caps. I'm glad for that. Seems like they are always so large compared to the ability of the circuit to slew their current needs. I'm never happy with them.

In this amp I have a pair of output trannys removed from an old Japanese receiver (I think ... it's been a while). Nice large-ish, good sounding pair at 5K:8R. These are coupled to an NOS pair of JAN RCA 1626's driven by a vintage black round plate RCA 6SL7.

The power supply tube is a great working Zenith engraved base that looks like an RCA but don't hold me to that on the 6X5. As with my other Darling, I opted for fast, low noise SS diodes as the main rectifiers feeding the DC to the 6X5 thus giving the slow warmup characteristic. The 6X5 breathes easily under this configuration. The power supply is dead quiet.

There are other new and vintage parts throughout this amp. While the power supply bypass's are all new (a total of 300uF) the coupling caps are extreme vintage Paper/Oil (PIO) caps that help give this amp its characteristic smooth sound. The bypass for the self bias resistor for the output tubes is provided by another excellent NOS vintage cap. In addition to these caps, there are decouplers strategically located around the amp to cut noise as well as hash filters in the heater circuit. There's no hum.

All vintage carbon comp resistors are used everywhere except for the vintage 500R self bias resistor in the output section. A bus ground system is used with a star type power supply feed. The input 6SL7 is decoupled from the main supply as in my previous Darling. The wonderful power tranny is the same I used in my previous version also. It's great in that it has 3 heater circuits which allows me the luxury of some additional 60Hz hum phase cacellation. It's my last tranny like this and I wish I had a 100 more.

I used a metalic blue for this one. I've always wanted to try it. Well, I don't have good things to say about using metalic paints, but the result is lovely. I left the output tranny cores the original color. I don't like painting this part unless I have to.

Gold input and output connectors, silver (old mil spec stuff) signal wire to the input grids, silver bearing solder and oversize copper ground and power planes. That sums up the construction. Built into my home made aluminum chassis. This will be the last 2.5" tall chassis I build as I am going to a 2.25" style for this class of amp.

The Sound This amp delivers excellent sound as one would expect from this configuration. The quality, imaging and depth are all there. The beautiful Triode bass detail and silky overall sound attest to the success of this design. Thanks for looking. Your comments are welcome. Critisms gracefully accepted. "Newbie" questions courteously answered. Email to: abraxasaudio at att dot net (this avoids the address scanners) May I build one for YOU? jef