I was able to snag a questionable factory schematic which got me started. One major goof-up: the 1M grid load for the 3rd stage was hooked up as a series resistor after the 2nd coupling cap so there was no grid load resistor at all. Oops! (The bright switch- actually the WAY TOO BRIGHT switch- would put a 220pF cap across that 1M resistor... no wonder.) Once I figured out a few fixes for the stock design I decided to try my hand at converting it into a SE TW Rocket.
It took a bit of persuading to get the Rocket circuit on that board- Dremel time! I think that what was the hardest was getting the tone stack right (with the recommended Mid control I read about here). I guess the big problem was that I've never wired up a Bass control as a potentiometer (it has almost always been a variable resistor) so when I first fired it up I found that the 1MA Bass control was backwards. No biggie if you have panel mounted pots with wires running between them but on a PCB I had to redo most of the work I had done on the tone stack, and traces I had cut previously would have come in real handy.
*** Important Note *** It is the CCW(cold) terminal of the Bass pot that connects to the CCW(cold) terminal of the Treble pot. (Most Bass pots are wired as a variable resistor so take care if you are reworking an existing PCB.) The layout drawings here do show the correct wiring.
I did swap the stock 250kA treble pot with the 1MA master volume pot because I think that the Rocket tone stack works best with the 1MA pot that KF used. I swapped it back for awhile seeing if I could get more volume out of the amp with the 1M MV pot (it didn't help so I swapped them one more time).
The sound was THERE but it wasn't working quite right so I studied the posts here on the Rocket as well as the schematic drawings of the Torn Sun Dark Star MKII amp which is a SE Rocket. I doubled up the two stages of V1 and put 1k5/22uF on the V1a cathode and 1k5/1uF poly for V1b. I tried running the 100k plate load in the Dark Sun but wanted more clean headroom so I replaced that with 56k. Unlike the DS I have only one input jack so I can't use their trick for connecting 1 grid with one plug and both grids with the other plug. So I connected both stages to the input jack. (If I was to add a switch to this amp it probably would be to select one or the other or both stages.)
I reconfigured 2 of the front panel switches to make this amp more versatile. I rewired the useless 5W/2W switch to bypass the DC-coupled gain stage 820R cathode resistor with a 1uF/50v poly cap. (I had first thought about using that switch to add a 500pF ceramic cap in parallel with the initial 500pF mica coupling cap.) I rewired the Bright switch to add a 500pF ceramic cap across the treble cap, which used a 50pF ceramic cap in parallel with a 50pF mica cap. With both switches engaged we have entered Marshall territory.
I used a 56k slope resistor because I had to back off the treble control to 1 or 2 with the official 100k slope resistor. The 250kA MV goes to a 0.022uF cap which I replaced with a polyester Orange Drop cap (as I did with the bass and mid caps) and the cap goes 4k7 grid stopper on the octal output tube.
I had some issues with the screen and cathode resistors on the output tube. When I was trying to get the Pentode/Triode switch to work better I replaced the 5K6/2W screen resistor with 1K/5W, which made the triode mode a little less muddy. After scrapping the P/T switch I left the 1K/5W resistor in there which seemed to work great until I got enough signal going to the 6L6 to get some volume. It sounded really great- a very high quality distortion which cleaned up really nice as I backed off the guitar volume control. However after maybe 20 minutes of it fully cranked the sound started to cut in and out, and it continued to do that even when I turned the gain back down to a clean setting. A quick look told me that the 250R 6L6 cathode resistor was badly discolored. I replaced it with a 470R/5W resistor and that pulsing cutting in and out crap was gone, but the notes were not sustaining very long. Not really a problem with a lap steel because I can get infinite sustain as long as I keep moving the bar back and forth. And with the note not sustaining you could mute the string pretty damn fast. But of course there was something wrong.
I did a few changes and I'm not sure exactly what worked and how, but I am happy with it. I first replaced the 1k/5W screen resistor with the PicoValve stock 5k6/2W, brought the 100k screen resistance on V1a/b down to 56K and added a 51pF cap between the grids and plates of V1. After doing all of that the amp was very stable, however it sounded a bit muffled compared to how it had been before. So I replaced the 5K6 screen resistor with 3K3 and replaced the 51pF squelch cap with 33pF. Voila! Everything is working really nicely now.
For tubes I have a GT5751M (my favorite preamp tube still in production) for V1 and a GT12AX7M for V2. The 6L6GC tube is a GT that I had around. (No I am not a GT salesman!
) The M as a tube number suffix means it is a Mullard copy/reissue from China (as compared to the Mullard branded tubes made in Russia). For some reason the GT5751M is still available but the GT12AX7M isn't. I had originally thought that the 12AX7's sold out much quicker than the less well-known 5751's and it was a matter of time before they were all gone. However they came out with a fresh batch of 5751's a few years ago that had paper labels instead of silk screening on the tube itself. So I have a hunch that it is a legal issue with EHX/Sovtek making the "official" Mullard 12AX7 reissues. FWIW I have A/B'd the GT5751M's with my all-time favorite preamp tube, the 5 Star GE 5751's, and they come pretty damn close. (I only have a few of the 5 Star and like to save them for testing- if a high gain circuit doesn't sound better with a 5 Star 5751 there is no need to try any of the lesser 5751's.) Pix and clips to follow.
Steve Ahola
Here is a link to the Dark Sun schematic:
https://tubeamparchive.com/download/file.php?id=10949
