Build name: Black box
This preamp is part of a bigger picture idea I had in mind. To build a couple of bare-bone preamps, no switching, no EQ and other nonsense, to get a decent graphic EQ pedal and to build digital logic for switching them. Concept of having digital circuits only control all-analog signal path was always interesting to me and this preamp is the first step in that direction.
I wanted to start with a high gain preamp that will cover Dream Theater grounds, so I started off with Soldano SLO 100 design. I removed the tone stack, FX loop, clean channel and the power amp. Later I found that it’s actually closer to the x88r preamp. Removing the clean channel left me with a spare triode in the first tube. I decided to wire it in parallel with the first gain stage. To compensate for the change, the first plate resistor is lowered to 120K so that each triode sees 240K (close enough to stock 220K value) and I put two 1.8K||1uF pairs on the cathodes as I didn’t have 900ohm resistors, nor 2uF caps. With those changes, parallel stage should be biased about the same as in the stock circuit. Benefits of the parallel stage should be lower noise, a bit less treble due to doubling of Miller effect and more complex tone, since two triodes are never the same. I also included cathode follower mods suggested by Merlin that’s supposed to reduce the stress on tube during startup without affecting the tone. Finally, I decided to use 12.6DC heaters, regulated and heavily filtered. Heaters are also elevated to 75V to make cathode follower tubes’ lives easier. Without the elevation cathode followers are operating at higher cathode-grid voltage than they should, according to data sheet – Vh-k(max)=180V.
Preamp – click to open full size
Power supply – click to open full size
Chassis layout – click to download PDF file
I found a nice sheet metal pre-painted box in the local store that seemed OK for the job. What’s nice about it is that every panel may be removed individually which makes it easier to install boards, and mess with the circuit later. Front and back panels are aluminum and the rest of it is steel.
For circuit boards I decided to use some pre-made generic turret board and tag board I had lying around. Making circuit boards is not my favorite part of making amps, especially drilling boards, so these boards saved me some times and nerves.
- 300-0-300VAC@40mA high voltage secondary
- 14-0-14VAC@2A heater secondary
I got both secondaries center-tapped because it’s easier to layout full wave rectifier with only two diodes and it doesn’t require virtual center tap to elevate the heaters. 14VAC are there to provide few extra volts before the regulator.
Both volume and gain pots’ wipers are connected to the board using shielded teflon wire. Shield is grounded on the board side and connected to the pin 3 on the pot. That way pot is grounded together with the stage it controls.
Click on a thumbnail to play the video on YouTube.
Click on an image to see more details.
This preamp turned out exactly as I wanted it. It gives awesome lead tone that I can shape later with my graphic EQ pedal. Leaving the 4th preamp tube from the SLO circuit makes the preamp more versatile. With gain pot I can control the amount of distortion that comes mostly from the cold clipper stage, but with volume pot over half the last stage gets overdriven. It gives different overdrive characteristics, since it’s warmer biased and followed by the cathode follower. Backing down on gain and cranking volume sounds warmer and more compressed than doing it the other way round. By playing with two knobs I can dial anything between warmer crunch tones and heavier distortion which is very cool.
- Lowered first coupling cap from 22nF to 10nF to tighten the low end just slightly.
- Lowered one of the two bypass caps on the input stages from 1uF to 0.47uF, thus making each of the parallel triodes “see” ~0.73uF which is awfully close to Marshall input stage bypass cap. This mod made the most noticeable effect on reducing the “flabbyness”.
- Added 470K resistor after the first coupling cap to ground. When I took the lead channel out of SLO/x88r I neglected the load of other channel(s) to the first stage. This compensates for it and reduces gain a bit (which is a good thing in this case!).