Rackmount Bass Preamp

Walter Harley, September 2000

Warning!

This page is old. If I designed this nowadays I'd do things differently. I've added some comments at the bottom, but basically, at this point this is more of a historical curiosity than anything else.


The grey thing in the rack above is the preamp I gigged through when I played rock. (9-2002: these days I mostly play folk and bluegrass, so I'm playing upright through an Eden WT-400, don't need the powerful rig.) Here's the schematic. I have tried various commercially available preamps; some didn't have the tonal warmth I was seeking, and others had too many bells and whistles and took up too much space in my rack. This preamp is probably not right for most people, but it fits my needs to a T.

Bells and Whistles

The preamp has four controls: input gain, high rolloff, distortion, and output volume. There's also a threshold LED that helps with setting the input gain. If the threshold LED is lighting up, there's enough signal to make the distortion kick in. On the back panel, there's a tuner output and a main output. The tuner output has signal even if the output volume is down, so you can tune silently.

Tone

The most noticeable difference between this and other preamps is that it has no EQ section. I play active basses with multiband EQ on the bass; every filter section adds noise and compromises transient response, so if I don't need it I'd rather not have it. Instead, there's a single lowpass filter that rolls off everything above the selected frequency, leaving the lower frequencies largely untouched. Active filter stages necessarily have a tradeoff between how smooth the frequency response is, how sharp the rolloff is, and how much phase distortion is introduced around the rolloff frequency; the filter in this preamp is designed to minimize phase distortion and maximize transient response at the expense of sharp rolloff. In other words, it tries to do a small amount well rather than a lot poorly. It works very nicely to reduce string noise and sit the bass better into the mix, without making the bass sound dead or sluggish.

Distortion

The preamp is completely solid state, to achieve high reliability and minimize size and weight. I tried to simulate the sound of a slightly overdriven tube amp with the distortion circuit shown here. It's a very mild distortion, almost more of a compression; and, like a tube amp, it is asymmetric, introducing primarily even harmonics. Turned up all the way, especially with the input gain cranked, it can sound pretty fuzz-y; but it's at its best somewhere in the middle, introducing just a bit of grind without really jumping out at you. I use a Stewart World 2.1 power amp, which is pretty sterile sounding and has way more headroom than I need, but with the bit of distortion from the preamp I get a nice warm tube-y sound. The distortion comes *after* the lowpass filter, so that the upper harmonics it adds don't get filtered back out. This is closer to the behavior of a real tube amp played loud, as opposed to a hybrid amp with an overdriven tube input stage and solid-state output stage.

Design Notes

There are a few problems with the design as shown here. First, there's no DI output. That was simple laziness on my part at the time I was building it; sooner or later I'll remedy that. Second, the tapers of the pots need work. For instance, the value of R21 is intended to make R20, a linear taper pot, behave a bit more like a real audio taper; but frankly, it didn't work. I need to find a better source of small, good-feeling, audio taper pots. Similarly, R10/R11 is a dual audio taper pot, but it should really be a reverse audio taper pot in order to make the knob make sense. As it is, fully clockwise is 200Hz highpass, and fully counterclockwise is wide open, which is counterintuitive. (Hey, *you* try to find an inexpensive precision dual reverse audio taper pot.) Another alternative would be to use a rotary 12-position switch and fixed resistors. Third, it would be nice to have a mute button, so I can switch basses or tune without anyone hearing. And finally, I constructed the unit in a salvaged chassis, including its power supply (see below); I don't know whether its transient capacity is really as good as it could be, although it's not been a problem.

More Design Notes

As mentioned above, this design is old and I've learned a few things since then. If you're interested in building it, here are some observations that might help.

The input section adds too much thermal noise and doesn't do anything to filter out RF. To fix this, change R1 to 1k; increase C1 to 0.1uF (use a polypropylene or polystyrene, not ceramic, cap). Change R2 to 1MEG, and put a 100pF ceramic cap in parallel with it. This will lower the input impedance to 1MEG from the 5MEG that is shown; in my experience, that is not a problem unless you're trying to amplify an unbuffered piezo pickup. If you are, try 10MEG instead of 1MEG for R2. Note that these changes will increase the gain by 3; I honestly don't remember why I felt that it was important to start by attenuating the signal, but if for some reason I was right, the place to do it is after that first opamp, not before it.

C3 and C4 should also be polypropylene or polystyrene, rather than ceramic. C2 and C7 should be bipolar electrolytics, not polarized ones as shown.

You can use an NE5532 instead of the LM833 for IC1, if you so desire. Or another OPA2134, or really just about any quality audio opamp. And just about any NPN and PNP transistors will do for Q1 and Q2.

Construction Notes

As you can see in the picture above, to build it I simply salvaged a piece of used rack-mount gear - in this case, an old BGW crossover. I used the existing power supply - regulated, +/-15V. I also used the existing holes for back-panel jacks, covering over the remaining holes with some brass mesh I had lying around. I periodically go to used gear stores and buy cheap rack-mount gear for exactly this reason, focusing particularly on gear with minimal front-panel controls and with lots of 1/4" and XLR back-panel jacks. The unit itself is built on a prototyping PC board with prefab pattern. It took less than a day to put together, as I recall.

Return to Cafe Walter