The Sounds of Bias
The debate about bias and its effect on tone has raged for decades. What follows will not end the debate, but it will demonstrate that the Blues Junior is biased unnecessarily hot and that the almost-imperceptible gain in dirty tone comes at the price of lost clean headroom. I did these tests after installing my adjustable bias mod in a late-2001 Revision A (cream-colored circuit board) Blues Junior.
Test Bed Setup
I have a digital oscilloscope (and an old analog scope), but I didn't relish the idea of taking dozens of screen shots with my digital camera. I decided to use two computers, one for the source of my test tones, the other to record the output, to in effect be a digital oscilloscope so I could analyze the waveforms and capture them here so you could see what I saw.
I used Syntrillium's Cool Edit 2000 (now, sadly, off the market) to generate the test tones. It has a powerful, flexible signal generator. I've also used it to analyze waveforms. When you sample at 192KHz you can zoom in on an individual cycle and see the rise, peak, and fall of the waveform. It has just two drawbacks: it's not interactive and preparing a waveform for Web publishing involves a lot of zooming, cutting, pasting, scaling, and then taking a screen capture.
My friend Luca R. introduced me to BIP Oscilloscope , a freeware, PC-based oscilloscope that uses your sound card for input. It has some limitations, but it turned out to be ideal for my purposes. Best of all, you can capture its screens and paste them into graphics programs such as Corel PhotoPaint or Adobe PhotoShop to prepare them for Web publication.
I used an 8-ohm dummy load--a 50-watt power resistor attached to an aluminum heatsink. The audio voltages at the speaker were too high for the line input on my IBM laptop, so I used an attenuator to bring them down to line level. A standard volume box--a 10K potentiometer in a project box, with in/out jacks is all you need. It doesn't have to be calibrated, because all we're interested in is the shape of the waveform, not the absolute values. The particular volume box I used was probably the most expensive volume box you can imagine: NHT/Vergence's Passive Volume Control (PVC), which uses premium components (an Alps stereo potentiometer, Neutrik connectors, and a heavy aluminum box) and costs $150. You can do the same thing with a 10K potentiometer in a project box for less than $20, but it won't have the silky feel and heft of the NHT box. If you build your own, you don't have to put jacks on the box. You can attach a cable with clip leads on one side and a cable ending in a 1/8 inch mini plug to connect to your PC.
Here's the test setup. The machine on the left is running Cool Edit as my signal generator, feeding tones into the BJr's input jack. Before testing, I used a sound level meter and the Fluke Scopemeter to set the signal level to approximately the same as my guitar with the tone and volume controls wide open. It's important to know that clipping levels in the amp under test will be roughly the same as aggressive guitar playing, even though there's a world of difference between a steady-state sine wave and the waveform from a guitar string. I normalized them at 110dB, 3 feet from the speaker.
The dummy load is plugged into the speaker jack. A speaker cord with clip leads on one end plugs into the Vergence volume box (atop the amp), and the output of the volume box goes to the line input of the second laptop. I ran a full-volume test and adjusted the volume box and input level on the line input channel so that the signal to the line input did not drive it into clipping.
Virtually all modern computers have full-duplex sound chips, so you can generate your test tones and test your amp with just one machine. I chose not to do this to eliminate any chance of crosstalk from the input to the output and to avoid the inconvenience of flipping from the signal generator to the scope to Corel PhotoPaint to capture the images. Using two machines also lets you monitor the output of the amp through the second machine's speakers.
This is the dummy load, an 8 ohm, 1%, finned aluminum 50W power resistor that I screwed to an aluminum heat sink. The Mouser part number is 71-RH50-8. These resistors cost less than $3 each, and if you need to handle more power you can get two 16s or four 33s and wire them in parallel. You'll want to get a much bigger heat sink, though, if you're planning on doing extensive tests of high-powered amps. This one gets warm to the touch after 15 or 20 minutes of higher power levels from the Blues Junior, but it doesn't flunk the "thumb test" and become too hot to touch.
I used thermal grease under the body of the resistor to aid heat transfer. In use, the heat sink gets warmer than the resistor body itself. If you're going for more power handling, you can also use a small fan to create some airflow over the heat sink. A CPU cooler would probably work well in this application, because they have to disperse a large amount of heat in a small space.
I tested the Blues Junior with it biased at the factory stock 40ma and at 25ma, and ran both 440Hz and 1KHz test tones through it. I set the volume and master volume at various levels, but left the tone controls in the midway position. Reverb was off, FAT switch out. The amp had stock Groove Tubes 12AX7s and new JJ Tesla EL84s. Ignore the actual height of waveforms; since the knobs were off, I didn't always get the shaft positions in exactly the same position each time. The focus here is the waveforms.
|Comment||25ma waveform||40ma waveform (stock bias)|
|Master 4, Volume 8, 440Hz
The BJr can produce a clean waveform with the volume control as high as 8. With that much preamp drive, however, going past 4 on the master causes power tube distortion. With 40ma bias, the signal is beginning to clip and is starting to become asymmetrical in the bottom lobes. It sounds clean, however. I changed time bases here to show the 40ma waveform better.
|Master 4 Volume 12, 1KHz
This setting drives the preamp into clipping and asymmetry. This is a typical master volume ploy, using the preamp for distortion while keeping the volume down to a reasonable level. 40ma bias causes additional clipping in the power tubes, but there's relatively little at 25ma. The 25ma sound is still pleasantly raunchy, without the clipping in the output stage.
|Master 11 Volume 4, 440Hz
Here's the opposite side of the coin, maximum clean tone in the preamp amplified by cranking the master volume. At 25ma you can see minor clipping; it still sounds pretty clean. At 40ma the clipped peaks are much broader. I expanded the time base for a better look, but you can hear the difference--more distortion with 40ma bias.
|Master 8 Volume 8, 1KHz
Both the preamp and amp are cranking pretty well here, and the BJr is making nice, chunky sounds, and the only difference is that the clipped peaks are a little broader on the 40ma sample.
You can find a difference between the sound of a cool-biased Blues Junior and the hot factory bias if you look--and listen--hard enough. The cooler bias sounds a little brighter, a little cleaner, with a little more headroom. It also sounds a little thinner. But paradoxically, it's a little louder. The emphasis throughout here is "little." As near as I could measure with my setup, the cooler-biased amp is 1dB louder. That's imperceptible.
The cleaner amp is louder because clipping is the enemy of power. When you drive a voltage all the way to the rail and hold it there, you've created a steady state--no magnetic lines of force are being cut, no current flows to the speaker. It may sound louder because it's impulse noise, but it isn't. Granted, the effect is worst on sine waves forced into clipping. Your playing covers many frequencies at many different levels, and only some of them get clipped.
So the upshot is this: Why toast your amp with too-hot bias if you can barely hear the difference, or if the difference can be erased with a small twist of the tone controls? Better yet, why not use that extra headroom, as small as it is, to revoice the tone stack so you don't have to run hot to get chunky sounds?
The other interesting result from this test is a clear demonstration that the Blues Junior generates distortion in both its preamp and power tubes. This is not immediately obvious from listening tests--it sounds like the preamp generates all the distortion and the power amp just makes it louder or softer. The distortion caused by the power section generally kicks in at such a high volume that you can't hear it clearly.
Questions? Comments? Email me.