A visitor to this website, Frank, from Germany, has built the fender amp and cabinet simulator. He has suggested some modifications based on his own listening tests. They appear in the main table of contents of this website. He has included some mp3 files at his site which is linked to here. They sound great!
Not shown on the schematic for the fender amp simulator is a pair of 68 ohm resistors in line with the power supply leads of op amp "6". (one for the positive lead and one for the negative lead). This will duplicate the effect of a tube rectifier like type 5Y3. This will give that "sag" that many guitarists seek.
Feel free to ask any questions you may have.
Table of Contents
The more desirable "Marshall" amps have a noticeable soft clipping rather like that of a diode clipper. As the signal strength progresses well beyond the threshold of soft clipping, hard clipping occurs. This can be simulated by preceding a diode clipper with an op-amp stage that does good hard clipping.
In the LXH2 "Marshall" simulator, the op-amp stage before the diode clipper has a +/- 9 volt split supply. The diode clipper is a pair of 1N914/4148 type silicon diodes in parallel in opposite directions. A 10K resistor feeds one end of the diode pair from the previous op-amp stage; the other end of the diode pair goes to ground. This provides a very wide range between the thresholds of soft and hard clipping, which we found to make a good "Marshall" tone. To have a narrower range between the soft and hard clipping thresholds to more accurately simulate most authentic vintage Marshalls, use more than one diode pair in series. To have a different hardness of soft clipping, use a different diode type. Silicon rectifier diodes generally clip more softly than 1N914 types, most red LEDs clip harder. "Low-current" type red LEDs (gallium phosphide red such as Radio Shack 276-044) clip rather hard.
"Fender" amps have only subtle soft clipping and primarily hard clipping. If you monitor the output of a "Fender" amp with an oscilloscope, you may be surprised to see the output clipped almost as hard as that of a solid state amp. This is typical of push-pull tube power amps with 6L6 output tubes.
This clipping can be simulated with an inverting op-amp stage, with a series pair of zener diodes in parallel with the feedback resistor. The 1N4733 5.1 volt 1 watt zener diodes work well. Lower voltage and lower wattage zener diodes often don't clip hard enough. To add a touch of soft clipping, add a series string of a large number of parallel pairs (opposite direction within each pair) of both silicon and germanium diodes. This is to achieve a small amount of "waveform compression" (the soft clipping) before hard clipping occurs. The number of diode pairs needed depends on the exact type and brand of diodes, since they can vary somewhat.
Crossover distortion is affected by any control used to adjust the bias of the output tubes. Amps generally sound better if this control is adjusted for a more positive/less negative bias, but without having excess idle current through the tubes or the output transformer. A higher bias favors less crossover distortion.
"Bias shifting" is included in some solid state tube simulating devices for a more realistic sound, even though the most desired tube amps have less of this effect.
As much as people talk about even harmonics and asymmetric distortion to produce them, push-pull tube amps with any control affecting output tube distortion symmetry (typically a bias symmetry control) are generally judged to sound best when output distortion characteristics are most symmetric. Asymmetric distortion generally leads to a rougher, more "hoarse" intermodulation effect when chords are played.
In "Marshall" amps, the output transformer has a bass response that rolls off gradually in the bass. In the "Marshall"-simulating LXH2 circuit, this is accomplished by adding a .1 uF capacitor in series with the 10K resistor in the diode clipper to achieve a distortion-dependent slight bass rolloff.
Some "Fender" amps have a slight treble attenuation from a loaded output transformer, which the negative feedback circuit normally corrects. The onset of distortion is accompanied or slightly preceded by a slight reduction in treble response, known as "darkening". This can be simulated in the LXH2 "Fender" simulator's distortion circuit by adding a capacitor, perhaps with some series resistance, across one or two of the diode pairs.
When a tube amp distorts, its output impedance is even higher, since lower output impedance is dependent on the negative feedback circuit functioning. The high output impedance of a distorting tube amp results in a bassier speaker sound, sometimes even resonant for closed box speaker cabinets.
To give a solid state amp a higher output impedance, simply add a resistor in series with the speaker cabinet. A resistor with a value near or slightly above the speaker's rated impedance usually makes the speaker sound best. This is not quite perfect, since the output impedance of a distorting tube amp is not constant and nonlinear, but this seems to work quite well.
Another trick is to use an equalization circuit, whose response resembles the impedance curve of the speaker, using resonant means at any frequency where the impedance rise is associated with a resonance. To adjust the bass characteristics of a closed box cabinet, one can use a parametric equalizer. Boost the resonant frequency of the cabinet, using a Q that matches the desired Q. The degree of boost in dB would be 20*log of the ratio of desired Q to the Q the cabinet has on the amp in use, OR whatever sounds best.
The LXH2 amp-simulating preamps have resonant bass-boosting/cabinet-simulating circuits. These circuits are made to simulate the low frequency characteristics of actual cabinets driven by amps of high output impedance, for use with headphones or non-guitar speakers. These circuits have been found to work well with guitar speakers on solid state amps and non-distorting tube amps, despite the combined low frequency ultimate rolloffs of the speaker cabinets and the cabinet simulating circuits.
This distortion is a gentle, assymetric "even harmonic" type of distortion. Although this generally does not detract from good tone, guitar speakers still sound good with a properly distorted signal at any power level. Our listening tests indicate it is not necessary to duplicate this distortion in order to get good or realistic guitar tone.
The more desirable forms of distortion require distortion like that of power tubes.
However, there are a few quirks some op-amps have.
1. Odd effects of input voltages going outside the "input common mode range". If the voltage at either input goes outside the usual working range, the op-amp may do strange things, including some sort of "clipping reversal". When well into hard clipping, "reversal" can cause the op-amp's output voltage to suddenly reverse, resulting in a bad distortion sound. To avoid this, have a mild voltage divider before the stage to protect from this. This is especially needed if the previous stage may clip and feeds the stage in question through a capacitor, since the flat part of a clipped waveform can get "tilted" by the capacitor and the leading portion of the "tilt" will have an extra high peak voltage.
2. If you have lots of gain, be sure to use capacitors to block DC before any offsets get amplified to significant levels.
3. If one op-amp stage in a multiple stage IC drives a heavy load, then the distortion characteristics of all stages in that IC may be modulated by the total output current or consumed supply current. Some op-amps have internal protection resistors in series with the IC's supply leads. Please also beware the protection is foolproof only at lower supply voltages.
4. The 324 op-amp has "crossover distortion". So may some other lower power consumption op-amps and some op-amps designed for single supply use. This distortion typically does not occur if the load is a resistor that terminates at one supply rail; problems are largely associated with zero crossings in output current.
5. Low power consumption op-amps may "slew" instead of clip, especially at higher frequencies. The op-amp should have a slew rate high enough to provide normal clipping, typically above a volt per microsecond. Low power op-amps may also have low gain at higher audio frequencies. You should have open-loop gain well above the wired gain at all audio frequencies for desired frequency response and for the negative feedback to reduce distortion while the op-amps are running "clean". Some low power type op-amps may simply have bad hard clipping characteristics.
Click on here forDisclaimer and Copyright Info.