|
|
 |
| VOLUME 9 | MAY 2007 | FREE |
By Terry Downs
One of the most basic requirements of any pedal (Stomp box) is the ability to switch between the original input signal, and the output of the effect. This can be viewed as a bypass switch, or effect in/out function. This was accomplished in the earlier days by using a mechanical push-button switch. Figure 1 shows a mechanical single-pole/double-throw (SPDT) switch that connects the output of the pedal to either the input signal or the output of the effect circuitry.
Figure 1.
This circuitry has two major disadvantages:
1. When bypassed, the signal is loaded by the input impedance of the effect, as well as the audio cable and amp. This typically results in high frequency loss.
2. Depending on the effect pedal output design, you may get a noticeable "pop" or "click" noise when switched. The output of an effect stage may have some DC bias present, which accounts for the "click".
Using a double-pole/double-throw (DPDT) switch, allows for a full bypass. Figure 2 shows a DPDT switch that can effectively switch between the inserted effect and a straight wire.
Figure 2.
This circuit has the same "click" noise problem as the SPDT switch. Another disadvantage is that when routed through several pedals including long wires, the cable capacitance begins to roll off high frequencies from the signal.
The insertion of a buffer amp (amp that isolates the output from the input) that has unity gain (gain of one) is often used. This is depicted on Figure 3. This helps drive the cable capacitance with no appreciable high frequency loss.
Figure 3.
Here are a few requirements that would be desirable for effect pedal bypass design.
1. No noise when being switched.
2. No signal degradation when bypassed
3. Controls a visual indicator, such as an LED, so the user can see if the effect is on or off
4. Assures being powered up in a bypassed state, despite the last state used
5. Uses an affordable, non-complex, momentary single-pole/single throw (SPST) switch element
The typical circuits used in most of the Boss or Ibanez pedals now accomplish all the desires above. A SPST momentary footswitch circuit can be made to define 2 stable states in a push-on/push-off manner. A bistable multivibrator is a circuit that has two stable states and has two complimentary outputs. One output is a higher voltage, while the other output is lower, and vice versa when switched. Figure 4 is a schematic of a symmetrical bistable multivibrator.
Figure 4.
The original form of bistable multivibrator made use of vacuum tubes invented by Eccles and Jordan in 1918. Be modest everyone, because a lot was accomplished before you were born. The example above is accomplished with general purpose bipolar transistors. With no current flowing in the collector of Q1, the voltage at the base of Q2 is biased on. The current in the collector of Q2 causes a voltage drop across the collector load resistor. This drop in turn lowers the voltage at the base of Q1 to continue to keep its collector current at zero. This condition of Q1 OFF and Q2 ON will be maintained as long as the circuit remains undisturbed.
If a sharp negative pulse is applied to the base of the ON transistor, its collector current decreases and its collector voltage rises. This rising voltage is coupled to base of the OFF transistor, causing some collector current to flow. The resultant drop in collector voltage is coupled to the base of the ON transistor. This turns the ON transistor off. The action is thus one of positive feedback, with nearly instantaneous transfer of conduction from one device to the other. There is one such reversal each time a pulse is applied to the base of the ON transistor. Normally pulses are applied to both transistors simultaneously so that whichever device is ON will be turned off by the action. The capacitances between the base of one transistor and the collector of the other play no role other than to improve the high-frequency response of the voltage divider network by compensating for the input capacitances of the transistors and thereby improving the speed of transition.
Figure 5 is the basic schematic for the Boss pedal bistable multivibrator. The footswitch transient action is simulated by using a transistor switched on by a voltage pulse source. The switch is pressed every second, and pressed down for one half second. The LED and zener diode are simulated with an array of diodes. The array of diodes simulate the zener diode and led status indicator...more on that later.
Figure 5.
The colored voltage probes match the colors on the plot in Figure 6.
Let's analyze the way the multivibrator switches from one state to the other.
Notice at the beginning, Q2 is on because the base voltage of Q2 (red) is 0.6V,
and the collector of Q2 (yellow) is near zero volts. The trace from the switch
is the salmon pink probe. It drops from 9 volts to zero volts when the switch
is pressed. When the negative going switch voltage (salmon) is passed to C3
and C4, a negative spike is put on the bases of Q1 and Q2. Both transistors
try to turn back on, but the collector of Q2 (yellow) has risen to about 4 volts
at 1.100027 seconds. The rising voltage is coupled to the base of Q1 (green)
which causes Q1 to turn on. Note that Q2s collector voltage (yellow) is now
rising up as Q2 is now off.
Figure 7 shows the two SPST switches required to bypass the effect.
Figure 7.
Figure 8 is the Boss pedal schematic with the audio switches added.
Figure 8.
Having complimentary outputs from the multivibrator facilitates use of solid-state switches to bypass the effect. Notice the collectors of Q1 and Q2 each bias a Junction Field Effect Transistor (JFET). These are depletion mode JFETS, which means they are on if no bias is present. Notice the audio signals from the input and effect output are capacitively coupled to a bias resistor that centers the audio around half the battery voltage (see the battery voltage divider R14 and R15 that creates 4.5V). When Q1 is on, J1 has a negative gate bias relative to its source. (I don't know why my PSPICE library used J as a reference designator for a JFET, but please bear with me). This causes J1 to be off. When Q1 is on, obviously Q2 is off. The 6V collector voltage of Q2 is isolated by D15. There is no negative bias on J2, so therefore it is on. The effect output appears at the output of the circuit. A buffer/emitter follower is usually placed after the text marked PEDAL OUTPUT to drive the cable. It is also biased back to ground as opposed to 4.5V bias.
One way to get noiseless audio switching is by changing the resistance of the JFET switch very slow. Slow enough that the rise-time is at a sub-audible rate. This is achieved by slowing down the bias to the JFET gates with a low pass filter. The series 1MEG resistors (R21 and R22) and the shunt 0.047µF capacitors (C8 and C9) make the gate voltage change very slow. Figure 9 plots the bias voltages on C8 and C9, along with the audio. It input is a sine wave. The output of the effect is a higher amplitude square-wave. Turn on the CRANK! Notice it takes about 250mS for the bias on the transistors to swap. The audio changes faster than that, but it is yet a nice slow transition.
Figure 9.
Ignore the diode array I needed to put in the PSPICE model for a moment and look at the Boss pedal circuit in Figure 10. When Q1 is on, the effect is switched in. Q1s collector is used to turn on the CHECK LED. Current flows through the zener, resistor and LED, thus illuminating the LED. When the battery voltage gets low enough so the LED and the 5.6V zener diode cannot conduct, the LED goes off. This indicates a low battery. Now you have the "effect on" LED indicator per the ultimate requirements stated above, along with a low battery indicator. What a deal.
Figure 10.
One of the other requirements we mentioned was that the pedal be in bypass every time it is powered up. It turns out that an ideal perfect symmetrical PSPICE model of a bistable multivibrator will not startup and operate properly. In the PSPICE model, I changed R8 from 56K to 55K and Q2 came on at power up. In the real world, due to tolerances of passive components and variations in semiconductors, one side of the multivibrator's collector will slew the base voltage of the other side up first and turn it on. Here is where we get some serendipity. Placing the low battery indicator network on one output of the multivibrator forces the other side to conduct first every time. Observe the startup characteristics in Figure 11.
Figure 11.
When current flows briefly in the LED (see the light green marker on LET network), the collector voltage of Q1 (blue) is raised quickly. This causes that rise to be coupled to the base of Q2 (red) and turn it on first. This biases off J2. Since Q1 is off, no negative bias is provided to J1 and the input signal appears on the output.
Moving the low battery network to the collector of Q2 would cause the effect pedal to power up in the "effect on" mode. Consequently, the LED would be ON when the effect is OFF.
I hope seeing the PSPICE analysis of these pedal circuits help you understand them better and may also help when troubleshooting in those areas.
Check out a large variety of effects pedal and other audio schematics at The Free Information Society.
Back to T's Technical Notes Index
©2007 Terry Downs Music™
