Now that I have a breadboarded pedal, before committing to the final version, I want to make sure I’m happy with it. Since the original parts are no longer manufactured, and I have a bunch of different transistors around I’ll try a few to see if I can tell the difference. I also want to understand the circuit a bit better to see how different components might affect the sound. At the end of this, I’ll be closer to having a pedal, I’ll be still missing some bits but I’m getting there. So the next step in building my DIY guitar effect: experimenting with the circuit.
Getting to Know the Circuit
If I want to know how to tune my creation I need to understand it first, or at least understand it well enough. I won’t go too much into theory, all I care about is whether the circuit will still work reasonably well if I just swap out transistors. Do I need to make any improvements based on input output characteristics and finally to commit to my final circuit design. I’ll go into more detail later on I hope to be rather brief here.
I’ll decompose the circuit into logical sections first, and then discuss each one of them separately.
Power and Input / Output
The original LPB-1 used only battery power but modern pedals commonly allow usage of both DC power adapter and battery power. The most common voltage for the power source is 9V, but higher voltages are not that uncommon. Input voltage is important for setting the working conditions of the amplifier section, and more importantly for the available headroom (maximum voltage on input/output that will not cause distortion).
Input and output jacks are not important for this discussion.
Gain section is what amplifies the signal. This is the heart of the effect if you will. The circuit is a common emitter amplifier with voltage divider biasing, probably the most common configuration for an amplifier using an NPN Bipolar Junction Transistor (BJT). In order to work as an amplifier the transistor needs to be in it’s active region. So the purpose of all of the resistors is to ensure correct working conditions for the transistor so it keeps working in the active region.
When it is in the active region, the transistor acts as an amplifier and when in common emitter configuration like here, gain of the amplifier is roughly R2/R4 = 10K/390 which is about 25 (this is about 28dB). This means that input signal is amplified 25 times, in reality it is a bit less than that but nobody will notice.
It turns out that the gain of the common emitter amplifier quite conveniently does not depend on the gain of the actual transistor used (the variation is very small for our purposes). This is very convenient because gain of the individual transistors can vary widely, even for the parts from the same batches.
Given the transistors I plan on trying out have relatively large gain, I should be able to happily use them without bother.
The gain of the amplifier is fixed, it cannot be adjusted. This in itself is not a problem. The effect was supposed to be driven directly from the guitar and most likely was supposed to go directly to the amp. With that in mind, it is not too crazy to have it fixed and just control volume with a potentiometer. The idea is to use volume knob of the guitar to control the signal level coming into the pedal and this way control the distortion to some extent.
Input impedance of the gain stage (and thus of the pedal) is low – around 35KOhms and this will cause a signal loss at higher frequencies (“tone suck”). Generally, to prevent this, modern pedals have higher input impedance in range of 1 Mega Ohms and higher.
Output impedance of the gain stage is roughly the same as collector resistor (R2), so around 10K, but due to potentiometer, it varies for the pedal and can be as high as 27.5K. This is not too bad, the aim is to keep it as low as possible.
Look here for more detailed analysis. I’ll add an article at some later stage with way too detailed analysis I’m sure.
C1 and C2 capacitors are there to only let AC signal through.
On the input, C1 blocks DC component of the signal if there is any and only lets through the guitar signal. This is important because any DC component in the signal could mess up operation conditions of the circuit.
Output coupling capacitor C2 blocks DC component of the signal that I introduced by biasing the transistor and only lets through amplified guitar signal.
This is how the waveforms look like:
The above diagram shows input waveforms on the left hand side. Waveform centered around 0V (in green) is what goes after input coupling capacitor C1. We get just the input signal without the DC component.
This changes after it goes past the voltage divider and is biased for transistor input – centered around 0.8V (in red). That already has a DC component in it.
On the right hand side are output waveforms, the grey one is the output at the transistor’s collector, just before the coupling capacitor C2 centered around 5V and in purple is the actual output after the coupling capacitor removed the DC component.
These capacitors could affect signal since they act as high pass filters and depending on the values of the capacitors and resistors involved they can block some of the low frequencies. In this circuit, cut-off frequency is 45Hz which won’t affect guitar’s sound.
In the end, I have just a simple potentiometer. The potentiometer is with logarithmic (audio) taper. Taper is just how the resistance changes when I turn the knob around. Linear potentiometers change resistance relatively uniformly, if the knob is turned a little, resistance changes a little, if the knob is turned more, resistance changes more. But, it is relatively similar change no matter what is the position of the knob.
Logarithmic potentiometers in the beginning, going clockwise, make very small difference in the resistance, but as you go towards the end of the rotation, the resistance changes way quicker for the same rotation of the knob. This is due to how we perceive loudness, we don’t perceive it linearly, but rather logarithmically (almost), and this type of potentiometer makes the increase/decrease of the volume feel more natural.
Here’s an excerpt from an Alps Alpine RK09L series potentiometer data sheet:
I highlighted in slight red the “audio” 15A potentiometer on the image above. It is quite different from the diagram for a linear potentiometer (1B in on the image). From the diagram we can see that resistance between pins 1 and 2 goes slowly up to 70% of the rotation of the shaft and then resistance change speeds up quite a bit.
I could go on about this, but I’ll leave that for some other time.
Finally, I can now do something concrete, all of the above was just to get intuition into what to expect.
I’ll use the following transistors in the experimentation that I had lying around:
|2N3904BU||100-400||General purpose amplifier and switch|
|BC337-40||250-630||Audio, power management, switching|
|BC108C||450-900||Low noise general purpose audio amplifiers|
|ZTX851||100-300||Emergency lighting circuits|
|PN2222ATA||100-300||General purpose amplifier and switch|
All of them are NPN BJTs which have somewhat different applications between them (based on their respective datasheets), but they are mostly general purpose. They have different gain range, can be anything in the range. They have different noise figures based on their datasheets but units and figures vary, some datasheets include graphs, some just a figure, so I didn’t include them here. From some experimentation ZTX851 is very low noise, so I thought to try it out.
With all of the above, I doubt I’ll notice much of a difference, but we’ll see. Since the output from the guitar is quite low, especially for single coils, I doubt single coil pickups will distort much. Humbuckers, maybe, probably chords strummed hard. If the pedal is not the first in the chain of effects, it would maybe be a different story.
Pinout of the Transistors
Since the cases of transistors that I have are slightly different I thought to include pinout diagram of all of them.
While 2N3904, PN2222A and BC337-40 are pretty much looking the same (2N3904 has straight leads in my case, the other two bent), ZTX851 and BC108C are different. I found ZTX851 slightly confusing because it looks a bit backwards from the previous three, but the markings are all pointing in the same direction (letters are on the longer face of the TO92 packaging and on the shorter face on E-Line).
BC108C is totally different packaging, it’s metal and has like a little tongue (not sure about the terminology) indicating where the emitter is.
Last minute addition
Before trying anything out I’ll wire up a switch to be able to switch the effect on and bypass it so I can hear the difference without stopping to rewire the breadboard.
I’ll also wire up couple of effects with different transistors side by side and switch between them to try hearing any difference.
I’m using a simple DPDT (double pole double throw) switch (not a stomp switch). This just means when it’s on guitar sound goes through effect, when it’s off guitar sound bypasses the effect altogether.
There is more to it when it comes to actually wiring up the effect. For this little experiment the following change is just fine:
DPDT switch schematic looks like this:
It really is two switches that working together. In position one pins 1 and 2 are connected and at the same time pins 4 and 5 are connected. In position two, pins 2 and 3 are connected and at the same time pins 5 and 6 are connected.
What we want is when in position one for example, input jack should be connected to the input of our effect and at the same time, output of our effect should be connected to the output jack. In position two, we want just input jack to be directly connected to the output jack. This is one way of doing it:
We’re finally ready.
As a bonus at the end of the video I did a Screaming Bird version. Essentially the same circuit as LPB with input/output capacitors changed from 100nF to 2nF making it a treble booster.