Thermionic Distortion (Friedman BE-OD)

12:45 PM

Today we're going to be looking at the Friedman BE-OD, a distortion pedal that was designed to emulate the Friedman Brown Eye Amplifier. If you're wondering how it sounds, just go check out the demo done by Pete Thorne here. While he makes everything he plays sound great, this pedal was really easy to dial in for various guitars because of the powerful EQ controls and internal gain trimmer. The gain itself is really nice, you can get a wide range of sounds from it, although you can't really get super low gain sounds unless you're rolling off the volume on your guitar. But if you're using this pedal then you're probably not looking for low gain anyway.

This pedal was built using the Thermionic Distortion PCB from PedalPCB. I really liked working with it because it's laid out really well and designed for top mounted jacks in a 125B sized box. The board version I have uses 1/8W resistors (my first time with those) but the updated one uses standard 1/4W. First, we're going to look at the theory behind the pedal and then I'll talk more about the build process and board itself.

Theory

It only seems natural to use the schematic from the Thermionic Distortion build doc, since that's the board I used, and the guys over at Pedal PCB were really cool and gave me the go ahead. I've gone and marked it up into 7 blocks that we'll be examining. While it looks like a lot, the building blocks are fairly simple. The signal path consists of 4 cascading gain stages like a tube amplifier might use, some power filtering, and various filters to shape your sound. So let's dive right in.

Power Supply

Let's talk about the power supply. This one might seem more intense than most pedals and that's for good reason. High gain pedals tend to get noisy, especially if they're not filtered properly. Here, D1 is acting as a reverse polarity protection diode in case the pedal is hooked up to the wrong power supply. The resistor that follows, R30, is a current limiting resistor that will blow if the pedal starts drawing too much current.

Then we have the voltage filtering. C20 is smoothing the majority of any DC ripple that existed in the power supply. It's followed by three 100nF caps which are used to filter the power between pins 8 and 4 of each of the ICs to prevent any noise in the signal path.

Lastly, there's a voltage divider created by R32 and R33 which is buffered by an op-amp in order to make sure each gain stage receives a constant current.

Input Gain Stage & Tight Control

This is a really simple stage. First we have an anti-popping resistor, R1, for when the pedal is switched on. The signal then goes through the input capacitor, C1, and through a low-pass filter, formed by C2 and R3, which removes any high frequency noise.

After all of that we get to the first gain stage. This gain stage is made using a non-inverting op-amp configuration which means we can calculate the gain with this equation:

Rf is our Feedback resistor; the one between the output and input of the op-amp. Rin is out Input resistor; the one that's connected to ground in this configuration. With this setup we get just a bit of amplification from this stage before we hit the Tight control. 

The Tight control is a simple variable high-pass filter. With the control at minimum we get a corner frequency of 152Hz. But as we roll up the potentiometer we reduce the resistance and at the maximum we get a corner frequency of 3388Hz which really cuts out a lot (all) of bass. 

After the tight control there is a low-pass filter formed by R8 and C4. Plugging those numbers into an RC filter calculator gives us a cutoff frequency of roughly 16kHz (mentioned by Javi Martin). This cuts out a lot of shrill treble frequencies while still giving the pedal a lot of tonal flexibility. 

2nd Gain Stage

The second gain stage is another non-inverting op-amp configuration. The difference this time is that there's now a potentiometer in line with the feedback resistor. When we do our gain calculation we get the following:

The other interesting part is that there's a cap in series with the Rin resistor. This forms a high-pass filter with a corner frequency of ~72Hz, so all of your bass will pass through. Then there's a set of LEDs used for soft clipping and a 47pF cap to smooth out some of the treble content. Overall, this looks a lot like the tube screamer clipping section. The major difference is that it uses LEDs (a common TS mod) and the corner frequency on a Tube Screamer is much higher, cutting out a lot of bass (~720Hz). This section doesn't need as high of a cutoff frequency because the tight control in the previous stage is allowing you to choose it yourself.

3rd Gain Stage

The signal first goes through a coupling capacitor to remove any DC offset from the previous stage. This capacitor also acts as part of a high pass filter in tandem with the overall impedance of the next stage to limit the amount of bass passing through.

Next we have an inverting op-amp configuration. This is actually simpler to calculate than the non-inverting op-amp. So let's take a look at the equation.
Now, if you're wondering why we have a negative amplification, it's because we're using an inverting op-amp configuration. It's flipping the phase of the signal, not reducing the amplification of our signal. So we really have an amplification of 10 for all intent and purposes.

There's also some more clipping diodes and another smoothing capacitor like the 2nd stage has. One cool thing is the additional 22k resistor. It's not a part of the gain calculation, it's actually acting as a "knee-softening resistor". It helps to smooth out the conduction of the diodes in the feedback loop, which makes the drive more tube-like.

4th Gain Stage

For the fourth gain stage we have another inverting op-amp configuration. This one's interesting because the feedback resistor is a combination of the 10k resistor and the 100k trim pot. The math is pretty similar to the previous gain stage:
That's really all there is to this stage. Nothing real crazy

Presence, Hard Clipping, and Treble Control

The Presence control is a simple RC low-pass filter, but it's really useful. The 2.2k resistor sets up a maximum corner frequency of 7238Hz. When the control is all the way down we have a total resistance of 12.2k which translates to a corner frequency of 1305Hz. You really get a lot of tonal variation out of that control.

Next we have a hard clipping phase. Again this is really simple, just two LEDs that run to the 4.5V bias reference (VREF). This is the kind of clipping you'd expect to see in a distortion pedal. We're using VREF instead of ground because our signal hasn't run through a coupling capacitor yet, so the AC part of the signal is still centered around 4.5V. By using LEDs the signal will clip at a higher voltage threshold which gives another character to the gain compared to some standard silicon diodes with a lower threshold. By varying the trim pot in the 4th gain stage you can change how these diodes clip. This is used to simulate the way a power amp distorts when changing the master volume.

Lastly, we have another low-pass filter that's forming the Treble control. Now since the Presence control is taming the high end before this stage, we don't really need to set up a maximum threshold. With the Treble control all the way down we have a corner frequency of 338Hz which is crazy. At this point the control is basically being used to cut off any excess treble that is not tamed by the Presence control.

Bass Control and Volume

The bass control is the trickiest part. It's active filter based on a bridged-T feedback network. Now that means something to someone, but that someone is not me. From what I've found, it involves very long equations and some complicated math that uses a bit too much greek for my liking. I'll break it down as best as I can.

A bridged-T feedback network is basically a low-pass filter mixed with a high-pass filter. It's usually a passive filter that creates a notch (cuts the frequencies) but when it's in a feedback loop it becomes a boost.

This filter is designed to boost the bass frequencies around 70hz. The two capacitors are used to form a ratio that determines how dramatic of a boost there is. The resistors are typically the same value; this is convention and helps in getting more predictable outcomes. They, in conjunction with the capacitors,help determine the frequency that is being boosted.

There's a lot of interplay between these components and it can change a lot by messing with one value. For the most predictable results you'll want to change values in pairs. For instance, changing the 22n to 10n and the 220n to 100n moves the frequency to about 150hz. If you leave those capacitors alone and change the resistors to 3.3k then you get a mid boost around 700hz (a factor of 10 higher). Generally speaking, decreasing the value the parts will shift the boost upwards in frequency and increasing the values will shift it downwards in frequency.

For a bit more information (and some complex equations) on how this works check out this site. Just know that the explanation is for a passive configuration. The active configuration is what turns it from a frequency cut (notch) to a boost. I simulated this section in LT Spice (a circuit simulation software) to figure out how it works and I highly recommend doing that before changing any part values. Simulating it will give you an idea of what to expect when you change a part and how it plays with the other pieces of the circuit

The potentiometer is used to limit how dramatic the boost is. At its lowest value it will boost 70hz the most because it's essentially out of the circuit. When it's all the way down, it's at its maximum value and is limiting how much the filter can actually do it's job which reduces the boost.

After all that complicated filtering, the signal runs through another cap and resistor and then to the volume knob which is just a standard voltage divider configuration. This is a pretty standard volume control for most pedals.

Pedal PCB Shoutout

As I mentioned earlier, I built this using a board from PedalPCB. Their boards are really well designed and are made for top mounted jacks, which I love. The build documents are really well done and easy to follow. They are constantly adding new stuff to their site so go check them out and let them know I sent you if grab something. 



Sources:

Jack Orman's RC Filter Calculator - http://www.muzique.com/schem/filter.htm
Op-Amp Basics - http://www.electronics-tutorials.ws/opamp/opamp_8.html
Soft Knee Resistors - http://www.diystompboxes.com/smfforum/index.php?topic=82764.0
Bridged-T Filters: - http://www.johnhearfield.com/RC/RC4.htm

You Might Also Like

14 comments

  1. Great write up. Do you happen to have a copy of your BOM ? Was thinking of doing this build myself. Thanks!

    ReplyDelete
  2. Amazing! Is there are bom link for this please?

    ReplyDelete
    Replies
    1. I don't. But I built the board from PedalPCB. They have a BOM in their build guide.

      https://www.pedalpcb.com/product/thermionic-distortion/

      Delete
  3. I have a question, what are the U1, U2, and U3 components in the power supply?

    ReplyDelete
    Replies
    1. It represents pins 4 and 8 of the three integrated circuits used, in this case TL072 op amps.

      Delete
  4. what kind of op amps are being used?

    ReplyDelete
  5. Hi, thanks for the superb analysis and congrats for the website!! Just a correction for 2nd gain stage section that I've just noticed: "This forms a HIGH-pass filter with a corner frequency of ~72Hz".

    ReplyDelete
    Replies
    1. Thanks for catching that Javi. I've updated the page and included a little blurb about the LPF that you pointed out. Thanks for contributing.

      Delete
  6. I mean, R9 + C8 from U1.2's feedback side form a HPF to ~72Hz and regarding the non-inverting input R8 + C4 would form a LPF to ~16kHz.

    ReplyDelete
  7. Se pueden usar todos potenciometros lineales?

    ReplyDelete
    Replies
    1. Tight: 100K Antilogarítmico
      Gain: 1M Lineales
      Pres: 10K Antilogarítmico
      Treble: 100K Lineales
      Bass 100K Antilogarítmico
      Volume: 50K Logarítmico

      Delete
  8. I'm doing the pcb, let's see how it works for me!

    ReplyDelete
  9. Lo hice pero me funciona muy bajo el volumen que puede ser?

    ReplyDelete
  10. Finally understood the Bridged t feedback and I feel like a god

    ReplyDelete