Simulating a pedal design

[james.watson]

Hi all,

I am new to this forum and pedal design so I apologise if my design gives anyone a stroke.

I am using yenka to simulate the design I have in mind (the project so far: https://drive.google.com/file/d/1QS39BlPJ2OoTRU3XHxFoagsgQ35FRILj/view?usp=sharing).

The basic idea is a fuzz-box, into a clean boost, into a clipping circuit with switchable diodes, into a 5 band parametric EQ.

I don't currently have the EQ stage on the diagram.

Thus far I have "borrowed" various sections of the circuit from various other pedals and the internet (the input buffer and power stage are nicked from a tube screamer, the clipping stage is straight from an MT-2).

The issue I am having is the waveform coming from the output looks very spikey and high frequency as opposed to the near square wave thats comes out of a nearby resistor.

Speaker output:


Resistor output:


The pedal is being fed a 62Hz signal to simulate the Low B of a 7-string.

I mainly just want to check, are my results correct and if not, what can I do to fix them?

And, what do you use to design and simulate pedals, if anything?

Many thanks in advance.

[PRR]

Welcome!!

I don't know "yenka", it seems to be a math package? Maybe someone here knows about circuit study with math software.

A spike waveform is not unlikely in a distortion scheme.

It may mean too much overload or EQ.

I don't quite understand your images; is the waveform hiding part of the circuit?

However one stray section caught my eye. Isn't this Positive feedback? Combined with any sort of differentiation and a square wave, spikes is what you get.

[grumpy old man] Learn to rough-out a circuit on a matchbook and sketch the expected results to one-thumb precision. SPICE does NOT know what you want to do, and I assume yenka doesn't either. Knowing roughly what you want at every important node speeds the too-human task of debugging the problem you set before the computer.[/grumpy old man]

{EDIT} ah, I see Yenka does have an electronics function. Then most warning against trusting SPICE wil apply to Yenka too.{/EDIT}

[james.watson]

A spike waveform is not unlikely in a distortion scheme.

It may mean too much overload or EQ.

The overload theory makes sense because the fuzz and the clean boost cascade into each other. Should I add some volume reduction between the fuzz and boost?

I don't quite understand your images; is the waveform hiding part of the circuit?

Yeah it is, I'll attach some more images so you can tell what's going on.

[grumpy old man] Learn to rough-out a circuit on a matchbook and sketch the expected results to one-thumb precision. SPICE does NOT know what you want to do, and I assume yenka doesn't either. Knowing roughly what you want at every important node speeds the too-human task of debugging the problem you set before the computer.[/grumpy old man]

I made a collage of the important bits and that's been what I'm transcribing, for lack of a better term.

[Vivek]

I was discussing simulation of MT-2 with ELECTRICDRUID

We found that the "Pre-distortion EQ" also distorts. So the MT-2 distorts at 3 places.

If one copies a part of a circuit on to another circuit, it helps to know wave shape, Vp, Vrms, and frequency response at the input of this stage in the original circuit. We all know the snippet wont act the same if the inputs are totally different in the new circuit.

[james.watson]

I was discussing simulation of MT-2 with ELECTRICDRUID

We found that the "Pre-distortion EQ" also distorts. So the MT-2 distorts at 3 places.

If one copies a part of a circuit on to another circuit, it helps to know wave shape, Vp, Vrms, and frequency response at the input of this stage in the original circuit. We all know the snippet wont act the same if the inputs are totally different in the new circuit.

You raise a good point.

I think I'm going to have to slice out parts of the circuit and test them individually and then concatenate the parts once they work how I want.

[james.watson]

I've started testing the individual parts and put the input buffer and power stage together.

Here's the resulting waveform:


Is that correct/desirable?

[PRR]

I don't know what most of that does. It looks like a power-chain top-left but where is the power? What is the bottom line connected to, if anything?

Are there any DC voltages here? A circuit must find a useful DC operating condition before it makes sense to put signal in.

The 'scope plot is scaled in MICROvolts.Do we even care?

Anyway there is a single transistor feeding 62Hz to an 8 Ohm speaker through 100nFd. This raises big "WTF?" to any seasoned audio technician. We expect to find >220uFd for poor bass and 1000uFd for good bass.

[james.watson]

I don't know what most of that does. It looks like a power-chain top-left but where is the power? What is the bottom line connected to, if anything?

yeah that is a power stage. it's currently feeding an input buffer.

Are there any DC voltages here? A circuit must find a useful DC operating condition before it makes sense to put signal in.

Sorry, that's my bad framing. There is a 9v battery just off-screen.

The 'scope plot is scaled in MICROvolts.Do we even care?

Yeah I can see it's in microvolts. as I mentioned, I'm new to pedal design so I'm not sure if that's meant to happen.

Anyway there is a single transistor feeding 62Hz to an 8 Ohm speaker through 100nFd. This raises big "WTF?" to any seasoned audio technician. We expect to find >220uFd for poor bass and 1000uFd for good bass.

That speaker is mostly there so I have something to measure voltage across for that screenshot. In reality that would be a fuzz circuit. Thank you for the cap values too, I shall make a note of that.

[Rob Strand]

You might find the waveforms look better with a higher impedance load or by decreasing the 1k on the emitter.

The 510k base resistor looks a bit high even for a 1k base resistor it's definitely going to be an issue for a smaller emitter resistor.

From design perspective there's all sorts of issues and the simulation will only show that.   Start with something you know is working  then make changes until it breaks then work out why.    If a design has issues they will show up on the simulator.  The designer then has to make changes to fix it.

[ElectricDruid]

Anyway there is a single transistor feeding 62Hz to an 8 Ohm speaker through 100nFd. This raises big "WTF?" to any seasoned audio technician. We expect to find >220uFd for poor bass and 1000uFd for good bass.

That speaker is mostly there so I have something to measure voltage across for that screenshot. In reality that would be a fuzz circuit. Thank you for the cap values too, I shall make a note of that.

I think PRR has identified the source of your original problem - the spiky looking waveform instead of a nice square wave. I think you inadvertently invented the highpass filter!

Making the cap bigger or the load resistance that follows it bigger both lower the cutoff frequency of the filter and allow your bass signal to pass as well as the spiky treble.