Help with picking a transformer for a guitar pickup simulator?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

carboncomp

Active member
Joined
Feb 22, 2022
Messages
39
Location
England
Working on a Buffer>Pickup simulator so I can have a buffer before vintage fuzz and boosts pedals that have a low input impedance and do not play nice with a buffer before them.

I have a feeling the answer to my question is going to be use what you're trying to simulate, just use a cheap guitar pickup!

As the transformer is trying to emulate the V,L,DC resistance of a real pickup, but a real pickup has a DC R of 6k to 20k, but the recommended transformer in the article iv been reading on pickup simulators the 42TM019 Audio Transformers only has a resistance of 600R on the Primary and Impedance 10k, so they recommend adding resistors in series, but while that does change the source impedance, its response is flat across the frequency range unlike that from the guitar. Then a 470pF capacitor emulates the capacitance from the tone control and guitar cord.

This is been a bit of a rabbit hole for me as I have only just started reading up on inductors and transformers and the rather complex nature of inductance being non-linear, affected by the size of the winding, cross-section, the magnetic flux, and so on. and feel that inductance is more important than resistance in this application, Right/Wrong?

And if so, is there a way to calculate the inductance form the data sheet in a meaningful way for this application?


So, is it better to go for a transformer with a higher resistance, got for inductance or Use an old pickup?





1660165454009.png
 
So let me get this straight. You want to create a circuit that simulates the complex output impedance of a guitar?

I think that would take some work using LTSpice.

But there is another way that is easier and probably better. You could use an actual guitar. Specifically, use the sleeve of the guitar's jack as the input and then get the output from the tip as usual. You just need a little die-cast box with three 1/4" jacks where one is isolated (meaning the sleeve is not connected to the grounds of any other jacks). Send your signal into one jack (this is the input) and it's tip to the sleeve of the isolated 1/4" jack for a cable to a guitar (note that there cannot be any devices between this jack and the guitar). Use whatever cable you want to simulate. Then take the tip of that jack (this is the signal returning signal from the guitar) and connect that to the tip of the other 1/4" jack for a cable to an amp or recording device (or add a fourth jack in parallel as a through connection to listen and record at the same time). The sleeve of all jacks except the isolated one are all connected together (these need to reach earth ground somehow such as through the sleeve of the jack to the amp / recording device or sleeve / shield of the input.
 
So let me get this straight. You want to create a circuit that simulates the complex output impedance of a guitar?
Basically, yes. The conventional thinking of pedalboards is to "sandwich" the pedal chain, So you have a Buffer at the start, and a buffer at the end to help preserve your higher frequencies. The trouble with this is vintage pedals just didn't take input impedance into account. So, for example, say you have a Vintage Fuzz with Z of 5k (keeping frequency out for my own simplicity, and saying the AC signal from the guitar is 1V) the pickup is loaded down to the point you lose over 50% of the signal:

Vout = Vin x Z/(Z2+Z1)
Vout = 1V x 5k /(5k+6K)
Vout = 0.45v

Compared with a modern pedal with a Z of 1M:

Vout = Vin x Z/(Z2+Z1)
Vout = 1V x 1M /(1Mk+6K)
Vout = 0.99v

And putting a buffer in front of a Fuzz just makes it sound terrible and you lose the ability to clean up the effect by rolling off the volume.

As most modern pedals now have a buffer built into them (usually a common drain JFET or Omiter follower BJT) this rules out doing things like adding an overdrive pedal before your fuzz, so most pedal boards have to go guitar > Vintage fuzz/treble booster > Buffer > rest of the pedals > buffer.

I would like to be able to have the convenience and kind of fun options of moving my Vintage Fuzz/Boosters around in the chain, so I still have the pedalboard sandwiched with buffers, but the fuzz does not have to be before the buffer.

Most people seem to get around this with a passive pickup simulator for example HERE is an article on the topic, and I'm just trying to work out what the best approach would be regarding the inductor in question.


I think that would take some work using LTSpice.
Good job I installed it last night and am working my way through youtube tutorials (A bit off-topic, but found one on adding a pot that was a game changer!).
But there is another way that is easier and probably better. You could use an actual guitar. Specifically, use the sleeve of the guitar's jack as the input and then get the output from the tip as usual. You just need a little die-cast box with three 1/4" jacks where one is isolated (meaning the sleeve is not connected to the grounds of any other jacks). Send your signal into one jack (this is the input) and it's tip to the sleeve of the isolated 1/4" jack for a cable to a guitar (note that there cannot be any devices between this jack and the guitar). Use whatever cable you want to simulate. Then take the tip of that jack (this is the signal returning signal from the guitar) and connect that to the tip of the other 1/4" jack for a cable to an amp or recording device (or add a fourth jack in parallel as a through connection to listen and record at the same time). The sleeve of all jacks except the isolated one are all connected together (these need to reach earth ground somehow such as through the sleeve of the jack to the amp / recording device or sleeve / shield of the input.
Think you, will look into this option, seems to be akin to a form of looping box if I understand correctly.
 
And putting a buffer in front of a Fuzz just makes it sound terrible and you lose the ability to clean up the effect by rolling off the volume.
The reason the Fuzz sounds terrible with a buffer is because the source impedance is too low. The input impedance of a Fuzz depends greatly on signal level. A Germanium emitter junction has a forward voltage of only 0.2V but the guitar can easily produce voltage higher than that (if you really bang on it just so, it will make 4Vpp+). So the input transistor is being tortured on and off. When the 0.2V forward voltage is reached, the Fuzz input impedance is very low (just the emitter resistance). When the signal swings the other way, the transistor is hard off and the input impedance is at least 100K depending on the Fuzz pot position. A forward conducting transistor also rectifies the input causing the coupling cap to charge up. This causes blocking distortion.

The solution to this is to add series resistance. This is what the volume control does and simultaneously reduces the level so that the first transistor is not being tortured on / off.

Another problem with Fuzz is high frequency content. Most Fuzz sound pretty horrible IMO. The ones that sound good have transistors that do not have the bandwidth to make high frequencies (and they have to have the right bias and leakage and gain and other properties that are elusive to characterize). You need to high cut those high frequencies. Then even the blocking distortion can sound good (like Spirit in the Sky which is a Fuzz builtin to a tele).

So looking at your circuit, I would just replace the coil with a 10K resistor. Then change C2 to be more like 4n or whatever value you need to high cut at about 4kHz. Experiment with different R and C to cut off around 4kHz. Try 82K / 470p. Etc. However, this will mess up the 470p on the output. You can drop the 10R1 (which is nothing compared to the series R) and adjust the 470p on the output to give you additional high frequency control when rolling off the volume. If done right, you'll have a two pole filter when the volume is rolled off.

However, you really need to filter the high frequencies on the output of the Fuzz. This is actually more important than what you're circuit is doing. Everything above 4kHz from the Fuzz is just nasty annoying garbage and needs to go. The best way to do this is to run it through and old Alnico speaker. If you run a fuzz into a power amp (not a guitar amp, a hi-fi amp) driving an old speaker, it will brick-wall filter the high frequencies and make the Fuzz sound way better. A speaker like a Jensen P12Q is perfect. It will completely crush everything above 4kHz. Obviously you can't do this in your pedal chain. But if you have a high cut pedal, that might work too.

When I do a fuzz, I just go from the guitar > Fuzz > buffer and then split the signal to reverb and then send both dry and wet to a hi-fi amp driving two old P12Q speakers. Put the dry in your face and reverb on the other side of the room.
 
Working on a Buffer>Pickup simulator so I can have a buffer before vintage fuzz and boosts pedals that have a low input impedance and do not play nice with a buffer before them.
Remember that fuzz pedals are ultimately tweaked by listening to an actual guitar and plugged in an amp.
So what you call "do not play nice" is probably what the designer wanted.
For example, the Brian May sound comes from a Rangemaster treble booster, which input impedance is about 300k at 100 Hz, but falls at 30kohms at 1kHz then to 15kohms at 4 kHz.
I know for sure that driving some overdrive pedals with a low-Z buffer does not result in sonic graal. Driving them with a high resistance is not adequate either. Some really want to see the impedance of a real guitar
So, is it better to go for a transformer with a higher resistance, got for inductance or Use an old pickup?
You're looking at an inductor of between 2 and 10 Henry and a resistance of 3 to 12k ohms, with a significant parallel parasitic capacitance (250p-1nF)
You won't find an inductor of this value with such a high series resistance, and probably not such a high parasitic capacitance, so you may use a standard inductor and add the series resistance and parallel capacitance, or you could use an old p/u The first option would offer you the possibility of exploring various options. You may also want to add the effects of the volume and tone control circuitry. Some pedals like to be driven from a guitar with the tone pot fully closed.
 
And if so, is there a way to calculate the inductance form the data sheet in a meaningful way for this application?
In the case of wide-band audio transformers, the nominal impedance is guaranteed at the lowest quoted frequency.
In aL R circuit, the -3 dB point happens when the impedance of teh inductor equals that of the resistor.
Using Z=L.2pi.f.
For example a winding rated at 10k with a quoted response of 30Hz at -3 dB should have an inductance of at least 53H.
Actually, for high quality xfmrs, the inductance is generally somewhat higher.
Here these xfmrs are rated for vocal range (300Hz-3.4kHz), so the inductance may be as low a 5.3H (for the 10k type), but the only way to know is to measure it. Beware that the common inductance test found on DMM's gives only an approximation, but considering the application, it's not a real issue.
If it confirms to this wet-finger approximation, it may be adequate, however you would have to add resistance and capacitance for proper emulation.
 
Back
Top