TUTORIAL Using an Active Buffer Can Greatly Reduce Hissssssss

JTEX

Well-known member
Have you ever wondered why your guitar makes more hiss (white noise) when you roll off the volume pot, compared to either all the way up or all the way down?

This is especially noticeable with high gain amps or pedals, and -- surprise! -- it comes mostly from your guitar, not the rest of your signal chain.

Any electrical conductor (such as a resistor, pot, wire, or the resistive component of a guitar pickup's windings) produces thermal noise, aka Johnson noise. It's caused by the thermal agitation of the electrons in the material, and it appears as a small random noise voltage across the resistor. The higher the resistance, the higher the noise.

An average pickup has a DC resistance of around 10kOhm. A common value for a passive guitar volume pot is 500k. This value is chosen to be much higher than the pickup to avoid heavily loading it, which would cause it to sound dark and lifeless. Higher resistance equals higher thermal noise, though, except for these special cases:

- When the pot is 100% up, it's basically a 500k resistor across a 10k pickup. Their combined (parallel) resistance is lower than 10k (about 9.8k), so the pot doesn't degrade the unavoidable, but fairly small amount of hiss produced by the pickup's 10k resistance.

- When the pot is all the way down, its wiper is shorted to ground (approx. 0 Ohms). A short makes no thermal noise, so this is your guitar at its quietest. Any hiss you still have in your signal chain at this point is not coming from the guitar.

But what happens when you back off the volume pot? You can think of it as two series resistors, the top one coming from the pickups and the bottom one going to ground. Their middle point is the pot's output (wiper). Let's say the pot is at 50%. We then have a 250k resistance from wiper to ground, and another 250k from wiper to the top. The top 250k is in series with the pickup. Their combined 260k resistance appears in parallel with the bottom 250k, for an equivalent 127 kOhm across the pot's output.

This is what the amp sees at this point: a 127k source resistance. This will make quite a bit more noise than the 10k of the pickup alone, or with the pot at max.

This is also what causes treble loss ("tone suck") with a long cable, due to the RC low pass filter formed by our 127k and the cable's capacitance, which can easily be on the order of 1nF or more. 127k and 1nF makes a nice low pass filter with a corner frequency of 1.25kHz. Dull, eh?

So, how can we fix this? Easy. We get rid of the 500k pot and use an active buffer right after the pickups, followed by a low value volume pot that will not cause significant hiss (or tone suck due to a long cable) when half the way down.

This is what active guitar electronics do, in a nutshell. Less noise/hiss when backing off the volume, and consistent tone at all volume settings, thanks to the low output impedance that doesn't care much about the cable capacitance.

Of course, the active buffer itself should have low self noise, so it doesn't contribute significant hiss to the unavoidable thermal noise from the pickups. Ideally, you'll want the buffer to have an equivalent noise that's lower than a 5k resistor, which is a typical resistance of a single coil pickup. A 5k resistor has a voltage noise density of about 9nV/rtHz. I won't explain this unit of measure here, but it's the same unit used to specify op amp voltage noise density.

To give an example, a TL072 has a voltage noise density of 18nV/rtHz -- the same as a 20k resistor, so it's not the greatest choice for a guitar input buffer if you care a lot about noise. LM4250, which I believe is what EMG uses as a buffer inside their active pickups, has it much worse, at about 40nV/rtHz (similar to a 100k resistor).

My current favorite op amp to use as an "activator" buffer is OPA207. It strikes a great balance between low voltage noise (about 8nV/rtHz) and low power (0.35mA, which is great for 9V battery power).

The absolute best I've seen is OPA2205, if you can find it. 7.2nV/rtHz for only 0.22mA. But it's a new part, pretty much unobtainium.

Check out the attached noise simulation, it's quite revealing.
 

Attachments

  • vol_pot_noise.JPG
    vol_pot_noise.JPG
    218.5 KB · Views: 49
Great write up! It's nice to have others contribute to the collective knowledge known as the Test Kitchen. Any other ideas you've got, keep em coming!
I'm happy to share some of what I've learned over the years. I was inspired to contribute when I saw how many tutorials are available on this site, and I believe you're responsible for most of them. Writing also helps me organize my thoughts. It's a big mess in there!
 
Didn't feel like starting a new topic, so I'll resurrect this one to share my findings about a pretty new opamp I had eagerly anticipated: OPA2205. I just dropped one into one of my Distiller filters for the first time. It lowered the noise floor from -99dBu (OPA2196) to almost -106 dBu. Big, big difference. It yields a dynamic range of almost 118dB from a 9V supply (the maximum unclipped output is +12dBu). Probably overkill for bass, but not for guitars, which often use lots of gain (this particular Distiller was ordered for use with a guitar).

Definitely my new favorite op amp for battery-powered (0.24mA) guitar-facing buffer/boost stages, along with the single version OPA205. Yes, it's bipolar, but with very low input bias current thanks to super beta, so really suitable for high impedance sources (guitar).

PXL_20240211_191903691.jpg PXL_20240211_191911168.jpg
 
Last edited:
Didn't feel like starting a new topic, so I'll resurrect this one to share my findings about a pretty new opamp I had eagerly anticipated: OPA2205. I just dropped one into one of my Distiller filters for the first time. It lowered the noise floor from -99dBu (OPA2196) to almost -106 dBu. Big, big difference. It yields a dynamic range of almost 118dB (the maximum unclipped output is +12dBu).
In a high gain setting, that's going to be very noticeable.
And at only ~$2 a pop, not bad!
Edit, $5. Was look at quantity pricing.
 
At some point though, the pickups' own Johnson noise becomes the limiting factor (from their 5k to 20k resistance). But OPA205 is about as good as it needs to be. Any further noise reduction would get lost in the pickups self noise (not to mention the residual EMI they still pick no matter how humbucking).
 
Here I used a dual op amp (2205) in a Sallen-Key tuneable low pass filter. But yes, technically its left side is the Axtivator, if we are to call it by a name instead of "just a good clean buffer that doesn't leave money on the table when it comes to noise or headroom" 😉
 
Back
Top