Hello guys!
I am working on a treble boost/fuzz circuit based on the Dallas Rangemaster. This is my first time building a transistor based pedal, as I've only worked with op amps before and I wanted to see if there were any tricks of the trade of designing transistor boost circuits I may be missing.
Here is my schematic and some info about my design:



I started with the Rangemaster circuit, switched it to an NPN silicon transistor (after looking at germanium prices lol). Since the germanium transistors typically only have a gain of 70-90 I added a gain control to the emitter of the transistor that lets you reduce the gain from the silicon transistors hfe gain of 200-800 down to as low as 10, that I'm calling the "tame" control.

Because of the high gain of the silicon transistors, when the tame control is at 0 it pushes into fuzz face hard clip territory. My design has a bit higher input impedance than the original since I wanted the volume knob cleanup to be a little less touchy, but it is still only 20kohm so cleans up very nice and sparkly. So it still does the thing of loading the strat pickups with the low input impedance of the 60s pedals just a little less touchy.

Here is an In/out transient waveform at max gain hard clipping:


And heres one with the tame control 40% up soft clipping:



the waveforms show it being having a slight asymmetric clipping but nowhere near that of the fuzz face. Audibly, turning down the guitar volume and turning up the tame control both sound the same. I also verified that in simulation, so I think the way I'll use this is gain maxxed and toying with the guitar volume.


I also added a full range control that switches to a larger input capacitor
Here is the frequency response in the normal treble booster based position:



And here is the response in the full range mode where it only really rolls off bass harmonics <50hz:



I built the circuit up on a breadboard and played with my trimpots and pots see if there were any sweet spots I was missing and I feel the circuit sounds best when biased exactly as in the schematic and I was very happy with how it sounds. Since I hadn't heard of a single transistor fuzz before I was worried I was about to find out why nobody does this but I thought it was very reactive and sounded cool. Has anyone heard of any other single transistor fuzz circuits?

It was quite noisey on the breadboard, not any unexpected noise just a high noise floor. A certain level of noise is expected from it being a fairly high gain pedal and from having many exposed leads on the breadboard, but I was wondering if there were any standard practices for reducing the noise that I have not done in the circuit.

I also would appreciate any advice on laying out the circuit on perfboard and how I should mount the perfboard in the enclosure. My only previous pedal build I went breadboard to PCB so I was able to mount it from the pots. I also used a ground plane on my pcb in that build so routing was super easy, doing this point to point routing on perfboard seems like I might have to build a few new neuron connections.

Appreciate any input, Thank you!
 

Your maths are good till step 2. 

Follow below formulas:

And "small signal analysis", as bonus  :

Quote from: camzilla on September 09, 2024, 09:52:37 PMIt was quite noisey on the breadboard, not any unexpected noise just a high noise floor. A certain level of noise is expected from it being a fairly high gain pedal and from having many exposed leads on the breadboard, but I was wondering if there were any standard practices for reducing the noise that I have not done in the circuit.

Use metal film resistors, replace the 50k trimpot at the input bias with a fixed metal film resistor.
But a treble booster will amplify hiss. It is not supposed to be silent.

And use a BJT of smaller r_bb' (Base spreading resistance) with proper bias for optimum(*) (lowest) noise for your particular source impedance..

(*) e_n falls while i_n rises with increasing Collector current..

Thank you guys for the responses! I did some more tinkering, and found out the main source of the noise was that I am using a breadboard backed with a metal plate and I did not ground the plate so it was picking up a lot of noise. Eliminating that dropped the noise to a subtle hiss.
Then I increased the bypass cap from base to collector to 570pf, limiting the frequency response a little bit more, but in a way that should be barely audible. It only attenuates 20khz by about 1.7 db

Quote from: antonis on September 10, 2024, 04:15:03 AMYour maths are good till step 2. 

I really appreciate your formula sheets! That is super useful documentation and I will go back through the math of the circuit again to try those methods. I believe that my math all lined up quite well with simulated values, so to clarify is that step 2 actually incorrect or just a different method? To get the 1.7 volt bias, I traced the path up from ground to the bias point from the emitter: I set the current so that the voltage across emitter resistor was 1V, then moving from emitter to base is effectively moving across a diode, so approximately 0.7 volt.

Quote from: antonis on September 10, 2024, 09:25:04 AMAnd use a BJT of smaller r_bb' (Base spreading resistance) with proper bias for optimum(*) (lowest) noise for your particular source impedance..

(*) e_n falls while i_n rises with increasing Collector current..

Also appreciate this. I did some research on this and it seems that there is not typically good documentation on the rbb and it would require some complicated setups to determine them. Definitely convinced me to use a socket instead of soldering directly so I can sub in some different transistors.

Quote from: Clint Eastwood on September 10, 2024, 04:35:07 AMUse metal film resistors, replace the 50k trimpot at the input bias with a fixed metal film resistor.
But a treble booster will amplify hiss. It is not supposed to be silent.

Makes sense. I am using all carbon and metal film resistors, no carbon comp at least so that pot would likely be the noiseyest resistive component. I'll see if I notice a difference swapping that out for the closest resistor value I have on hand. If it doesnt contribute audibly, I might leave it to be able to adjust the bias.

Thank you both for your input! I think I will soon begin putting this on a soldered breadboard.

Quote from: camzilla on September 10, 2024, 05:37:23 PMhank you guys for the responses! I did some more tinkering, and found out the main source of the noise was that I am using a breadboard backed with a metal plate and I did not ground the plate so it was picking up a lot of noise. Eliminating that dropped the noise to a subtle hiss.
Then I increased the bypass cap from base to collector to 570pf, limiting the frequency response a little bit more, but in a way that should be barely audible. It only attenuates 20khz by about 1.7 dB

A 560pF cap between c and b will present quite a high input capacitance to the previous stage, due to the Miller effect.

It's more than likely this pedal will be the first in the signal chain.  The input capacitance will increase the HF roll-off when the guitar volume is backed off.   Some of these effects might be implicity considered already.   However, if a buffer is placed before the pedal then there will be considerable brightening in comparison to the guitar.   You can find a better mid-ground by adding say a 10k to 22k (maybe higher) in series with the input line, just before the cap.  Then after that perhaps re-tune the cb cap.

Quote from: Rob Strand on September 10, 2024, 05:55:32 PMA 560pF cap between c and b will present quite a high input capacitance to the previous stage, due to the Miller effect.

It's more than likely this pedal will be the first in the signal chain.  The input capacitance will increase the HF roll-off when the guitar volume is backed off.   Some of these effects might be implicity considered already.   However, if a buffer is placed before the pedal then there will be considerable brightening in comparison to the guitar.   You can find a better mid-ground by adding say a 10k to 22k (maybe higher) in series with the input line, just before the cap.  Then after that perhaps re-tune the cb cap.

Thanks!
I just breadboarded with the BC capacitor and it caused a very significant high rolloff but I had no idea why. It did completely solve noise issues at the 570pf value but took away the point of a treble booster. Thank you for this! Its odd that the simulations wouldn't show this.
I just did some research on this effect and the methods of compensating for it seem to be large circuit changes that I don't think I want to use on this circuit, like adding a positive feedback buffer across the capacitor as this video suggests : https://www.youtube.com/watch?v=czk0I3ga3LQ



Is this effect as simple as, at max gain of 100, with a 47pf base collector cap, the capacitance would appear to be 100*47pf=4.7nf ? This is how I interpreted the effect and it seems to explain what I heard, even with a 47pf bc cap the tone was very full at maximum gain. Here is what simulation says the response would be with a 4.7nf bc cap which would be the bc cap value at gain 100 accounting for miller effect:



Because of the tone changes, I think I will eliminate this cap from the circuit at the cost of noise. I originally added this capacitor under the rule of thumb to reduce the frequency range of an amplifier to the minimum amount that you care about for best noise performance. I am assuming that the noise is due to high frequencies not being attenuated since the capacitor eliminated the noise. I will try a simple RC filter on the input to see if this improves the noise without significant tone change. Let me know if there is anything I am missing in this analysis   

Also don't think I am completely following on your input resistor suggestion. I understand that the BC capacitance and the guitar volume knob effectively form an RC so I can see the rolloff, but I'm not completely understanding how adding a series resistor would modify that behavior.

I am not experienced with vintage style pedals so I am not sure exactly how I want the rolloff behavior to be. The only drives I've used are a boss blues driver, drive from my amps, and the op amp style muff you and Antonis helped me get running last summer (huge thanks to you guys and this forum). I was hoping to land roughly in the rangemaster,fuzzface region of rolloff so I could start forming my opinions and become really annoying about my them.

Appreciate the info and advice!

Here is where my circuit is at currently with a few updated values:

Quote from: camzilla on September 10, 2024, 10:00:35 PMThanks!
I just breadboarded with the BC capacitor and it caused a very significant high rolloff but I had no idea why. It did completely solve noise issues at the 570pf value but took away the point of a treble booster. Thank you for this! Its odd that the simulations wouldn't show this.

The simulation won't show the roll-off unless you add source impedance to the source.   In reality the source has an impedance.  For a guitar pickup the source impedance is itself a circuit.  For example, see figure 13 from,
https://www.theguitar-blog.com/wp-content/uploads/2012/07/Analysis-Electric-Guitar-Pickups.pdf

QuoteI just did some research on this effect and the methods of compensating for it seem to be large circuit changes that I don't think I want to use on this circuit, like adding a positive feedback buffer across the capacitor as this video suggests : https://www.youtube.com/watch?v=czk0I3ga3LQ

" class="bbc_img" loading="lazy">

Is this effect as simple as, at max gain of 100, with a 47pf base collector cap, the capacitance would appear to be 100*47pf=4.7nf ? This is how I interpreted the effect and it seems to explain what I heard, even with a 47pf bc cap the tone was very full at maximum gain. Here is what simulation says the response would be with a 4.7nf bc cap which would be the bc cap value at gain 100 accounting for miller effect:

There's no point of compensating for it.   The cap is added for deliberate filtering of audio and/or for filtering RF.  It's the designers decision to add it.   If you compensate it out then you have a more complex circuit which no longer performs its desired function!

QuoteBecause of the tone changes, I think I will eliminate this cap from the circuit at the cost of noise. I originally added this capacitor under the rule of thumb to reduce the frequency range of an amplifier to the minimum amount that you care about for best noise performance. I am assuming that the noise is due to high frequencies not being attenuated since the capacitor eliminated the noise. I will try a simple RC filter on the input to see if this improves the noise without significant tone change. Let me know if there is anything I am missing in this analysis

It's often a good idea to have some capacitance to ensure RF is filtered out.  The transistor itself have some built in cb capacitance, say around 2pF.   

Adding an additional cb capacitance increases the input capacitance and filtering.   As per the Miller effect the input capacitance depends on the amplifier gain.  In many cases we can add a capacitance from the input to ground to perform filter instead of a cb capacitance.   Design the filtering in this way means the capacitance is largely independent of the amplifier gain - if you make the transistor gain variable this could be a good decision.

The point I was making in the previous post is the filtering of the audio depends on the source impedance and the input capacitance of the amplifier.   If the source impedance changes by a large factor then the filtering will also change.    You can see this when comparing the response of the amplifier when driven by a buffer to when it is driven by a pickup.   By adding some series resistance you can reduce this variation in filtering because for the buffer gain instead of the source being zero ohm, providing no filtering, it now has some impedance and provides some filtering.    In other words it behaves a little closer to the pickup case.

Your simulations didn't include a source impedance so you simulations don't show any filtering of the audio signal.  Whereas when you build the unit and test it with a real guitar you can hear the filtering is present.    That large discrepancy in behaviour will become real if you drive the input of the amplifier stage from a buffer. 

Quote from: camzilla on September 10, 2024, 05:37:23 PMto clarify is that step 2 actually incorrect or just a different method?

Just for "academic" purpose:
(if you're OK with your particular bias, it should be more than fine..)

On step 3. you want VRc = (Vcc-Vbias)/2 where you should want to be (Vcc+VRe)/2, for symmetrical swing (in case of AC grounded Emitter)..

So, you resulted in Collector voltage level at 3.65V where it should be at 5V..!!
You can realize the signal output "asymmetry"..
(Collector can swing positive 5.35V but only 2.65V negative - V_CEsat ingnored..)

Quote from: Rob Strand on September 11, 2024, 01:45:42 AMThe simulation won't show the roll-off unless you add source impedance to the source.   In reality the source has an impedance.  For a guitar pickup the source impedance is itself a circuit.  For example, see figure 13 from,
https://www.theguitar-blog.com/wp-content/uploads/2012/07/Analysis-Electric-Guitar-Pickups.pdf

Thanks for the documentation! I knew roughly the attributes of pickups but that is very in depth and informative. Do you think that I should get into the habit of modelling these or are they just factors to keep in mind during the design process?

Quote from: Rob Strand on September 11, 2024, 01:45:42 AMThere's no point of compensating for it.   The cap is added for deliberate filtering of audio and/or for filtering RF.  It's the designers decision to add it.   If you compensate it out then you have a more complex circuit which no longer performs its desired function!

I wanted to compensate for it because I believed that this bypass cap was solving my noise problem at the cost of cutting high end from my circuit, however now that I have the circuit built and boxed and battery powered, the noise is only present when powered by my pedal power supply. This capacitor was eliminating the noise only because it was drastically limiting the frequency response since I didn't account for miller effect when picking the value, so that was my misunderstanding. I appreciate the insight. Definitely a good thing to know. Seems like a good way to cut high end based upon gain if the gain is adjustable on it, so I'll put that in my back pocket.

For this build, I ended up putting no BC capacitor since I don't have a low value enough capacitor to not significantly alter the tone over the gain range. I did leave a socket across there so I could easily pop one in if I get some lower value, say 5 or 10pf, in a next parts order. I put a 47pf to ground instead to hopefully bypass RF and prevent oscillation.

Quote from: antonis on September 11, 2024, 05:04:59 AMJust for "academic" purpose:
(if you're OK with your particular bias, it should be more than fine..)

On step 3. you want VRc = (Vcc-Vbias)/2 where you should want to be (Vcc+VRe)/2, for symmetrical swing (in case of AC grounded Emitter)..

So, you resulted in Collector voltage level at 3.65V where it should be at 5V..!!
You can realize the signal output "asymmetry"..
(Collector can swing positive 5.35V but only 2.65V negative - V_CEsat ingnored..)

When I calculated VRc I was thinking of it as the drop across Vrc, meaning that I set my collector voltage at 5.35 volts which is believe is in line with what your saying . I did this mostly following along with math and did not completely understand, and your explanation has helped me understand it conceptually, that the signal can only swing from Vcc to Vbias. Thanks!
I believe that asymmetrical clipping was a result of forgetting to update my bias resistors when I changed my Rc and Re values to values which I had in my parts stash, do the circuit was definitely misbiased in those waveforms but was correct on my breadboard. here are the current waveform, without that error (and hopefully without any others)
Max gain:


It looks much more symmetrical. It still has a slight asymmetry and I belive thats due to the input waveform being large enough to fully bias the base emitter junction on the postive cycle of the waveform as this is my base voltage waveform:



So I think this is expected behavior of a properly biased circuit. Would love to hear if that isn't proper for academic purposes at least




I have finished my build and boxed it and I'm very happy with the result! I like the volume knob reactivity of this type of circuit. It feels like the blues driver and my amp dont react quite as drastically to volume knob changes.

My only gripe is that there is a high pitched noise present in the pedal when powered with my pedal power supply. It is cheap and off brand but supposedly an isolated power supply nonetheless. Currently I have a 100uF and a 22nF capacitor across the supply as close to the transistor 9V connections as I could get on my perfboard setup. I tried adding a 100 ohm resistor in series with the supply before the caps to form a lowpass filter to clean the supply but it did not reduce the noise at all so I did not bother adding it to the perfboard. Very interested in any other PSU noise reduction strategies, for now I'll keep using it with a battery. Heres the current schematic in case its helpful:



Really appreciate all the info you guys have given me along the way!

Quote from: camzilla on September 13, 2024, 04:24:35 AMI wanted to compensate for it because I believed that this bypass cap was solving my noise problem at the cost of cutting high end from my circuit, however now that I have the circuit built and boxed and battery powered, the noise is only present when powered by my pedal power supply. This capacitor was eliminating the noise only because it was drastically limiting the frequency response since I didn't account for miller effect when picking the value, so that was my misunderstanding. I appreciate the insight. Definitely a good thing to know. Seems like a good way to cut high end based upon gain if the gain is adjustable on it, so I'll put that in my back pocket.

The filtering of you hear on the audio signal is also filtering the noise.  It's the same mechanism.   Often the game is to filter some of the noise without filtering too much of the signal.

If you compensate out the cap it will compensate out the noise removal function as well.

As far as your original video goes.   The idea behind compensating out the capacitance is to remove capacitances that aren't put there by the designer.   It's more the opposite of what you are doing.   Even without the an added cb capacitance there is a cb capacitance inside the transistor which you cannot control.  This affects and limits the high frequency performance.  The RF people try to null that out to increase bandwidth (or reduce input capacitance).    ie. it stretches the transistor technology.

Quote from: camzilla on September 13, 2024, 04:24:35 AMMy only gripe is that there is a high pitched noise present in the pedal when powered with my pedal power supply. It is cheap and off brand but supposedly an isolated power supply nonetheless. Currently I have a 100uF and a 22nF capacitor across the supply as close to the transistor 9V connections as I could get on my perfboard setup. I tried adding a 100 ohm resistor in series with the supply before the caps to form a lowpass filter to clean the supply but it did not reduce the noise at all so I did not bother adding it to the perfboard. Very interested in any other PSU noise reduction strategies, for now I'll keep using it with a battery. Heres the current schematic in case its helpful:

The noise comes from the power rails and gets into the audio via Rc1/Rc2.  If you add a resistor say 100 ohm (upto say 1k) between the power input and C6 (100uF)  it will form a low pass filter and remove the noise on the power rail.  You will find many posts on this forum about supply filtering, RC filtering and low pass filters.

Quote from: Rob Strand on September 13, 2024, 04:48:08 AM

Quote from: camzilla on September 13, 2024, 04:24:35 AMMy only gripe is that there is a high pitched noise present in the pedal when powered with my pedal power supply. It is cheap and off brand but supposedly an isolated power supply nonetheless.

The noise comes from the power rails and gets into the audio via Rc1/Rc2.

I'm not sure if OP refers on 100/120 Hz by "high pitched" noise or is using some kind of switching PS..

Quote from: camzilla on September 13, 2024, 04:24:35 AMSo I think this is expected behavior of a properly biased circuit. Would love to hear if that isn't proper for academic purposes at least

This is the behavior of a circuit properly biased for this particular behavior..

Quote from: Rob Strand on September 13, 2024, 04:48:08 AMThe filtering of you hear on the audio signal is also filtering the noise.  It's the same mechanism.  Often the game is to filter some of the noise without filtering too much of the signal.

If you compensate out the cap it will compensate out the noise removal function as well.

As far as your original video goes.  The idea behind compensating out the capacitance is to remove capacitances that aren't put there by the designer.  It's more the opposite of what you are doing.  Even without the an added cb capacitance there is a cb capacitance inside the transistor which you cannot control.  This affects and limits the high frequency performance.  The RF people try to null that out to increase bandwidth (or reduce input capacitance).    ie. it stretches the transistor technology.

Understood

Quote from: Rob Strand on September 13, 2024, 04:48:08 AM

Quote from: camzilla on September 13, 2024, 04:24:35 AMMy only gripe is that there is a high pitched noise present in the pedal when powered with my pedal power supply. It is cheap and off brand but supposedly an isolated power supply nonetheless. Currently I have a 100uF and a 22nF capacitor across the supply as close to the transistor 9V connections as I could get on my perfboard setup. I tried adding a 100 ohm resistor in series with the supply before the caps to form a lowpass filter to clean the supply but it did not reduce the noise at all so I did not bother adding it to the perfboard. Very interested in any other PSU noise reduction strategies, for now I'll keep using it with a battery. Heres the current schematic in case its helpful:

The noise comes from the power rails and gets into the audio via Rc1/Rc2.  If you add a resistor say 100 ohm (upto say 1k) between the power input and C6 (100uF)  it will form a low pass filter and remove the noise on the power rail.  You will find many posts on this forum about supply filtering, RC filtering and low pass filters.

As I said, this was my first instinct and I did try this on my breadboard setup and it did not change the noise at all. I can try it again in the morning as a sanity check but adding a series resistor should be pretty simple.

Quote from: antonis on September 13, 2024, 05:12:35 AM

Quote from: Rob Strand on September 13, 2024, 04:48:08 AM

Quote from: camzilla on September 13, 2024, 04:24:35 AMMy only gripe is that there is a high pitched noise present in the pedal when powered with my pedal power supply. It is cheap and off brand but supposedly an isolated power supply nonetheless.

The noise comes from the power rails and gets into the audio via Rc1/Rc2.

I'm not sure if OP refers on 100/120 Hz by "high pitched" noise or is using some kind of switching PS..

It definitely sounds higher pitched than 60/120hz, I but I could be higher harmonics of that due to it being a treble booster boosting like 50db. Here is the psu I am using if this helps: link

Expirimenting with it now, I noticed that he noise dissapears when the input jack is unplugged while the pedal is being powered by the power supply. While the pedal is on, the psu noise should still be present on the output right? Does this make it more likely that it is a feedback or ground loop error?

1. Are IN & OUT GNDs directly wired..??
(i.e. via a piece of wire..)

2. Did you try to shield your breadboard circuit..??
(e.g. via a grounded wrapped aluminium sheet..)

P.S.
After a rough noise analysis simulation, your circuit exhibits a good S/N ratio, starting from 120dB at 80Hz and going down to 70dB at 50kHz and up..

Quote from: antonis on September 13, 2024, 06:09:31 AM1. Are IN & OUT GNDs directly wired..??
(i.e. via a piece of wire..)

2. Did you try to shield your breadboard circuit..??
(e.g. via a grounded wrapped aluminium sheet..)

P.S.
After a rough noise analysis simulation, your circuit exhibits a good S/N ratio, starting from 120dB at 80Hz and going down to 70dB at 50kHz and up..

My input and output jack grounds are directly wired, but they are also non-insulated jacks so they are also both connected to chassis ground. This is a potential ground loop. The In/Out jacks are the only points of grounding on the enclosure. Generally the random internet forum post research I did seemed to think that this small of a ground loop would not matter. Is it better to have one jack wired and one grounded through the bare aluminum chassis or have both wired with the small ground loop?

Quote from: camzilla on September 13, 2024, 05:28:46 PMIs it better to have one jack wired and one grounded through the bare aluminum chassis or have both wired with the small ground loop?

Practically, I have never seen it make any difference whatsoever.

It appears that the noise is external to my circuit and perhaps a product of the power supply.

I have added the 100ohm resistor in series with the supply before the filtering. I also experimented with unwiring the ground on one of the output jacks to see if the ground loop was the problem. This added noise and did not solve any so I rewired the ground on both outputs, in the future I will invest in insulated jacks for peace of mind though it seems this is pretty standard practice and probably not the culprit of any noise.

The high pitched whining noise only appears with the power supply not with the battery, but also the noise only appeared when my guitar cable was plugged in. Furthermore, with my guitar plugged in using a 6 inch patch cable the noise is not present.
I do not know how using the power supply would make the circuit more vulnerable to external noise, but the noise only appears, using the PS with a long cable. To me this signals that it may be a product of some combination of noise pickup from guitar pickups and the cable, and a weird grounding scheme or nonisolation in the PS?

I realized that the high pitch noise I am experiencing is only present when a guitar is plugged in,  otherwise the circuit just has an ambient hum at a much lower volume than the signal. The circuit has a lot of noise when my unshielded single coil guitar is plugged in, and only the high pitched whine that I was describing yesterday when my les paul style humbucking guitar is plugged in.

To summarize:

I believe I have ruled out the possibility of any ground loop noise and have done everything I can for PSU ripple - series resister, large electrolytic cap, small film bypass cap, RC filter cutoff of about 1hz

with battery- no noise ever
with PSU - only noise when using a long cable, else noise is equivalent to battery ckt

This seems a very fuzzy situation to me. Curious to hear any thoughts

Quote from: camzilla on September 13, 2024, 05:30:02 AMt definitely sounds higher pitched than 60/120hz, I but I could be higher harmonics of that due to it being a treble booster boosting like 50db. Here is the psu I am using if this helps: link

Expirimenting with it now, I noticed that he noise dissapears when the input jack is unplugged while the pedal is being powered by the power supply. While the pedal is on, the psu noise should still be present on the output right? Does this make it more likely that it is a feedback or ground loop error?

If you are using a plug-in breadboard try connecting the aluminium base plate to the circuit ground.  If not get some kitchen foil and place it under the circuit, add an insulator between the foil and the circuit, then connect the foil to circuit ground.

Mains powered supplies can promote capacitively coupled noise.   Normally when you connect the pedal to an amplifier which has a mains earth connection this type of noise is largely removed.   If the amp or audio interface device doesn't have a mains earth connection it's possible for this type of noise can show up.