Hello everybody. This particular box is somewhat noisy in its stock form. My intuition says a large part is coming from the first stage. I see a 1M elevating resistor before the JFET gate. Would this add significant noise to the circuit, even though it is not directly in the signal path?
Attachments
Fuzz-boxes have a lot of gain in them, so it's not entirely unexpected that they might hiss a lot. This one is the infamous "metal zone", i.e., it is designed to have an entirely ridiculous excess of voltage gain. Hiss is almost inevitable.
In a well-designed audio amp, the first stage invariably generates most of the noise, simply because the first stage has voltage gain, and so the signal level is much bigger from that stage onwards. That means the stronger signal swamps out noise from subsequent stages.
Unfortunately, this circuit violates that design principle - the input stage is a JFET source-follower, with a voltage gain of slightly less than unity (1). So this stage adds noise, without making the signal any bigger - a bad idea, particularly when followed by a ton of gain from subsequent stages!
Guitars need an input impedance of around 1 megohm from the circuit they're plugged into. This was not available with the bipolar transistor op-amp used in the MT-2. The poor design with the unity-gain input stage was most likely driven by the unavailability of a low-noise, FET-input op-amp when this pedal was designed. Instead, they used a JFET to get the required input impedance, but made the bad decision to set it up for unity voltage gain.
Speaking of op-amps, what sort of op-amp is in the circuit? Is it a 4558? That's not what the schematic shows, but online, I found other owners of the product who found 4558 op-amps in the pedal. (Edit: apparently it's a version of the op-amp with 8 pins in a single row, which is going to make op-amps swaps almost impossible; you'll need some sort of adaptor board to go from a standard 8-pin DIL to an 8-pin SIL layout.)
The 4558 dates from the dawn of audio-op amps, and it's noisier than some later op-amps. You might get a slight improvement in noise by swapping to more modern op-amps instead. But de-soldering and removing op-amps from a small PCB with narrow, delicate copper tracks is a quite a delicate operation, and might kill your pedal stone dead if anything goes wrong during the process. (And any audible reduction in noise will be slight at best.)
That resistor will add noise when the input is open-circuit. But as soon as you connect a signal source to the input, the source impedance of that signal appears in parallel with the 1M, lowering it. If there is another stomp-box in the chain before the fuzz box, the impedance might be lowered a lot (but the noise of that preceding stomp-box is now likely to dominate.)
What happens to the noise level if you insert a 1/4" mono plug wired to short-circuit the input? This will bypass all noise from that 1M resistor, so what remains will be noise from the input JFET and the subsequent op-amps.
Rather than fiddle with the delicate innards of the MT-2, I would try using a very quiet (low noise) clean gain pedal between guitar and MT2. If you can get a signal gain of even a few dB ahead of the MT2, from a more modern pedal with better noise performance of its own, that should cut down the noise from the MT-2 itself (you can turn the output level down more, cutting the noise with it.)
But I have a suspicion you're always going to have to live with noise when you use a metal distortion pedal - it's the nature of the beast. Which is why noise-gates were invented, and are invariably used by guitarists who play metal and other genres that use extremely high-gain signal chains.
-Gnobuddy
Last edited:
Good stuff to think about.
"What happens to the noise level if you insert a 1/4" mono plug wired to short-circuit the input? This will bypass all noise from that 1M resistor, so what remains will be noise from the input JFET and the subsequent op-amps."
This I will try. There is a slight increase in hiss when I plug the guitar in. I don't remember which volume setting increased it, though. I do know it has crappy, carbon pots in it.
The opamps are NJM4558L and I am restricting any noise work to the first stage.
Ultra low-noise JFETs are almost impossible to find in small quantities, like the 2SK184. I found the LSK389 (a dual) on eBay for reasonable price.
I agree with the noise gate idea, but I would rather do without it, and I will not be using all the gain available from this particular box. The Jensen C10Q speaker is not high gain friendly so this is adding to the problem, although they sound very sweet without the fuzz box.
Fromel Electronics makes a kit that improves the MT-2 in several ways and is cheap. It consists of film caps in lieu of the ceramics, MOSFETs to replace the clipping diodes, and EQ mods that make the unwieldy unit more managable. They do not mention noise reduction, but if I can get a good sound without too much drive, it may ameliorate the hiss due to the lower settings.
Since the JFET is nothing but a buffer, your idea to replace that with an opamp sounds like the best idea. There are some out there with a noise figure that of a 1k resistor. I would replace the 10k input resistor with a metal foil type, and add a 800-1k just before the opamp input.
Thanks again for the good ideas.
"What happens to the noise level if you insert a 1/4" mono plug wired to short-circuit the input? This will bypass all noise from that 1M resistor, so what remains will be noise from the input JFET and the subsequent op-amps."
This I will try. There is a slight increase in hiss when I plug the guitar in. I don't remember which volume setting increased it, though. I do know it has crappy, carbon pots in it.
The opamps are NJM4558L and I am restricting any noise work to the first stage.
Ultra low-noise JFETs are almost impossible to find in small quantities, like the 2SK184. I found the LSK389 (a dual) on eBay for reasonable price.
I agree with the noise gate idea, but I would rather do without it, and I will not be using all the gain available from this particular box. The Jensen C10Q speaker is not high gain friendly so this is adding to the problem, although they sound very sweet without the fuzz box.
Fromel Electronics makes a kit that improves the MT-2 in several ways and is cheap. It consists of film caps in lieu of the ceramics, MOSFETs to replace the clipping diodes, and EQ mods that make the unwieldy unit more managable. They do not mention noise reduction, but if I can get a good sound without too much drive, it may ameliorate the hiss due to the lower settings.
Since the JFET is nothing but a buffer, your idea to replace that with an opamp sounds like the best idea. There are some out there with a noise figure that of a 1k resistor. I would replace the 10k input resistor with a metal foil type, and add a 800-1k just before the opamp input.
Thanks again for the good ideas.
Low noise guitar pu amplification requires low noise CURRENT input devices. So you may forget any op-amp with bipolar input that offer low VOLTAGE noise but high noise current. I fiddled around extensively with these things so I can say the best bet here is a TL071 op-amp with a gain of 3-5V/V. There is no further room of improvement using discrete low noise JFETs.
@Gnobuddy: I know and accept your bias against hard rock guitars. And I do not accept noisy stuff as well. So I built a "hot" pre-amp with excessive gain - and extreme low noise. Maybe later on I will deliver some sound sample...
@Gnobuddy: I know and accept your bias against hard rock guitars. And I do not accept noisy stuff as well. So I built a "hot" pre-amp with excessive gain - and extreme low noise. Maybe later on I will deliver some sound sample...
Agreed. The venerable 4558 actually isn't bad when it comes to voltage noise! Unfortunately that doesn't help - for the input stage, we need an op-amp with a good noise figure when the source impedance is tens to hundreds of kilo-ohms, because of the nature of the electric guitar.
I guess you could say that as soon as you develop preferences for any type of music, you have a bias against the types you don't like as much.
In my case, I enjoy the sound of chords with some colour in them - major sevenths, major sixths, suspended seconds and fourths, and so on. These simply don't work with heavily distorted amps, and the two-note "power chords" that are the staple of more distorted genres don't hold my interest for long.
I look forward to it!
-Gnobuddy
Opamp selection
How about the LME49710? Does anyone think this source is a good risk?
2pcs LME49710HA ADAPTER TO-99 | eBay
How about the LME49710? Does anyone think this source is a good risk?
2pcs LME49710HA ADAPTER TO-99 | eBay
The setup while recording the sound clip was this one:
Download link:
Dropbox - crossroads_TASCAM_0110.mp3 - Simplify your life
Download link:
Dropbox - crossroads_TASCAM_0110.mp3 - Simplify your life
Attachments
Last edited:
Sound Clip Evaluation
Well, it sounds dead quiet. You recommended the TL071 op-amp with a gain of 3-5V/V. This is too much boost to place as the first stage of the MT-2; it will have to be attenuated to unity gain. Is it necessary to have this gain to keep the noise low?
Oh, skip the question about the Chinese TO-99 supplier (bad idea). If I go with the LME49710, the SOIC package will be mounted on a PDIP adapter. This is small and makes attaching the wires easy.
Well, it sounds dead quiet. You recommended the TL071 op-amp with a gain of 3-5V/V. This is too much boost to place as the first stage of the MT-2; it will have to be attenuated to unity gain. Is it necessary to have this gain to keep the noise low?
Oh, skip the question about the Chinese TO-99 supplier (bad idea). If I go with the LME49710, the SOIC package will be mounted on a PDIP adapter. This is small and makes attaching the wires easy.
To address your last question first: Yes, if you don't have gain in the first stage, you add noise without increasing the signal voltage...so the second stage is as noisy as ever, and the first stage has made things even worse by adding additional noise!
This is exactly why having the (slightly less than) unity-gain JFET source follower is not a great idea from the noise point of view.
Back to your problem: why not simply lower the gain of the first op-amp stage (3A) in the MT-2? All you'd have to do is increase the value of the 1k feedback resistor (RD41) from the non-inverting input to ground. Increase that to 4.7k, and you drop the gain by a factor of about five. Now you can insert the TL071 with a gain of 5 in front, and end up with about the same signal level out of op-amp 3A.
-Gnobuddy
That is in *shunt* with the signal source. Approximating guitar as 20k, and another 10k RF protection, the hiss voltage of the 1Meg is attenuated 30:1. Hiss would be lower if you made it 10Meg, but the difference would be inaudible.
Mostly, it's just got a LOT of gain. The devices and circuits are not lowest-hiss, but not far off. Taking gain out one place and putting-back another is unlikely to change things much.
Guitar should be full-up most of the time; ample source signal is the first rule of "low noise".
There is a remote chance that JFET or opamp have been zapped by excess level or static shock. This is an occasional problem on much older gear, and this one is not young anymore.
Yes, but you are right, and I got it all wrong.
Between the tiny schematic and the ancient thrift-store computer monitor I use at work (because it's better than the official one!), I managed to completely miss the first op-amp, surrounded as it is by clouds of JFETs and resistors and caps.
I agree with PRR that you won't get a huge noise improvement by shuffling gain around. But it's easy enough to get whatever small improvement can be had by using an external clean-boost pedal between guitar and MT-2.
You could tack-solder a resistor in parallel with that 220k NFB resistor to lower the gain, so you can easily remove it if you want to return the pedal to stock.
Who knows, that MT-2 might be worth enough for you to retire on some day, if you leave it stock. Stranger things have happened!
-Gnobuddy
@PRR:
"That is in *shunt* with the signal source. Approximating guitar as 20k, and another 10k RF protection, the hiss voltage of the 1Meg is attenuated 30:1. Hiss would be lower if you made it 10Meg, but the difference would be inaudible."
With the guitar plugged in and not outputting anything, the filter cap is isolating the the instrument, so I'm having trouble seeing the parallel resistance in this case.
"Mostly, it's just got a LOT of gain. The devices and circuits are not lowest-hiss, but not far off. Taking gain out one place and putting-back another is unlikely to change things much."
The pedal has a reputation for being noisy, due to its very high gain nature and lower than Hi-Fi component quality. Deleting the first JFET and jiggling the gains around is inexpensive and not that time consuming. The effort would be worth it to me if a 5% hiss reduction is achieved.
Guitar should be full-up most of the time; ample source signal is the first rule of 'low noise'.
This is impossible for those who wish some amplitude dynamics. Technique is enough for "mind-blowing shredders".
"There is a remote chance that JFET or opamp have been zapped by excess level or static shock. This is an occasional problem on much older gear, and this one is not young anymore."
The seller admitted the pedal came in with almost zero voltage on the 4.5v leg. and the unit had zero output. He replaced all the opamps which solved the problem. He also surmised the unit had the wrong power supply connected which blew the blocking diode. The box in question is one of the older models with a through-hole PCB. Since 2017 they use SMT.
@Gnobuddy:
"Yes, but you are right, and I got it all wrong." Don't sweat it, we all do it.
"I agree with PRR that you won't get a huge noise improvement by shuffling gain around. But it's easy enough to get whatever small improvement can be had by using an external clean-boost pedal between guitar and MT-2." I would rather incorporate the "clean boost" technique into the MT-2.
"You could tack-solder a resistor in parallel with that 220k NFB resistor to lower the gain, so you can easily remove it if you want to return the pedal to stock." This is a truism.
@mandu: "A noise gate will keep the silent passages quiet for the worst case." I'm actually looking for a level of distortion not far from the MXR Distortion Plus, which did not require a noise gate, but had no EQ. The MT-2 is unwieldy in stock form but more versatile. The Fromm kit corrects this and improves quality. Speaking of the Fromm kit, I received this response to my question as the whether their kit reduces hiss:
"Yes the kit will help to lower some of the noise in the circuit. We upgrade a number of blocking caps with the kit and that can help a low with hiss and noise."
This does not sound too comforting. I think he meant to say, "..can help a little with hiss and noise."
"That is in *shunt* with the signal source. Approximating guitar as 20k, and another 10k RF protection, the hiss voltage of the 1Meg is attenuated 30:1. Hiss would be lower if you made it 10Meg, but the difference would be inaudible."
With the guitar plugged in and not outputting anything, the filter cap is isolating the the instrument, so I'm having trouble seeing the parallel resistance in this case.
"Mostly, it's just got a LOT of gain. The devices and circuits are not lowest-hiss, but not far off. Taking gain out one place and putting-back another is unlikely to change things much."
The pedal has a reputation for being noisy, due to its very high gain nature and lower than Hi-Fi component quality. Deleting the first JFET and jiggling the gains around is inexpensive and not that time consuming. The effort would be worth it to me if a 5% hiss reduction is achieved.
Guitar should be full-up most of the time; ample source signal is the first rule of 'low noise'.
This is impossible for those who wish some amplitude dynamics. Technique is enough for "mind-blowing shredders".
"There is a remote chance that JFET or opamp have been zapped by excess level or static shock. This is an occasional problem on much older gear, and this one is not young anymore."
The seller admitted the pedal came in with almost zero voltage on the 4.5v leg. and the unit had zero output. He replaced all the opamps which solved the problem. He also surmised the unit had the wrong power supply connected which blew the blocking diode. The box in question is one of the older models with a through-hole PCB. Since 2017 they use SMT.
@Gnobuddy:
"Yes, but you are right, and I got it all wrong." Don't sweat it, we all do it.
"I agree with PRR that you won't get a huge noise improvement by shuffling gain around. But it's easy enough to get whatever small improvement can be had by using an external clean-boost pedal between guitar and MT-2." I would rather incorporate the "clean boost" technique into the MT-2.
"You could tack-solder a resistor in parallel with that 220k NFB resistor to lower the gain, so you can easily remove it if you want to return the pedal to stock." This is a truism.
@mandu: "A noise gate will keep the silent passages quiet for the worst case." I'm actually looking for a level of distortion not far from the MXR Distortion Plus, which did not require a noise gate, but had no EQ. The MT-2 is unwieldy in stock form but more versatile. The Fromm kit corrects this and improves quality. Speaking of the Fromm kit, I received this response to my question as the whether their kit reduces hiss:
"Yes the kit will help to lower some of the noise in the circuit. We upgrade a number of blocking caps with the kit and that can help a low with hiss and noise."
This does not sound too comforting. I think he meant to say, "..can help a little with hiss and noise."
@voltwide: I like your approach with your design, especially the noise reduction. I had an idea myself that is similar in its effect. I called it a "soft-gate", meaning it did not completely shut off the signal, but slowly attenuated it to near zero as the notes/chords decayed. The reason being that it would not be that noticeable.
There are pedals that sport some kind of "smart" circuitry at significant expense. But I'm not going to buy one only to find out I don't like it.
For me, the best and first assault would be in the inherent reduction of hiss, and then go with the gate only if necessary.
The only reason I am messing with this pedal is due the nature of the Fender Vibrolux Reverb. I had used the MXR Distortion Plus with a Super Reverb and it worked flawlessly, except with much limited distortion when compared with the MT-2.
"Caps rolling is what people do who have no clue." Fromm does not state the kit reduces any kind of noise. It's advertized as a performance enhancing mod.
There are pedals that sport some kind of "smart" circuitry at significant expense. But I'm not going to buy one only to find out I don't like it.
For me, the best and first assault would be in the inherent reduction of hiss, and then go with the gate only if necessary.
The only reason I am messing with this pedal is due the nature of the Fender Vibrolux Reverb. I had used the MXR Distortion Plus with a Super Reverb and it worked flawlessly, except with much limited distortion when compared with the MT-2.
"Caps rolling is what people do who have no clue." Fromm does not state the kit reduces any kind of noise. It's advertized as a performance enhancing mod.
Last edited:
I remember that you posted some details of your Tramp about a year ago.
The sound-clip you posted has a distortion level appropriate to blues or early ("classic") rock. But metal demands much, much, higher gain, and much harsher distortion.
For metal, you keep adding gain until the guitar stops sounding like a guitar, and begins to sound like a circular saw cutting through a tin roof, and then you add another 20 dB of gain for good measure.
I have a hypothesis that the music we love as adults echoes the ambient sounds we heard around us as children. For instance, classical Western music echoes the sounds of nature, with horns that sound like the bellowing of cows or antelope, the viol family that sounds like the human voice, etc. Some music of the time is quite explicit about this - Vivaldi's "Four Seasons", say, or Vaughan William's "The Lark Ascending". A few hundred years ago, that's what people heard around them as they grew up - the sounds of nature.
Classic rock echoes nature too, but a more aggressive side of nature. Thunder, growling predatory animals, things like that.
But IMO metal echoes the sounds of a big city in the mid to late 20th century: jackhammers tearing up the streets, big trucks accelerating with deafeningly loud growling engines, shrieking metallic train brakes, metal dumpsters being banged about and dragged along the concrete by the trash truck, et cetera. The sounds of tortured metal, in other words. The fact that the music is called "metal" isn't a coincidence!
So the purpose of a metal distortion pedal is to take a musical instrument, and make it sound like the worst part of inner-city Chicago in the 1980s. You need a lot of gain to do that!
-Gnobuddy
It's a very old idea. Nearly fifty years ago, in the early 1970s, Phillips designed and sold their "Dynamic Noise Limiter", a circuit that rolled off treble frequencies, but only when the input signal was below a certain strength. It was intended to reduce playback hiss from the company's audio Compact Cassettes, and it worked quite well. (Then Dolby B came along, and took over much of the market.)
Google should turn up the DNL's schematic; I built one in my teens, and it did quite a good job of reducing cassette hiss as best I can recall. The circuit wasn't complex, being from the dawn of solid-state audio circuitry. But it used silicon epitaxial transistors and small-signal silicon diodes, so it didn't suffer from the problems of previous circuitry using germanium transistors.
Changing the value (capacitance) of a cap can change the frequency response, which can be audible. But substituting one type of cap for a different (invariably more expensive) cap of the same value makes no audible change at all, despite legions of technically ignorant audiophiles making such claims.
Lots of people still do it, and usually convince themselves that they hear a change - which is very easy to do. We humans very routinely fool ourselves into hearing and seeing things that don't actually exist (Google "optical illusions" or "auditory illusions" for plenty of examples), so this sort of subjective impression is rarely meaningful, unless it's the result of a proper series of double-blind listening tests, followed by a proper statistical analysis to confirm that the test results are unlikely to have occurred by sheer chance.
I know nothing about Fromm or the kit in question - if the kit supplies different cap values from the original pedal, it might change the sound in a way that some musicians might prefer. If it only substitutes more expensive caps of the same value, though, then all it does is siphon money out of your wallet.
-Gnobuddy
The coupling cap isolates the instrument at DC - but the whole point of a coupling cap is that at the frequencies of interest, it couples the input signal through (because its reactance is infinite at DC, but falls as ~1/frequency .)
I said the same thing as PRR in post #2 - i.e., that the 1M resistor is shunted by the source impedance of the input signal.
Remember, people sign up at diyAudio with varying levels of expertise in electronics. Some have essentially zero knowledge, others come in with Master's degrees in engineering and years of experience in the field. Most are nearer to the former than the latter end of the spectrum.
So, until I know better, I assume very little previous knowledge when responding to posts from a user I haven't encountered before. Not everyone realizes that it's easier to tack-solder an additional resistor in parallel, than to desolder and remove an existing resistor on a PCB and replace it with a different value. Not everyone even realizes that either thing accomplishes the same end-result in a negative feedback network.
So, it's a truism for you (good!) But not necessarily for everyone who posts here.
-Gnobuddy
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.