Cybercow
Well-known member
- Build Rating
The Little Green Scream Machine is an excellent Tube Screamer circuit and modding it is a treat. All I did was add some clipping element switching with a 6-way rotary switch using mostly not-so-standard clipping devices.
Found these nifty little 9mm sized 6-way rotary switches on SmallBear and they had a link to a shared OSH Park PCB for mounting switching option components. Wanted to have a means of ear-balling the sonic difference between 6 different diode clipping options.
In this compound clipping configuration are the following diodes\LEDs used:
(That's about $60 worth of clipping components alone. The LM313H reference diode is a special little device that basically has two diodes in a single 2-pin can. The LM313H cannot be subjected to any DC voltage - much like the Nanolog devices cannot be exposed to DC voltages. And the LM313H can are about twice as expensive as the Nanolog devices.)
After getting the circuit completed and enclosed, I got it fired up and started checking out the six different clipping options. At first, I quickly ran thru the selections to listen for any significant differences. Obviously the Fvd differences were most noticeable volume-wise and the LEDs were the loudest. All Fvd values were acquired from the respective datasheets where testing was not possible with a Peak Atlas DCA75. (The LM313H, Nanolog N2 & N3 devices cannot withstand direct DC voltages.)
With no sine wave generator at the time, I was unable to provide the circuit with a pure sine wave to get any scope readings of the clipped wave forms. So this dissertation is an aural exploration only. I used a ’79 Strat with switchable dual hum-bucker & single bridge pickup (with the Volume set to max) thru a Vox MV50 AC set mostly to a clean signal with the gain set to 1/3 and the Volume set to 1/2 - pumping a pair of 12’ Marshall speakers.
I found nothing extremely “special” about ANY of the devices described here. Even the Nanaolog N2 & N3 devices, with their crazy high price and “mojo”, super-secret quantum tunneling physics marketing-speak held nothing sonically spectacular for me that couldn’t be achieved with some creative pre and\or post filtering in most any dirt circuit. The LM313H was a fluke find almost 3 years ago and only recently did I apply it in a clipping circuit to see how they sound (behave). They are a super-specialized reference diode. And like the Nanolog devices, are just price prohibitive.
In conclusion . . . . I recommend experimenting, use sockets in testing, and settling on using what sounds good to you. For me, this experiment taught me little more than how throw money at a question instead of investing in a solution. What I take away from this little project is keep it simple and do a little testing before wrapping up a build; as I found no point trying to get fancy with regard to signal clipping with diodes. There are already plenty of inexpensive yet creative ways to provide a nice assortment of clipping options for most any circuit.
Meanwhile, the following images are the photojournalist of the complete build . . . .
Thanks for looking. Happy building!
Found these nifty little 9mm sized 6-way rotary switches on SmallBear and they had a link to a shared OSH Park PCB for mounting switching option components. Wanted to have a means of ear-balling the sonic difference between 6 different diode clipping options.
In this compound clipping configuration are the following diodes\LEDs used:
- LM313H Reference Diode (Fvd ~0.67v symmetrical)
- MA856 (Fvd ~0.7v to ~1.0v symmetrical - “King of Tone” fame)
- 3mm Red/Green 2-pin LED (Fvd ~2v & 1.7v asymmetrical)
- 5mm Red/Blue 2-pin LED (Fvd ~2.7v & 1.9v asymmetrical)
- Nanolog N2 (Fvd ~0.7v symmetrical)
- Nanolog N3 (Fvd ~1v symmetrical)
(That's about $60 worth of clipping components alone. The LM313H reference diode is a special little device that basically has two diodes in a single 2-pin can. The LM313H cannot be subjected to any DC voltage - much like the Nanolog devices cannot be exposed to DC voltages. And the LM313H can are about twice as expensive as the Nanolog devices.)
After getting the circuit completed and enclosed, I got it fired up and started checking out the six different clipping options. At first, I quickly ran thru the selections to listen for any significant differences. Obviously the Fvd differences were most noticeable volume-wise and the LEDs were the loudest. All Fvd values were acquired from the respective datasheets where testing was not possible with a Peak Atlas DCA75. (The LM313H, Nanolog N2 & N3 devices cannot withstand direct DC voltages.)
With no sine wave generator at the time, I was unable to provide the circuit with a pure sine wave to get any scope readings of the clipped wave forms. So this dissertation is an aural exploration only. I used a ’79 Strat with switchable dual hum-bucker & single bridge pickup (with the Volume set to max) thru a Vox MV50 AC set mostly to a clean signal with the gain set to 1/3 and the Volume set to 1/2 - pumping a pair of 12’ Marshall speakers.
- The LM313 reference diode has the typical clipping harshness of silicon diodes with what sounds like a bit more bottom end available than the other silicon devices in this selection group. An unusual device in that it is ridiculously expensive for use as a clipping diode configuration. One device has several components, sounding much like two diodes, back-to-back like a standard symmetrical clipping diode pair - only one device is required for clipping. (I didn't test for any actual "symmetry".) Sound-wise, it’s nothing to write home about and are a prohibitive choice due to their expense. (When I found mine over a year ago, they were about $25 each. They are bit less expensive now, but a bit harder to find.)
- MA856 diodes are known for their explicit use in genuine “King of Tone” pedals and are supposed to be “rare”. However I found 10 of them at an easy price (about $1 each) and tried them out in this build. This clipping option (MA856) is the only one that required 2 elements to comprise the back-to-back diodes to form a symmetrical clipping option. To my ears, with the given setup in this Tube Screamer circuit, they are also standard silicon sounding clipping devices that seem to exhibit nothing special about them. Still, they do their job well.
- The 3mm red\green 2-pin dual-color LED has a measured Fvd of 2.0v and 1.7v respectively for each color. This provides an instant asymmetrical clipping diode configuration with a single element. Two diodes (LEDs) in a single 2-pin package. Louder than than four of the other clipping options in this build, these provide a slightly different set of harmonics in their clipped result. About the same price as regular LEDs, but only one is required.
- The 5mm red\blue 2-pin dual-color LED has a measured Fvd of 2.7v and 1.9v respectively for each color. These also provide an instant asymmetrical clipping diode configuration with a single element. Two diodes (LEDs) in a single 2-pin package. With the highest Fvd of all the devices in this clipping selection group, this clipping device selection is the loudest and the volume difference between this and the 3mm red\green 2-pin dual-color LED is clearly audible. I find this another good asymmetrical clipping device for use for clipping in a feedback loop with an opamp. About the same price as regular LEDs, but only one is required.
- The Nanolog N2 dedicated clipping device has a specified Fvd of ~0.7v and was (aurally) noticed immediately when switching from the 5mm red\blue 2-pin dual-color LED. Also immediately aurally noted was a drop in the noise floor. While not playing anything thru the circuit with the N2 selected, I noticed a drop in the slight noise floor of the Tube Screamer; as if the highs were turned way down - or that a low-pass filter was suddenly engaged. While playing my Strat thru the circuit with this clipping option selected, the rich harmonics of clipping were still apparent, but also with a somewhat diminished extreme upper harmonic content - AND - a notable clarity of of the bottom end. While this circuit build does not include any germanium clipping device options, the N2 tends to be very “germanium” sounding, with a softer clipping knee. (Can I “hear” a ‘clipping knee’? - No. I cannot. But I’ve heard enough Tube Screamers with Ge and Si clipping to know the difference.) Again, this is an expensive device for clipping - coming in at ~ $16 per device.
- The Nanolog N3 dedicated clipping device has a specified Fvd of ~1.0v and is noticeably a little bit louder than the N2. The N3 also exhibits a definite drop in the noise floor in the highs, but still renders some sweet clipping harmonics while playing. Sonically similar to the N2, the N3 is just a little bit louder, presumably due to its higher Fvd. This Nanolog device is priced the same as the N2.
I found nothing extremely “special” about ANY of the devices described here. Even the Nanaolog N2 & N3 devices, with their crazy high price and “mojo”, super-secret quantum tunneling physics marketing-speak held nothing sonically spectacular for me that couldn’t be achieved with some creative pre and\or post filtering in most any dirt circuit. The LM313H was a fluke find almost 3 years ago and only recently did I apply it in a clipping circuit to see how they sound (behave). They are a super-specialized reference diode. And like the Nanolog devices, are just price prohibitive.
In conclusion . . . . I recommend experimenting, use sockets in testing, and settling on using what sounds good to you. For me, this experiment taught me little more than how throw money at a question instead of investing in a solution. What I take away from this little project is keep it simple and do a little testing before wrapping up a build; as I found no point trying to get fancy with regard to signal clipping with diodes. There are already plenty of inexpensive yet creative ways to provide a nice assortment of clipping options for most any circuit.
Meanwhile, the following images are the photojournalist of the complete build . . . .
Thanks for looking. Happy building!
Attachments
Last edited:
