Melon Soda Boost  [documentation]

Original effects with schematics, layouts and instructions, freely contributed by members or found in publications. Cannot be used for commercial purposes without the consent of the owners of the copyright.
Post Reply
User avatar
asbestosaurus
Information
Posts: 42
Joined: 27 Mar 2020, 00:55
Has thanked: 50 times
Been thanked: 94 times

Post by asbestosaurus »

Okay here's the Melon Soda Boost.
melon soda front.jpg
melon soda side.jpg
The whole idea was to put a tone control before a gain stage rather than after, where it usually is. The result is fairly versatile, from treble boost to almost-fuzz and even reaching a nice 'pushed-fender'y kind of grit. I was initially inspired to use part of soda can as a faceplate because of the colour match from the dark green lettering to the Demon TS enclosure.
Melon Soda Boost schem.PNG
the convention I use for pots is that (unless specified) the wiper moving rightward or upward is the lug 1-to-lug 3 direction or (almost always) clockwise pot rotation.

Fairly standard JFet CS input section. Drain and Source resistors calculated from an interactive GeoGebra page I developed.

Next is a Modified Big Muff Tonestack with a 3-way switch to control mids. I've included screengrabs from the Duncan Tone Stack Calculator to show frequency response. Mid boost in light blue, flat in green and scoop in red. 0%-50%-100% traces for each, maybe a little crowded but I didn't want to upload too many pictures.
Melon Soda Tone Sweep.PNG
Then comes the hybrid Mosfet/PNP gain stage. This one is really fun, it's essentially a discrete IGBT with DC feedback for stable biasing. The overdrive character is hard to describe, and different from a lot of others I tried, especially at high gains. I was hoping to achieve a blend of Mosfet and Germanium characteristics and I think it captures that pretty well.

*Since first posting this I've adjusted the emitter and source resistors slightly and also shorted the 47k to lug three of volume for just a bit more output.*

Lastly a plain emitter follower for a nice low output impedance, perfect for pushing the input of an amp or other drive pedals.
The whole thing is fed voltage through a capacitance multiplier to stop supply noise being coupled into the signal, since discrete transistors have no inherent supply rejection.

I'd love to hear your thoughts and if there's any ideas here you find interesting.
Last edited by asbestosaurus on 06 Oct 2021, 10:45, edited 2 times in total.

User avatar
Jarno
Resistor Ronker
Information
Posts: 358
Joined: 12 Nov 2008, 10:18
my favorite amplifier: Something nice
Completed builds: Alembic-like state-variable and sallen-key filter preamps
Lovepedal Eternity
Phase 100
Brown source
Fuzz Face
Flipster
Alembic F2B (tube preamp)
Opamp and FET buffers
Loads of speakercabinets and ampracks
Busy building a modular synth (ssm2044 vcfs, preamps, ADSR's, VCO's, VCA's)
Tables
Bookshelves
Basses
So many things! :D
Location: Rosmalen, NL
Has thanked: 27 times
Been thanked: 79 times

Post by Jarno »

Nice idea on the frontpanel :D
And also, interesting idea with the capacitance multiplier, class A indeed, no PSRR
"It crackles....., but that's ok"

User avatar
stolen
Breadboard Brother
Information
Posts: 129
Joined: 04 Mar 2021, 09:54
Has thanked: 44 times
Been thanked: 55 times

Post by stolen »

asbestosaurus wrote: 05 Oct 2021, 05:43 Then comes the hybrid Mosfet/PNP gain stage. This one is really fun, it's essentially a discrete IGBT with DC feedback for stable biasing. The overdrive character is hard to describe, and different from a lot of others I tried, especially at high gains. I was hoping to achieve a blend of Mosfet and Germanium characteristics and I think it captures that pretty well.
Whoop whoop whoop new clipper dropped :D! Many thanks, made my day, wanna try asap :3

User avatar
ppluis0
Diode Debunker
Information
Posts: 915
Joined: 14 Jul 2010, 18:33
Location: Argentina
Has thanked: 127 times
Been thanked: 299 times

Post by ppluis0 »

Hi asbestosaurus,

Interesting idea to filter supply noise, but the capacitance presented at emitter is the base capacitance multiplied by the hfe of the transistor, so is irrelevant (IMHO) to put an additional 100uF at this point :hmmm:

Also, clever idea to cut a soda can to attach as faceplate :applause:

Cheers,
Jose

User avatar
mauman
Resistor Ronker
Information
Posts: 419
Joined: 17 Jun 2017, 20:01
Location: Texas, US
Has thanked: 423 times
Been thanked: 183 times

Post by mauman »

I like the tone stack, it fixes everything that I don't like about the BMP "tilt" control. It's an elegant solution and quite versatile!

User avatar
bow_and_error
Breadboard Brother
Information
Posts: 56
Joined: 13 Jun 2019, 23:13
Has thanked: 95 times
Been thanked: 41 times
Contact:

Post by bow_and_error »

asbestosaurus wrote: 05 Oct 2021, 05:43 Then comes the hybrid Mosfet/PNP gain stage. This one is really fun, it's essentially a discrete IGBT with DC feedback for stable biasing. The overdrive character is hard to describe, and different from a lot of others I tried, especially at high gains. I was hoping to achieve a blend of Mosfet and Germanium characteristics and I think it captures that pretty well.
Definitely the first time I've seen an IGBT-based guitar pedal! Can you go into a bit more detail about that circuit block and how it sounds?

User avatar
dylan159
Resistor Ronker
Information
Posts: 267
Joined: 18 Jul 2018, 14:02
my favorite amplifier: Chempion!
Completed builds: Oh boy...
Has thanked: 97 times
Been thanked: 535 times
Contact:

Post by dylan159 »

Interesting one! I can see that some thought went into it, especially the care about power supply noise and output impedance and it gave us something cool and new.
Some questions:
  • What's the reasoning behind D1 and D2? Maybe some extra clipping if the input is driven harder by another circuit? Otherwise the gate diode will be happy to conduct limited by the 1k and 10k if the gate is driven positive and reverse breakdown won't happen unless the gate is driven 25V negative
  • Likewise the LED. I appreciate the thought of protecting the gate from breakdown, but it should be safe within at least +-20V Vgs, so it's either impossible or unlikely the LED will conduct, but it doesn't do harm either.
  • Did you notice scratching on the volume pot? Sure, the base current is low with an high beta transistor and it doesn't get amplified, but given the large resistance of the pot I'd still like to know your experience.
bow_and_error wrote: 06 Oct 2021, 04:35 Definitely the first time I've seen an IGBT-based guitar pedal! Can you go into a bit more detail about that circuit block and how it sounds?
As you can see it's basically a nmos-pnp complementary feedback pair. It saturates like a bjt but has the infinite gate resistance of a mosfet (altough it has the capacitance too). Sound? who knows, maybe some simulations can shed some light. I've checked it myself quickly because I find it interesting and while it doesn't seem to do much for gain compared to just the mosfet (the pnp is basically a dc coupled follower), headroom is reduced a bit.
Personal blog/archive: https://bentfishbowl.wixsite.com/electronics
Find me at https://discord.gg/bMuhX4TkZM Audio Electronics discord server.

User avatar
asbestosaurus
Information
Posts: 42
Joined: 27 Mar 2020, 00:55
Has thanked: 50 times
Been thanked: 94 times

Post by asbestosaurus »

dylan159 wrote: 06 Oct 2021, 08:29 Interesting one! I can see that some thought went into it, especially the care about power supply noise and output impedance and it gave us something cool and new.
Some questions:
  • What's the reasoning behind D1 and D2? Maybe some extra clipping if the input is driven harder by another circuit? Otherwise the gate diode will be happy to conduct limited by the 1k and 10k if the gate is driven positive and reverse breakdown won't happen unless the gate is driven 25V negative
  • Likewise the LED. I appreciate the thought of protecting the gate from breakdown, but it should be safe within at least +-20V Vgs, so it's either impossible or unlikely the LED will conduct, but it doesn't do harm either.
  • Did you notice scratching on the volume pot? Sure, the base current is low with an high beta transistor and it doesn't get amplified, but given the large resistance of the pot I'd still like to know your experience.
diodes are there just for my peace of mind. ive gotten away without them in other builds and they are mostly optional here.

no scratching on the volume pot, not sure about thermal noise although its quiet enough for me. most of the damage is done by the MOS/Ge stage anyway, both are renowned for their less than stellar noise performance. its dc coupled to the output of the last stage (which is biased to about a pn drop above half supply voltage) to save on parts for a voltage divider, decoupling caps etc. the pot is as big as it is to minimise loading on the emitter in combination with the dc feedback. i found a 100k had less output swing.

User avatar
asbestosaurus
Information
Posts: 42
Joined: 27 Mar 2020, 00:55
Has thanked: 50 times
Been thanked: 94 times

Post by asbestosaurus »

bow_and_error wrote: 06 Oct 2021, 04:35
Definitely the first time I've seen an IGBT-based guitar pedal! Can you go into a bit more detail about that circuit block and how it sounds?
dylan159 has already added some helpful technical details. I'll add that it works better with a GE transistor with lower hfe probably sub 80 or so and lower leakage if possible.

I'm not worried about the loss of headroom since its more of a colour booster and there are definitely many better ways to get a high headroom clean boost.

the DC feedback network might be better with trimpot but in breadboarding i found equal resistors (both 220k) got it where i wanted. the absolute size of these isn't critical but bigger means less load on the emitter so 100s of kOhms is good.

I'm not great with adjectives but its a real squishy kind of gain. MOSFET input capacitance and GE saturation mean its definitely not a harsh distortion.

User avatar
asbestosaurus
Information
Posts: 42
Joined: 27 Mar 2020, 00:55
Has thanked: 50 times
Been thanked: 94 times

Post by asbestosaurus »

ppluis0 wrote: 06 Oct 2021, 01:19 Hi asbestosaurus,

Interesting idea to filter supply noise, but the capacitance presented at emitter is the base capacitance multiplied by the hfe of the transistor, so is irrelevant (IMHO) to put an additional 100uF at this point :hmmm:

Also, clever idea to cut a soda can to attach as faceplate :applause:

Cheers,
Jose
The extra 100u is a vestige from an earlier (noisier) prototype without the cap multiplier. since i reused the same veroboard layout i couldn't be bothered removing it. The extra cap is included in places like the BD-2 schem and also the Sound AU page on the same topic so i figured it doesn't hurt to leave it there.

User avatar
asbestosaurus
Information
Posts: 42
Joined: 27 Mar 2020, 00:55
Has thanked: 50 times
Been thanked: 94 times

Post by asbestosaurus »

mauman wrote: 06 Oct 2021, 02:06 I like the tone stack, it fixes everything that I don't like about the BMP "tilt" control. It's an elegant solution and quite versatile!
thanks. i tried my best to make it as symmetrical as possible. if on-on-on switches are hard to come by (they are for me, in DIY quantities) a more flexible solution is to use two independent on-on switches, this also gives you the bonus of another flat setting with a higher centre frequency.

User avatar
dylan159
Resistor Ronker
Information
Posts: 267
Joined: 18 Jul 2018, 14:02
my favorite amplifier: Chempion!
Completed builds: Oh boy...
Has thanked: 97 times
Been thanked: 535 times
Contact:

Post by dylan159 »

asbestosaurus wrote: 06 Oct 2021, 09:17 diodes are there just for my peace of mind. ive gotten away without them in other builds and they are mostly optional here.

no scratching on the volume pot, not sure about thermal noise although its quiet enough for me. most of the damage is done by the MOS/Ge stage anyway, both are renowned for their less than stellar noise performance. its dc coupled to the output of the last stage (which is biased to about a pn drop above half supply voltage) to save on parts for a voltage divider, decoupling caps etc. the pot is as big as it is to minimise loading on the emitter in combination with the dc feedback. i found a 100k had less output swing.
Hehe damage indeed, same way I'd say it. :) I have to say though, the jfet input stage helps a lot with noise, being relatively quiet and the first thing in the circuit, since everything done at the input is super important for overall noise. I also do like to save parts.
About the loading, if you approximate the output impedance to 22k (which seems close enough in simulation), the x10 rule of thumb gives you a 220k minimum load, which the top bias resistor shunted by the 33u already reaches, so yeah a big volume pot is necessary, even if you increase the resistors in the bias divider.
asbestosaurus wrote: 06 Oct 2021, 09:31 dylan159 has already added some helpful technical details. I'll add that it works better with a GE transistor with lower hfe probably sub 80 or so and lower leakage if possible.

the DC feedback network might be better with trimpot but in breadboarding i found equal resistors (both 220k) got it where i wanted. the absolute size of these isn't critical but bigger means less load on the emitter so 100s of kOhms is good.
Simulation seems to agree that hfe isn't important, and neither is supply voltage. thanks to emitter degeneration and dc feedback this stage should work the same no matter what you throw at it. The loss of headroom can be beneficial since the goal is distortion.
asbestosaurus wrote: 06 Oct 2021, 09:53 thanks. i tried my best to make it as symmetrical as possible. if on-on-on switches are hard to come by (they are for me, in DIY quantities) a more flexible solution is to use two independent on-on switches, this also gives you the bonus of another flat setting with a higher centre frequency.
I haven't tried but I'm pretty sure you can get there with an on off on switch too, changing some values around and either shorting series caps or connecting them in parallel. At most the order will be mixed.
Personal blog/archive: https://bentfishbowl.wixsite.com/electronics
Find me at https://discord.gg/bMuhX4TkZM Audio Electronics discord server.

Post Reply