Novice Driver: A minimalist 2-knob overdrive

[Fancy Lime]

Hi there,

I recently posted the Angry Grouse (https://www.diystompboxes.com/smfforum/index.php?topic=126139.0), where I found that the tone and clipping stages can be combined into one surprisingly well without loosing much effectiveness, at least for bass or treble cuts (bass boost is the most problematic). That got me thinking, why not use this principle in a minimalist opamp overdrive and also make use of the fact that the gain control can be setup to act as a simultaneous treble control? I present, the Novice Driver:



EDIT: In response to duck_arse's good point below: Plus is right on C2, plus is left on C5, and plus is top on C8. I'll add this info to the next iteration of the schematic.

The way it is set up here, it has a flat frequency response down to 22Hz (no TS-like bass cut/mid hump) but this can of course be adjusted by lowering C2 (e.g. C2 = 68nF for a TS-like response). The treble response is controlled by C4. Without going into too much technical detail, what C4 does is effectively fixing the high frequency gain independently of the gain setting, above a certain frequency and a certain gain setting. In practical terms that means that the amount of overtones (distortion products) cannot increase further if gain is increased, once that limit has been reached. However, increasing the gain further will still increase the sustain. It all sounds rather complicated, but the bottom line is: if we choose C4 right, we can maintain a pleasant amount of treble that will not change too dramatically if we change the gain knob, which was the whole point of the exercise.

I rather like the sound of this one. Nice beefy alternative to a DOD 250. I think this is a nice circuit for beginners (hence the name) because it is super easy to breadboard and changes to certain values have a simple and obvious effect on the sound, so it is rather educational to play with them. The clipping diodes can obviously be changed to whatever catches your fancy, the option given here is just one example, which I personally liked best here. Use NOS Soviet germanium diodes or MOSFET body diodes or both for extra mojo points. You can obviously add switches to change C2 and C4.

The opamp can be any single opamp or one unit of a dual or quad. The way the values are chosen in the diagram assumes very high input impedance. This works well with pretty much any FET-input opamp (TL071, LF356, OPA134, CA3130, TL072, LF353, OPA2134,...) or with very high input impedance BJT opamps (I tested an OP07). For lower input impedance devices, the following changes should be made: R2=470kΩ, R8=R9=33kΩ, C8=4.7μF. The rest should be fine. The you can use a 741 or 4558 for more additional mojo points. It will not sound audibly different (apart from the lower input impedance), though but who has that ever stopped. If you add the appropriate compensation cap, you can even use an LM308! If you do that, you definitely want to go for the metal can version. Point is, the circuit cares very little what opamp you throw at it, and you'll need golden ears to hear a difference, so use whatever you have. If you use a dual though, do not forget to stabilize the unused half by connecting the non-inverting input to the junction between R8 and R9 and connecting the output to the inverting input directly.

I'll try and get a 2-layer PCB layout done for this but I suck at that, so it may be a while. If someone wants to give some novices a leg up by providing a perf or strip layout, that'd be awesome! I'm embarrassingly bad at those, too.

Please have fun and drive responsibly!

Cheers,
Andy

[duck_arse]

[ Novices are known for their questions - like which way does the positive go on those caps? cause they'll probably use electros for the 1uF and up's. maybe add (+) to the schematic. my 2 cents werth. ]

[cab42]

Hi there,

If someone wants to give some novices a leg up by providing a perf or strip layout, that'd be awesome!

Thanks Andy,

I just did a quick vero layout while waiting for the pasta to cook. It's Done with pen and paper, so I need to redraw it in DIY Layout Creator.

I hope to be able to do it later tonight.

Carsten

[jdoughty]

Hows this for a perfboard diagram (Novice here)? Do I have the right type of caps? I changed any of the 1u+ to electrolytic, but that may have been a mistake based on the the discussion above.

Btw, love the idea of a beefier DOD 250, so this may have to happen...

[Fancy Lime]

Hows this for a perfboard diagram (Novice here)? Do I have the right type of caps? I changed any of the 1u+ to electrolytic, but that may have been a mistake based on the the discussion above.

Btw, love the idea of a beefier DOD 250, so this may have to happen...

That looks great, thanks! However, there is something wrong with D2 and D3. Those need to be connected differently. C7 is the stabilization cap for the opamp. This should be placed physically close to the V+ pin (7) of the opamp for best results, although in s circuit like this it rarely makes much difference, in my experience.

BTW, does anyone know of something already in existence with that topology? The Boss OD-1 is not far off if you forget the second stage and some of the odd component value choices. But is there anything closer?

Cheers,
Andy

[antonis]

Could I use 47pF for C3  or do I have to search for an expencive 50pF 1% 500V mica cap.. 

[Mark Hammer]

I'm a fan of the first-issue Boss OD-1, which this circuit is sort of a reduced version of, using a fixed lowpass filter, rather than the tone control found later in the SD-1.  It doesn't really take more than a single op-amp to deliver pleasing overdrive.  Te low and higher end of the bandwidth always needs to be tailored to the individual user and rig, but once one has settled on that, there's no real need for more features or more extensive circuitry.  I suppose that's one of the perks of DIY: commercial pedals have no idea what the end-user will need, so circuits are fancier and intended to be more tweakable.  But if you know what you need and want, and can adapt the circuit to achieve that, you don't need any more.

As for 50pf, chances are pretty good that if you could source a 50pf cap, it may well measure 47pf anyway.

[Fancy Lime]

Could I use 47pF for C3  or do I have to search for an expencive 50pF 1% 500V mica cap.. 

Always use mica caps for the RF filter! Ceramic caps are the Devil's Tic-tacs!    If you use ceramics you loose so many mojo points, you will be reborn as an acoustic bass player!

Andy

P.s.: since this is supposed to be a novice friendly thread, I feel obligated to clarify that the above was what passes for humor around these parts. Using a ceramic cap is perfectly fine here. The value is not critical, anything between 10pF and 100pF is probably fine. In fact I used a 47pF ceramic on the breadboard but given the very large (20% or so) error on these things, I rounded to 50pF in the schematic.

[cab42]

The Novice Driver on vero:



As I just did the layout a few moments ago, it is not verified, but I have checked the nodes, and it looks fine. But I would appreciate if some one looked it over.

I could have shaved off the bottom row, but that would mean that the 9V would enter the board on the right side, but I preferred to have it just below the ground input.

I guess I could use a 741? I have  quite a few of those.

EDIT: I just realised I could have moved ground to H18, 9V to I18, put a trace cut in I16 and placed D4 vertically from I17 to I15. Then the bottom row is redundant and 9V and ground still close to each other.

EDIT2: Image updated with information on standing diodes orientation. Also forgot the potentiometers in the BOM (rev 1)

[snk]

Thank you for the veroboard layout (and, of course, Fancy Lime for the circuit) !

A couple suggestions to save up some space on the layout and make it 16 columns wide :
- move R5 on column12;
- put the cut on [C/16] under D1 (column 14);
- move R6 column 13,
- now move C5 to row 15.

If we want to have it even more compact (down to 15 columns wide) :
- Move C6 to column 12 and R7 to 13;
- Now C5 : keep one leg on row E, but move it to column 14.
- The formerly upper leg of C5 could go down to row G (C6 isn't there anymore, as it was moved to col. 12). The [C/16] cut could go to row G, as well as the volume 3 wire (C5 would be rotated 180°).

[iainpunk]

nice write up, i do have to be skeptical tho, the C4 capacitor does take out treble before clipping, and this softens the distortion character/sound, but the treble created by the clipping isn't filtered by this right? or am i just confused and seeing problems that aren't there?

talking about MOJO
lvl1) soviet Germanium
lvl2) Mosfet body diodes
lvl3) ca3019
lvl3) snip the magic diodes out of an original Klon Centaur

cheers, Iain

[cab42]

Thank you for the veroboard layout (and, of course, Fancy Lime for the circuit) !

A couple suggestions to save up some space on the layout and make it 16 columns wide :
- move R5 on column12;
- put the cut on [C/16] under D1 (column 14);
- move R6 column 13,
- now move C5 to row 15.

If we want to have it even more compact (down to 15 columns wide) :
- Move C6 to column 12 and R7 to 13;
- Now C5 : keep one leg on row E, but move it to column 14.
- The formerly upper leg of C5 could go down to row G (C6 isn't there anymore, as it was moved to col. 12). The [C/16] cut could go to row G, as well as the volume 3 wire (C5 would be rotated 180°).

Thanks! I have no idea how I could miss the R5

I might make a snk'ed compact layout.

[Fancy Lime]

@Carsten
Awesome! Thanks!

nice write up, i do have to be skeptical tho, the C4 capacitor does take out treble before clipping, and this softens the distortion character/sound, but the treble created by the clipping isn't filtered by this right? or am i just confused and seeing problems that aren't there?

talking about MOJO
lvl1) soviet Germanium
lvl2) Mosfet body diodes
lvl3) ca3019
lvl3) snip the magic diodes out of an original Klon Centaur

cheers, Iain

Actually, I don't think that what C4 does can really be understood from the before/after angle in frequency domain. But it is rather simple in time domain. Just think what the R||D||C arrangement does to the waveform. C4 limits how fast the voltage can ride or fall much like a compensation cap limits the slew rate. Less steep flanks mean less sharp corners mean less high order overtones if you Fourier transform it back to frequency domain. It also means that the stage cannot reproduce fast transients and therefore high treble coming from the instrument. So in a way, C4 cuts treble before and after the clipping but really more simultaneously with it, if that makes any sense. Just think in time domain and it all becomes clear.

Andy

[jdoughty]

Fixed the diodes and moved C7 alot closer, though its still not RIGHT next to the pin.

[jdoughty]

Fixed the C7 problem!

[duck_arse]

The Novice Driver on vero:

I think D3 and D4 both show as "ambiguous" for polarity, perhaps even both backwards, taking grey as the Kathode, as implied by the other two diodes.

[cab42]

I think D3 and D4 both show as "ambiguous" for polarity, perhaps even both backwards, taking grey as the Kathode, as implied by the other two diodes.

You are absolutely right! I was going to write a note in the layout, but it seems that I forgot.

Thanks

EDIT: Done. Image in post above has been updated

[Fancy Lime]

OK, so I finally got the PCB layout to adhere to standards I think I can kind of live with. It's not pretty but it fits in a 1590A and can be adapted to work as a single or double layer with only a little bit of awkwardness.

The question now is, how do I get the layout out of KiCad and into this thread in a way that is most useful for those who etch at home? It cannot be difficult but KiCad is not always the most self-explanatory. Does anyone here know? I personally plan to have the dual sided version produced via Aisler.net. I'll be happy to share the relevant files for that as well, as soon as I figure out what those are and how to export them.

Cheers,
Andy

[Fancy Lime]

So here is my pcb layout for self-etching. It's stuffy and inelegant and I'm sure I violated a lot of rules of good layouting but at least it is reasonably compact, eh? The layout is, as of now, unverified. Please let me know if you have built and verified it or if you find errors or general bad practices. I have no experience with pcb layouting and still need my dumb mistakes pointed out to me.

backside for toner transfer (print in original size!):



And the front for component placement:



The component values are to be taken from the schematic in the first post of this thread. Please let me know if further information is needed. I hope I'll find the time to write a little mod guide, explaining what components need to be changed and how, for certain changes in sound (tighter, brighter, darker, TS808 without tone stack...).

Cheers,
Andy

[Fancy Lime]

So...

I finally got around to doing a better layout. I think... At least this one does not need a jumper and has a better ground routing. I used a dual opamp because a) I figured most people have more duals in their drawers than singles, and b) the pinout makes it easier to layout (for me at least). I also realized that for the layout in the previous post, the transfer graphic need to be flipped before transfering. Best to just forget that previous post and use this new version. Here the graphics can be printed as is and transferred right away.

Copper side (print in original size on laser printer in highest contrast setting for toner transfer):



And the front for component placement:



The whole thing should still fit into a 1590A, if I measured correctly. Last but not least, the schematic for this version:



You may notice that C2, C5, and C8 are polar in the schematic but box-shaped in the layout. That is so that either (non-polar) boxed film caps or electrolytics will fit and to mark the polarity in the schematic. You may also notice a pad marked "X". This can be used for a number of interesting mods that make this whole thing a lot less basic than it is in its, well, basic form.

Suggested mods using X pad:
1) Discreet tone control mod: Change C4 to 470pF. Add an on/off/on switch between the X and g2 pads that switches in either of two caps in the two on positions. Or use a rotary switch for more positions. Suggested values to try: 470pF, 1nF, 2.2nF, 4.7nF.

2) Diode mod: Change D1 to red LED. Add an on/off/on switch between the X and g2 pads that switches in a single Si diode in parallel with D1 or two antiparallel Si diodes. Or leave D1 as a Si diode and use Ge diodes on the switch. Or leave out D1 completely and replace D2 and D3 with a pair of back to back 2.7V Zener diodes and use whatever you want on the switch. Sky's the limit. Clipping diodes are for having fun!

3) Dual ground leg mod: Change R4 to 2.2kΩ and C2 to 100nF. This changes the frequency response to something close to a Tube Screamer. Now run a 1kΩ resistor, a rev-log (C) taper 10kΩ pot wired as a variable resistor (by tying legs 1 and 2 together you will have a variable resistor between 1 and 3), and a 10uF cap in series from the X pad to ground. The order of the components does not matter but make sure that the negative side of the cap points toward ground. Now you have a "body" or "fat" control. Call the whole thing Tiny Tim if you do that You can use a normal log pot as well but then the bass will increase as you turn the pot left (and you need to tie 2 and 3 together instead of 1 and 2). Maybe call it "tight" control then.

4) Tone pot mod: Change C4 to 220pF or so and add a 50kΩ rev-log (again as a variable resistor) in series with a 4.7nF cap between the X and g2 pads. Now you have a classic tone control although I prefer the discreet version of mod 1) in this circuit.

Please let me know if you find errors or if you have suggestions for more mods or improvements.

Have fun,
Andy