OKKO Dominator

HEAD · Post by **HEAD** » 28 Jan 2011, 16:14

bajaman wrote:BF245A

Seems to me rather likely as Mr. Okko is from Germany (and even lives in the same town as me) and BF245s are pretty common here in Germany. I already wanted to suggest this type.

Post by **Saruman** » 28 Jan 2011, 17:13

Corrected version of schematic and layout.

bajaman · Post by **bajaman** » 28 Jan 2011, 21:56

The BF245A and the 2SK30ATM have very similar responses in my simulations

The BF245A has slightly higher voltage gain, but you would most likely to be hard pressed to hear any difference in the sound when using the 2SK30ATM.
Forget about J201, MPF102, 2N5457 etc - they will NOT work properly in this circuit

Rnfr - the 10k resistor (R4) is simply to load the input fet stage down gain wise - perhaps to avoid clipping the jfet buffer following it

Without R4 the input stage has twice the peak to peak output voltage swing and the same assymetrical clipping characteristics

The drain biases at 9.3 volts dc with no R4 fitted.
cheers
bajaman

RnFR · Post by **RnFR** » 29 Jan 2011, 00:00

ah ok, much simpler than I was making it out to be. i'm still showing my ignorance!

thanks for the clarification, mr. baja.

if the proper FET isn't on hand, would it be possible to use a more common lower gain device, and then decrease Rs? or would the whole asymmetrical aspect go out of balance?

bajaman · Post by **bajaman** » 29 Jan 2011, 03:42

if the proper FET isn't on hand, would it be possible to use a more common lower gain device

?
tell me what fet you would like to use instead and I will try adjusting the source and drain components to suit it.

cheers
bajaman

bajaman · Post by **bajaman** » 30 Jan 2011, 01:30

Okay - I did some further simulations with other fets - no way i could get the same response overload curves with the J201, but i managed to get pretty close to the BF254A with the 2SK117 BUT I had to change the 10k drain and source resistors to 820 ohms and 68 ohms respectively (R4 also 820 ohms), and the bypass cap to 150uf.
The BF245A draws 146uA of current in the Okko schematic, however the 2SK117 draws 1.91mA (1910uA

) - 13 times the current draw

The simulations were both using a 10.8v dc power supply

I shall try some simulations using a different power supply voltage in the next few days, but so far the simulations show that the input fet type is critical to the Okko dominator's response characteristics.
cheers
bajaman

bajaman · Post by **bajaman** » 30 Jan 2011, 02:02

Hello Saruman
Could you verify the value of R18 for me please - unsolder one end and measure with a multimeter if necessary, although because it is capacitor coupled you should be able to get an accurate reading without unsoldering perhaps.

My simulations show that with 10k there is a notch in the frequency response curve at approximately 400Hz, but with 47k, this notch occurs a lot higher at 900Hz

This filter (r16,17,18,19,20 and the level control,and C11,12,13 and 14) is the main tone shaper after the gain stage and rolls off a lot of the high end harmonics (fizz)

cheers
bajaman

RnFR · Post by **RnFR** » 30 Jan 2011, 03:33

2SK117 works for me.

would 2n5457 be easier? i just want to hear what that thing is doing.

yeah, the RC network you mentioned above is where i would guess that most of the "sound" of this box is coming from(besides that f'ed up JFET

). something along those lines will work in many situations.

The Rotagilla · Post by **The Rotagilla** » 30 Jan 2011, 05:26

I just listened to a sample of this pedal. Jeebus this thing can be brutal!

bajaman · Post by **bajaman** » 30 Jan 2011, 06:05

yes - no diode clipping to compress the shit out of your sound

hbo · Post by **hbo** » 30 Jan 2011, 11:37

Here's a vero layout for you all. It's unverified as of yet, but I'm fairly certain it's correct according to the revised schematic posted by saruman. I even tried to retain the component names (but there were a few duplicate ones).

Comments and suggestions welcome.

guiddruid · Post by **guiddruid** » 30 Jan 2011, 11:46

Given the position of the gain control, is the distortion primarily OP1 clipping?

roseblood11 · Post by **roseblood11** » 30 Jan 2011, 13:10

@hbo: that looks great, many thx for sharing! I had started to make a layout with a tl074 in hopes of making the board a bit smaller, but yours looks really good.

Lawnchair · Post by **Lawnchair** » 30 Jan 2011, 18:05

Bravo! This is a fun study.

bajaman · Post by **bajaman** » 30 Jan 2011, 19:19

Given the position of the gain control, is the distortion primarily OP1 clipping?

yes - the OPA2604 ( which is arranged as a very high gain stage similar to the Marshall Bluesbreaker or Analogman KOT ) clips very hard - even at only 30% of the gain pot rotation, unless you turn your guitar's volume control right down

I seem to remember this two stage high gain configuration being first used by Douglas Self as an active gain control for a stereo hifi control preamplifier article in Wireless World magazine in the mid 70s.
The input fet section has only a very small effect on the sound of this pedal I would feel - although it does introduce some nice second harmonic distortion, the clipped op amp following introduces considerably more odd order harmonic distortion and the following filter rolls off a very large proportion of these harmonics.
cheers
bajaman

bajaman · Post by **bajaman** » 30 Jan 2011, 19:21

I wonder how this pedal would sound using a TLC2262 dual rail to rail low current draw mosfet opamp instead of the OPA2604

bajaman · Post by **bajaman** » 30 Jan 2011, 20:37

Saruman
How sure are you that R10 is 22k and R13 is 1k

Changing R10 to 220k and R13 to 10k totally changes the mix of second harmonic fet and odd order harmonic op amp clipping characteristics

cheers
bajaman

roseblood11 · Post by **roseblood11** » 31 Jan 2011, 00:28

Can´t the Mid-Switch be done easier?

How does it switch the resistors in the original?
R26 or R27 or R28?

Then I would leave the 1M always in the signal chain and add a 320k in parallel (= ~240k) or a 10k (= ~10k).
(316k and 10,1k would be accurate, but who cares)
This could be done with a simple 1pdt on-off-on switch...

DimebuGG · Post by **DimebuGG** » 31 Jan 2011, 01:08

roseblood11 wrote:Can´t the Mid-Switch be done easier?

How does it switch the resistors in the original?
R26 or R27 or R28?

Then I would leave the 1M always in the signal chain and add a 320k in parallel (= ~240k) or a 10k (= ~10k).
(316k and 10,1k would be accurate, but who cares)
This could be done with a simple 1pdt on-off-on switch...

A better way I think would be using a 1M pot so it can have a variable frequency selection.

roseblood11 · Post by **roseblood11** » 04 Feb 2011, 14:35

A better way I think would be using a 1M pot so it can have a variable frequency selection.

agree...
And DimebuGG, didn´t we hear you mumble: "Mmmhh, this needs a pcb layout..." ??

---

Is the veroboard layout verified?
@Harold: Could you please post a picture of the mirrored layout with all the components removed, except for the cuts and the jumpers? As it´s a quite big pcb, this might help people to place the cuts correctly.
Like this: http://forum.musikding.de/cpg/displayim ... fullsize=1

OKKO Dominator [traced]