Problems biasing Joe Davisson's Obsidian-M

[Dave_B]

Hi all,
I breadboarded this BS250-based circuit last night and ran into the same issue SpheriX did back in January. 

I measured between -8.16 and -8.46 at the drains of the three BS250's.  The sources measured between -.2 and -.27.  The battery voltage is -9.24.  I checked the components as they went in, and double-checked them tonight.  All the resistors are well within tolerance and the caps have the negative side connected to the sources.

The thing is, it sounds reasonably close to Aron's sample.  All the controls worked and it was very playable.   Does it make any sense that it should be functioning to the point that it's really enjoyable?

For the record, here's the circuit I used: http://www.diystompboxes.com/analogalchemy/pedals/obsidian.html
And here's the original thread: http://www.diystompboxes.com/smfforum/index.php?topic=28825.0

[RDV]

Replace those 3.9k with 10k trimpots and find the voltge that sounds best to you, replace with the apprpos fixed resistors, though if you like it as is, you could just leave it.

RDV

[Dave_B]

Thanks RDV, I'll give that a shot and report back. 

[Dave_B]

I guess the numbers are so far off I wouldn't have thought to try it.  I ended up with 15k resistors in place of the 3.9k's.  Now all three stages are in the 4.5-5.2v ballpark.  It sounds fine, and actually got a little louder as I rebiased the first stage.  I played it through my Bassman for over an hour and it's a lot of fun as long as I don't max out the gain and volume.

Questions:
Does it make sense that the resistors should be that high? 
Do mosfets vary that much? 
At the 8v bias point, would the signal be basically half-rectified? 

[Dave_B]

If I may, just one more question on this thing before I start soldering.  Joe's page says to adjust both the 3.3m and the 3.9k to set the bias.  I didn't read it correctly at the time, so I only changed the 3.9k.  I ended up with a relatively whopping 15k in it's place.  I don't have any large value resistors at the moment, so I'm hoping someone can shed some light on how the parts interact, and if I should be bumping the 3.3m to a 4.7m or something, then re-trim the drain resistor to a more reasonable value.

The circuit does a low-volume, low-frequency, 'motorboating' thing when I crank the gain past 1 o'clock.  I'm not sure my bias job is contributing to that.

Obviously I don't know how to set gain or biasing on a Mosfet (or , as I'm sure someone will point out).

[WGTP]

When I breadboarded it, I had to change both resistors to get my BS250's to work.  I don't recall now what values I used.  They must vary quit a bit.  Didn't have any motorboating issues, and I had it running wide open.   

[Joe]

try a 100uF cap across the battery

[Dave_B]

try a 100uF cap across the battery

Thanks, Joe.  I tried that with no luck.  It's all soldered up as of 12:30am CST this morning, including the 100uf cap.  The truth is, there's so much gain by the time I get to that point, I don't need to crank it that high anyway.  If it starts to really bug me I may end up adding a resistor to the top leg of the gain pot to 'trim' it out.

[KORGULL]

bellyflop wrote:

If I may, just one more question on this thing before I start soldering.  Joe's page says to adjust both the 3.3m and the 3.9k to set the bias.  I didn't read it correctly at the time, so I only changed the 3.9k.  I ended up with a relatively whopping 15k in it's place.  I don't have any large value resistors at the moment, so I'm hoping someone can shed some light on how the parts interact, and if I should be bumping the 3.3m to a 4.7m or something, then re-trim the drain resistor to a more reasonable value.

I have exactly the same question. Does it effect the circuit operation/sound if you raise the value of only the 3.9K alot, as opposed to adjusting both the 3.9K and the 3.3M to achieve the proper voltage? Does it really matter which resistor(s) gets adjusted in this case? I have mine biased right, but like bellyflop, I only raised the 3.9K to 15K. There is quite a bit of noise and not really much difference in sound from when it was misbiased. I remember somebody saying that their Obsidian pedal was very quiet and I'm wondering if the way I biased it could have anything to do with all the noise.

[Dave_B]

Is yours boxed up Korgull?  Mine was slightly prone to picking up noise but that was fixed when I boxed it.  I still have a problem where  I can't crank the gain without 'motorboating' (the only way I can describe it).  I tried Joe's suggestion of a 100µf across the supply.  No go.  The good news is, there's so much gain on this pedal I don't really need to turn it up that high.

It's funny that I didn't realize how inappropriate this pedal was for my pop band until I took it to practice last week.   

[KORGULL]

It's still on the breadboard, so I hope the noise is somewhat minimized once I box it up. I think it will be OK if I don't crank the gain all the way up - like you said. Beyond the halfway point it doesn't seem to add much gain (only noise) anyway. As soon as I finish the layout I guess I'm just gonna box 'er up and forget about it (for now at least).

[WGTP]

I seem to recall an older design by Joe that had .001uf caps going to +9v at the output of each stage in parallel with the 3.9K resistor, that could help the noise.  Start with the last stage.

Increasing the resistors between stages will lower overall gain.   

[Dave_B]

Increasing the resistors between stages will lower overall gain.   

That's good to know.  Thanks, WGTP.  I love playing through the O-M at home, but at rehearsal it's become the dirty-look pedal. 

[WGTP]

I think the Obsidian was designed to be a dirty look pedal.  Is that a bad thing?   :icon_biggrin 

That would be -9v if using the BS250 P channel version. 

[Dave_B]

That would be -9v if using the BS250 P channel version. 

Yep.  No matter how much I knew that, I still wired up the jacks wrong, frying one of the BS250's.  Soldering at 1am on a worknight.  Bad plan.