proposed dr. boogey layout -- seeking comments

[gaussmarkov]

i guess we're virtually ready to issue this layout.  i think i will take the letters "gm" off of this one and replace them with "diy" to indicate corporate authorship. 

one final question:  i am pretty sure we can take out 2 columns and make this layout 23 x 16.  is it worth it?  the 2 big islands of ground fill on the left and right would each lose a column.  would this significantly change the functioning of the circuit?

cheers, gm

[Victor]

Just a little suggestion for avoiding the beast from screaming..and what about pcb mounting jacks?

I agree. And maybe pcb mounting pots, also? To eliminate all that crossing wires that leads into "oscillation city" 

A little room for the battery and for the switch, it's enough, I believe.

[gaussmarkov]

Just a little suggestion for avoiding the beast from screaming..and what about pcb mounting jacks?

I agree. And maybe pcb mounting pots, also? To eliminate all that crossing wires that leads into "oscillation city" 

A little room for the battery and for the switch, it's enough, I believe.

builds with the pcb mounted jacks and pots are very cool.  i wonder if such designs shouldn't be left to each builder, because the way things fit in the enclosure varies so much from person to person?  because i post the eagle files for these layouts, it is possible for anyone to load them in the eagle layout editor and make such adjustments without having to lay the whole circuit out again.  so i am going to pass on these suggestions, though i am sympathetic. 

also, any thoughts about replacing the trimmers with sockets for fixed resistors?  i suggest this for two reasons:  (1) trimmers are space hogs and (2) trimmers are noisy.  with sockets for the drain resistors, you can still adjust until you get the right values.  this seems like something worth considering for all of these high gain circuits.

[gaussmarkov]

fwiw, here is a smaller (23 x 16) layout (0.1" grid).  there are some incidental improvements in this version.  the potential problem that i see is that the treble signal is getting close to the bypass cap for Q2.  at least that's not the primary signal path.  and there's still some space.

[John Lyons]

After all the headache with the trimmers on the tornado I think I'm converted to fixed resistors for good now! I just tack on a pot, measure the voltage and then put in a close valued resistor. Depends on the circuit but the next high gainer I build will be "fixed resistored".
Either that or buy better Trimers ($$$)

John

[matchless]

gaussmarkov - what value for c20?  47uf?   thank you.

[John Lyons]

Both Caps at the power supply should be 100uf. It's not totally necesary but gauss is trying to build a bullet proof version here.

Gaussmarkov
Seems like the revisions are set for now. I wouln't mind seeing the larger of the two PCBs used but I don't know if it's an issue. The space between Q2 and Q5/Q6 looks a little close to me. It's all so confusing isn't it!
I need to build another so when you decide just let me know when you get a chance to post the Transfer patern.

Thanks for the work again!

John

[gaussmarkov]

Both Caps at the power supply should be 100uf. It's not totally necesary but gauss is trying to build a bullet proof version here.

exactly.  thanks, john! 

Seems like the revisions are set for now. I wouln't mind seeing the larger of the two PCBs used but I don't know if it's an issue. The space between Q2 and Q5/Q6 looks a little close to me. It's all so confusing isn't it!
I need to build another so when you decide just let me know when you get a chance to post the Transfer patern.

i just posted the new version:  dr. boogey.  this is the smaller layout, because i ended up moving the tone pots around as you suggested much earlier and this gave me a chance to move a cap and introduce a guard trace.  john, if you want the other version after all, please let me know.

everyone:  coupling caps are in this version, too.  this required quite a bit of rejuggling to stay inside 23 x 16.  and it led to the new pot order.

man i hope this is an improvement, gm 

[Kornell]

  Great work. Thanks!

[John Lyons]

This newest version looks good. In moving around the parts It looks better to my eyes.

Some things I did to mine and I like the ways it sounds. Quiet and squeal free with all 201s
20pf replaced with 120pf
Shielded wire from in/out jacks to switch and from switch to in and out of board. If you have very short leads and they are routed out of the way and kept from the tone controls the shielding won't be necesary.
I used a 500K log pot for the gain control on one out of control build. Still had all the gain but not of the oscillation.
220pf caps across gate and source for Q1, Q2 and Q4.


Thanks gauss!!

John

[Ardric]

Hi again, and sorry about the delay.

The last March 14th version seems to have introduced a new problem.  C21, the new input coupling cap, prevents R1 from biasing Q1's gate.  It won't work as drawn unless C21 is replaced by a jumper, or by adding another 1M from the C21/R19 junction to gnd for bias.  R1's purpose is then to prevent the opposite side of C21 from floating, DC-wise, to prevent clicks and pops.

Personally, I've had good luck omitting an input cap with ROG's JFET simulations.  None of my guitars or pedals seem to have a problem with it so far.  I guess best would be to put the extra cap and resistor on the layout, but make their use optional.

Regarding the power supply, I was thinking more like the +9 end of R17 would connect to the R16/C20/VB node.  That means Q1 and Q2's power goes through R16, C20, R17 and C14, filtered twice.  I'm still hoping someone will chime in with reasons for or against this approach, since it's used by the actual amps but rarely seen in pedals.  Maybe if the R's and C's are sized appropriately we can get a little tube rectifier sag simulation going on too.

I'd move C15 on the schematic from Q2's drain to the power supply drawing, parallel to C14.  That's really what it's doing.  It would also be nice to connect common gnd nodes on the schematic together with wires before bringing them to a gnd symbol, so what's common with what before returning to the star gnd is seen at a glance.

We've got a few grounds to account for... the circuit gnd, the input jack gnd, the output jack gnd, and the box/shield gnd.  Where's the star when it's all boxed up?  Is it the input jack sleeve?  Perhaps a couple extra gnd pads on the board would be nice to have available if needed.  Heck, we could also leave room for a LED dropping resistor for those so inclined.  It would be wired to PWR before the filter resistors, then to a pad to the wire to the LED's anode, so pulling the other end to gnd lights it.

Lastly, I can't help frowning at the 1MA volume pot on the output.  The whole 5-knob tonestack/volume output deal, actually.  I wonder if at least one of those tonestack knobs could be replaced with preset resistors to free box space.  I'd also prefer a buffered output of some kind, post volume control.  But there's a lot of people happy with the original, so it's only fair I give it a shot.

I've got a bag of J201's sorted out and ready to go.  I'm looking forward to getting a new build underway.  Thanks to everyone for contributing.

[John Lyons]

Some good points Ardric

The 1M pot at the end is a bit foolish. In the amp version it's fine because at that point the impedance is still high and being fedd to two more stages, (phase inverter and the output tubes) I brought this up in another thread but yeah.... The Volume pot should be 100Klog. This lowers the ompedance signifinantly and pwdal after the DB won't load down the tone. I tried a delay after the DB and it was super muddy. 100k has plenty of volume.

The input cap does need a pulldown resistor at both sides of the cap to work and keep the first FET biased correctly.

Putting the resistors bettween the power supply fileter caps may work but the 9v supply is so small anyway.
Smoothing the last bit or posible ripple out of the 9V is pretty common with a small resistor and a large cap. 1000uf  sometimes.
Putting the resistor between the caps may work as well but I think Gaussmarkov was trying to balance out the power distribution so the later stages had a clean source of ground and power.

C15 is a curious one because in the amp version the cap is localized because of the r/c filters. But in the pedal veersion here the cap may as well just be across +9v and ground anywhere in the circuit as the power supply is the same all through the circuit. Nothing gained by putting it close to Q2.

The grounding is a goos point. We've come up with a fairly elaborate star grounding system but we've left out a big part of the grounding circuit. input/output and chassis ground. Since most input/output grounds are connected to ground at the chassis I guess thats a little bit of a mute point but the star system is dependant on a single point to the chassis and localizes grounding point with one connection to ground at the star. Maybe we should just make a couple more pads at the main star point on the board and wire jack grounds there?

What kind of buffer are you thinking Ardric? The Source follower Q4 lowers the impedance....

John

[gaussmarkov]

ardic, thanks for those comments.  

i have made a critical change and accomodated two others.  first, the needed 1M resistor to ground is in place for proper grounding biasing.  second, i have reconfigured the power supply section so that VA gets filtered twice.  third, i have placed extra pads on the ground rail.

i will make the schem clearer later.  those are all helpful suggestions and, honestly, it had not yet dawned on me to make the star-grounding approach clear on the schematic.

as ardic said, it seems reasonable to leave the rest as is, given the general happiness with the dr. boogey.  that said, i think additional discussion would be great.  i'm just getting this off again so that those who want to go ahead this weekend and try the version we have can do so.  everything is posted in the same place as above:  http://gaussmarkov.net/index.php?page=layouts#drboo

over and out, gm

[Pushtone]

I brought this up in another thread but yeah.... The Volume pot should be 100Klog. This lowers the ompedance signifinantly and pwdal after the DB won't load down the tone. I tried a delay after the DB and it was super muddy. 100k has plenty of volume.

Read that thread. Makes perfect sence.
I was planning and referencing your post in a DB build report.
Thanks, a very good tip for the DB.

Didn't you also use 500k instead of 1meg on the gain?
Thats another tip I went with based on the exsisting DB threads.
PLENTY of gain!

I even used linear pots on the Mastr and gain and it works good.
The taper dosent seem to matter too much with this one. Smooth response across the dial.

[John Lyons]

I used a 500K gain pot for my first one which was a little out of control at first. But then I built a second one for a friend and used all the stock values at first and was prepaired to be blown away with squeal but it was well behaved and not a peep of noise or oscillation.
I guess my wrestling with the first one forever got my layout issues worked out and got lucky... or just had a good combonation of j201s....

I've been using 100k linear pots for the volume and tacking on a 25K resistor from the wiper to ground to simulate an audio taper. Really, for the output volume the taper isn't a big deal, you set it so it sounds right.... taper shmaper!

For the gain control that's interesting that you got a nice sweep with the liner because even with the audio pots I get a bit of a gain jump right off the bat. Barely cracked open it gets a good OD sound but then a hair later it's "for those about to rock!" Maybe a linear taper 1M pot with a 330K or even 470K across wiper and ground it would get some mid drive levels and then ramp up pretty good to the high octane stuff.

John

[Pushtone]

you set it so it sounds right.... taper shmaper!

John

LOL,  we need more of that attitude in general around here.

I can get some low gain sounds out of it. At least low gain for a DB.
Which is super crunchy but without the violin sustain of the higher setting.
I guess the 500k GAIN pot is helping that along some.

[John Lyons]

It IS funny how the low gain stuff it pretty crunchy just not as sustained. Pretty much the same tone just a bit thinner but with the same crunchy edge. I do like how the sound isn't squishy or rubbery sounding at all though. Reducing or omitting the bypass cap on the gain control will give you more low end meat to the low gain settings.

One thing that I wonder about is the third stage and the 3.9K source bypass resistor. In the Dual Rec amp the value is 39K! which is called a "cold clipping" stage as it biases the stage super "cold" and with asymetrical clipping in spades. I wonder why the DB uses 3.9K as oppossed to the 39K of the original? One of those things I've been meaning to try out. Can't just be a typo they (electrictabs)  went with....

John

[MartyMart]

Great work Gauss - looking V good
I can also recommend the 100k vol pot, I use whar ever I have around, log/lin
250k with a 220k across it - you name it !
John, I spotted the 3k9/39k thing a while ago when making one and I did try it.
This sets up the bias VERY differently and it lost a lot of gain that way, it did "work"
and may be a closer approximation to the real deal !!
It would certainly syop any gain/howl problems

MM.

[mojotron]

...
This sets up the bias VERY differently and it lost a lot of gain that way, it did "work"
and may be a closer approximation to the real deal !!
It would certainly syop any gain/howl problems
...

I tried that too and got the same result, the tone was kind of thin and uninspiring. That change did not affect my squeal issues on my first build (that was a really troubled build anyway - too many experiments killed the thing) - I went back to the 3.9k value.

...
just a bit thinner but with the same crunchy edge.
...

.
I tried bumping up the values of the "Miller Caps" (270pF, 330pF) and the tone got less thin without reducing the high-end too much, but the 220pF values worked better with my gear.

[John Lyons]

Hmmm ok. so it thinned the sound out to have the 39k in there? Did you notice a change in the clipping or asymetry?
Can you comment on it more Marty?

Running through a '73 Twin Reveb ("blackfaced') The DB to me is a bit bright and I run the treble about 12 oclock, Mid at about 9 or 10, bass full up or a bit less, and the presence about 12 o clock. The twin is a bright amp though. (bright switch off of course).
Maybe I'll try the higher miller caps or even a snubber cap across one of the later stages trimmers. .001 or so...

John