New at runoffgroove.com: AZABACHE

Started by stm, August 07, 2012, 04:19:31 PM

Previous topic - Next topic

WGTP

Once again, great stuff.  Thanks for the explaination of the diode/jfet clipping combination.   :icon_cool: :icon_cool:
Stomping Out Sparks & Flames

Chris S

I know this is kind of heresy but, I've be trying to not build pedals and play guitar instead (2 small kids limited time, not actually that great at playing). Run Off Groove I have to build this. Thanks.

YouAre

Quote from: Mark Hammer on August 10, 2012, 10:21:48 AM

And since the topic has been raised, do any of you thousands out there have anything to say about running ROG circuits offer supply voltages higher than 9V?  I know that some of the Shaka Braddah versions from Aron and Doug were run as high as 33V.  What happens to an English Channel or Matchbox at 18V or 27V?


:Raises hand:

I've been working to modify the Professor Tweed to work as a clean to slightly dirty preamp with the Tiny Giant amplifier.

I'll post a separate thread dedicated to it with more detail (so as not to clutter up this thread), but here are some of my findings.

When running at higher voltages, I didn't find myself sticking to the V+/2 for the collector voltage. I ran it a little higher, for some more cajones to the sound. I had to find the right balance between volume, clarity, note decay, and noise. That said, starting at half the power supply voltage is a great starting point. I usually raised it from there.

The problem arose when I was using the 19v of my power supply for the TG. A bias point of 12v was too high for the ~16v rail to rail opamp preamp of the TG. If the voltage swing brought me above the opamp clipping point, I got nasty rail clipping. The problem was solved when I used the V+ of the opamp as the supply for the Professor Tweed. I did like the sound of the higher voltage better, but obviously stability is more important.

Hope that helps/was cohesive. A thread will follow once I can whip up a respectable schematic.

Mark Hammer

That's a) very useful, and b) very brave.

All too often, around here, we tend to lean towards the tried and true, focussing on successful replication, and treat difficulties as a sign of failure.  Exploratory attempts that push the envelope with the otherwise familiar are VERY important.  It's how the other 26,000 of us learn.

So...thanks!  :icon_biggrin:

YouAre

Quote from: Mark Hammer on August 12, 2012, 06:07:29 PM
That's a) very useful, and b) very brave.

All too often, around here, we tend to lean towards the tried and true, focussing on successful replication, and treat difficulties as a sign of failure.  Exploratory attempts that push the envelope with the otherwise familiar are VERY important.  It's how the other 26,000 of us learn.

So...thanks!  :icon_biggrin:

I'll agree with B, not necessarily A.  :P

It's an extension of how I am with the commercial pedals I used to buy. Not a single crybaby, boss, ibanez, proco pedal stayed stock on my board in high school. Now I look at DIY designs as a building block process. I'll break each design down into it's individual sections, and see how I can mix and match to create my own designs.

One of my methods of getting rid of rattiness from FET clipping was my placement of the Master Volume. A master volume at the end of the pedal is great for taming down the volume of all output, but it means that everything is running on ten when it hits that last FET. Rattiness. So I figured, why not copy tube amp design more, and move the volume control in front of the "power tube" FET, and after the phase inverter. My results were much cleaner sounding, more graceful overdrive.

The drawback was that I lost out on the effectiveness of the negative feedback loop, which is to be expected. When you convert most amps to a post phase inverter master volume, it will work to decrease the effects of a NFB loop, or at least the controls in it. This is what I've gathered from reading Randall Aiken's articles. I have not yet tested this in a tube amp, but the principle did apply to this pedal.

Please note: This was NOT a stock professor tweed. There is an extra gain stage, and a scaled TMB tone stack. It's rather different than the Professor Tweed from which it was derived, but my goal was to create a clean to gritty preamp that took drive pedals well.

So I guess the point of this was....placement of the MV has a huge effect, because it affects the signal hitting the last stage. That's likely where the clipping is worst, but that's just a guess.

Hope that helps!

Mark Hammer

It helps....especially by highlighting how much the distinctive sound of many amps has often been a question of where in the signal a volume control, and/or tonestack, was situated.  One ought to keep in mind that tonetsacks are often attenuators, just like a volume pot; they just happen to be frequency-selective attenuators.

So, one will see amps where the tonestack and volume pot come right after the first 12AX7 stage, and others where the volume may come after that first stage and the tonestack further downstream, or where that order is reversed, or even where the volume, middle and bass occur earlier than the treble.  And so on.  Tinkering with the location of these various controls, relative to gain stages that precede and follow, can lead to different sorts of voices, some more "classic", and some more distinctive.

lopsided

#26
I have just finished doing this on vero so here are a couple of first impressions. Have yet to try it with a big amp, and it might still have some mistakes, so this is all preliminary.
I like the nature of the overdrive tones it produces, It gets quite dirty and I like it for "garage" like rhythms.

Having four jfet to choose and four trimmers to adjust, I find it quite hard to tune it up. With the given drain voltages mine had quit unpleasant fizzy undertones, and this tremolo like sustain.
I had to go as high as 7V in the third stage to get the sound I liked. There is a lot of to tweak and fiddle with that it's almost overwhelming - to get the right tone.

It seems very bright, I actually found only the low third of the tone pot to be usable.
I like how the switches affect the sound.

I will write more after I play with it some more.

B Tremblay

Did you make any substitutions for components? Will you please provide your FET pin voltages? Can you please share the layout you used?

The fizzy undertones and tremolo-like sustain you describe are indicative of one or more problems with your build.
B Tremblay
runoffgroove.com

lopsided

hey, thanks for being ready to help me out!
the only substitution I have made was the 2,2K resistor from gain pot to ground, I did not have one and sticked in a 1,5K. This should lower the minimum gain range by a hair, but I think this is not a problem.

My voltages when drains set as by the schematic go:
v+ 9,27
q1
6,07
0,70
0,00

q2
5,05
0,30
0

q3
5,04
0,33
0

q4
5,01
0,50
0,00

q5
8,98
5,68
5,40

I did it based on a layout I made, went through it several times, but still might have missed something.

for some reasons the DIYLC did not render some of the labels complete, but the respective resistors are 33K, 100K and 1m

I will give it some more looks tomorrow, but if you spot something suspicious I will appreciate if you let me know.

Thanks!
Jakub

B Tremblay

At first glance, the voltages of Q5 base and emitter are odd.  There should be a voltage drop closer to 600mV between the two.
B Tremblay
runoffgroove.com

Ronan

#30
I built this circuit recently, here's some voltages for comparison:

Battery voltage 8.98V
Q1 source 6.02, drain 0.568
Q2 source 5.08, drain 0.486
Q3 source 5.00, drain 0.363
Q4 source 5.01, drain 0.430
Q5 collector 8.72, base 6.00, emitter 5.41

Edit: I did look at your layout and couldn't see anything wrong around Q5, the layout looks good.

lopsided

thank you both for the responses guys!
Quote from: B Tremblay on August 28, 2012, 08:03:39 PM
At first glance, the voltages of Q5 base and emitter are odd.  There should be a voltage drop closer to 600mV between the two.

I have noticed that, but I assumed that it is just some interaction with the multimeter. I have experienced a smaller BE voltage before on working builds. Not saying that it could not indicate a problem though. With the transistor out I get perfect 3K3 from emitter to ground. Base is little more sophisticated to check this way, but so far I found continuity to all the right components only.

I have tried switching the Q4 and after readjusting the trimmer I think I got rid of most of the fizzyness I have described. But on the upper half of the gain pot, when I struck a chord, I still get this wha-wha-wha tremolo-like sustain after half a second or so. Not sure if it is supposed to be so. Listening to the soudclips, I think I am maybe getting a tad more distortion of it,, but this comparison might be misleading.
Quote from: Ronan on August 29, 2012, 07:07:31 AM
I built this circuit recently, here's some voltages for comparison:
Thanks, I think that except the Q5, the little differences at sources can be ascribed to the high variance of jfets.
How would you describe the tone control Ronan? That's another part of my build I want to recheck again. I find it very bright even with the bright switch off, for me the ideal center of the pot would be at about 8 or 9 o clock of how I am getting it now. Could be still my mistake, and also admit that my general preference lies in rather darker tones.

B Tremblay

Quote from: lopsided on August 29, 2012, 09:54:43 AM
I have tried switching the Q4 and after readjusting the trimmer I think I got rid of most of the fizzyness I have described. But on the upper half of the gain pot, when I struck a chord, I still get this wha-wha-wha tremolo-like sustain after half a second or so. Not sure if it is supposed to be so.

The sound you describe is absolutely not how it should be.  I suggest rechecking the components and solder joints.  The low drop of Q5 B-E is still odd.  Have you tried an audio probe to determine where the sustain begins having that odd effect?
B Tremblay
runoffgroove.com

Ronan

Quote from: lopsided on August 29, 2012, 09:54:43 AM
How would you describe the tone control Ronan? That's another part of my build I want to recheck again.

I run it with the tone control just a little bit left of centre. The circuit is a little bit on the bright side but that is the intended design.

I also had a problem in the beginning with Q5 base to emitter voltage being too low, 0.28V, and the circuit did not sound right, overdrive/distortion was a bit harsh and too much treble. I replaced Q5 and that fixed it. I have also cranked it to max gain with fairly high output humbuckers, and not noticed any unusual behaviour. I would replace Q5 first up, and see what happens, if you have no spare 2N5088, a 2N5089 or BC546/547/548/549/550 would be OK just to test, check the pinout first.

stm

Yes, the pedal was designed to be quite bright as goal was to give a "fendery" vibe to darker guitar/amps.  Some additional information that may be useful:

With both the BRIGHT and SCOOP switches off, the neutral tone position lies between 10 and 11 o'clock.  In general I like it brighter, so 12 o'clock makes perfect sense to my rig.  I also like dialing full treble with the bridge pickup for a chiming "voxy" tone.  If you prefer darker tones, it is a simple matter of bypassing the 47k resistor close to the TONE control with a wire.  This will shift the neutral tone position to 12 o'clock and will allow dialing darker tones.  Bright tones won't be affected by this mod and will still be available.

As for the final transistor (Q5), any medium-high to high gain NPN transistor should do (but not darlingtons!).  Make sure hfe equals or exceeds 200 at least.  The odd base-to-emitter voltage indicates some problem, maybe with the pinout, maybe with the transistor itself, so I'd try a different transistor just to be sure.

lopsided

Thank you all guys for the responses and tips.
I will go through all of it over the weekend, will let you know how it went.
J.

stm

#36
Some additional info that might be useful:

I still have my prototype on a breadboard.  After maybe two months playing around with it, on day I noticed there was an ugly sort of white noise below the notes, especially during note decay.  This lasted several days until I got an oscilloscope and found Q5 was oscillating at a high frequency (inaudible) and when audio was present this high frequency produced nasty artifacts.  To make a long story short, I realized that the 220u bypass capacitor (V+ to GND) was placed at the beginning of the breadboard and Q5 was at the other end.  I picked up the first cap I had at hand (a 47uF cap) and placed it bewteen Q5:C and the GND point close to the 3k3 emitter resistor and voila!, oscillation gone.

This may or may not be related to your problem, however it is interesting to notice that your ground trace goes a long way around the board until it gets to Q5.  In addition, depending on the type of oscillation and digital voltmeter, you may get odd voltage readings that could explain why your base and emitter voltages are not following the "one diode drop" relationship.

A lesson I've learned many times is that even in relatively simpler circuits sometimes trace layout and/or lack of bypass capacitors do matter.  The problem is that most times they don't, so it is easy to lower your guard and disregard this as a potential cause of trouble.

Hope you sort it out soon, I'm really enjoying this circuit.

mordechai

Quote from: stm on August 07, 2012, 04:56:23 PM
Azrael, the clipping voltage of the diodes happen to be slightly less than the voltage required to make a J201 clip in the common source configuration without source bypass capacitor.  This means the three J201 stages amplify but are not really overdriven.  It's the diodes who actually produce the overdrive.  Nevertheless each J201 stage does introduce a nice amount of 2nd harmonic, so the overall sound is rich in both odd and even harmonics.  As the gain is turned up, one, two or the three J201 stages reach the point where the diodes are overdriven, so the harmonic content increases gradually as the signal gets hotter.  Having said that, the clipping characteristic of the 1N4148 or 1N914 diodes is softer than a JFET that hits the supply rails, so the usual nasty fizziness is avoided.  A key aspect of the sound in this circuit lies in distributing gain and clipping among several low-gain stages, each with some high-frequency filtering, instead of having a single high-gain clipping stage. This arrangement proved to cleanup very well when the guitar volume knob is turned down.

If I understand you correctly, the clipping diodes both keep the JFET from clipping, but since they produce their own clipping, the clipped signal passes through the JFET, which amplifies it and introduces nice 2nd order harmonics, and this repeated process is what leads to the stronger, but still soft, clipping in the overall signal by the end of the signal path.  Do I have that right?

So, here's an academic question: if you have a JFET that mildly clips a signal, and THEN that signal passes through clipping diodes, the character of the clipping would be significantly increased but the signal volume would be cut a bit, right?  Also, would the diodes adversely affect the 2nd order harmonic introduced by the JFET?

Ben N

Back from the dead... Sebastian, Brian, any reason not to split the tone stack into separate T & B controls, a la AMZ?
  • SUPPORTER

B Tremblay

Go for it! Let us know how it works out.
B Tremblay
runoffgroove.com