Vintage Phaser Issue
1 - 13 of 13 Posts
So, every phaser is predicated on the assumption that each and every phase-shift stage is unity gain. That way, when one adds some feedback from the end back to the beginning, there won't be any unexpected teensy little bit of amplification that gets amplified and reamplified, turning into oscillation (audible howling). But, 5% tolerance resistors being what they are, one can't absolutely guarantee that what gets fed back to the input is NOT just that teensiest bit greater amplitude than what you fed in.
Since much of the guitar signal "lives in the basement", as I like to say, the greatest risk of unintended oscillation will come from lower frequencies when feedback is applied and peaks around the notches are exaggerated. So one of the strategies that phaser designs use is to trim the bass in the feedback path. That results in very pronounced resonances at the upper part of the sweep, and as the sweep approaches the bottom, it starts to sound like the Feedback has been dialed way back. It's not a drop in volume, per se, just risk management.
Make sense?
Since much of the guitar signal "lives in the basement", as I like to say, the greatest risk of unintended oscillation will come from lower frequencies when feedback is applied and peaks around the notches are exaggerated. So one of the strategies that phaser designs use is to trim the bass in the feedback path. That results in very pronounced resonances at the upper part of the sweep, and as the sweep approaches the bottom, it starts to sound like the Feedback has been dialed way back. It's not a drop in volume, per se, just risk management.
Make sense?
Joined
·
12,980 Posts
Hmmm. That said, could that high pass filter be tuned a bit with either a pot or a change of R and/or C? Probably not worth it to spend tons of time researching - perhaps just identifying and then shotgunning a couple different resistor valuess in the circuit might get a bit more low resonance. Time consuming but maybe worth it for the tone connoisseur.
Certainly.
Keep I mind that the cheapest and simplest risk management strategy involves what amounts to a single-pole 6db/oct filter, which usually needs to start higher up to do anything, because it is such a shallow slope (i.e., has a very gradual effect). One needs to stick a DC-blocking cap in the feedback path anyway, just to keep things civil. All the "risk management" strategy involves is calculating what the cap value ought to be.
One could conceivably use a 2 or even 3-pole passive filter, yielding a much steeper rolloff, and thereby permitting the user/designer to set the rolloff point much lower. That would involve more parts, more board space and more calculations. But it could be done. I've just never seen anyone do it...and I must have a few hundred phaser schematics in my collection.
I will say that when one doesn't try to keep the feedback bass content in check, it sounds awful.
Keep I mind that the cheapest and simplest risk management strategy involves what amounts to a single-pole 6db/oct filter, which usually needs to start higher up to do anything, because it is such a shallow slope (i.e., has a very gradual effect). One needs to stick a DC-blocking cap in the feedback path anyway, just to keep things civil. All the "risk management" strategy involves is calculating what the cap value ought to be.
One could conceivably use a 2 or even 3-pole passive filter, yielding a much steeper rolloff, and thereby permitting the user/designer to set the rolloff point much lower. That would involve more parts, more board space and more calculations. But it could be done. I've just never seen anyone do it...and I must have a few hundred phaser schematics in my collection.
I will say that when one doesn't try to keep the feedback bass content in check, it sounds awful.
Joined
·
12,980 Posts
Yea, I'm thinking more like just tweaking a bit. Either to compensate for the 5% or 10% components or just go 10% off spec to move the knee point a bit.
And while a 2 or 3 pole filter sounds like a good idea (although larger and more complicated), wouldn't that introduce more phase problems, something you maybe don't want in a phaser? Just spit-balling here, as I don't know.
And while a 2 or 3 pole filter sounds like a good idea (although larger and more complicated), wouldn't that introduce more phase problems, something you maybe don't want in a phaser? Just spit-balling here, as I don't know.
One sensible approach is to simply use 1% resistors. When I get inquiries from folks who want to make 8, 10, and 12-stage phasers, I remind them that the more stages you have, the greater the risk that one or more stages will introduce a tiny bit of gain, so risk is partly managed by keeping the number stages to a minimum (I think 6 pretty well nails it for guitar). But many modulation effects, whether flangers or phasers will include a small trimpot inside to set the maximum feedback. The panel-mounted feedback/regen/emphasis control won't introduce anything beyond that maximum, but you'll know that you can dime that control without running into oscillation.
I wouldn't doubt that some pedals have such a trimmer set a little conservatively, and a little more emphasis can be coaxed out of them. Typically the way they are set is that the adjuster maxes the feedback pot, dials the trimmer until they get howling, and then ease the trimmer back until the howling stops. It's entirely possible to overshoot sometimes.
But bear in mind that's overall feedback level not any sort of frequency-specific adjustment.
I'd be curious to know if the perceived loss in level of adcandour's phaser also occurs when the feedback control is set to zero. In other words, if my hunch is correct, then the apparent change in level occurs because when the sweep goes down low, there isn't quite as much emphasis of peaks at the low end. If there is no emphasis whatsoever, then there should be no apparent different in level at any point in the sweep because nothing is getting favoured.
Of course, if it's a Phase 90, then all bets are off because the feedback in those is fixed.
Joined
·
8,894 Posts
Sorry I haven't been able to check back throughout the day. I'm not ashamed to same I'm pretty confused.MJF$#
Is the only knob of the phaser called a "feedback" knob? I will turn it right down and finger tap the tapping part of eruption in the morning. It's more noticeable then. The tone is great, it's just that the sound almost disappears.
Is the only knob of the phaser called a "feedback" knob? I will turn it right down and finger tap the tapping part of eruption in the morning. It's more noticeable then. The tone is great, it's just that the sound almost disappears.
Joined
·
4,855 Posts
LMFAO @adcandour. I'm confused as well ....from a non tweaking standpoint sort of. is it acting different than it did, or is it a recent issue? Did it sound different before?
Joined
·
8,894 Posts
Okay, here's the second clip. I did the intro to Panama instead - once with and once without. Ignore the timing and extra chugging 
How is it that it just doesn't sound good? It doesn't disappear in this clip, but it's sound absolutely terrible - especially compared to the 'without' portion. Is it just because it's vintage or something?
[soundcloud]
How is it that it just doesn't sound good? It doesn't disappear in this clip, but it's sound absolutely terrible - especially compared to the 'without' portion. Is it just because it's vintage or something?
[soundcloud]
Joined
·
8,894 Posts
Listening to it again, you can hear (imo) some really great tone buried in the phaser portion (even better than the 'without'), but then it gets to the low part of the sweep again and sounds like my mom taking a shit. Normally, I'd just get rid of the pedal, but I'm working on an all-vintage MXR board at the moment, and really want this to work out.
1) Guess it's a matter of taste but it doesn't sound bad or wrong to me.
2) Phasers create notches. So when the notches are inserted down in the bass, it WILL suck the guts out of your signal, no differently than if someone was deliberately turning down the Bass knob on your amp. So, possible what you dislike is simply normal phaser behaviour.
3) I assume you are using a Phase 90, by your description. The one knob is Speed. Many other phasers will include multiple controls, and I was assuming them in my earlier posts. All issues of the Phase 90 include a fixed amount of feedback, replacing what would be a variable control on some other units with a single preset. Different P90 issues used different amounts of feedback in that preset, and some players prefer one issue over another for that reason. MXR/Dunlop saw a market, and has included a push-button option to select two different feedback presets on recent P90-derived products. Someissues had a wider sweep than others, as well. Typically, people like a wide sweep with more feedback when the sweepis slow, and less feedback/sweep-width when setting the speed for a faster bubbly rate. EHX addressed that in the Small Stone by having one switch that simultaneously altered speed, feedback, width, and sweep-waveshape all at once. Clever design.
4) The nicest-sounding phasers have a smoother "turnaround" - how they reverse direction at the high and low end of the sweep. The smoothness of the turnaround is a function of how well-matched the FET transistors used as control elements are. It can happen that if they are not well-matched, one or more will crap out at the extreme of a sweep, and effectively say "You guys carry on without me. I can't go any farther. I'll catch you on the way back." The phaser essentially changes from 4-stage to 3-stage or whatever, yielding a less intense effect. That can happen at either end of the sweep. The 2N5952 FETs that MXR used, and continue to use, were selected because they are very easily matched, and I gather MXR sells enough of these puppies that they have a system in place for matching large quantities of them - lord knows they've had 45 years to do so. Long story short, if you had built this phaser, I might have reason to suggest maybe better matching was needed, as a source of your complaint. But since it is a commercially-built unit, I highly doubt it is that.
5) There is a trimmer inside on the board that sets the range of the sweep. Chances are pretty good it is optimally set and glued in place. But if it isn't, you might consider tinkering with it to see if they set it to YOUR tastes at the factory. Just make note of where it started out so you can return it to that setting. Changing the setting will not fry anything. What you WILL notice is that at the extreme settings there will be no audible phasing, while in between extremes will be settings ranging from low and gurgly to higher and swirly. Myself, I like to make it a panel-mount control, but that's going too far for your needs.
2) Phasers create notches. So when the notches are inserted down in the bass, it WILL suck the guts out of your signal, no differently than if someone was deliberately turning down the Bass knob on your amp. So, possible what you dislike is simply normal phaser behaviour.
3) I assume you are using a Phase 90, by your description. The one knob is Speed. Many other phasers will include multiple controls, and I was assuming them in my earlier posts. All issues of the Phase 90 include a fixed amount of feedback, replacing what would be a variable control on some other units with a single preset. Different P90 issues used different amounts of feedback in that preset, and some players prefer one issue over another for that reason. MXR/Dunlop saw a market, and has included a push-button option to select two different feedback presets on recent P90-derived products. Someissues had a wider sweep than others, as well. Typically, people like a wide sweep with more feedback when the sweepis slow, and less feedback/sweep-width when setting the speed for a faster bubbly rate. EHX addressed that in the Small Stone by having one switch that simultaneously altered speed, feedback, width, and sweep-waveshape all at once. Clever design.
4) The nicest-sounding phasers have a smoother "turnaround" - how they reverse direction at the high and low end of the sweep. The smoothness of the turnaround is a function of how well-matched the FET transistors used as control elements are. It can happen that if they are not well-matched, one or more will crap out at the extreme of a sweep, and effectively say "You guys carry on without me. I can't go any farther. I'll catch you on the way back." The phaser essentially changes from 4-stage to 3-stage or whatever, yielding a less intense effect. That can happen at either end of the sweep. The 2N5952 FETs that MXR used, and continue to use, were selected because they are very easily matched, and I gather MXR sells enough of these puppies that they have a system in place for matching large quantities of them - lord knows they've had 45 years to do so. Long story short, if you had built this phaser, I might have reason to suggest maybe better matching was needed, as a source of your complaint. But since it is a commercially-built unit, I highly doubt it is that.
5) There is a trimmer inside on the board that sets the range of the sweep. Chances are pretty good it is optimally set and glued in place. But if it isn't, you might consider tinkering with it to see if they set it to YOUR tastes at the factory. Just make note of where it started out so you can return it to that setting. Changing the setting will not fry anything. What you WILL notice is that at the extreme settings there will be no audible phasing, while in between extremes will be settings ranging from low and gurgly to higher and swirly. Myself, I like to make it a panel-mount control, but that's going too far for your needs.
1 - 13 of 13 Posts
Join the discussion
