This Week on the Breadboard: Phase 90
- Thread starter Chuck D. Bones
- Start date
Chuck D. Bones
Circuit Wizard
My money's on Cooder.
Chuck D. Bones
Circuit Wizard
I'm exploring one other tweak - the duty factor on the LFO. It's not exactly 50% in the Phase 90 or XC Phase because the LFO opamp doesn't clip symmetrically. When I replaced the zener with a 5Vreg, that affected the duty factor a little bit. What I'm trying to find out is does it matter?
cooder
Well-known member
My build report is scheduled for end of January.... haha... (not sure if 2024 or 2025)....
Chuck D. Bones
Circuit Wizard
If you're asking about the duty factor, I'm seeing it on the scope. With the circuit shown above, it runs just under 60%, which is not necessarily a bad thing. With the addition of one resistor, I can tilt the DF one way or the other. I'm exploring it now.Are you seeing this in the sim or on the scope? Throw parts at it.
Aentons
Well-known member
I have a lingering question from yesteryear that I never had a chance to explore. Wondering if you can provide any additional sight.
Chuck D. Bones
Circuit Wizard
In my opinion, the mixer stage does not require compensation for stability. Rolling the high-freq gain off above 20KHz is always a good idea since there is no musically useful information above 20KHz. A 47pF cap from B to C on Q5 would accomplish that.
[EDIT]
Upon further reflection, the mixer stage is connected to the output jack and depending on the loading from cable capacitance or the next device in the chain, Q5 might have stability issues. IMO, the better fix is to isolate Q5 from the external load by installing a 10K resistor between C6/R20 and the stomp switch.
[EDIT]
Upon further reflection, the mixer stage is connected to the output jack and depending on the loading from cable capacitance or the next device in the chain, Q5 might have stability issues. IMO, the better fix is to isolate Q5 from the external load by installing a 10K resistor between C6/R20 and the stomp switch.
Last edited:
Aentons
Well-known member
I found this thread talking about the additional caps on the later block logo version which the reissues were based on. Sounds like it was mostly done to deal with hiss but I've always wondered if it was intentional to get a more vintage sound. I have a reissue that I think sounds better than 95% of the diy pedals based on the script that don't include the caps. I don't know the cap values that were used but mine is definitely in the audible range.
"The pre-emphais/de-emphasis does have a purpose. It reduces noise/hiss. The TL062 devices can be noisy although original LM741's aren't exactly low noise. The pre-emphasis/de-emphasis frequencies on that unit aren't particularly aggressive. They are just enough to shave off some hiss.
What I can't remember is how well the pre-emphasis and de-emphasis match. If they don't match it can affect the tone a small amount - most people wouldn't pick it up."
This page mentions the extra caps as well.
Chuck D. Bones
Circuit Wizard
The only significant noise sources in a Phase 90 are the opamps. TL072 has a very low noise floor and does not contribute enough noise in a Phase 90 to be noticeable. Bear in mind that every stage in the phase 90 has unity gain.
The other use for pre-emphasis / de-emphasis is to lower the low-freq content going into the phase shifter stages to reduce distortion. In a guitar signal, the low freq content is much higher amplitude than the high-freq content. That means that the lower freq content is the main source of distortion in FET-based and OTA phase shifters. LDR-based phase shifters don't have an issue with distortion.
You are correct that if the pre-emphasis / de-emphasis are mismatched, it will effect the overall freq response. Given that a phase shifter's freq response is anything but flat, it would be hard to notice small variations in the end-to-end freq response.
I am leery of schematics posted in the DIY forums (including this one) because once an incorrect schematic gets out there, it's like a groupie with an STD backstage at an AC/DC concert. Pretty soon, everyone has it.
The other use for pre-emphasis / de-emphasis is to lower the low-freq content going into the phase shifter stages to reduce distortion. In a guitar signal, the low freq content is much higher amplitude than the high-freq content. That means that the lower freq content is the main source of distortion in FET-based and OTA phase shifters. LDR-based phase shifters don't have an issue with distortion.
You are correct that if the pre-emphasis / de-emphasis are mismatched, it will effect the overall freq response. Given that a phase shifter's freq response is anything but flat, it would be hard to notice small variations in the end-to-end freq response.
I am leery of schematics posted in the DIY forums (including this one) because once an incorrect schematic gets out there, it's like a groupie with an STD backstage at an AC/DC concert. Pretty soon, everyone has it.
Chuck D. Bones
Circuit Wizard
OK, I really mean it this time. It's done. Finie. Fertig, Finito.
The switch on the left is for shifting the duty factor. Green knob: SPEED; White knob: DEPTH; Yellow knob: COLOR; Switch: RESONANCE. The army of JFETs just below the knobs are matched sets: 2N5484, 2N5952, 2SK30A-Y, BF244A. They all work. I have BF244B installed in the circuit. They have a higher Vp (around 2.1V) which yields more headroom.
Changes from the last schematic:
Added R50 to isolate Q5 from the next device in the chain.
Reduced R18 to give Q5 a little more headroom.
Added R51. This is routine in my circuits now.
Added R52 to bring the LFO duty factor close to 50%. The sweep sounds more "balanced" to me.
Reduced R24 for a little more sweep when DEPTH is dimed.
Put the Bias Trim (TR1, R47 & R48) back in.
Added R46 which controls the range of the COLOR knob.
Added C11 to block DC from the LFO signal, which makes DEPTH & COLOR less interactive.
Replaced D1 & R101 with a 78L05 regulator. More stable than a zener. C10 should be 1uF tantalum or film. Do not use a larger cap, it can make the 78L05 oscillate.
Replaced the RESONANCE pot with an ON-OFF-ON switch.
RESONANCE has three settings: Script, Block Logo and 1/2-way in between. I had a knob for RESONANCE, but there were only three useful settings so I replaced the knob with a switch. I have all 47nF caps installed in the phase shifters. With the right DEPTH and COLOR settings, it does a pretty convincing UniVibe tone. As such, I don't see the need for stagger-tuning. Those of you who are breadboarding this can try it both ways and form your own opinion.
NOTE: schematic corrected
If what you want is a stock Phase 90, then by all means build the XC Phase.
If you want a 1-knob phase shifter that sounds like a Phase 90 but has more headroom, then build this and jumper all three pins of the DEPTH control, jumper C11, change R24 to 3.9M, change R25 to 1M, omit R46 and the COLOR control and change R52 to 2.2M.
If you want a sort-of UniVibe, then use the stagger-tuned capacitor values. The RESONANCE switch, R27 & R28 are optional.
If you want a hopped-up Phase 90, then build it as shown, but with C2-C5 = 47nF.
The switch on the left is for shifting the duty factor. Green knob: SPEED; White knob: DEPTH; Yellow knob: COLOR; Switch: RESONANCE. The army of JFETs just below the knobs are matched sets: 2N5484, 2N5952, 2SK30A-Y, BF244A. They all work. I have BF244B installed in the circuit. They have a higher Vp (around 2.1V) which yields more headroom.
Changes from the last schematic:
Added R50 to isolate Q5 from the next device in the chain.
Reduced R18 to give Q5 a little more headroom.
Added R51. This is routine in my circuits now.
Added R52 to bring the LFO duty factor close to 50%. The sweep sounds more "balanced" to me.
Reduced R24 for a little more sweep when DEPTH is dimed.
Put the Bias Trim (TR1, R47 & R48) back in.
Added R46 which controls the range of the COLOR knob.
Added C11 to block DC from the LFO signal, which makes DEPTH & COLOR less interactive.
Replaced D1 & R101 with a 78L05 regulator. More stable than a zener. C10 should be 1uF tantalum or film. Do not use a larger cap, it can make the 78L05 oscillate.
Replaced the RESONANCE pot with an ON-OFF-ON switch.
RESONANCE has three settings: Script, Block Logo and 1/2-way in between. I had a knob for RESONANCE, but there were only three useful settings so I replaced the knob with a switch. I have all 47nF caps installed in the phase shifters. With the right DEPTH and COLOR settings, it does a pretty convincing UniVibe tone. As such, I don't see the need for stagger-tuning. Those of you who are breadboarding this can try it both ways and form your own opinion.
NOTE: schematic corrected
If what you want is a stock Phase 90, then by all means build the XC Phase.
If you want a 1-knob phase shifter that sounds like a Phase 90 but has more headroom, then build this and jumper all three pins of the DEPTH control, jumper C11, change R24 to 3.9M, change R25 to 1M, omit R46 and the COLOR control and change R52 to 2.2M.
If you want a sort-of UniVibe, then use the stagger-tuned capacitor values. The RESONANCE switch, R27 & R28 are optional.
If you want a hopped-up Phase 90, then build it as shown, but with C2-C5 = 47nF.
Last edited:
Chuck D. Bones
Circuit Wizard
Pin 7 has a square wave on it. We want something resembling a triangle wave, which can be found on the + end of C8. The only sticking point is that we can't load C8. Since we're driving the gates of JFETs, that's not a problem. The only load on C8 is R24 + R25 + VR2 + C11 which is a high enough impedance that the LFO will work. The Phase 100 has the same LFO circuit, but uses a transistor to buffer the signal coming from the timing capacitor.
Have you read my articles on LFOs?
Last edited:
Chuck D. Bones
Circuit Wizard
2 many knobs?
Chuck D. Bones
Circuit Wizard
To be fair, it only mentions the Phase 90 LFO in passing.Oh damn I haven’t done my homework!
I gotta read your article! Thank you
Sorry, looks alright to me!???
