FYA Scream Queen
- mirosol
- Resistor Ronker
Behold! 41 seconds of research and development from FYA engineering department bring you the completely new and exciting overdrive design that's never seen before!
I was sorta minding my own business on my desk and it hit me. I have a bag of NJM JRC4558BL chips i bought a year or a few ago. Used just one to swap a burnt chip from one Ibanez Ping-Pong delay, but the rest have been just sitting in the drawer. Not that many diy layouts out there with SIP chips, but the urge to do something with said chip became overwhelming. In the heat of the moment, i didn't have time to draw a schem, but i drew a layout. While drawing, i figured that Clay Jones' and Briggs' "son of screamer" idea would have otherwise been ok to play around with, if it weren't for the obvious issue with the input impedance. In other words, the TS styled clipping amp will need a buffer or any other method of impedance transformation to keep it from loading pickups and squeeling like a walkie-talkie. Next i thought about adding a transistor to act as input buffer. Nah. I'm not that big of a fan for TS styled tone control, so that needed to go anyway. And that out of the picture, the other half of the opamp was freed to take care of the input buffering.
So. We have our basic power supply section with absolutely nothing special in there, just series polarity protection and standard vref network. Next up is the input buffer. This buffer design is not exactly conceived behind closed doors, so credit is mainly due to amz. Now, this buffer ia a great sounding little thing. As noted on the schematic (i drew up after building one), one can build his/hers with the basic true bypass. Or. One could take another cap (10u will suffice) from the buffer output and use that as buffered bypass (with something like 10K resistor shunt to ground). This would allow the use of SPDT on-on stomp. Or DPDT to accomodate LED in the switching. I'd like to call this cap-splitting a Bajaman buffered bypass method. Anyway. Next up is the heart of the pedal. The clipping amp. Which is straight up loan from the TS. Lower roll off for the high pass and slightly, no, considerably more gain than standard TS has. The diodes (2x4148 + 2x4001) were just a hunch of what might work. Now, it's just the simplest tone control that works very well, and the volume control. 200K linear, because why not.
So the design method in a nutshell - notice a part i haven't used a lot - draw up a build doc to see how it would work on a simple od circuit - build it - play with it - retire from the day and draw the schem and post it here.
But there is one more thing i didn't say anything about yet. After i built the circuit, it fired up. I started with tone and gain at CCW, volume maxed. Nice boosted signal with very cool slight breakup that follows the playing dynamics in a beautiful manner. Open up the tone and whoah. The similarity with Gainster is apparent. Gain has very smooth and well working sweep. No dead area on the controls. Plus the highest gain settings goes nicely to the lead tone territory, while still having glass-like clarity. Not muddy, but the mids are still strong. I'm very, very happy with the way this thing sounds.
I'm calling this the "Scream Queen Overdrive". And i love it. And i'm somewhat proud how it came to be. Wonder if i should do a second version with +/- swing at some point.
I was sorta minding my own business on my desk and it hit me. I have a bag of NJM JRC4558BL chips i bought a year or a few ago. Used just one to swap a burnt chip from one Ibanez Ping-Pong delay, but the rest have been just sitting in the drawer. Not that many diy layouts out there with SIP chips, but the urge to do something with said chip became overwhelming. In the heat of the moment, i didn't have time to draw a schem, but i drew a layout. While drawing, i figured that Clay Jones' and Briggs' "son of screamer" idea would have otherwise been ok to play around with, if it weren't for the obvious issue with the input impedance. In other words, the TS styled clipping amp will need a buffer or any other method of impedance transformation to keep it from loading pickups and squeeling like a walkie-talkie. Next i thought about adding a transistor to act as input buffer. Nah. I'm not that big of a fan for TS styled tone control, so that needed to go anyway. And that out of the picture, the other half of the opamp was freed to take care of the input buffering.
So. We have our basic power supply section with absolutely nothing special in there, just series polarity protection and standard vref network. Next up is the input buffer. This buffer design is not exactly conceived behind closed doors, so credit is mainly due to amz. Now, this buffer ia a great sounding little thing. As noted on the schematic (i drew up after building one), one can build his/hers with the basic true bypass. Or. One could take another cap (10u will suffice) from the buffer output and use that as buffered bypass (with something like 10K resistor shunt to ground). This would allow the use of SPDT on-on stomp. Or DPDT to accomodate LED in the switching. I'd like to call this cap-splitting a Bajaman buffered bypass method. Anyway. Next up is the heart of the pedal. The clipping amp. Which is straight up loan from the TS. Lower roll off for the high pass and slightly, no, considerably more gain than standard TS has. The diodes (2x4148 + 2x4001) were just a hunch of what might work. Now, it's just the simplest tone control that works very well, and the volume control. 200K linear, because why not.
So the design method in a nutshell - notice a part i haven't used a lot - draw up a build doc to see how it would work on a simple od circuit - build it - play with it - retire from the day and draw the schem and post it here.
But there is one more thing i didn't say anything about yet. After i built the circuit, it fired up. I started with tone and gain at CCW, volume maxed. Nice boosted signal with very cool slight breakup that follows the playing dynamics in a beautiful manner. Open up the tone and whoah. The similarity with Gainster is apparent. Gain has very smooth and well working sweep. No dead area on the controls. Plus the highest gain settings goes nicely to the lead tone territory, while still having glass-like clarity. Not muddy, but the mids are still strong. I'm very, very happy with the way this thing sounds.
I'm calling this the "Scream Queen Overdrive". And i love it. And i'm somewhat proud how it came to be. Wonder if i should do a second version with +/- swing at some point.
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
- FiveseveN
- Cap Cooler
Information
Am I missing something? It's a non-inverting amplifier, so ~5 MΩ input impedance for the 4558, right?mirosol wrote:the TS styled clipping amp will need a buffer or any other method of impedance transformation to keep it from loading pickups
And if I'm not, the buffer would probably be of more use at the output, to keep whatever follows from interacting with the tone control. It also means a tiny bit less noise going into the gain stage.
You also already have a decoupled, lower-noise reference voltage source, so why the 1 M biasing divider?
Ignorance more frequently begets confidence than does knowledge. (Charles Darwin)
- induction
- Resistor Ronker
I guess I'm missing something too. The gain and the input impedance don't appear to be related to the output impedance of the previous stage at all. RG's analysis seems to confirm this. And the input impedance of the circuit could be made just as high without the input buffer by changing the biasing resistors.
Not saying it doesn't sound great, but I don't understand your explanation. What am I missing?
Not saying it doesn't sound great, but I don't understand your explanation. What am I missing?
- mirosol
- Resistor Ronker
There is always a chance i'm dead wrong. Have been before. This would be a decent moment to delete a post as i'm not that comfortable with posting shit. Even if my intentions were good. You know what they say about the road to hell..
In other words. The behaviour of the SoS / SoCJ was the motivation for the buffering seen here. More of a gut feeling and a result of hands on experience than strong knowledge on theory. Buffered pedal in front of the SoS or SoCJ cures the behaviour and this behaviour isn't present on this circuit. Which makes me think that my original assumption about impedance being the key factor for the squelch isn't correct. It's more likely something else, and i'm not aware what the real cause is. But i do know it's cured by placing a buffer before the TS clipping amp.
In other words. The behaviour of the SoS / SoCJ was the motivation for the buffering seen here. More of a gut feeling and a result of hands on experience than strong knowledge on theory. Buffered pedal in front of the SoS or SoCJ cures the behaviour and this behaviour isn't present on this circuit. Which makes me think that my original assumption about impedance being the key factor for the squelch isn't correct. It's more likely something else, and i'm not aware what the real cause is. But i do know it's cured by placing a buffer before the TS clipping amp.
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
- induction
- Resistor Ronker
User phatt has steadily maintained that feeding a high impedance source signal to a stage with large voltage gain leads to noise and/or hiss. So it's certainly possible that the source impedance is important to the amplifying stage, even if it doesn't affect gain. Perhaps the buffer does some kind of signal preconditioning that limits the amplification of noise in the clipping amp. Even if that noise is outside the audible range, it can interfere with the audible signal. Just a guess. I'm in over my head here, theory-wise, but that's not saying much. Maybe somebody else knows.
Anyway, as long as it sounds good...
Speaking of which, got any clips?
Anyway, as long as it sounds good...
Speaking of which, got any clips?
- mirosol
- Resistor Ronker
The output impedance of the input buffer is less than 100 ohms. In my limited understanding this means that the source signal is very low in terms of impedance. I'm still not convinced that the 10K to vref at the + input doesn't affect the clipping amp input Z.
If anyone has an explanation of why SoS/SoCJ tends to squeel and this goes away by having a buffer in front of it, i'd be glad to hear it. Should never assume, but i seem to do it anyway. Don't have Sos/SoCJ circuit at hand, but i'm beginning to wonder if swapping the 10K at + input for way higher value would cure the behaviour on those.
As for clips. Not at the moment. Plus my motivation to do aything with this circuit is kinda low at the moment due to severe remorse of sharing it.
If anyone has an explanation of why SoS/SoCJ tends to squeel and this goes away by having a buffer in front of it, i'd be glad to hear it. Should never assume, but i seem to do it anyway. Don't have Sos/SoCJ circuit at hand, but i'm beginning to wonder if swapping the 10K at + input for way higher value would cure the behaviour on those.
As for clips. Not at the moment. Plus my motivation to do aything with this circuit is kinda low at the moment due to severe remorse of sharing it.
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
- induction
- Resistor Ronker
Exactly. I'm guessing that's what's making it behave better than the SoS, etc. I'm just not sure exactly why.mirosol wrote: The output impedance of the input buffer is less than 100 ohms. In my limited understanding this means that the source signal is very low in terms of impedance
It does. It drops it to about 10k. That's pretty low for pickups, but not crazy low compared to the output impedance of the buffer. It's the ratio that's important. (More precisely, Zin_clipping/(Zout_buffer+Zin_clipping). Voltage dividers, y'know.)I'm still not convinced that the 10K to vref at the + input doesn't affect the clipping amp input Z.
Those circuits already swap that resistor for a 470k to make up for the missing input buffer. The input impedance ends up about the same as the buffered TS.Don't have Sos/SoCJ circuit at hand, but i'm beginning to wonder if swapping the 10K at + input for way higher value would cure the behaviour on those.
Don't beat yourself up. If it sounds good, it is good. Your filtering is different from what I've seen on a TS before, and the tone control is different. Empires have been built on less original circuits. Not bad for 41 seconds worth of RnD. I'd like to hear it if you change your mind. I'd breadboard it, but all my stuff is packed up in storage for now.As for clips. Not at the moment. Plus my motivation to do aything with this circuit is kinda low at the moment due to severe remorse of sharing it.
- mirosol
- Resistor Ronker
Now that i think of it, this may be the real cause. If the impedance is high (around 470K) at the noninverting input, this would mean more noise, right? This noise could be the key factor in squelch and audible oscillation - not the "pickup loading". And also, since the buffer in front of the clipping amp cures the behavior here as well as for SoS, the issue has to be, at least in my mind, still an impedance mismatch between a guitar and a clipping amp. Versus a very low output impedance buffer and the same clipping amp. Impedance mismatch in terms of how the following clipping amp performs, not so much issue that affects earlier stuff in the signal path. In other words, i believe Induction gave me the correct answer right away, it just took me a few days to get it. Of course other method of curing this could be to simply up the feedback loop's cap value. This does however have a strong downside of easily eating a lot of highs and making overall tone muddy.induction wrote:User phatt has steadily maintained that feeding a high impedance source signal to a stage with large voltage gain leads to noise and/or hiss.
So if i'm correct (please do tell me if i'm not and why), everything else in my description was and still is correct except the note about pickup loading.
What comes to redundancy, i'm not sure if one extra resistor at BOM is worse than not having the two dividers completely isolated from each other.
Now that the circuit doesn't seem that bad and i don''t feel nearly as stupid as before, i might reconsider recording a clip. Maybe.
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
I've built the Wampler Ecstacy with a common vref and the sound was crap. Then I' built it with separate dividers as in the original and the sound was then as the original (as much as I can judge). There is of course the chance that I have screwed something else in my first build (though I inspected it several times), but the coincidence is more than suspicious.mirosol wrote:What comes to redundancy, i'm not sure if one extra resistor at BOM is worse than not having the two dividers completely isolated from each other.
- Frank_NH
- Solder Soldier
Could elaborate on this drd? I too built a vero version of the Ecstasy and didn't like it either. Is there something to having separate voltage dividers for the bias voltages such a those shown here (though this is just a simulation schematic)?drd wrote:I've built the Wampler Ecstacy with a common vref and the sound was crap. Then I' built it with separate dividers as in the original and the sound was then as the original (as much as I can judge). There is of course the chance that I have screwed something else in my first build (though I inspected it several times), but the coincidence is more than suspicious.mirosol wrote:What comes to redundancy, i'm not sure if one extra resistor at BOM is worse than not having the two dividers completely isolated from each other.
- mirosol
- Resistor Ronker
I don't have much to add to the isolated reference voltage discussion. Vref is in many cases used as virtual gound, as is done here with the 3K9 and 220n. Even though the Vref should have no signal, just voltage, it still can carry AC signal. Not sure about how and why Ecstasy behaves as it does, but i've seen a few other new designs that separate Vrefs for different jobs. Nothing new about that and i believe there is good reason for several Vref networks.
Back to the earlier subjects.. I still have no theory to back me up, but i just built up a basic true bypassed layout of the Hotcake - Which is another board notorious for squeeling like a pig when presence and gain are maxed - and tried to fiddle with the resistor from Vref to noninverting input. This resistor is at 1M on the schematic. First i simply swapped it for 220K. Notably lower amount of audible high pitch oscillation. Still not good enough, but better tham 1M. Lowered it down to 47K and that was even better, but not good enough. 10K. Yup. No audible oscillation at all. No matter what pickups i was running, singles, humbuckers, hot or vintage. I was rather worried that lowering this resistor this low would mean pickup loading and bad, or lazy tone. No. Didn't hear anything like that. Tried the circuit with and without my old trusty TU-2 in front of it. No audible difference.
With this little empirical test, i'm fairly certain that the issue is with the high impedance at the noninverting input causing noise - amp this noise getting amplified by the driver amp. Once this resistor value gets lowered enough, the noise/hiss that was turned into full blown oscillation is completely gone.
In addition, here's a stripboard layout for DIP chip version...
Back to the earlier subjects.. I still have no theory to back me up, but i just built up a basic true bypassed layout of the Hotcake - Which is another board notorious for squeeling like a pig when presence and gain are maxed - and tried to fiddle with the resistor from Vref to noninverting input. This resistor is at 1M on the schematic. First i simply swapped it for 220K. Notably lower amount of audible high pitch oscillation. Still not good enough, but better tham 1M. Lowered it down to 47K and that was even better, but not good enough. 10K. Yup. No audible oscillation at all. No matter what pickups i was running, singles, humbuckers, hot or vintage. I was rather worried that lowering this resistor this low would mean pickup loading and bad, or lazy tone. No. Didn't hear anything like that. Tried the circuit with and without my old trusty TU-2 in front of it. No audible difference.
With this little empirical test, i'm fairly certain that the issue is with the high impedance at the noninverting input causing noise - amp this noise getting amplified by the driver amp. Once this resistor value gets lowered enough, the noise/hiss that was turned into full blown oscillation is completely gone.
In addition, here's a stripboard layout for DIP chip version...
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
@Frank,
Yes, that's how I built it the second time. I didn't find it exceptionally good, but it is certainly a very usable overdrive.
@Miro,
Thanks for the info. I didn't have the sqealing problem with the Hot Cake, but it is a nice info. I doN't know though why this is happening
Yes, that's how I built it the second time. I didn't find it exceptionally good, but it is certainly a very usable overdrive.
@Miro,
Thanks for the info. I didn't have the sqealing problem with the Hot Cake, but it is a nice info. I doN't know though why this is happening
- mirosol
- Resistor Ronker
Can you recall which layout you used? Maybe that layout was addressing the issue with some fix? Or it was with the original bybass method?drd wrote:I didn't have the sqealing problem with the Hot Cake
+m
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
I used my own pcb layout and I have built the newer version with the Presence control and not the old two-pot version. As I see from my schemos they are quite different in this regards: the old version connects to vref also on the non-inverting input with a 220k resistor.
- mirosol
- Resistor Ronker
http://mirosol.kapsi.fi/varasto/circuit ... D-CLIP.mp3
Accompanied by Fender Blacktop Stratocaster, halved bridge humbucker and lead by Tokai Love Rock model, neck humbucker, modern style at ~11K. Played into my trusty ol' bench amp, the Laney VC15-110 (armed with WGS ET10, not heard here though) with signal taken out of the effects loop and fed through Joyo JDI-01 w/ 4x12 sim -> Behringer shitmix from 2005 -> M-Audio Audiophile 24/96 -> Ardour4, no added effects.
All controls (vol/gain/tone) dead center for the Strat and gain maxed + tone at 3/4 for the LP. IMO - Strong mids, lows are present and quite tight. Highs are crisp and ringing. It's still the "TS on steroids" tone with nothing that special in there. Still, i like it.
Accompanied by Fender Blacktop Stratocaster, halved bridge humbucker and lead by Tokai Love Rock model, neck humbucker, modern style at ~11K. Played into my trusty ol' bench amp, the Laney VC15-110 (armed with WGS ET10, not heard here though) with signal taken out of the effects loop and fed through Joyo JDI-01 w/ 4x12 sim -> Behringer shitmix from 2005 -> M-Audio Audiophile 24/96 -> Ardour4, no added effects.
All controls (vol/gain/tone) dead center for the Strat and gain maxed + tone at 3/4 for the LP. IMO - Strong mids, lows are present and quite tight. Highs are crisp and ringing. It's still the "TS on steroids" tone with nothing that special in there. Still, i like it.
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
- mirosol
- Resistor Ronker
DIP layout now verified too.
http://tagboardeffects.blogspot.com/
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic
http://mirosol.kapsi.fi/
"No such thing as innocence" -Iron Chic