4-Channel Preamp Design for Guitar
This is an experimental design for a solid-state preamp made using op-amps and is intended to result in a guitar preamp that is more versatile than typical. I start with the input stage, which is simply a normal voltage amp using an OPA2134 op-amp, chosen because it is reported to be one of the cleanest op-amps around, having noise figures better than most other op-amps on the market, and which was purpose designed for professional audio applications in which low noise is of primary concern.
The circuit is also a distribution amp, though gain is rather low. Yet, the gain should be sufficient to send the input signal to four separate channels. I have shown it with input and output jacks on purpose. Each sub-circuit of the project is shown this way so as to emphasize the versatility of the project. Making each sub-circuit’s input and output user accessible, say via a patch-bay in the back of the completed project’s enclosure, allows the user to patch the sub-circuits together as preferred, and to treat each in/out interface as an effects loop. An Input Level control has been included to tame large input signals that result from hot pickups and stomp-box effects, if needed. The gain here can be set from unity to 25, and each output has its own volume control.
As can be seen below, each channel is enabled independently of the others. I suggest using Channel 1 as a direct channel, though with a gain stage and active tone-control, so the input signal can keep up with the signals from the other three channels.
There is no digital circuitry in this project, although the switching IC (a CD4066B quad bilateral switch) is often used in digital equipment. Rather, the switching IC is only there to perform as a quiet switch the user employs to enable a given channel.
The genius of this IC in audio applications is that it makes doing quiet switching easy, though simple passive pop-eliminators (shown later) must be used on the footswitch lines run in a non-shielded 5- or 6- conductor cable. One line carries a +5V control voltage, with no need for a ground wire, while the other lines are the returns in which push-on / push-off footswitches are installed. Builders of this project must design and construct their own footswitch array, but there are many sources for enclosures that can be used for that purpose. Or a suitable replacement 4-switch array made for an existing guitar-amp could be used, though wired according to the needs of this project.
Pinout diagrams for all the ICs in this project are given at the end of this post.
What follows are the sub-circuits designed to be used in the various channels, but which are presented in no mandatory order, since they can be connected in any order within a given channel, and any sub-circuit can even be used independently for some purpose other than including it with the others in this project. In other words, any given output can be used as an effect send, or high-Z output to an amplifier or to recording gear, but which signal need not necessarily be returned.
Shown next is a 3-band tone-control in which each control adjusts the amplitude of an actual band, not just the center of each band. The Bass band has frequencies < 1 kHz, the Mid band is from 500 Hz to 5 kHz, and the Treble band has frequencies > 3 kHz.
It is recommended that four of these circuits are built, using four quad op-amps, in case the user wants one in each channel, but they need not all be used. Below is a gain stage of which four are also recommended, though using only one quad op-amp (with one gain stage for the direct channel).
Here therefore are three different distortion circuits, labeled Fuzz 1, 2, 3, but which can be used in any order, though four of each circuit are not needed. Rather, the purpose is to use one each in a channel of its own while keeping the direct undistorted signal in Channel 1 (though it too should be given a gain stage and 3-band tone-control). Two of the Fuzz circuits use diodes to get distortion while the third uses overdriven LF353 op-amps. The OPA2134 op-amps are here to provide clean amplification, but the LF353, also a dual JFET-input op-amp, is used here because of its inferior noise figures. This op-amp was one of the earliest versions of JFET-input op-amps and should therefore be fairly easy to cause to clip with sufficient input signal voltages.
At this point let me explain the operation of Fuzz Unit 1, shown below. An input level control is provided in case a Gain Stage is used before it in its channel, or if it is to be used independently. Gain is fixed fairly high; around 220. It drives a diode circuit that uses germanium diodes, said to produce a distortion sound similar to that of overdriven preamp tubes. In fact, they would convert a sine-wave into a square-wave, but whose edges are more rounded than what would happen with silicon diodes. LEDs are also said to produce good distortion, and a passive diode distortion circuit using both types has been included with this project.
The only other control for Fuzz Unit 1 is an output level control, which adjusts the fuzz signal’s amplitude going to the input of a buffer, which is there simply to serve as an active circuit placed between the diodes and any given succeeding circuit.
Up next is Fuzz Unit 2, which uses diodes in a feedback loop rather than as a shunt on the signal path. It cannot be predicted as to which method of obtaining distortion will sound better than the other. Only building them and testing them will reveal the answer to that question. Yet, both circuits were based on preamps in proven guitar-amps.
On the other hand, Fuzz Unit 2 has more adjustability than Fuzz Unit 1, sporting not only input and output volume controls but separate “Intensity” and “Fuzz” controls. And below is given Fuzz Unit 3, which has no diodes but relies on overdriven op-amps to get distortion, as said. It has four volume controls, labeled in order Drive, OD1, OD2, and Level. [OD = overdrive.] Drive sets the circuit’s input level, OD1 and OD2 control the interstage signal voltages, and the Level control sets output voltage. However, there is also a Feedback control, employed in case the user wants to tame the distortion down while still getting some distortion from the circuit. It is connected backwards so the circuit will sound as if it is getting louder as the control is turned clockwise.
Here are a couple of passive distortion circuits. If built, for best sound they should be driven by one of the Gain Stages. How they work is that the diodes cause clipping of the input signal, though there is some loss due to the diodes actually constituting shunts to ground -- hence the need for an active device driving each. And the best distortion tone, emphasizing even-order harmonics, is obtained when a given diode array has one diode on one side but two on the other. It does not matter which side has the lone diode since what matters is that there is an imbalance. Of course, any setting except when all diodes are disabled will produce distortion. The user can set switches according to their personal taste. In some cases, for instance, harsher distortion, emphasizing odd-order harmonics, may be desired, instead of the sweetness of even-order harmonics.
I recommend using the 3-Band Tone controls at the end of each channel, between the final sub-circuits and the Mixer, so that the tonality of each channel can be adjusted at the inputs of the Mixer, shown below.
The op-amp circuit in the Mixer is configured as a basic summing amp, where each of its inputs has a separate volume control, but there is only one Mix Volume control on the output. This is meant to be the final output of the project and is suitable for the input of any guitar-amp or any other equipment that can accepts a high-Z guitar signal, including another preamp, a stomp-box effect, PA mixing console, recording gear, etc.
Shown below are the pinout diagrams for the ICs used in this project, along with the pop-eliminator circuit, with four required, and which can be placed within the project enclosure or in the remote footswitch array’s enclosure, as desired, as long as a given pop-eliminator is between its footswitch and the Switching IC. There is also a sketch of how to properly connect each IC to its respective supply-voltage leads.
Here is a block diagram indicating the suggested initial connections scheme, though the user can connect individual subcircuits in any sequence or in any channel.
Last but not least is the power supply. This circuit is of my own design but is based on long-standing power-supply topologies. Major components are specified below.
EOF
This is an experimental design for a solid-state preamp made using op-amps and is intended to result in a guitar preamp that is more versatile than typical. I start with the input stage, which is simply a normal voltage amp using an OPA2134 op-amp, chosen because it is reported to be one of the cleanest op-amps around, having noise figures better than most other op-amps on the market, and which was purpose designed for professional audio applications in which low noise is of primary concern.
The circuit is also a distribution amp, though gain is rather low. Yet, the gain should be sufficient to send the input signal to four separate channels. I have shown it with input and output jacks on purpose. Each sub-circuit of the project is shown this way so as to emphasize the versatility of the project. Making each sub-circuit’s input and output user accessible, say via a patch-bay in the back of the completed project’s enclosure, allows the user to patch the sub-circuits together as preferred, and to treat each in/out interface as an effects loop. An Input Level control has been included to tame large input signals that result from hot pickups and stomp-box effects, if needed. The gain here can be set from unity to 25, and each output has its own volume control.
As can be seen below, each channel is enabled independently of the others. I suggest using Channel 1 as a direct channel, though with a gain stage and active tone-control, so the input signal can keep up with the signals from the other three channels.
There is no digital circuitry in this project, although the switching IC (a CD4066B quad bilateral switch) is often used in digital equipment. Rather, the switching IC is only there to perform as a quiet switch the user employs to enable a given channel.
The genius of this IC in audio applications is that it makes doing quiet switching easy, though simple passive pop-eliminators (shown later) must be used on the footswitch lines run in a non-shielded 5- or 6- conductor cable. One line carries a +5V control voltage, with no need for a ground wire, while the other lines are the returns in which push-on / push-off footswitches are installed. Builders of this project must design and construct their own footswitch array, but there are many sources for enclosures that can be used for that purpose. Or a suitable replacement 4-switch array made for an existing guitar-amp could be used, though wired according to the needs of this project.
Pinout diagrams for all the ICs in this project are given at the end of this post.
What follows are the sub-circuits designed to be used in the various channels, but which are presented in no mandatory order, since they can be connected in any order within a given channel, and any sub-circuit can even be used independently for some purpose other than including it with the others in this project. In other words, any given output can be used as an effect send, or high-Z output to an amplifier or to recording gear, but which signal need not necessarily be returned.
Shown next is a 3-band tone-control in which each control adjusts the amplitude of an actual band, not just the center of each band. The Bass band has frequencies < 1 kHz, the Mid band is from 500 Hz to 5 kHz, and the Treble band has frequencies > 3 kHz.
It is recommended that four of these circuits are built, using four quad op-amps, in case the user wants one in each channel, but they need not all be used. Below is a gain stage of which four are also recommended, though using only one quad op-amp (with one gain stage for the direct channel).
Here therefore are three different distortion circuits, labeled Fuzz 1, 2, 3, but which can be used in any order, though four of each circuit are not needed. Rather, the purpose is to use one each in a channel of its own while keeping the direct undistorted signal in Channel 1 (though it too should be given a gain stage and 3-band tone-control). Two of the Fuzz circuits use diodes to get distortion while the third uses overdriven LF353 op-amps. The OPA2134 op-amps are here to provide clean amplification, but the LF353, also a dual JFET-input op-amp, is used here because of its inferior noise figures. This op-amp was one of the earliest versions of JFET-input op-amps and should therefore be fairly easy to cause to clip with sufficient input signal voltages.
At this point let me explain the operation of Fuzz Unit 1, shown below. An input level control is provided in case a Gain Stage is used before it in its channel, or if it is to be used independently. Gain is fixed fairly high; around 220. It drives a diode circuit that uses germanium diodes, said to produce a distortion sound similar to that of overdriven preamp tubes. In fact, they would convert a sine-wave into a square-wave, but whose edges are more rounded than what would happen with silicon diodes. LEDs are also said to produce good distortion, and a passive diode distortion circuit using both types has been included with this project.
The only other control for Fuzz Unit 1 is an output level control, which adjusts the fuzz signal’s amplitude going to the input of a buffer, which is there simply to serve as an active circuit placed between the diodes and any given succeeding circuit.
Up next is Fuzz Unit 2, which uses diodes in a feedback loop rather than as a shunt on the signal path. It cannot be predicted as to which method of obtaining distortion will sound better than the other. Only building them and testing them will reveal the answer to that question. Yet, both circuits were based on preamps in proven guitar-amps.
On the other hand, Fuzz Unit 2 has more adjustability than Fuzz Unit 1, sporting not only input and output volume controls but separate “Intensity” and “Fuzz” controls. And below is given Fuzz Unit 3, which has no diodes but relies on overdriven op-amps to get distortion, as said. It has four volume controls, labeled in order Drive, OD1, OD2, and Level. [OD = overdrive.] Drive sets the circuit’s input level, OD1 and OD2 control the interstage signal voltages, and the Level control sets output voltage. However, there is also a Feedback control, employed in case the user wants to tame the distortion down while still getting some distortion from the circuit. It is connected backwards so the circuit will sound as if it is getting louder as the control is turned clockwise.
Here are a couple of passive distortion circuits. If built, for best sound they should be driven by one of the Gain Stages. How they work is that the diodes cause clipping of the input signal, though there is some loss due to the diodes actually constituting shunts to ground -- hence the need for an active device driving each. And the best distortion tone, emphasizing even-order harmonics, is obtained when a given diode array has one diode on one side but two on the other. It does not matter which side has the lone diode since what matters is that there is an imbalance. Of course, any setting except when all diodes are disabled will produce distortion. The user can set switches according to their personal taste. In some cases, for instance, harsher distortion, emphasizing odd-order harmonics, may be desired, instead of the sweetness of even-order harmonics.
I recommend using the 3-Band Tone controls at the end of each channel, between the final sub-circuits and the Mixer, so that the tonality of each channel can be adjusted at the inputs of the Mixer, shown below.
The op-amp circuit in the Mixer is configured as a basic summing amp, where each of its inputs has a separate volume control, but there is only one Mix Volume control on the output. This is meant to be the final output of the project and is suitable for the input of any guitar-amp or any other equipment that can accepts a high-Z guitar signal, including another preamp, a stomp-box effect, PA mixing console, recording gear, etc.
Shown below are the pinout diagrams for the ICs used in this project, along with the pop-eliminator circuit, with four required, and which can be placed within the project enclosure or in the remote footswitch array’s enclosure, as desired, as long as a given pop-eliminator is between its footswitch and the Switching IC. There is also a sketch of how to properly connect each IC to its respective supply-voltage leads.
Here is a block diagram indicating the suggested initial connections scheme, though the user can connect individual subcircuits in any sequence or in any channel.
Last but not least is the power supply. This circuit is of my own design but is based on long-standing power-supply topologies. Major components are specified below.
EOF
Last edited:
Some comments:
You seem to have forgotten the need for providing DC-bias for most of the positive opamp inputs.
What is missing is a resistor from ground to positive input (IC1a, all IC4 +inputs, IC5a, IC6a, IC7a, IC8a and IC8b) - without this, the opamp outputs will be rather unpredicable).
The potentiometer at the input amp (100K) is rather small compared with the normal 470K - 1M input impedance usually preferred for passive guitars.
Cheers
Martin
You seem to have forgotten the need for providing DC-bias for most of the positive opamp inputs.
What is missing is a resistor from ground to positive input (IC1a, all IC4 +inputs, IC5a, IC6a, IC7a, IC8a and IC8b) - without this, the opamp outputs will be rather unpredicable).
The potentiometer at the input amp (100K) is rather small compared with the normal 470K - 1M input impedance usually preferred for passive guitars.
Cheers
Martin
Yes. And the 3-band EQ are all in parallel, so the bass filer is filtering the highs and mids as well, and the EQ frequencies depends on the source impedance. If you move the caps to after the series resistors then this has a better chance of working. The "nc" end of the preamp gain pot needs to be connected to the slider and/or a bypass fixed resistor so that DC feedback is never lost when the slider loses contact. The lack of DC blocking caps on the bottom of feedback networks means that DC offset will be amplified.
Let me repeat what Martins said about the need for very high input impedance for electric guitars, ie 1 Meg Ohm, not less. Guitar pickups are massive inductors and a load like 100K or less will kill all the treble.
Also CD4066 inputs cannot be left open. Like the op-amp, they need pull-down resistors.
Let me repeat what Martins said about the need for very high input impedance for electric guitars, ie 1 Meg Ohm, not less. Guitar pickups are massive inductors and a load like 100K or less will kill all the treble.
Also CD4066 inputs cannot be left open. Like the op-amp, they need pull-down resistors.
Last edited:
Thanks for the tips diyMartin. I will make the changes and replace the diagrams with updated versions as soon as I can.
This is very helpful. I was expecting the pots on the inputs to provide bias but I see now that the capacitors are in the wrong places. I have seen pots used for biasing before, so I should either move the caps or else just add some biasing resistors, as you suggested.
And I will look into the 3-band EQ and other issues pointed out by steveu. Rest assured, all said problems will be addressed.
My purpose in posting was to garner interest in my ideas, but some of the diagrams are old, and were not inspected thoroughly enough before being posted. I greatly appreciate everyone's help turning this into a viable project.
Look for updated diagrams to appear soon.
This is very helpful. I was expecting the pots on the inputs to provide bias but I see now that the capacitors are in the wrong places. I have seen pots used for biasing before, so I should either move the caps or else just add some biasing resistors, as you suggested.
And I will look into the 3-band EQ and other issues pointed out by steveu. Rest assured, all said problems will be addressed.
My purpose in posting was to garner interest in my ideas, but some of the diagrams are old, and were not inspected thoroughly enough before being posted. I greatly appreciate everyone's help turning this into a viable project.
Look for updated diagrams to appear soon.
OK. I just replaced most of the previous diagrams with updated versions, after confirming my mistakes and researching how to fix them. However, I have yet to review and restudy the quad bilateral switch. But I'm on it. And thanks again.
Confirmed. The previous diagram for the CD4066B was incorrect. After researching the matter, and finding an appropriate schematic online, I have upgraded the earlier diagram and replaced it with what I believe is a viable one. The entire project should now be getting close to being valid. If there are any more issues, feel free to point them out. And thanks again. You guys rock!
Hello Kurtus,
Looking at your schematics I think that you are not aware of what constitutes the input impedance a connected guitar will see.
Take the first schematic in your post #1. The value of the input level pot is 1M. But right after the coupling cap of 1 uF there is a resistor to ground of 10K. This causes the input impedance a connected guitar will see with the input level pot fully opened to be 1M in parallel with 10K, so only 9901 Ohm, which is way too low.
Looking at your schematics I think that you are not aware of what constitutes the input impedance a connected guitar will see.
Take the first schematic in your post #1. The value of the input level pot is 1M. But right after the coupling cap of 1 uF there is a resistor to ground of 10K. This causes the input impedance a connected guitar will see with the input level pot fully opened to be 1M in parallel with 10K, so only 9901 Ohm, which is way too low.
Last edited:
To Robert,
I see what you mean. I got that value from a guitar preamp found online, but I'm now guessing that such a circuit is off too. My tendency is to agree with you, so I will be changing that spec. However, on another preamp circuit for guitar, a proven design, I have seen that resistor specified as 47k, and on another as 100k. So, what value would you recommend? I'm thinking up to 500k, but I will defer to your experience.
By the way, searching online for guitar preamps using split-supply op-amps, I found that 100k is the most common, but I have also found them up to 470k, thus confirming that my guess at 500k is not too far off. On the other hand, I noticed some schematics that use a 50k or 100k pot there, and no separate biasing resistor at all (thus relying on the pot's element to provide the bias). So, I'm now thinking a range of values between 50k and 500k would work, though I'm not sure what effect a higher or lower value within that range will have on tonality.
Meanwhile, I will be changing that resistor in my diagrams to an in-between value, such as 270k. But I will also look forward to your next reply.
I see what you mean. I got that value from a guitar preamp found online, but I'm now guessing that such a circuit is off too. My tendency is to agree with you, so I will be changing that spec. However, on another preamp circuit for guitar, a proven design, I have seen that resistor specified as 47k, and on another as 100k. So, what value would you recommend? I'm thinking up to 500k, but I will defer to your experience.
By the way, searching online for guitar preamps using split-supply op-amps, I found that 100k is the most common, but I have also found them up to 470k, thus confirming that my guess at 500k is not too far off. On the other hand, I noticed some schematics that use a 50k or 100k pot there, and no separate biasing resistor at all (thus relying on the pot's element to provide the bias). So, I'm now thinking a range of values between 50k and 500k would work, though I'm not sure what effect a higher or lower value within that range will have on tonality.
Meanwhile, I will be changing that resistor in my diagrams to an in-between value, such as 270k. But I will also look forward to your next reply.
Last edited:
I have updated my diagrams again, taking Robert's advice, but only for sub-circuits I designed myself. The input bias resistor values for proven circuits I based two of the Fuzz circuits on I left unchanged from the original schematics (which I referenced in their respective diagrams). Hopefully, my project is getting close to viability, but I am open to any further suggestions.
All the best to everyone who has taken an interest in my design.
All the best to everyone who has taken an interest in my design.
More on the 3 band EQ. If you connect the voltage output of the 3 op-amps to the zero-point inverting (current) input of the 4th op-amp, you will get constant clipping, mostly noise. Theoretically, a zero-impedance output into a zero-impedance input is infinite current, but of course, actual impedances are not quite zero, and op-amps have about a ~10mA limit so it won't burn up, but the gain will be "uncontrolled". You are also connecting op-amp outputs together which is another no-no for similar reasons. You need resistors from each voltage source to the zero-point mixing input. For unity gain in the mixing stage, the input resistors are the same value as the feedback resistor.
Steve U,
I concur. I actually thought of that but did not implement it before posting the diagram you commented on. My bad.
I have redone the diagram according to your advice and replaced the previous one with the updated version. It should work now.
I concur. I actually thought of that but did not implement it before posting the diagram you commented on. My bad.
I have redone the diagram according to your advice and replaced the previous one with the updated version. It should work now.
This post is probably so much of topic that I should post elsewhere - but given my reflection is related to exactly this topic, here goes.
First of all: Thank you, gijser!
I had some odd feelings about the mismatch between the elaborate schematics accompanied by references to prior work and the rather basic level of electronics knowledge of the OT.
Nevertheless I jumped on and tried - as did other members - to help making the design work if implemented.
One fear I nurtured was that my comments would be received badly (as in: "I DID NOT ASK FOR YOUR OPINIONS!!!" 🙂
If this is actually (as I with the comment from gijser suspect to be true) the more or less random output from an AI that is presented as some real persons work to the forum, I feel rather misused and almost as being the involuntary participant in a psycological test.
Maybe I should just use this example as a training exercise in identifying AI generated stuff - in the hope that I next time will have my filters up (just as I have for digital fraud).
Cheers, Martin
First of all: Thank you, gijser!
I had some odd feelings about the mismatch between the elaborate schematics accompanied by references to prior work and the rather basic level of electronics knowledge of the OT.
Nevertheless I jumped on and tried - as did other members - to help making the design work if implemented.
One fear I nurtured was that my comments would be received badly (as in: "I DID NOT ASK FOR YOUR OPINIONS!!!" 🙂
If this is actually (as I with the comment from gijser suspect to be true) the more or less random output from an AI that is presented as some real persons work to the forum, I feel rather misused and almost as being the involuntary participant in a psycological test.
Maybe I should just use this example as a training exercise in identifying AI generated stuff - in the hope that I next time will have my filters up (just as I have for digital fraud).
Cheers, Martin
Why are you guys coming up with this AI stuff? I assure you I am real. My photo is real. And you can read about me at my blog.
https://kurtusrichtersblog.com
I don't understand the reference to AI with respect to my posts. Seems like a phobia to me.
I am just a retiree who now has time to return to studying audio electronics and have been posting old diagrams to see if I could get help making them viable. What makes you think this has anything to do with AI generated junk?
https://kurtusrichtersblog.com
I don't understand the reference to AI with respect to my posts. Seems like a phobia to me.
I am just a retiree who now has time to return to studying audio electronics and have been posting old diagrams to see if I could get help making them viable. What makes you think this has anything to do with AI generated junk?
By the way. my music is real too.
https://audiomack.com/hkurtrichter/album/
And if you think that is AI generated, then there must be something wrong with you.
But you need not worry about me any longer. I don't stay where I'm not welcome. Goodbye.
https://audiomack.com/hkurtrichter/album/
And if you think that is AI generated, then there must be something wrong with you.
But you need not worry about me any longer. I don't stay where I'm not welcome. Goodbye.
Well, I didn't say it was junk. And if I am wrong, I sincerely apologize. Maybe I am getting paranoid... or maybe I am not real myself 😬
Don't.
You are very welcome.
These are troubling times full of potential misunderstandings.
Hello Kurtus,
I fully agree with Netlist that these are troubling times full of potential misunderstandings.
I would hate to see you leave this forum.
And we share love for self-made music and electronics. See:
Soundcloud - Robert Gribnau
Soundcloud - Rude de Dud
OTL800 - Robert Gribnau
But I also have to be honest: You have to realize that electronics is not about 'glueing together' some partly circuits and expect that the result is what you thought it would be.
It takes some time to get to know/understand electronic circuits. It took me more than 10 years to even begin to understand them a little bit. And after some 25 years I am still learning new things all the time.
But I am sure that some of us will need less time than it took/takes me. Perhaps you will be one of them!
Long story short: Please keep posting on this forum. It will make your and our understanding of electronics flourish.
I fully agree with Netlist that these are troubling times full of potential misunderstandings.
I would hate to see you leave this forum.
And we share love for self-made music and electronics. See:
Soundcloud - Robert Gribnau
Soundcloud - Rude de Dud
OTL800 - Robert Gribnau
But I also have to be honest: You have to realize that electronics is not about 'glueing together' some partly circuits and expect that the result is what you thought it would be.
It takes some time to get to know/understand electronic circuits. It took me more than 10 years to even begin to understand them a little bit. And after some 25 years I am still learning new things all the time.
But I am sure that some of us will need less time than it took/takes me. Perhaps you will be one of them!
Long story short: Please keep posting on this forum. It will make your and our understanding of electronics flourish.
Have you taken the Voigt Kampff test?
All kidding aside, I think the OP cited that some of the design elements were from other sources, and that the overall design was something he was trying to improve and ultimately achieve.
The engaging thing overall, was parallel versus sequential analog signal processing. It gives rise to a lot of options.
Personally I don't know what to say about generative AI..or much of what we humans do being iterative. That's maybe a subject for the Lounge Forum.
Cheers