Microphone schematics by chat GPT

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
It's not hard to imagine if they are trained on all the great designs from the last 70 years that they come up with some viable, interesting, and sonically great schematics.

It's only a matter of time. I thought this kind of AI explosion wouldn't happen for another 2-3 years; it's really been the last 2-3 months.

One phrase: exponential growth.
If AI can only design by reconfiguring preexisting schematics it failed as AI.
 
It's not hard to imagine if they are trained on all the great designs from the last 70 years that they come up with some viable, interesting, and sonically great schematics.

It's only a matter of time. I thought this kind of AI explosion wouldn't happen for another 2-3 years; it's really been the last 2-3 months.

One phrase: exponential growth.
AI can't even hands or feet convincingly yet, they always come out like some sort of mangled claw mess, so no way in hell AI will be designing functional circuits anytime soon.
 
Tested the schematic…it’s working! Never heard such a warm open sound before! Neumann should have a look. Maybe these AIs dont have the biases human designers have, pun intended..
 
Last edited:
ChatGPT is not AI, it's glorified chat bot. It has done all of it's training on text, and it's doing remarkable job at text generation. I'm sure if there were a similar black box trained exclusively on all the available schematics it would generate quite useable circuits. AI term gets thrown around a lot, media makes big deal about it, but we are not there yet.
 
It's not hard to imagine if they are trained on all the great designs from the last 70 years that they come up with some viable, interesting, and sonically great schematics.

This would only happen if they are given the correct parameters by which to judge what makes a specific "great design" great.

This is the problem with "AI". (I use the term AI in the loosest possible term, these ARE NOT AI in the classic sense, they are knowledge based systems with limited self learning.

They are able to process a lot more information a lot faster than humans, but they lack the ability to filter in ways humans can.

They do learn to filter according the needs of the Humans who use them.

As the people who use these "AI" tend to be non-experts and largely ignorant about the subjects they ask about (why else would they ask in the first place?); there is inherent feedback loop that will promote generic "0815" (it is german term denoting ubiquity and unoriginality - it comes from the german WW1 0815 Machine Gun).

In fact, once AI can filter like humans AND extrapolate from the filtered results to genuinely new ideas, we Humans are obsolete and will be placed as batteries in the Matrix. If that has not already had.

One phrase: exponential growth.

Exponential growth of garbage.

Because:

Garbage in -> No Filtering -> Garbage out -> Human looking at the result cannot tell garbage from real resources -> Filter Weighting is pushed towards garbage -> return to beginning.

This increase of Garbage is great for areas that already are pure and unadulterated Garbage to start with. Like the new RuZZian Foreign Policy, which is clearly ChatGPT generated.

Thor
 
AI will develop working circuits with ease.

Real AI, that can pass a Turning Test and operate similar to a real human in handling information will be able to.

Current AI are kind of like mainland chinese engineers. They can find a working circuit in a publication and can copy this circuit, they will even be able to make an auto routed PCB and place an order with JLCPCB and get a working PCB.

And for 99% of "Designs" this will probably be good enough.

Thor
 
Real AI, that can pass a Turning Test and operate similar to a real human in handling information will be able to.

Current AI are kind of like mainland chinese engineers. They can find a working circuit in a publication and can copy this circuit, they will even be able to make an auto routed PCB and place an order with JLCPCB and get a working PCB.

And for 99% of "Designs" this will probably be good enough.

Thor
And AI will use an intentionally flawed electrolytic capacitor design to make capacitors that are used by the zillions and fail. Oh my, that's already occurred without AI but the flawed electrolytic capacitor design is in the database AI uses to design electrolytic capacitors. What goes around comes around, karma.
 
Looks like ChatGPT is really bad at drawing schematics, I asked to show me a tube based voltage to current convertor. It drew this enormous unreadable ascii schematic with a rectangle for the the tube. I told it that its schematic was unreadable so it drew another which was identical to the first. I aked it to draw a non-test schematic and it drew the same thing again. I told it it had drawn the same text one again and I wanted a non-test one so it drew me a pic and posted a link to it at Imgur but the link was 404. IMHO it's pretty dumb. Artificial yes, intelligent no.

Cheers

Ian
 
IMHO it's pretty dumb. Artificial yes, intelligent no.

Cheers

Ian
yeah, until you ask it to tell you your license plate and it gives it to you, along with your address and your underwear brand and size..... not so dumb then....
 
I've been reviewing chatGPT for writing software and other tasks, such as reviewing manuals, and it has been REALLY impressive. I believe it is going to absolutely revolutionize a lot of aspects of how people work, learn, and interact with technology. It's improving rapidly - just from the previous released versions (3.5->4).
One thing I really don't like about it is the overconfidence it displays, or a bias to trying to sound right. When it is completely wrong, but it doesn't give any clues (ie 'I'm not sure but I'll take a guess'), it REALLY loses credibility, if you've asked it something you have expertise in.
There are a lot of easy tasks it can already do but it will also fail at complex tasks. I described it to someone as having the intelligence of a child with the ability to converse well and instantly read and digest (and parrot) information.
So on to the topic of hand, I asked it to review the schematic of a U67, and was unimpressed with it. First without giving it a link to the correct schematic it started talking about a FET and then gave a simplified schematic that wasn't a U67.

With the correct schematic to review, it did this. Able to explain some simple things, but not the hard things.

ChatGPT:
The Neumann U67 is a condenser microphone that was first introduced in the 1960s and has since become a classic microphone in the recording industry. The schematic you provided appears to be an accurate representation of the internal circuitry of the U67.

Here is a breakdown of the different components and how they work:

Capsule: The U67 uses a dual-diaphragm capsule, which consists of two metal plates that vibrate in response to sound waves. These vibrations are converted into an electrical signal, which is then sent to the amplifier circuit.

Impedance Converter: The signal from the capsule is first sent through an impedance converter, which matches the high impedance of the capsule to the low impedance of the amplifier. This helps to ensure that the signal is not degraded or lost as it travels through the circuit.

Amplifier: The U67 uses a tube amplifier to boost the signal from the capsule. The tube provides a warm, natural sound that is highly sought after in recording.

Polar Pattern Selector: The U67 has a switch that allows the user to select between three different polar patterns: omni-directional, cardioid, and figure-eight. This switch changes the way the sound is picked up by the capsule, and can be used to achieve different recording effects.

Low Cut Filter: The U67 also has a switch that activates a low cut filter, which reduces low-frequency noise and rumble. This is useful when recording in noisy environments or when dealing with low-frequency sources such as wind or traffic noise.

Output Transformer: The final stage of the U67's circuit is an output transformer, which matches the low impedance of the amplifier to the higher impedance of the output cable. The transformer also adds some coloration to the sound, which contributes to the U67's unique character.

Overall, the Neumann U67 is a highly sophisticated and well-designed microphone that has stood the test of time. Its combination of a dual-diaphragm capsule, tube amplifier, and versatile polar patterns and filters make it a popular choice for recording everything from vocals to acoustic instruments.


ME:
What does the tertiary winding of the output transformer do?

ChatGPT:
The tertiary winding of the output transformer in the Neumann U67 microphone is used to provide additional tonal shaping and coloration to the sound.

The primary function of an output transformer is to match the low impedance of the amplifier circuit to the higher impedance of the output cable, allowing the microphone signal to be sent to a recording device or mixing console. In the case of the U67, the output transformer also adds some characteristic coloration to the sound, which contributes to the microphone's unique sonic character.

The tertiary winding is an additional winding on the output transformer that is connected to a separate set of contacts on the microphone's output connector. This allows the user to switch between different levels of output gain, which affects the tonal balance and coloration of the signal.

In essence, the tertiary winding is another way that the U67's designers allowed for tonal shaping and customization of the microphone's sound. By switching between different gain levels, the user can dial in the ideal amount of coloration for their specific recording needs.

ME:
No, the tertiary winding is used to provide negative feedback to the amplifier

ChatGPT:
You are correct, I apologize for the confusion in my previous response. The tertiary winding of the output transformer in the Neumann U67 microphone is indeed used to provide negative feedback to the amplifier circuit.

Negative feedback is a technique used in electronic circuits to improve the overall performance of the circuit by reducing distortion and increasing stability. In the case of the U67, the tertiary winding is connected to a feedback network that provides a portion of the output signal to the input of the amplifier circuit. This feedback signal is out of phase with the original signal, which causes the amplifier to reduce its gain in a specific frequency range. This results in a flatter frequency response and reduced harmonic distortion, which can lead to a more natural and transparent sound.

The use of negative feedback in the U67 is one of the factors that contributes to its well-regarded sound quality and versatility. By adjusting the amount of negative feedback, the user can fine-tune the microphone's response to suit a variety of recording scenarios and achieve the desired sonic character.


ME:
Why didn't you know that?

ChatGPT:
As an AI language model, I do not "know" things in the way humans do. My responses are generated based on patterns in the data that I was trained on. However, I am designed to be able to learn and adapt to new information, so I can improve my responses over time as I receive feedback and corrections.

In this case, my previous response was incorrect, and I appreciate you bringing it to my attention. Your correction has helped me to improve my understanding of the Neumann U67 microphone and its circuitry.

It does much better with coding.
 
And AI will use an intentionally flawed electrolytic capacitor design to make capacitors that are used by the zillions and fail. Oh my, that's already occurred without AI but the flawed electrolytic capacitor design is in the database AI uses to design electrolytic capacitors. What goes around comes around, karma.
Can you be a little more specific about it. You seem to have privileged information...
 
And this is interesting - with prompts it gets to the right answer. But for some reason it doesn't try to figure it out until pushed

ME:
In that schematic what voltage rating should the capacitors have?

ChatGPT:
In the schematic of the Neumann U67 microphone, there are several capacitors shown with different values and types, so the voltage ratings of the capacitors will vary depending on the specific capacitor and its location in the circuit.

Generally speaking, the voltage rating of a capacitor should be higher than the maximum voltage it will be exposed to in the circuit. This is to prevent the capacitor from breaking down and failing due to excessive voltage.

To determine the appropriate voltage rating for a capacitor in the U67 circuit, you would need to look at the maximum voltage levels present in the circuit and choose a capacitor with a voltage rating higher than that level. The maximum voltage levels will depend on various factors, such as the operating voltage of the power supply, the voltage swings in the amplifier stages, and any voltage spikes or transients that may occur in the circuit.

In general, it's a good idea to choose capacitors with a voltage rating at least 1.5 times higher than the maximum voltage they will be exposed to in the circuit, to provide a safety margin and ensure reliable operation.

ME:
You do not know the answer?

ChatGPT:
I apologize if my previous response was not helpful. In the schematic of the Neumann U67 microphone, the voltage ratings of the capacitors are not specified. However, the voltage rating of a capacitor is typically chosen based on the maximum voltage that the capacitor will be exposed to in the circuit.

To determine the appropriate voltage rating for a capacitor in the U67 circuit, you would need to consider the maximum voltage levels present in the circuit and choose a capacitor with a voltage rating higher than that level. The maximum voltage levels will depend on various factors, such as the operating voltage of the power supply, the voltage swings in the amplifier stages, and any voltage spikes or transients that may occur in the circuit.

In general, it's a good idea to choose capacitors with a voltage rating at least 1.5 times higher than the maximum voltage they will be exposed to in the circuit, to provide a safety margin and ensure reliable operation.

Me:
The schematic shows the maximum voltage for the B+ voltage and the heater voltage. You can answer my question by looking at the schematic

ChatGPT:
I apologize for the confusion in my previous responses. I have reviewed the schematic of the Neumann U67 microphone and can provide more specific information on the voltage ratings of the capacitors shown.

The B+ voltage in the schematic is shown as +250V, which indicates that any capacitors connected to this voltage rail should have a voltage rating higher than 250V. Similarly, the heater voltage in the schematic is shown as +6.3V, which indicates that any capacitors connected to this voltage rail should have a voltage rating higher than 6.3V.

However, it's worth noting that the voltage rating of a capacitor should not just meet the maximum voltage in the circuit, but should be chosen with a safety margin to account for any voltage spikes, transients, or other unpredictable events that may occur in the circuit. A good rule of thumb is to choose a capacitor with a voltage rating at least 1.5 times higher than the maximum voltage it will be exposed to in the circuit.

In summary, based on the schematic of the Neumann U67 microphone, any capacitors connected to the +250V B+ voltage rail should have a voltage rating higher than 250V, and any capacitors connected to the +6.3V heater voltage rail should have a voltage rating higher than 6.3V. However, it's recommended to choose capacitors with a higher voltage rating to provide a safety margin and ensure reliable operation.
 
About AI or AGI, maybe it's fun to talk about, but get real - chatGPT is estimated to have 1 trillion parameters - impressive - but the human brain has 86 BILLION neurons. There is a limiting big O() of training. And human knowledge still hasn't ever been able to make or really understand computational automata (the computational architecture of the brain). It's fascinating that the most incredible technology is so poorly understood by us

It is another development in technology that will be distruptive and amazing, but not wise or intelligent the way humans are. It has clearly been trained to put on a show of it for us though. It uses conversational signifiers in discourse to make us see it as AI.
 
Over the years I have played with a few chatbots and chatGPT is definitely the best of the lot. That said, its scope seems to be rather limited and it demurs when asked for an opinion. Others I have tested are often willing to offer opinions and holding opinions is one marker of intelligence.
I asked it how to perform a DOS attack on quantum crypto (BB84 protocol) and it told me that is illegal. I pointed out that I did not ask it to perform an attack, just how it could be done. It then explained that BB84 is able to detect tampering with the message so it would be impossible to decode the message. Once again, I stated that I did not want to decode the message, merely prevent its delivery. It fell back on a restatement that the message could not be decoded.
The interesting thing, to me, is that it did not answer the question I asked. I have gotten bad answers from other chatbots, but they never answered a different question than the one asked.
It's got a way to go.
 
Come now, fellers ... this is nothing really new -- they have been using artificial intelligence of one kind or another in government for years !! :) JHR
 
Back
Top