6sX7
Thank you for your question.I understand that a 3.5-way design would indeed be simpler and would provide strong low-end support with easier implementation.
In my case, I chose a 4-way structure because I wanted more precise control over the lower midrange and midrange regions.
By fully splitting the duties between the woofer, midwoofer, and midrange, I aim to achieve cleaner separation, tighter imaging, and a more defined soundstage — which are important goals for me in this project.
That said, I fully agree that a 3.5-way approach has many advantages in terms of simplicity and bass reinforcement.
Depending on the results of real-world testing, I am open to experimenting with a 3.5-way setup later if it turns out to better suit the overall system balance.
I truly appreciate your suggestion and it helps me think through the project from multiple angles!
How would you be able to judge that without having knowledge yourself?
AI is incredibly good at making you believe what it says even if it's nonsense.
I'm not trying to be negative, just save you from blindly believing AI.
Yes, my current design is actually a true 4-way system.
Each driver covers a distinct frequency band without significant overlap:
I fully understand that a 3.5-way system would simplify the crossover and boost bass headroom, but for this project, a
chieving maximum detail and clean transitions between drivers is my main goal.
I really appreciate your suggestion though — it definitely made me think carefully about different design paths!
Each driver covers a distinct frequency band without significant overlap:
- The woofer handles below ~370 Hz,
- The midwoofer covers ~390 Hz to ~707 Hz,
- The midrange handles ~681 Hz to ~2.8 kHz,
- And the tweeter takes over from ~2.8 kHz upward.
I fully understand that a 3.5-way system would simplify the crossover and boost bass headroom, but for this project, a
I really appreciate your suggestion though — it definitely made me think carefully about different design paths!
"It was able to make me believe that" 😊
A 4 way with a single 7" bass driver (that is identical to the midbass driver, by the way) seems strange to me.
Anyway, good luck!
If I took money while I'm supposed to be doing other work, that would be moonlighting. 😉
If you have access to additional amplifiers or design resources allowing you to do a mix of active / passive design, I would highly recommend a holistic approach, where the upstream signal chain is also accounted for. The sketch I posted earlier should be quite flexible as a rough starting point.
E.g.: a class-D amplifier could be used for the woofers, and a capacitor-coupled class-A for the mid-range and tweeter. One neat thing is that there are many amplifier designs (check out the Pass Labs sub-forum) that use an output capacitor by default. The output capacitor is usually sized to pass 20Hz, but you could use higher quality polypropylene caps because the minimum frequency is much higher, in the case of the C90+C25.
Even a simpler 'bi-amping' setup without any active filtering could be beneficial: the class-D amp is not loaded by the mid+tweeter, and the class-A is not loaded by 4-ohm at bass frequencies.
Not all AI systems inherently understand audio design.
Some degree of training or guidance is needed to help them interpret and apply audio concepts correctly.
Some degree of training or guidance is needed to help them interpret and apply audio concepts correctly.
Thank you for your honest feedback and encouragement! 😊
I completely understand your point — using two identical 7" drivers in a 4-way layout is definitely an unconventional approach.
My idea was to split the load carefully:
- One driver focuses purely on deep bass,
- While the other helps reinforce the low-mid region without overextending either one.
Thanks again, and I hope to share more updates as the project develops!
Thank you for the great insights and suggestions! 😊
I hadn't considered a hybrid active/passive approach seriously before, but now I see the advantages you mentioned — especially for optimizing amplifier loading and maximizing overall system performance.
Using a class-D amplifier for the woofers and a class-A (possibly capacitor-coupled) amplifier for the midrange and tweeter sounds like a very promising idea.
I will definitely explore bi-amping possibilities as the project progresses, and I really appreciate you pointing me towards the Pass Labs designs as well.
Thanks again for sharing your experience — it’s really inspiring for someone like me who's still learning!
@CrossoverCrafter I don't mean to be rude, but are you an AI bot? Surely only an AI would have such naive faith in AI, and such a relentlessly sunny disposition when errors are pointed out.
It seems almost like someone is feeding our answers to the AI and then returns the AI comebacks here ;-)
Maybe its our first lone AI member - it went here and registered to be able to check some questions it got from someone...
//
Maybe its our first lone AI member - it went here and registered to be able to check some questions it got from someone...
//
Are this loudspeaker DIY?
Useless. Big time.
Useless.
Wishful thinking.
I am strongly suggesting to you to ditch AI and learn the basics in a old -fashioned way, and after that you can make measurements and design crossovers.
Sorry, I'm getting a little help from AI 😅
Audio design is a lot harder than I thought...
Last edited by a moderator:
This was the part I was curious about, and now I’ve found the answer.
The parts marked in red boxes are circuits that were suggested by AI (ChatGPT) during the design discussion.
These elements (mostly additional capacitors and Zobel networks) were not present in the original crossover design, but were proposed by AI to improve impedance stability, frequency response, and phase alignment.
I am still learning and carefully considering each suggestion before final implementation.
Your expert feedback would be greatly appreciated to verify whether these additions make sense in a real-world application.
Let's calm down a little.
I simply wanted to get some advice from experienced people to check whether the AI’s suggestions were correct or not.
I apologize if there was any misunderstanding.
For now, I think it’s better to assume that the AI might not have been fully correct.
Thank you very much for your feedback — I really appreciate it.
I will explore different approaches based on your advice, and if I need further help in the future, I’ll make sure to reach out.
I simply wanted to get some advice from experienced people to check whether the AI’s suggestions were correct or not.
I apologize if there was any misunderstanding.
For now, I think it’s better to assume that the AI might not have been fully correct.
Thank you very much for your feedback — I really appreciate it.
I will explore different approaches based on your advice, and if I need further help in the future, I’ll make sure to reach out.
Are the frequency measured or calculated. Can you show the measurements, so we can see slopes etc. 390-707 is a very narrow band. I don't think you gain anything. Use both woofers to cover same range, or as 3½ way. Using one woofer for 390-707 will not clear up anything imo. You can use Vituixcad to see how you schematic works in close to real life simulation
Last edited:
Sorry about that, but as we have mentioned it was clearly apparent to us that it shouldn't be possible to put emphasis on smooth response and phase under these conditions with the limited information that was used. While many of us could make a reasonable estimate as a starting point, we know it would take time and testing to become confident.