New speaker design by NOOB

[abbey road d enfer]

I think of it as autotune, you throw whatever at it, fine tune and it sounds ok - but it loses authenticity.

Then you should revise your thinking. You have shown evidence of your many technical shortcomings, and your dominant trait of character: you don't want to accept opinions that are not in accordance with your preconceived ideas. Good luck with that.

 

[Adam Frandsen]

Then you should revise your thinking. You have shown evidence of your many technical shortcomings, and your dominant trait of character: you don't want to accept opinions that are not in accordance with your preconceived ideas. Good luck with that.

All I said was it feels non authentic to me, and it is just cognitive bias perhaps, but that is my reality. I never said I was a technical or mathematical wizard, I clearly stated I am new to all of the DIY aspects of audio design. I do know I am not interested in binary/switching amplifiers or DSDs for my own build. I never said they cannot sound good, it just is not as aesthetically pleasing to me. Class G and Class H interests me as well as the Eigentakt and GaN, and I would likely be very satisfied with either system, but I likely will go with Class A all the way and passive crossover. The plan is to build my own speakers, preamp, power amp and riaa - may take a while. For the RIAA this: Hifi ECC83 MM RIAA Turntables Ear834 Tube Phono Amplifier Board Pcb | eBay for the preamp this: Balanced BOZ Nelson Pass Preamp PCB Board Set With Power Supply - Custom Class A | eBay, maybe this for the dac: DUAL ES9038PRO DSD DAC OLED Display ES9038 HIFI DAC USB Bluetooth5.0 APTX Player | eBay - I also would like to add that I am very thankful for all the info I am provided with here. My sE5600a MKII repair ended up a great success, and while that was new for me, and some people discouraged me from continuing the project, I managed to find several mistakes and correct them. (With help of course from several users).

 

[ccaudle]

I think of it as autotune, you throw whatever at it, fine tune and it sounds ok - but it loses authenticity.

What does it mean for a speaker crossover to lose "authenticity," and why is a crossover implemented with inductors, resistors and capacitors more "authentic" than a crossover implemented with op-amps, resistors, and capacitors?

 

[Adam Frandsen]

What does it mean for a speaker crossover to lose "authenticity," and why is a crossover implemented with inductors, resistors and capacitors more "authentic" than a crossover implemented with op-amps, resistors, and capacitors?

I was referring to active crossovers with DSPs. I don’t want anything digital or binary in my signal path is all. Authenticity is a personal standard, and they are scientifically proven to have stronger effect on the perception than say a scientific measurement. I still want to make something as good as possible.

 

[Ricardus]

Authenticity is a personal standard, and they are scientifically proven to have stronger effect on the perception than say a scientific measurement. I still want to make something as good as possible.

Ummmm.... WHAT??

 

[Adam Frandsen]

Ummmm.... WHAT??

What part of it was unclear? I don’t mean authenticity in the general sense, but was referring to what I meant by it in my previous post. Authenticity meaning true to the source first of all and not having anything digital or binary in my signal path. I don’t mean for playing digital music of course, but analogue recordings on LP. Now whether or not I can truly detect the difference is completely up to my reality, which is different from everyone elses.

 

[ccaudle]

I was referring to active crossovers with DSPs.

That is fine, which is why Abbey responded with his suggestion for "a simple 2-way electronic x-over," no DSP required and avoids many of the difficulties with passive crossovers.
Highly recommended as good background on considerations for speaker design, active crossover design, general acoustics is the website created by the late Sigfried Linkwitz (still maintained by a friend he personally chose to carry on after his death):
Linkwitz Lab
The non-obviously named "Conversations with Fritz" link on the side is a series of conversations with someone who wanted to build a home design speaker inspired by one of Mr. Linkwitz's designs, and walks through a lot of the acoustic and system design considerations for that.

 

[Adam Frandsen]

That is fine, which is why Abbey responded with his suggestion for "a simple 2-way electronic x-over," no DSP required and avoids many of the difficulties with passive crossovers.
Highly recommended as good background on considerations for speaker design, active crossover design, general acoustics is the website created by the late Sigfried Linkwitz (still maintained by a friend he personally chose to carry on after his death):
Linkwitz Lab
The non-obviously named "Conversations with Fritz" link on the side is a series of conversations with someone who wanted to build a home design speaker inspired by one of Mr. Linkwitz's designs, and walks through a lot of the acoustic and system design considerations for that.

Awesome, and yes, I may have missed that. I am not opposed to that, only I want to have a 4-way speaker with the crossover more or less in the range of 200hz, 900hz-1000hz and 7000-10000hz, and as far as I undertsand that would require 4 amps/4 channels of amplification. I want to use two tweeters, because that is how my brain visualises the best reproduction of those frequencies including soprano high C. The 1100hz crossover was chosen for various reasons that are no longer valid, and it will be lower. I will recalculate everything, probably several times, which is why I seek knowledge here amongst other places. I will choose drivers that allow for some flexibility in terms of extension of a flat frequency response around either side of my crossover points, so small calculation errors will have less effect. I will also buy the UMM-6 or UMIK-2 for measuring the drivers, once they are in place (this is far down the road from now). According to parts express the UMM-6 can measure down to 5hz, which I think will be useful, as I would like my speakers to be able to reproduce lowest notes of say a Bösendorfor Grand (16.5hz) Pipe organ etc..

 

[ccaudle]

s far as I undertsand that would require 4 amps/4 channels of amplification.

For a fully active implementation yes, although the load of a single speaker is easier to drive than a complex multi-way speaker, so the individual amplifier channels can have somewhat reduced requirements compared to a single channel driving all four drivers.
Also re-read Abbey's earlier post:

I've designed several systems with a simple 2-way electronic x-over, where the mid and high used a passive x-over. You get most of the advantages of active x-over and the relatively low cost of passive.

He was suggesting the possibility to use active cross-over for the low frequency driver(s) where the components are largest and most expensive in a passive crossover, but still use passive crossover for the higher frequency drivers.

I want to use two tweeters, because that is how my brain visualises the best reproduction of those frequencies

Mental models should always be guided and refined by actual physical behaviors. The size of the wavelengths at those frequencies means multiple drivers can have a problematic spatial response.

 

[Adam Frandsen]

For a fully active implementation yes, although the load of a single speaker is easier to drive than a complex multi-way speaker, so the individual amplifier channels can have somewhat reduced requirements compared to a single channel driving all four drivers.
Also re-read Abbey's earlier post:


He was suggesting the possibility to use active cross-over for the low frequency driver(s) where the components are largest and most expensive in a passive crossover, but still use passive crossover for the higher frequency drivers.



Mental models should always be guided and refined by actual physical behaviors. The size of the wavelengths at those frequencies means multiple drivers can have a problematic spatial response.

I see that now and apologise for missing the point of his post, that does indeed seem like a good solution to the problem of high cost caps etc., and what he mentioned about parasitic behaviours. I would still need two amps… but if the result is significantly better or rather if it is significantly easier to reach a good result, this may still be the way to go.

 

[RoadrunnerOZ]

This is not super complicated. Grab your favorite free analog simulator and check it out. You can find the relevant parameters in the capacitor datasheets.



As I said before, be suspicious of any advice with no numbers. Sometimes the divided schools of thought are differences between those with experience designing electronics, using simulation, calculation, and test equipment, and people who do not. No need to put much stock in circuit design advice from the latter.

Personally not having been involved in building or designing speaker and crossover systems but only in fixing them when they break, it’s fascinating too get the opinions of those that have - plus looking at the widely varying opinions online in the hifi world. I have designed many recording studios and implemented the installation and setup of the equipment as well as working in them as a recording or mixing engineer.
I have noticed the big difference in PA vs studio monitor systems vs home hifi. This is way more noticeable when listening to well recorded material. Having spent countless hours recording and mixing I’ve noticed that no matter how good the specs are on paper that some speakers just cause listening fatigue - maybe due to the tweeter design.
Having been at the mixing consoles for a Pink Floyd concert (fixing one of the SSL’s which had blown some channels before and as the show started - plus I got to stick around for the rest of the concert on standby) I was super impressed by the depth and quality of the sound - this was a good as any recordings I’ve heard on a top end hifi (including of the same album as the concert).
All the crossovers were electronic and these had been tuned to the system and the concert hall. No hum, no hiss, no feedback with kilowatts of power and a massive array of speakers and amplifiers.
Most active studio monitors use internal electronic crossovers, the subs have variable crossover points to allow matching to the upper range monitors plus what is fed to the upper range. These are tuned to the monitors and the room, as well as the monitors having their own adjustments - trimmed using a reference mic at the listening position and a sweep generator going through the whole record/mix down path or using monitor specific software. A lot of studios use room compensation software to match the whole system tuned to the room as part of the signal path to the monitors. Some active monitors have their own custom software to measure the room effect on the drivers and compensate the EQ and level of the individual stages as well as crossover, level and phase control for subs.
Adam I think if you are building your own speakers and cabs the investment in some electronic crossovers would pay off - then you can bi-amp or tri-amp or however many speakers you want to use and trim the system to its best possible level and EQ and match this to your room. The amp power requirements will vary according to each speaker being driven. You can’t easily do this with passive crossovers.

 

[Adam Frandsen]

That is fine, which is why Abbey responded with his suggestion for "a simple 2-way electronic x-over," no DSP required and avoids many of the difficulties with passive crossovers.
Highly recommended as good background on considerations for speaker design, active crossover design, general acoustics is the website created by the late Sigfried Linkwitz (still maintained by a friend he personally chose to carry on after his death):
Linkwitz Lab
The non-obviously named "Conversations with Fritz" link on the side is a series of conversations with someone who wanted to build a home design speaker inspired by one of Mr. Linkwitz's designs, and walks through a lot of the acoustic and system design considerations for that.

Really interesting content, especially about orchestral recordings and how Linkwitz wanted to design a dip in the 3-4khz to account for the errors in recordings vs real life experience of the sound, due to the ear canals construction. For sure I would want to implement this in my design, incl. a switch for it for different type recordings as suggested. I also understand that both Linkwitz and Harmann has done research on how in general the frequency curve should not be straight because of emphasised bass in the room, caused by I guess reflections etc.. This I would also like to implement, but not sure how it would work together with the equaliser circuit. Julian Krause has a very nice video of how to use the UMIK-1 to equalise speakers, there he provides values for a target frequency curve, that is more true to what it should actually be as opposed to flat. I was also gonna use zobel circuits to do impedance compensation and l-pads to account for the extra sensitivity of the tweeters, I was even gonna do a baffle step compensation of about 3-4db for the bass, but then I end up with so many components that I am not sure if it will be too much or even work well together…

 

[Ricardus]

What part of it was unclear?

Seriously? That paragraph I quoted was completely un-parsable.

 

[Adam Frandsen]

Seriously? That paragraph I quoted was completely un-parsable.

They refers to personal standards

 

[pucho812]

It takes a lot to build a speaker and refine it. I did a few years at westlake when I first started in pro audio. They make some of the finest if not the finest studio monitors on the planet. It requires a lot of disciplines of which you have minimal to zero experience in. Do what you want because you usually do not listen to people who have any experience in the matters at hand. Please don’t cry for help as a noob and then proceed to argue every point that doesn’t match what you want. It’s frustrating to us, No different than when you did the mic. I’m not saying don’t build speakers, but I am saying take the advice of those who have been down those roads, it will make for smoother sailing.

 

[Adam Frandsen]

It takes a lot to build a speaker and refine it. I did a few years at westlake when I first started in pro audio. They make some of the finest if not the finest studio monitors on the planet. It requires a lot of disciplines of which you have minimal to zero experience in. Do what you want because you usually do not listen to people who have any experience in the matters at hand. Please don’t cry for help as a noob and then proceed to argue every point that doesn’t match what you want. It’s frustrating to us, No different than when you did the mic. I’m not saying don’t build speakers, but I am saying take the advice of those who have been down those roads, it will make for smoother sailing.

Give me an example of me not listening, there is a difference between not listening and stating what kind of speaker I want to build. I misread one reply and immediately upon realizing corrected myself.Apart from that I have taken in all information and continue to read about the subject.

 

[pucho812]

Give me an example of me not listening, there is a difference between not listening and stating what kind of speaker I want to build. I misread one reply and immediately upon realizing corrected myself.Apart from that I have taken in all information and continue to read about the subject.

Do you want examples from this thread or in general.
Off the top of my head, a big glaring example is the suggestion of using dsp for the crossover vs making an analog crossover. Dsp is common in modern design. It allows you to really tinker and tailor the crossover to the cabinet and speakers chosen. However you insist upon an analog crossover design without rhyme or reason other than you want it. You seem to be missing theory on how to get there and when helpful tools are suggested they are dismissed.
This is no different then when you insisted on building a new mic cable for the se tube mic when it was advised to confirm you needed a new one or any other number of things that went on in that thread. I still did not hear the mic pre and post work which I would love to hear how it turned out vs being told it turned out.
Other things that were suggested included building a speaker kit. A kit is a great way to get into building speakers but again you scoffed at the idea.
Now you are trying to pick an argument with me purely because I pointed out the obvious.
If you wish to really get into this and go from the ground up with a new design, there are no shortcuts, there is a lot to learn. I think you would best use your time reading books on crossover design and speaker building. We as a whole can recomend you a lot of good books. We can even comment on what designs you bring to the table.
I can say from my westlake days that in their monitors the crossovers were very complex but that is one piece of why those speakers sound so good to most folks.

 

[JohnRoberts]

Loudspeaker design is hard, even for professionals.

One way to learn is to make sawdust and see what doesn't work.

I built my own design speakers back in 70s. I used my DIY speakers for about a decade.

Since then I buy (or trade for) commercial speakers that don't suck.

JR

 

[Adam Frandsen]

Loudspeaker design is hard, even for professionals.

One way to learn is to make sawdust and see what doesn't work.

I built my own design speakers back in 70s. I used my DIY speakers for about a decade.

Since then I buy (or trade for) commercial speakers that don't suck.

JR

Not opposed to that at all. And as I said, I am still reading and learning. I have incredible commercial speakers already, but the fun and joy in having my own, would outweigh any potential loss in sound - I hope. In fact I hope and expect to make a decent build, once I am ready to get started.

 

[JohnRoberts]

Not opposed to that at all. And as I said, I am still reading and learning. I have incredible commercial speakers already, but the fun and joy in having my own, would outweigh any potential loss in sound - I hope. In fact I hope and expect to make a decent build, once I am ready to get started.

Not to repeat myself but speaker design is hard. You may be underestimating that difficulty.

Over my career I have worked with a handful of skilled loudspeaker design engineers, and had the honor of knowing a few giants. Most had decades of experience that informed their design choices.

By all means build something. You will learn from that experience.

JR