Hello,
My first post after many years of lurking. I have been sketching up a pair of speakers with the drivers and come pretty far on how I would like to have it setup. Purifi 10" low(s) and bliesma mid / hi, sealed cab. However I would like to have the crossover part fully analog and no dsp on the signal path. Basicly I seem to have re-sketched MUM-10 speaker but without the dsp and hypex amps...
Both the skill and learning time needed to have proper crossover is way over my resources so I was wondering if there are any services available to design an crossover for a specific set of drivers in a specific cabinet.
My first post after many years of lurking. I have been sketching up a pair of speakers with the drivers and come pretty far on how I would like to have it setup. Purifi 10" low(s) and bliesma mid / hi, sealed cab. However I would like to have the crossover part fully analog and no dsp on the signal path. Basicly I seem to have re-sketched MUM-10 speaker but without the dsp and hypex amps...
Both the skill and learning time needed to have proper crossover is way over my resources so I was wondering if there are any services available to design an crossover for a specific set of drivers in a specific cabinet.
Last edited:
Sorry I dont understand your question. I clarified my post a bit. Basicly looking to commission for a design and possibly manufacturing an analog crossover for a certain set of drivers in a specific cabinet. Hope I make more sense now 🙂
Last edited:
Ideally you'll need to provide in-box impedance & on-baffle (preferably on & a range of off axis) frequency / phase responses to whoever you commission a design from. A purely theoretical design off processed data from the data sheets is possible if you can give full dimensions & box loading details, & although it's obviously not ideal, it's a lot better than simple textbook values & can often be surprisingly good with reliable / consistent drivers -my late friend Colin & I did a few experiments out of interest along those lines & within reason, results were quite decent.
Assuming you want to proceed on a commercial basis, Tony at www.humblehomemadehifi.com offers services like this, as does www.solen.ce -although it's not on his site, I believe Stefan at www.kjfaudio.com is intending to do something similar too, and I'm sure there are others that I don't know about. All three also provide construction too, AFAIK. Madisound used to provide similar design services, but if memory serves, they had to stop when a member of staff retired. A
Assuming you want to proceed on a commercial basis, Tony at www.humblehomemadehifi.com offers services like this, as does www.solen.ce -although it's not on his site, I believe Stefan at www.kjfaudio.com is intending to do something similar too, and I'm sure there are others that I don't know about. All three also provide construction too, AFAIK. Madisound used to provide similar design services, but if memory serves, they had to stop when a member of staff retired. A
Last edited:
Thank you very much. Other idea I have is to do the initial configuration in DSP and then apply that data to the analog design. Dont know if its totally dumb thinking or could it work to a certain extent.
Oh geez. I just built my own custom crossover for my Satori drivers this year. It did not take long to learn the software.
I did find that Boxsim from Visaton was easier to learn than Vituaix. Mostly because I couldnt figure out how to model baffle step reduction in Vituaix. It is pretty easy in Boxsim
1) download the Frd/zma files from Purifi for your selected woofer.
2) download the freq response sheet from Bliesma for mid and tweeter, and use a converter to make the FRD/ZMA files
3) download boxsim, and ask for a bit of help setting up your project in it. You create a project with 3 drivers. Edit the T/S parameters for each. And load those FRD and ZMA files. Set the drivers to have been measured on a DIN baffle. (And click the box on the tweeter for it being in its own closed enclosure ). And Boxsim will automatically calculate baffle step reduction.
4) build a crossover. You can start with a simple 2nd order crossover on every driver. You can also set up your box dimensions and driver placement in Boxsim.
4) build and listen. Tweek the Crossover. Listen more. Repeat for a couple months as months (as I have found).
It helps to have an external crossover in a box right next to each speaker. Much easier to tweek.
I did find that Boxsim from Visaton was easier to learn than Vituaix. Mostly because I couldnt figure out how to model baffle step reduction in Vituaix. It is pretty easy in Boxsim
1) download the Frd/zma files from Purifi for your selected woofer.
2) download the freq response sheet from Bliesma for mid and tweeter, and use a converter to make the FRD/ZMA files
3) download boxsim, and ask for a bit of help setting up your project in it. You create a project with 3 drivers. Edit the T/S parameters for each. And load those FRD and ZMA files. Set the drivers to have been measured on a DIN baffle. (And click the box on the tweeter for it being in its own closed enclosure ). And Boxsim will automatically calculate baffle step reduction.
4) build a crossover. You can start with a simple 2nd order crossover on every driver. You can also set up your box dimensions and driver placement in Boxsim.
4) build and listen. Tweek the Crossover. Listen more. Repeat for a couple months as months (as I have found).
It helps to have an external crossover in a box right next to each speaker. Much easier to tweek.
Trust me, I would love to IF I had the time learn and do it. However I dont and I know its one of these rabbitholes that I cannot afford to get sucked into right now in my life 😀
From the HH Hifi website on Crossover Design:
"At Humble Homemade Hifi we offer all stages of design work in regards to loudspeakers crossovers.
This can vary from basic simulation only based crossovers to full custom design using real-life measurements...
For this basic crossover design we charge EUR. 130,- excluding VAT. This price is for a design with a maximum
of three drivers.
Please note that a crossover design based on simulations only won't give as good results as one based on
real-life measurements of the actual loudspeakers.
A full crossover design on the actual speakers is an extremely time consuming process that not only involves
lots of measuring and simulating but also building prototype crossovers and extensive listening sessions.
This is of course more expensive and involves having the actual speakers here on site. "
End of quote.
I wonder what would the underlined sentence parts cost actually and would you have to specify driver
brand and model prior to estimating it?
edit: let us not forget, here, forum members used to provide that very simulation service to many, free of charge. 🙂
"At Humble Homemade Hifi we offer all stages of design work in regards to loudspeakers crossovers.
This can vary from basic simulation only based crossovers to full custom design using real-life measurements...
For this basic crossover design we charge EUR. 130,- excluding VAT. This price is for a design with a maximum
of three drivers.
Please note that a crossover design based on simulations only won't give as good results as one based on
real-life measurements of the actual loudspeakers.
A full crossover design on the actual speakers is an extremely time consuming process that not only involves
lots of measuring and simulating but also building prototype crossovers and extensive listening sessions.
This is of course more expensive and involves having the actual speakers here on site. "
End of quote.
I wonder what would the underlined sentence parts cost actually and would you have to specify driver
brand and model prior to estimating it?
edit: let us not forget, here, forum members used to provide that very simulation service to many, free of charge. 🙂
Absolutely, although I got the impression it was more a question aimed at a commercial scope.
Re the underlined parts quoted from Tony -no doubt quite a lot, but it will presumably vary on a case by case basis depending on the specifics of the design, & also assumes that the built speakers (sans filter) are physically there with him to work with. Distance design would clearly be simulation only, either with processed manufacturer or 3rd party driver data, or (better, obviously) with quality measurements supplied to him, and having an approximate idea of the rest of the system / room & client preferences / objectives. I assume that would also be the case for Solen, KJF etc., in a similar way to how it would be here or on other forums -likely just on a more personalised / detailed basis.
Re the underlined parts quoted from Tony -no doubt quite a lot, but it will presumably vary on a case by case basis depending on the specifics of the design, & also assumes that the built speakers (sans filter) are physically there with him to work with. Distance design would clearly be simulation only, either with processed manufacturer or 3rd party driver data, or (better, obviously) with quality measurements supplied to him, and having an approximate idea of the rest of the system / room & client preferences / objectives. I assume that would also be the case for Solen, KJF etc., in a similar way to how it would be here or on other forums -likely just on a more personalised / detailed basis.
Last edited:
Interesting, Looked it up found this.
Finally something straight forward that makes sense. Is the bliesma mid/hi drivers pictured? or would it just follow the same arrangement.
Center to Center is tightly arranged and are offset on the baffle for flattest response. Pretty good way to do it, or the way to do it.
If your concept includes or if your ok with it " looking" that way. More than ideal and suggest it for relatively painless design.
Look pretty fun to me.
Load up your drawings/concept and list the drivers model/ part numbers.
Long as the baffle size and driver positions are confirmed or eventually confirmed
something accurate is possible.
Check out also the MUM-10MB which is to me even better looking a bit more compact as it only operates in one orientation. MB comes from 'Meter Bridge' as its designed to be mounted on top of a mixer or a desk rack. That was very close to my initial sketches. I aesthetically prefer horizontal speakers.
I'm going to disagree with the poster @WhiteDragon above, but confirming what @Lojzek posted earlier:
In order to properly design a crossover it is best to have multiple measurements of each driver taken after the drivers have been mounted in the final cabinet. Without measurements, the crossover will be OK or possibly even "good" but will probably not be as good as it can be. You will never be able to tweak the crossover to work out the frequency response irregularities and optimize the sound. Simulations are only an approximation, unless you have some serious software tools. With very good drivers like you have chosen it would be better to get actual measurements using a high quality calibrated microphone, performed by someone who has more experience in that sort of thing.
Yes. Like many other things in life, you can always find someone who will tell you what you want to hear so they can take your money!
Maybe an alternate way would be to try and find another DIY person near you who can help you out?
Agreed. Although in fairness to Tony, Solen, Madisound, KJF etc., they're all competent professionals & I doubt any of them are likely to tell people what they want to hear. If anybody tries to railroad a client in a specific direction without any basis or explanation, then they would be well advised to ask for that, and if they're not happy with the response, to take their business elsewhere. As I suggested above, if for whatever reason they're having to work distance, they're much more likely (like, say, yourself & most members here) to spend some effort getting as much context as possible can about client requirements, system etc., and discuss with them in advance possible options, different compromises etc., and only then, when the client is happy with the chosen direction, actually proceed with any design work. Preferably with a good set of in-box measured impedance, and on / off axis frequency & phase responses, & if those aren't available, then with the best possible data from manufacturers & reliable 3rd parties, processed with the best possible tools to a reasonable approximation -it being made clear (as Tony does) up front that the filter design is just that, and more limited in outright performance compared to one done with a comprehensive measured data set.
Last edited:
The most likely partner as they have 80% of the available brands to develop your dream speakers. But who knows that it suits you in both price and preference?
https://www.lautsprechershop.de/intro_en.htm
https://www.lautsprechershop.de/intro_en.htm
Thanks for the input, much appreciated. Measurements are not a problem se as I have good contacts and people on that side of the field. As you state later, what is missing is a person who has both experience and skill to read the measurements and adopt my personal preferences to the analog design.
One of the main things I would like to accomplish is the very tight phase allignment Ive gotten used to with my PSI Audio 25's. Theyre are absolutely sublime on that department. This I believe takes a lot of skill to get right in analog domain.
Yup and computer aided design and measurements make awareness and alignment of phase
far more intuitive. Aside from physical alignment, which isn't being requested.
The drivers behavior or phase will be what it is, filter order plays a role in that adjustment.
Pretty normal straight forward stuff. If there is a fight with phase. There is the usual tricks.
Many times the drivers work rather well, magic lol. If not then yes the experience plays a role.
Again the design is so straight forward, it was a relief too see. Doesn't look like a fighter.
far more intuitive. Aside from physical alignment, which isn't being requested.
The drivers behavior or phase will be what it is, filter order plays a role in that adjustment.
Pretty normal straight forward stuff. If there is a fight with phase. There is the usual tricks.
Many times the drivers work rather well, magic lol. If not then yes the experience plays a role.
Again the design is so straight forward, it was a relief too see. Doesn't look like a fighter.