DIY Scanspeak tweeter + Purifi midwoofer (active XO)

[Geertidow]

Dear fellow forum members,

Someone I know buids custom speakers. I really like his design and works and the custom subwoofer he has build me sounds excellent (10", only hifi listening).
He has a pair of speakers build using a fancy scanspeak tweeter and purifi midwoofer.
The crossovers are done with a dsp from hypex.

However, I am always reluctant with DIY speakers since bigger factories have so much more room for design, measurements etc.
But also overhead et cetera.

How wrong can you go with a well build (sturdy) cabinet with two top of the shelf drivers (using active xo with hypex DSP)?

Of course I can test them at home and I am highly sufficient using DSP control software.
Furthermore, I am also extremely happy with my Focal Chora 916, but as always I'm intruiged by such projects and I like supporting smaller entrepreneurs.

 

[Penelinfi]

The main things may be tweeter and woofer directivity - you will likely still have the mismatch with a regular tweeter. Doesn't mean it can't sound good, but you may have to test different crossover slopes or work with having different dispersion around the crossover.
There's also port design if using that.

If you enjoy playing around with it and like the look of the speaker then it should do fine

 

[Salt]

What's the meaning of "fancy" regarding the tweeter?

 

[Geertidow]

What's the meaning of "fancy" regarding the tweeter?

That it's a very expensive tweeter. Nothing more.

 

[Sokel]

That it's a very expensive tweeter. Nothing more.

If it's ellipticor,it is fancy (I mean the looks)

 

[Penelinfi]

Aircirc models also seem fancy

 

[ThatSoundsGood]

How wrong can you go with a well build (sturdy) cabinet with two top of the shelf drivers (using active xo with hypex DSP)?

You can always go wrong with just about anything. But, if you know what you're doing, you can build your own speaker for far less than major manufacturers and you can get much better results. Many speaker manufacturers are designing to a price point and using inferior parts and process to meet a certain consumer demand. On the other hand, some manufacturers are selling overpriced speakers with lots of snake oil involved. There are many DIY designs out there that crush a lot of these manufacturers for a fraction. My recommendation for DIY is that you do it to explore and enjoy the process. It definitely takes some experience plus trial and error to get a good product unless you use someone else's proven design. Unless you understand how crossovers work, speaker directivity, enclosure principles, etc. having DSP won't make the process that much easier. It might mask some problems but you still have to understand how it works and how to utilize measurements to get a good result. So, again, you ca go very wrong and you can go very right depending on how much you're willing to put into it.

 

[Jim Taylor]

So, again, you ca[n] go very wrong and you can go very right depending on how much you're willing to put into it.

Excellent description of not just audio, but of everything in life!

Jim

 

[Geertidow]

You can always go wrong with just about anything. But, if you know what you're doing, you can build your own speaker for far less than major manufacturers and you can get much better results. Many speaker manufacturers are designing to a price point and using inferior parts and process to meet a certain consumer demand. On the other hand, some manufacturers are selling overpriced speakers with lots of snake oil involved. There are many DIY designs out there that crush a lot of these manufacturers for a fraction. My recommendation for DIY is that you do it to explore and enjoy the process. It definitely takes some experience plus trial and error to get a good product unless you use someone else's proven design. Unless you understand how crossovers work, speaker directivity, enclosure principles, etc. having DSP won't make the process that much easier. It might mask some problems but you still have to understand how it works and how to utilize measurements to get a good result. So, again, you ca go very wrong and you can go very right depending on how much you're willing to put into it.

Thank you for your advice. I do not have a lot of knowledge on speaker cabinets (enclosure principles) and directivity, the advantage of this speaker is that it's already built. I would not mind experimenting a bit with the DSP on it, so I would have to trust the builder in this regards.

 

[ThatSoundsGood]

Thank you for your advice. I do not have a lot of knowledge on speaker cabinets (enclosure principles) and directivity, the advantage of this speaker is that it's already built. I would not mind experimenting a bit with the DSP on it, so I would have to trust the builder in this regards.

I think that's a perfect place to start experimenting with DSP. A couple of notes: be cautious with the tweeter when you're measuring it without any filter in front of it. Low frequency information can damage many tweeters (especially planar, ribbon and AMT's). Keep your frequency sweeps at 600Hz and above for the tweeter. When doing near field measurements on the woofer, be careful to not place the microphone too close since the woofer's excursion could cause damage to the woofer or microphone. Aside from that, take many measurements. Play with the polarity and different slopes a lot. I use a rotating platform underneath my speakers so I can get off-axis measurements. Have fun with it. It's wild to see the different slopes and how the drivers sum (or don't sum).

 

[Geertidow]

I'll see what I can manage.

 

[kemmler3D]

How wrong can you go with a well build (sturdy) cabinet with two top of the shelf drivers (using active xo with hypex DSP)?

You can't really make total garbage from those components, but you may not make anything so good, either.

Purifi drivers have incredible specs, but they aren't super forgiving when it comes to enclosure volume, you need to be sure things are the right size.

And any cabinet for a good speaker needs good bracing, good seals, good internal damping, etc.

You can play with the crossover a lot using Hypex HFD so you have plenty of opportunity to get that right, but the spacing matters. It's actually best to use a waveguide, because the purifi directivity starts to go off around 2khz at the highest, which often leaves you in a situation where you are doing a trade-off between tweeter distortion and directivity errors in the midrange.

These are the sorts of things that are tricky to sort out, but you only need to do it once per design. This is the advantage of mass-produced speakers.

Measuring speakers is also simple, but not trivial to do right. You need to do it a lot to optimize a design.

Then you have stuff like the Ascilab speakers which are coming out soon, which sort of prove my point. They've done all the hard work already and they don't ask a huge premium in return.

Does your friend have any measurements of his speakers, or does he simply eyeball it? The amount of measuring you do is often the difference between fair, good and great speaker designs.

On the other hand, with HFD you can do wacky stuff like ultra-steep crossovers and FIR phase correction, so you might be able to do "unreasonable" things to correct any design mistakes your friend might have made. It's a pretty flexible system from what I understand.

 

[DWPress]

I would hesitate if the builder hasn't done any measurements themselves to confirm the driver matching for a potentially smooth XO in the cabinet design and decent sensitivity. My speakers are DIY active 3-way XO + 2 subs and over the past 32 years they've evolved with going digital miniDSP 4x10 ten+ years ago (and way beyond that now), different drivers, different angles and tilt, speaker stuffings BUT over all this time they have been UGLY by most standards.

Measurements, a greater understanding how individual speaker drivers interact together and with my room plus some selective room treatment and I'm quite content BUT the midrange driver selection I trialed just a couple weeks ago was never going to work despite what the published specs said. Ugly is beautiful in my case.

 

[Geertidow]

These are helpful responses. I'll stick to stuff I'm comfortable with.

 

[Jukka]

If you pay for work, you can go hugely wrong.

If going custom, I'd say that approve a design that has been simulated properly and extensively (at least frequency response of all drivers individually and combined, counting baffle step and box dimensions, driver excursions, impedance, phase, power handling and whatever you'd look at commercial speaker). Actual measurements will be different from simulation, but if a builder can model and simulate properly, he probably knows what he is doing and chances of success are higher.

 

[Short38]

The Madisound web site has a number of resources for the diy community. Scroll to the bottom of their web site for the list.