DIY Loudspeakers, How to pick drivers

[hulminator]

As a fairly proficient EE/ME, I'd like to try my hand at building my own loudspeakers (I have the opportunity to contract out the cabinets at a great price). As I'm perusing the various driver suppliers, I notice that there seem to be discrepancies in the way different manufactures take parameters (specifically the frequency response). For example, if I look at similar 10" woofers from SEAS and ScanSpeak, there is a huge difference in bass response. My assumption is that they've measured differently, but I'm then left wondering how to pick drivers that fit my design. Am I just out of luck if I don't have a modeling program I can throw the Thiele/Small parameters into?

 

[lsiberian]

If you are using Windows you can use my designer to generate a box for you. I don't see why you should reinvent the wheel.

As an engineer you should know that specifics are required to make comments. Which drivers specifically are we talking about. Seas and Scanspeak make many drivers.

I recommend you build a 2-Way with a Dayton Reference RS180 and a Dayton RS28A You'd crossover around 1800hz. You can build your own chipamp and active crossover which might be a lot of fun for an EE or use a passive crossover.

 

[TheWarrior]

As an engineer you should know that specifics are required to make comments.

Its like you're sitting up on a hill with a troll trebuchet.... one after the other!

 

[hulminator]

Thanks for the reply. The reason for the lack of specifics is that it was a rather general question. I'm trying to get an idea of the response ranges of the various drivers out there and it would seem that companies measure in very different ways. I don't currently have software i can plug all the values into.

 

[its phillip]

What makes you say they are measuring in different ways? Obviously different drivers are designed for different things and will have different parameters.

 

[hulminator]

When I compare two nearly identical 10" woofers in the same price bracket from two reputable manufacturers, and one rolls off at 100hz while one is flat down to 40, it makes me think they've measured the frequency response differently.

 

[its phillip]

Well, it would appear they aren't nearly identical then...they're designed for different things. What drivers are you comparing?

 

[Dennis Murphy]

When I compare two nearly identical 10" woofers in the same price bracket from two reputable manufacturers, and one rolls off at 100hz while one is flat down to 40, it makes me think they've measured the frequency response differently.

Hi Seas measures on a realistically sized baffle that shows you the baffle step--so there will be a sharp rise from 400 Hz or so on up to 1 kHz, and this could look like a lack of bass. Other manufacturers measure on a much larger "infinite" baffle that will avoid a baffle step rise, and the bass will be much more in line with the rest of the response until the inevitable break up mode, which will occur in very different places depending on the cone material.

 

[Dennis Murphy]

Hi Seas measures on a realistically sized baffle that shows you the baffle step--so there will be a sharp rise from 400 Hz or so on up to 1 kHz, and this could look like a lack of bass. Other manufacturers measure on a much larger "infinite" baffle that will avoid a baffle step rise, and the bass will be much more in line with the rest of the response until the inevitable break up mode, which will occur in very different places depending on the cone material.

I checked the Scanspeak measurements on Maisound, and they do in fact use an infinite baffle to measure their woofers. So the bass response will look much stronger than on the Seas plots, even though they might be quite similar when mounted on a typical baffle.

 

[hulminator]

http://www.scan-speak.dk/datasheet/pdf/26w-8534g00.pdf

H1316-08 CA26RE4X

Similar prices, similar drivers, I was assuming they were meant for similar applications.

 

[Dennis Murphy]

http://www.scan-speak.dk/datasheet/pdf/26w-8534g00.pdf

H1316-08 CA26RE4X

Similar prices, similar drivers, I was assuming they were meant for similar applications.

Well, you would tell whether they're meant for similar applications by examining the Thiele-Small Parameters (in particular Vas, Qts, and sensitivity) and modeling it in box optimization software. You can't really tell that much from a published response curve, except to check for where breakup modes occur. The Scan plot appears to be made with measurement program that transitions from anechoic to room mode as it goes down below 300 Hz. That's why you see all of the peaks and dips in the bass response. The Seas measurement is apparently made outside of a box, so there is bass cancellation. But you can see the predicted box response by the dashed line. I'm not sure exactly what assumptions underlie the dashed-line predicted response

 

[GranteedEV]

As a fairly proficient EE/ME, I'd like to try my hand at building my own loudspeakers (I have the opportunity to contract out the cabinets at a great price). As I'm perusing the various driver suppliers, I notice that there seem to be discrepancies in the way different manufactures take parameters (specifically the frequency response).

I wouldn't even worry about manufacturer plots beyond "can this work in the design?". You will want to measure paramters - both electrical and acoustic, yourself. On that note, I recommend gettng any necessary tools and software now. Some software is free for DIYers, by the way - including Jeff Bagby's PCD.

For example, if I look at similar 10" woofers from SEAS and ScanSpeak, there is a huge difference in bass response. My assumption is that they've measured differently, but I'm then left wondering how to pick drivers that fit my design. Am I just out of luck if I don't have a modeling program I can throw the Thiele/Small parameters into?

Get a modelling program (very basic, but the free WinISD pro is fine) or we can help you out by running it through our own (I use Martin J King's mathcad worksheets).

Your best bet is to start a thread on the DIY forum of audioholics where we can help you out.

Thanks for the reply. The reason for the lack of specifics is that it was a rather general question. I'm trying to get an idea of the response ranges of the various drivers out there and it would seem that companies measure in very different ways. I don't currently have software i can plug all the values into.

Like I said, winISD is free and easy to install.

For what it's worth, there's a lot of factors that mfgr spec is useless for - especially in terms of large signal testing.

As far as the scan and seas woofers go, you can see that they're wholly pretty similar, but the seas has a lower q (and with it a higher sensitivity). So the scanspeak probably goes deeper.

For a 10" Woofer in that mid-price bracket though, you should strongly consider the Dayton RSS270 and the Usher 1001B as well.