Thanks Andy. I understand that, although I wasn't really very clear, I meant to ask more from the point of view of drivers that will work with the SPL requirements. For example, let's say hypothetically, if I was working with an SPL requirement of say 89dB (e.g. I'd found two 4 ohm 89dB woofers that I planned to put in series), then would the Audax still be a good contender, or should I look at other drivers around the same money instead?
Yeah, that does seem to be a problem, although I can still get hold of one Audax HM100Z0, that said, if I accidentally damaged it, then that could mess things up a fair bit.
That is a good point. These days, I write software and I'm kind of thinking of writing a bit, that given a drivers FR calculates the optimum baffle dimensions and shape within a set range of dimensions... but there is probably something out there that does that already.
I take your point and have noticed drivers that look like they have a bit of a lumpy response, and yet people say they are great. That said, I'm curious, with that particular driver it looks nice up to around 400Hz, would people normally expect to cross it higher than that? If so, would a wide notch filter at around 1500Hz be the thing to use to tame the bump and then cross at a higher frequency? Or would it be better to apply a 1st order low pass filter at 400Hz using the natural rise in the woofers response to cancel out the drop off from the filter so that the mid can be crossed higher?
For a center you want the best horizontal dispersion you can get so everyone on the couch gets the same sound. Around the crossover point you get lobing (dispersion getting smaller) due to 2 drivers producing the same frequency and interfering with each other. For a horizontal MTM this is a big problem since you will get horizontal lobing, in order to minimise this effect the center to center distance between tweeter and mid should be smaller than 1/4 the crossover wavelength (almost impossible with normal drivers). For a W(TM)W with tweeter on top of the mid you sidestep this problem by making vertical lobing which is not as big of a problem. If you put a small faceplate tweeter as close as possible on top of a 4" mid the height will be similar to a 6.5" woofer so you minimise the height of the center.
Hmmm... I started to write a similar response to earlier ones but deleted it. If I may be blunt, you have set yourself the task of designing a 3 way centre channel (nice challenge - no problems) but seem to be trying to decide the details of which drivers to use before you have put together the overall design or wrapped your head around what driver parameters are important to the design.
What is the overall design to be? Are you going to offload the bass to the mains and/or subwoofers or do you want deep bass from the centre speaker? This affects the size of the woofers and their parameters, whether the speaker is sealed or ported and how much height you have for the tweeter + midrange. How important is it that the centre can match future main speakers? This might mean for example, the same tweeter and midrange but 2 x 8" woofers with a port for the mains and 2 x 6.5" sealed for the centre with the deeper bass handed over to the mains and/or subs.
How to wrap your head around what is important? I would suggest designing a centre speaker on paper using standard drivers from, say, the SEAS or SB Acoustics ranges. Put aside details, just learn what fits together and what does not and why. This would mean something like:
- driver performance in ported and sealed cabinets -> WinISD or PCD
- baffle diffraction -> Edge or PCD
- rough crossover -> PCD, WinPCD, XSim
It is not a huge task but it will take a few evenings.
Another approach might be to dive in and build something ASAP and learn from your mistakes in which case opting for £200+ drivers seems rather unwise. SB Acoustics for example make high value for money drivers in £15-40 range: £17 tweeter, £27 midrange, 2 x £31 midwoofers = £106 total for drivers. I should stress I have made no attempt to determine how well they work together just noted the prices.
What is the overall design to be? Are you going to offload the bass to the mains and/or subwoofers or do you want deep bass from the centre speaker? This affects the size of the woofers and their parameters, whether the speaker is sealed or ported and how much height you have for the tweeter + midrange. How important is it that the centre can match future main speakers? This might mean for example, the same tweeter and midrange but 2 x 8" woofers with a port for the mains and 2 x 6.5" sealed for the centre with the deeper bass handed over to the mains and/or subs.
How to wrap your head around what is important? I would suggest designing a centre speaker on paper using standard drivers from, say, the SEAS or SB Acoustics ranges. Put aside details, just learn what fits together and what does not and why. This would mean something like:
- driver performance in ported and sealed cabinets -> WinISD or PCD
- baffle diffraction -> Edge or PCD
- rough crossover -> PCD, WinPCD, XSim
It is not a huge task but it will take a few evenings.
Another approach might be to dive in and build something ASAP and learn from your mistakes in which case opting for £200+ drivers seems rather unwise. SB Acoustics for example make high value for money drivers in £15-40 range: £17 tweeter, £27 midrange, 2 x £31 midwoofers = £106 total for drivers. I should stress I have made no attempt to determine how well they work together just noted the prices.
Hi Andy, thanks for your advice. I think you are right, although I just want to buy and play, it would be far more sensible for me to simulate a bunch of driver, cabinet and crossover options first. But before that, as jReave pointed out, I should probably finalise my AV stand design so that I know for sure how much room I have to play with.
As you say, it will probably be a few days before I have anything useful to reply back with. Besides, I still have a lot of reading left to do from all the links that have been posted... so if nobody hears much from me for a few days it will be because I'm reading all the links and learning to use the sim tools. 🙂
As you say, it will probably be a few days before I have anything useful to reply back with. Besides, I still have a lot of reading left to do from all the links that have been posted... so if nobody hears much from me for a few days it will be because I'm reading all the links and learning to use the sim tools. 🙂
Last edited:
As a PS I noted that Troels Gravesen actually has designed a centre channel using 2 x 6.5" woofers, a 4" midrange and a 1" tweeter from the SEAS standard range. He has a 91 dB sensitive speaker using an 86 dB midrange. How did he do it?
My guess is this can happen if you use the mid for a narrow range and low order slopes, the sensitivity then gets boosted by the tweeter and woofer. Also maybe bafflestep although I thought it decreased sensitivity in the lows instead of raising it in the highs.
If you want to get started fast, building a proven design and trying to understand why choices were made is a very good way to learn and guaranties a very good performing speaker. (also notice he cuts into the tweeter face-plate to minimize CC distance 🙂
If you want to get started fast, building a proven design and trying to understand why choices were made is a very good way to learn and guaranties a very good performing speaker. (also notice he cuts into the tweeter face-plate to minimize CC distance 🙂
Last edited:
In a nutshell he seems to have used the mid driver as a filler. But I'll take a deeper look when I'm home.
There are a number of factors at play here:
1) Baffle diffraction will indeed add a 1-2dB peak right in the midrange (see attached for a random eg).
2) In a 3-way, besides the overlap of the tweeter and the mid adding maybe another dB to the mid (and note the rather high xo point between the woofer and mid, about 850Hz, in those Troel's designs), often the mid will gain just another touch of SPL simply due to the combination of LP and HP filters "squeezing" the FR upward.
3) Look at the summed FR in those Seas CC's: it is not uncommon for a speaker to have a small, wide 1dB or so 'valley' right in the midrange in the on-axis response and/or for the tweeter to be pulled down a little as well.
4) Don't forget, it is also not unusual for any given driver to have a FR in reality that deviates somewhat from what the manufacturer has published.
But in terms of choosing drivers in the design stage, you want to have the flexibility in the mid and tweeter that will allow you to play them louder in order to get the right balance with the woofers, just in case.
Response Modeler is actually quite good for this. Import the FR of drivers you are looking at, set up your baffle and driver position in the Diffraction module at the bottom, and then the program will quickly combine them together for you. Repeat either with a different driver or with a different baffle configuration. You can even add in simple textbook shaped xo filters to see how that will look.
Attachments
With woofers, look at the SPL level at about 200Hz. What you do then is bring the rest of the woofer's response down with your xo to match this level. And you want to do this actually after you include baffle diffraction losses. This is why I've been saying you have to know what baffle effects your AV stand are going to introduce before final decisions on woofer and mid choices. Plus as already mentioned, in actual usage, you'll get a little more SPL out of your mid than what it specifies.
Height of the box is the major consideration. Center to center spacing applies equally to a CC as well as the mains but I think all the tweeters you have been considering can be crossed low if needed. And I think baffle diffraction effects on the tweeter are usually worse (or at least different) when it is very, very close to the top edge. That would be another reason therefore to go with a larger height. Play around with a diffraction program to see for yourself.
You can't really tell much of anything about what a driver will do once it's put in a box just by looking at its FR. The manufacturer's FR is usually done on a very large baffle without a box, called an infinite baffle measure. To know what a driver will do in a sealed or vented box of different sizes and with different amounts of power, you must enter the driver's TS parameters into a box modeling program. I prefer Unibox. Others like WinISD. Download and play around - that will teach you plenty.
Thank you very much jReave for your posts, as always they are really helpful. I'll download Unibox and WinISD and take a look at them both. I'll not get a chance to do anything tomorrow, but hopefully I will on Wednesday night. For now though, I have drawn a scale model of the AV stand that I hope to build. It is 1.75m wide and 0.65m deep. Here are some pictures:
Yes, that's more or less what I pictured. I've built something similar on 2 occasions. One low like that to sit below a projection screen but with the 2 center poles extended vertically to hold the CC a little higher and a current prefab AV console type thing with the tv attached similarly with square aluminum rods.
If the shelves get filled as you've shown, then the whole front of the AV unit should act more or less like 1 large front baffle. That means probably no significant baffle step loss above 100Hz. Use the Edge to simulate with your exact dimensions.
Personally I always find myself slouched down when watching tv or in other words my head is always tilted upward. So a higher tv works better for me. ymmv of course.
If the shelves get filled as you've shown, then the whole front of the AV unit should act more or less like 1 large front baffle. That means probably no significant baffle step loss above 100Hz. Use the Edge to simulate with your exact dimensions.
Personally I always find myself slouched down when watching tv or in other words my head is always tilted upward. So a higher tv works better for me. ymmv of course.
- Status
- Not open for further replies.