Hello people,
I'm looking to order the following plate amp for a DIY Shiva sub (EBS fb = 18.1 Hz, predicted f(3) ~ 21Hz)
http://www.users.bigpond.net.au/redgumaudio/rgsw2.html
I'm worried the 6dB xover will be too shallow. I only have a very basic 2-channel receiver - and am not able to attenuate the main's low freq to integrate.
Would I be better off investing in a plate amp with 12dB (or greater) xover slope? I'm sorry I can't tell you what the in-room response of the bookshelf mains speakers are (eg. no idea what f(3) point is) - which would probably help you alot.
If room effects is also needed to answer this question - it is about 23x16 feet, 8 foot high ceiling, wood floor over concrete.
Thanks,
David.
I'm looking to order the following plate amp for a DIY Shiva sub (EBS fb = 18.1 Hz, predicted f(3) ~ 21Hz)
http://www.users.bigpond.net.au/redgumaudio/rgsw2.html
I'm worried the 6dB xover will be too shallow. I only have a very basic 2-channel receiver - and am not able to attenuate the main's low freq to integrate.
Would I be better off investing in a plate amp with 12dB (or greater) xover slope? I'm sorry I can't tell you what the in-room response of the bookshelf mains speakers are (eg. no idea what f(3) point is) - which would probably help you alot.
If room effects is also needed to answer this question - it is about 23x16 feet, 8 foot high ceiling, wood floor over concrete.
Thanks,
David.
Speaking from my experience with the Dayton Titanics--a very similar driver--I think you'll find that flat response will require you to set that 6 dB/oct slope to the minimum setting (20 hz) in order to offset a rising response. This will get you flat response up into the mid-hundreds. At that point, you'll need to contrive an external crossover in order to bring about an actual low pass so that you can mate with your satellites.
This is a common problem with long throw woofers. It will also answer the nagging question in the back of your mind as to why on Earth anyone would have such an outlandish thing as a 6 dB/oct @ 20Hz setting. (Aside from the fact that the designer might have been drunk at the time...)
To see graphs of something like this, take a look at the XVR1 manual at www.passlabs.com.
It's the price you pay for a driver like that. You can have deep bass or flat bass, but not both at the same time. At least, not without doing a little frequency tailoring.
Grey
This is a common problem with long throw woofers. It will also answer the nagging question in the back of your mind as to why on Earth anyone would have such an outlandish thing as a 6 dB/oct @ 20Hz setting. (Aside from the fact that the designer might have been drunk at the time...)
To see graphs of something like this, take a look at the XVR1 manual at www.passlabs.com.
It's the price you pay for a driver like that. You can have deep bass or flat bass, but not both at the same time. At least, not without doing a little frequency tailoring.
Grey
Thanks Grey,
I asked the manufacturer - and they said they chose a crossover that was the least intrusive in terms of signal processing.
What you mentioned makes perfect sense. If the xover didn't start from 20Hz then it would be difficult to get suitable attenuation at 80Hz and above.
I'm worried I'll end up with a "dip" between 40 and 80 Hz. AS it is - Adire Audio's room acoustics simulator in their version of lspCAD expects I'll get a serious dip around 60Hz.
Either way - I may as well stop fluffing about and take the plunge - buy it - install it - and if it doesn't work - put it down to experience. After all - it is only money 🙂
Dave.
I asked the manufacturer - and they said they chose a crossover that was the least intrusive in terms of signal processing.
What you mentioned makes perfect sense. If the xover didn't start from 20Hz then it would be difficult to get suitable attenuation at 80Hz and above.
I'm worried I'll end up with a "dip" between 40 and 80 Hz. AS it is - Adire Audio's room acoustics simulator in their version of lspCAD expects I'll get a serious dip around 60Hz.
Either way - I may as well stop fluffing about and take the plunge - buy it - install it - and if it doesn't work - put it down to experience. After all - it is only money 🙂
Dave.
I'd be a bit worried about the -6dB crossover a bit too, but then I'm kind of biased because I come from the pro-audio end where people set crossovers at at least -24db and usually -48dB rolloffs.
Not quite sure what the manufacturer means by "least intrusive" either - even if the phase response of a first order filter is superior to that of a fourth order, the combined effect when connected to a driver is generally superior with a higher order filter (I think there's something on Mr. Linkwitz's site about this to do with driver impedence and stuff).
Personally I'd be a lot happier with at least a 2nd order, but at the end of the day it's how it sounds that counts.
Not quite sure what the manufacturer means by "least intrusive" either - even if the phase response of a first order filter is superior to that of a fourth order, the combined effect when connected to a driver is generally superior with a higher order filter (I think there's something on Mr. Linkwitz's site about this to do with driver impedence and stuff).
Personally I'd be a lot happier with at least a 2nd order, but at the end of the day it's how it sounds that counts.
Thanks Redeye,
Yeah - I don't understand that either - here is my reason
20Hz 6db Down
40Hz 12db down
80Hz 18db down
160Hz 24db Down
320Hz 30db down
740Hz 36db down
@ 95db (loudest I'll probably play it for any length of time) The sub is still singing along @ 59db @ 740Hz!
Whereas a 12db slope will be 36db down @ 80Hz - (starting frm 20)!
I'm surprised since I can't find *any* other plate amp with only a 6db xover
The the manufacturer claims the 6db slope is least instrusive in terms of phase delay - I thought the delay due to the low bass from a sub had a fair amount of delay anyway (hence having phase controls!)
Dave.
Yeah - I don't understand that either - here is my reason
20Hz 6db Down
40Hz 12db down
80Hz 18db down
160Hz 24db Down
320Hz 30db down
740Hz 36db down
@ 95db (loudest I'll probably play it for any length of time) The sub is still singing along @ 59db @ 740Hz!
Whereas a 12db slope will be 36db down @ 80Hz - (starting frm 20)!
I'm surprised since I can't find *any* other plate amp with only a 6db xover
The the manufacturer claims the 6db slope is least instrusive in terms of phase delay - I thought the delay due to the low bass from a sub had a fair amount of delay anyway (hence having phase controls!)
Dave.
Is there any way you can get to the line level? The reason I am asking is that there are Passive Line Level Crossovers that seem made to order for this situation, but not if you are stuck dealing only with speaker level signal.
Though there are only illustrations for 1st and 2nd order, author Peter Rettweiler makes it clear that higher orders are certainly achievable. Besides, you can add these to the 6 dB crossover you already have and get as high as a third order filter just using the author's illustrations.
This page is from Planet 10's TLine site, www.t-linespeakers.org
Here is the specific page:
http://www.t-linespeakers.org/tech/filters/passiveHLxo.html
Though there are only illustrations for 1st and 2nd order, author Peter Rettweiler makes it clear that higher orders are certainly achievable. Besides, you can add these to the 6 dB crossover you already have and get as high as a third order filter just using the author's illustrations.
This page is from Planet 10's TLine site, www.t-linespeakers.org
Here is the specific page:
http://www.t-linespeakers.org/tech/filters/passiveHLxo.html
Thanks kelticwizard.
My receiver has no line level / pre-out. I've started a post on the SS forum - but I really need a schematic of the receiver in question to do it.
I'll go with the speaker outs - and if there is still too much high frequency energy going to the sub - put in a low pass filter with steeper slope.
Thanks,
Dave.
My receiver has no line level / pre-out. I've started a post on the SS forum - but I really need a schematic of the receiver in question to do it.
I'll go with the speaker outs - and if there is still too much high frequency energy going to the sub - put in a low pass filter with steeper slope.
Thanks,
Dave.
I don't think I managed to get my point across. Let me take another stab at it.
I think you'll find that the Shiva's response curve is anything but flat. The curve you see when you run a simulation in WinISD or whatever is nowhere near an accurate portrayal of what you'll get in real life. The best you can hope for is a decent calculation of cabinet volume.
After you build your cabinet, test the resulting sub with a real time analyzer. Prepare yourself for a shock. You thought you were going to get flat response from, say, 200 hz down to somewhere in the 20s. Nope. What you'll see is a very large, broad hump--nothing like flat. If you were to pick a frequency somewhere down around 30Hz, you'll probably find that 60Hz is 6 dB higher and 120 is 6 dB higher still (+12 compared to 30Hz). Looked at from the other direction, you're getting a long, slow (6 dB/oct) rolloff that you hadn't planned on down to the f3 of the driver/cabinet, at which point it'll start rolling off with an additional 12 dB/oct.
The fact that people get online and profess to have had religious experiences after playing their new Shiva subs tends to indicate that they don't know what true low end sounds like. Loud at 80-100Hz isn't true low end. Organ pipes playing a low C...now <i>that's</i> true low end.
The 6 dB/oct rolloff on that plate amplifier will not give you the response you're planning on in the real world. It's not going to be -18 dB at 80Hz for instance. It's going to be roughly flat, i.e. no attenuation. Ditto for 160...and for 320...etc. Somewhere up in the mid hundreds, the Shiva's response will quit rising, at which point the 6 dB/oct slope will begin to have an effect. Combine that with the fact that the Shiva's response is dropping, and you'll end up with a 12 or 18 dB lowpass, but at a way, way higher frequency than you expected.
To get a lowpass filter function at a reasonable frequency, you will have to add an external crossover.
Right about now you're saying,"Well, then I'll just buy a different plate amp--one with a sharper cutoff."
You'll get more of a rolloff, and it will be more-or-less at the frequency that you want, but it will be roughly one pole (i.e. 6 dB/oct) less than you think you're getting. A 12 dB/oct crossover will give you 6, an 18 will give you 12, etc. Worse yet, the sub's response will not be flat. You'll still have that 6 dB/oct slide down to the 20Hz region. To get flat response, you'll need to run an external 6 dB/oct lowpass starting at some really low frequency...like 20Hz.
It's not the plate crossover's fault, it's the driver's. If you buy a driver of this sort, you need to know going in that you're going to have to twist its tail to get flat response.
It can be done. Just be prepared to fiddle a bit.
Grey
I think you'll find that the Shiva's response curve is anything but flat. The curve you see when you run a simulation in WinISD or whatever is nowhere near an accurate portrayal of what you'll get in real life. The best you can hope for is a decent calculation of cabinet volume.
After you build your cabinet, test the resulting sub with a real time analyzer. Prepare yourself for a shock. You thought you were going to get flat response from, say, 200 hz down to somewhere in the 20s. Nope. What you'll see is a very large, broad hump--nothing like flat. If you were to pick a frequency somewhere down around 30Hz, you'll probably find that 60Hz is 6 dB higher and 120 is 6 dB higher still (+12 compared to 30Hz). Looked at from the other direction, you're getting a long, slow (6 dB/oct) rolloff that you hadn't planned on down to the f3 of the driver/cabinet, at which point it'll start rolling off with an additional 12 dB/oct.
The fact that people get online and profess to have had religious experiences after playing their new Shiva subs tends to indicate that they don't know what true low end sounds like. Loud at 80-100Hz isn't true low end. Organ pipes playing a low C...now <i>that's</i> true low end.
The 6 dB/oct rolloff on that plate amplifier will not give you the response you're planning on in the real world. It's not going to be -18 dB at 80Hz for instance. It's going to be roughly flat, i.e. no attenuation. Ditto for 160...and for 320...etc. Somewhere up in the mid hundreds, the Shiva's response will quit rising, at which point the 6 dB/oct slope will begin to have an effect. Combine that with the fact that the Shiva's response is dropping, and you'll end up with a 12 or 18 dB lowpass, but at a way, way higher frequency than you expected.
To get a lowpass filter function at a reasonable frequency, you will have to add an external crossover.
Right about now you're saying,"Well, then I'll just buy a different plate amp--one with a sharper cutoff."
You'll get more of a rolloff, and it will be more-or-less at the frequency that you want, but it will be roughly one pole (i.e. 6 dB/oct) less than you think you're getting. A 12 dB/oct crossover will give you 6, an 18 will give you 12, etc. Worse yet, the sub's response will not be flat. You'll still have that 6 dB/oct slide down to the 20Hz region. To get flat response, you'll need to run an external 6 dB/oct lowpass starting at some really low frequency...like 20Hz.
It's not the plate crossover's fault, it's the driver's. If you buy a driver of this sort, you need to know going in that you're going to have to twist its tail to get flat response.
It can be done. Just be prepared to fiddle a bit.
Grey
Excellent! Thanks.
I suspect that graph was made with a crossover in the circuit, since the Shiva wouldn't normally (lowpass) rolloff anywhere near the 100Hz region; probably more like 1kHz. However, the slope from there on down is exactly what I was trying to describe.
Note that it's a simple 6 dB/oct rise, i.e. 40Hz looks to be pretty close to 6 dB higher than 20Hz, etc. And since the graph shows three different Qs, you can see that you can't fix it by varying the cabinet volume.
Now, if you were to use a 6 dB/oct <i>lowpass</i> starting around 20Hz, the falling crossover slope would combine with the rising driver response to give you flat response. But you'd have nothing remotely resembling a crossover point at any frequency that you'd regard as normal for a subwoofer. Once you've got that rise tamed, you can do pretty much anything you want with it.
Then, Grasshopper, you'll discover the true meaning of deep bass.
Grey
I suspect that graph was made with a crossover in the circuit, since the Shiva wouldn't normally (lowpass) rolloff anywhere near the 100Hz region; probably more like 1kHz. However, the slope from there on down is exactly what I was trying to describe.
Note that it's a simple 6 dB/oct rise, i.e. 40Hz looks to be pretty close to 6 dB higher than 20Hz, etc. And since the graph shows three different Qs, you can see that you can't fix it by varying the cabinet volume.
Now, if you were to use a 6 dB/oct <i>lowpass</i> starting around 20Hz, the falling crossover slope would combine with the rising driver response to give you flat response. But you'd have nothing remotely resembling a crossover point at any frequency that you'd regard as normal for a subwoofer. Once you've got that rise tamed, you can do pretty much anything you want with it.
Then, Grasshopper, you'll discover the true meaning of deep bass.
Grey
Thanks Grey / jag,
I certainly don't want a "one note bass" sub. I was originally choosing between the Shiva and Peerless XLS drivers. I liked the construction characteristics of the peerless (rubber surrounds are proven) but the Shiva modelled the best and had the best writeups.
Adire's shiva whitepaper lower Q graphs have a much smoother response - yet still a noticeable hump. I've gone for about Q=0.56 - dual vented / sealed enclosure.
I wish modelling programs were better a simulating in room response. Here's what I'm expecting - based on Adires lspCAD:
(see graph below)
I realise the cabinet won't make up for the Shiva's "hump". I will research pc card attached mics for some near field and in room response measurements. If I can manage a mic (electronically challenged am I) - then I'll graph and post the actual response.
Should get the driver and plate amp in the next 2 weeks, construction will take me about another 4 - 6 weeks (can't work on it weekends - by order of the boss).
Dave.
I certainly don't want a "one note bass" sub. I was originally choosing between the Shiva and Peerless XLS drivers. I liked the construction characteristics of the peerless (rubber surrounds are proven) but the Shiva modelled the best and had the best writeups.
Adire's shiva whitepaper lower Q graphs have a much smoother response - yet still a noticeable hump. I've gone for about Q=0.56 - dual vented / sealed enclosure.
I wish modelling programs were better a simulating in room response. Here's what I'm expecting - based on Adires lspCAD:
(see graph below)
I realise the cabinet won't make up for the Shiva's "hump". I will research pc card attached mics for some near field and in room response measurements. If I can manage a mic (electronically challenged am I) - then I'll graph and post the actual response.
Should get the driver and plate amp in the next 2 weeks, construction will take me about another 4 - 6 weeks (can't work on it weekends - by order of the boss).
Dave.
Attachments
Dave Bullet said:
Dave, have you ordered the Shiva yet?
Reason I am asking is that apparently Eminence makes a 6 ohm, single voice coil subwoofer that seems to have almost exactly the same specs as the Shiva, and which might be cheaper over there-the Lab12.
Even though the Lab 12 seems to be slightly more expensive than the Shiva in the USA, outside the USA it might be cheaper since Eminence has worldwide distribution. The Lab 12 is considerably cheaper than the Shiva in Britain, for instance.
About the only major difference I could discern, aside from the 6 ohm and single voice coil, is that the Shiva has a slightly better Xmax-16 mm to the Lab 12's 13mm. But 13 mm is still a lot.
Even though Eminence advertises this as being specifically designed for horns, (with an admission that it could be used for other enclosures), the specs indicate that it really is virtually a 6 ohm, single voice coil version of the Shiva.
When modelled, both subs seem to give precisely the same performance in the same boxes.
Anyway, I just thought you should be aware of the option, if not now, then for the future.
Here is the Eminence page for the Lab 12 specs.
http://editweb.iglou.com/eminence/eminence/pages/products02/lab12/lab12.htm
Here is the New Zealand distributor:
http://editweb.iglou.com/eminence/eminence/pages/support/inquiry/wp.htm
Thanks kelticwizard,
I have already ordered the Shiva. Agree the Lab12 has similar specifications (but 2 x as much magnet) - foam surround.
I will try and find out local pricing out of interest - but the Shiva should be in the belly of a jumbo by now.
If it all turns to custard - the speaker will make a nice planter box for my wife.
Thanks for your help,
Dave.
I have already ordered the Shiva. Agree the Lab12 has similar specifications (but 2 x as much magnet) - foam surround.
I will try and find out local pricing out of interest - but the Shiva should be in the belly of a jumbo by now.
If it all turns to custard - the speaker will make a nice planter box for my wife.
Thanks for your help,
Dave.
.. BTW - the Eminence Lab 12 says it is designed for a "horn loaded enclosure" - what does one look like for a subwoofer?
Thanks,
Dave.
Thanks,
Dave.
Dave Bullet said:
Huge, for one thing. And they don't go that low. While there were folded horn subwoofers going as low as 30 Hz, these things were simply enormous.
Many horn "bass bins" are very, very large and do not go that low. Why have them? Because much of the bass in rock is around 70 Hz or above, for one thing. The other thing is-next opportunity you get to hear bass through a horn, try it. Clarity, detail-all the attributes you would ascribe to mid and high range get used up by horn bass bins-even if they don't go that low.
If a horn could be invented that used up the space of a vented or sealed box and had the same low frequency cutoff, there would be nothing but horn enclosures very quickly.
I suppose the Shiva caught my eye because of the great DIY following - I hadn't even heard of the Lab 12 (I haven't previously found it when searching for a NZ distributor of 12" + drivers).
Some other things it looks like it has over a Shiva (not that these necessarily affect performance):
Cast basket
Double the magnet size (160oz vs. 80oz)
The spec sheets however don't have any power handling information.
Horn loaded enclosures - is the reason a horn has to be so big for a sub is to get some "directivity" of such low bass frequencies? I mean with wavelengths so long - you need a long "channel" in order to direct them (otherwise they bend / flop around the edges). You can see why I never became a physics teacher. I must Call Stephen H regarding floppy wave theory 🙂
Dave.
Some other things it looks like it has over a Shiva (not that these necessarily affect performance):
Cast basket
Double the magnet size (160oz vs. 80oz)
The spec sheets however don't have any power handling information.
Horn loaded enclosures - is the reason a horn has to be so big for a sub is to get some "directivity" of such low bass frequencies? I mean with wavelengths so long - you need a long "channel" in order to direct them (otherwise they bend / flop around the edges). You can see why I never became a physics teacher. I must Call Stephen H regarding floppy wave theory 🙂
Dave.
A horn enclosure is, in essence, a mechanical transformer. The intent is to take a comparatively small volume of air being violently jiggled at the driver and change that to a much larger volume of air moving somewhat more gently at the mouth.
(Just for the record, note that panel loudspeakers like electrostats and planars accomplish this mating with the room air without the use of a horn.)
The size thing with the horns comes in two parts:
--The area of the mouth of the horn (for even 30Hz or so, you're talking a mouth the size of an entire wall--not an exaggeration). This isn't really open to negotiation, it's set by the wavelength of the lowest frequency you intend to reproduce.
--The other is the length of the horn, which is open to a little bit of fudge factor, but not much. If you start trying to cheat Mother Nature the distortion of the horn climbs rapidly. (Distortion in horns is already a problem.) You'll also see people debating the merits of the curve used to accomplish the flare; tractrix, exponential, conical, etc.
One thing frequently overlooked with horns is that the bigger they are, the harder they are to time align with the other drivers. By the time you get the drivers right, the mouth of any reasonably low frequency bass horn is in your lap.
It's not hard to find pictures of horns that actually do take up the wall of a room, in which case you literally are sitting <i>inside</i> the horn.
It's a whole 'nother mindset. People who like horns <i>really</i> like them a lot. Like everything else, they involve a set of compromises. Whether that's practical for you is your choice.
Grey
(Just for the record, note that panel loudspeakers like electrostats and planars accomplish this mating with the room air without the use of a horn.)
The size thing with the horns comes in two parts:
--The area of the mouth of the horn (for even 30Hz or so, you're talking a mouth the size of an entire wall--not an exaggeration). This isn't really open to negotiation, it's set by the wavelength of the lowest frequency you intend to reproduce.
--The other is the length of the horn, which is open to a little bit of fudge factor, but not much. If you start trying to cheat Mother Nature the distortion of the horn climbs rapidly. (Distortion in horns is already a problem.) You'll also see people debating the merits of the curve used to accomplish the flare; tractrix, exponential, conical, etc.
One thing frequently overlooked with horns is that the bigger they are, the harder they are to time align with the other drivers. By the time you get the drivers right, the mouth of any reasonably low frequency bass horn is in your lap.
It's not hard to find pictures of horns that actually do take up the wall of a room, in which case you literally are sitting <i>inside</i> the horn.
It's a whole 'nother mindset. People who like horns <i>really</i> like them a lot. Like everything else, they involve a set of compromises. Whether that's practical for you is your choice.
Grey
Thanks Grey,
Well... the Shiva arrived yesterday. I also decided to order the 6dB plate amp is on its way (hey the manufacturer e-mailed me when it was dispatched, insured it for me and asked me to contact them to let them know it transported ok when it arrives). - Now that's service!
I also got a metal grille - to stop kids stuffing pens etc... through the cone (or whatever else they might dream up)
buying MDF tomorrow - I'm being adventurous and using my own enclosure design.
dave.
Well... the Shiva arrived yesterday. I also decided to order the 6dB plate amp is on its way (hey the manufacturer e-mailed me when it was dispatched, insured it for me and asked me to contact them to let them know it transported ok when it arrives). - Now that's service!
I also got a metal grille - to stop kids stuffing pens etc... through the cone (or whatever else they might dream up)
buying MDF tomorrow - I'm being adventurous and using my own enclosure design.
dave.
- Status
- Not open for further replies.