The TPA3116D2 is apparently only 50 watts if the load is 4 ohms. The TI data sheet doesn't state the impedance with it's max power specs, but the max supply voltage suggests this. I'm fairly confident that there are very good and reliable switchmode poweramps out there these days, but I'm not personally sure which ones they are. I've heard some horror stories of switchmode amps blowing up, taking the speakers with them, and in at least one case it allegedly started a house fire... (hard to believe, but I did read that somewhere - probably one of the earlier designs).
The TPA3116D2 is fully protected from blowups and shorts, etc. It has smart diagnostic that shuts itself down when you do something stupid, or even plugging in hot RCA inputs without powering off, or run without a heatsink (T>150C trips shutoff). In PBTL bridged mode you can get 100 watts/ch into 2 ohms. The max suplply voltage should not exceed 24 volts, and below 21 volts for best SQ.
I was also thinking of the IRS2092 amps like this one:
assmbled IRS2092 Class D 2.0 audio power amplifier board 300w+300w speaker protection on Aliexpress.com
I have not listened to one myself but hear they are very clean. They need a dual rail supply though so most people go with linear and use huge transformers and caps and rectifiers. So that is added cost.
Some have built in rectifiers and you only need supply AC from transformer like this one (350 watts into 8 ohms - which I believe is what Roger Russel recommend as the minimum power to driver the 25 Vifa array):
IRS2092 top Class D amplifier Mono Audio Power amplifier board Class D AMP board Assembled Amp board-in Consumer Electronics on Aliexpress.com
I was also thinking of the IRS2092 amps like this one:
assmbled IRS2092 Class D 2.0 audio power amplifier board 300w+300w speaker protection on Aliexpress.com
I have not listened to one myself but hear they are very clean. They need a dual rail supply though so most people go with linear and use huge transformers and caps and rectifiers. So that is added cost.
Some have built in rectifiers and you only need supply AC from transformer like this one (350 watts into 8 ohms - which I believe is what Roger Russel recommend as the minimum power to driver the 25 Vifa array):
IRS2092 top Class D amplifier Mono Audio Power amplifier board Class D AMP board Assembled Amp board-in Consumer Electronics on Aliexpress.com
Hi guys, thanks again!
First off, xrk971, I would like to use these for listening to music but also watching movies. So, I do need them to fill in double duty. I know that critical music listening and the blasting an imperial space cruiser does not fall in the same category, but we often have to make sacrifices and combine our desires into a single unit. So, single (or multiple) amps are not easy to deal with, since I would still need an external movie track processor. On that, when I first started in the home movies, there used to be a bunch of preamp processors... now, I can't even find one! Ok, scrap that. I found some... most well over $1,000! the most interesting was the Emotiva UMC-200. Seems to have some good EQ. I was looking at the Oppo 103, but it doesn't have EQ.
I know I am at a crossroad here. Do I push on with this design and get completely involved (which usually means spending money...) or do I drop it for a more traditional WMT.
My receiver has a MD-R tape loop that I could use, but that would only be for music listening, and big problems would arise when watching movies, as it would bypass the active EQ.
I then looked into integrated receivers offerings these days, I found the Marrantz SR5008 that looks like a good price vs specs ratio. It has Audyssey XT which offers auto or 8 bands EQ.
I also found the nanoAVR 8x8, which goes in between the HDMI out of the source and the HDMI in of receiver. It allows control of all signal streams independently like curve EQ, shelving, compressor, ... before entering the receiver. Looks nice. nanoAVR 8x8 | MiniDSP
So, lots to digest and decision to make on my part.
First off, xrk971, I would like to use these for listening to music but also watching movies. So, I do need them to fill in double duty. I know that critical music listening and the blasting an imperial space cruiser does not fall in the same category, but we often have to make sacrifices and combine our desires into a single unit. So, single (or multiple) amps are not easy to deal with, since I would still need an external movie track processor. On that, when I first started in the home movies, there used to be a bunch of preamp processors... now, I can't even find one! Ok, scrap that. I found some... most well over $1,000! the most interesting was the Emotiva UMC-200. Seems to have some good EQ. I was looking at the Oppo 103, but it doesn't have EQ.
I know I am at a crossroad here. Do I push on with this design and get completely involved (which usually means spending money...) or do I drop it for a more traditional WMT.
My receiver has a MD-R tape loop that I could use, but that would only be for music listening, and big problems would arise when watching movies, as it would bypass the active EQ.
I then looked into integrated receivers offerings these days, I found the Marrantz SR5008 that looks like a good price vs specs ratio. It has Audyssey XT which offers auto or 8 bands EQ.
I also found the nanoAVR 8x8, which goes in between the HDMI out of the source and the HDMI in of receiver. It allows control of all signal streams independently like curve EQ, shelving, compressor, ... before entering the receiver. Looks nice. nanoAVR 8x8 | MiniDSP
So, lots to digest and decision to make on my part.
Keep Pushing on with your project!!
After you get them done and running you may find just a little low end boost is all that you need providing that your current amplifier does have some sort of EQ built into it.
That is all the Linkwitz Transfom filter does, But only it is taylored with your drivers parameters within a enclosure of a particular size.
If you where to choose a different Box volume in WInlsd and recalculated the Linkwitz Transform it would come up with a totally different value of parameters for the filter.
Yes, it is hard these days to find Hi-Fi equipment that allows you complete flexibility without the hassles of digital signal being involved in these times.
Aside from building or own amplifier, as long as you can get a signal out from your preamp there are many standalone amplifier options that you can choose from to run them.
But, I don't know what you have available in your area.
jer
After you get them done and running you may find just a little low end boost is all that you need providing that your current amplifier does have some sort of EQ built into it.
That is all the Linkwitz Transfom filter does, But only it is taylored with your drivers parameters within a enclosure of a particular size.
If you where to choose a different Box volume in WInlsd and recalculated the Linkwitz Transform it would come up with a totally different value of parameters for the filter.
Yes, it is hard these days to find Hi-Fi equipment that allows you complete flexibility without the hassles of digital signal being involved in these times.
Aside from building or own amplifier, as long as you can get a signal out from your preamp there are many standalone amplifier options that you can choose from to run them.
But, I don't know what you have available in your area.
jer
Cool! There's a store that stacks a bunch of them just a few blocks away.
Assuming I brace it well, it shouldn't resonate too much, I hope.
I may have found a way to deal with preamps. I use my MacBookPro a lot for streaming content and discovered a way to alter system wide sounds, including curve EQ and a bunch of extra settings (check it out if you are on a Mac, it's called "Hear"). Also, my line out port is coupled with an optical out, so I can use that into my AV receiver to get 5.1 surround. That would cover my needs for a preamp and a way to EQ my source audio into my existing receiver.
Since I also have a minidisplayport, I can convert that to HDMI and send it to my projector.
Looks like this project is back on track!
Assuming I brace it well, it shouldn't resonate too much, I hope.
I may have found a way to deal with preamps. I use my MacBookPro a lot for streaming content and discovered a way to alter system wide sounds, including curve EQ and a bunch of extra settings (check it out if you are on a Mac, it's called "Hear"). Also, my line out port is coupled with an optical out, so I can use that into my AV receiver to get 5.1 surround. That would cover my needs for a preamp and a way to EQ my source audio into my existing receiver.
Since I also have a minidisplayport, I can convert that to HDMI and send it to my projector.
Looks like this project is back on track!
A line array in a pipe could be a great idea IMO.
One of the things to look out for with this idea is allowing the driver diaphram to get too close to the center of the pipe where it would be very resonant. If you cut a big section out of the pipe for the driver mount plate, Since the plate is flat, the center of it will be closer to the center of the pipe. If you then rear mount the drivers because you're worried about the tiny vents on the back side of the drivers not breathing well enough, now the diaphrams are even closer to the center of the pipe where it would ring like a bell.
I'd use a pipe that was a bit larger diameter, maybe 8 inch diameter and use Vifa TC9 or Peerless TG9 three inch drivers instead. These drivers are relatively cheap, and have all the features that you want for a high end line array.
If the drivers are big enough and have a reasonable Xmax spec, such that you can take them down to at least 80 - 100HZ, you may get significantly less of the typical boominess due to room resonances. This is because of the fact that the lower mid energy will be emitted spread out not just left to right, but up and down as well. Reflection cancellations from each driver, will be largely filled in by the other drivers which will each have a slightly different cancellation frequency due to floor and ceiling bounce, relative to where the listener sits. Most rooms I've measured have problems in the 80HZ - 300HZ area, causing that irritating boominess. If you can fix that, you'll want to turn the bass way up and bath in it.
Roger Russel runs his $19000 line arrays (which I believe use the Peerless TG9 drivers) all the way down to 25HZ (with active EQ) and I asked him specifically in an email if there was any boominess and he said no. "no noticable boominess". I've been wanting to build some of those ever since, but I've already got too many speakers... :-/
One of the things to look out for with this idea is allowing the driver diaphram to get too close to the center of the pipe where it would be very resonant. If you cut a big section out of the pipe for the driver mount plate, Since the plate is flat, the center of it will be closer to the center of the pipe. If you then rear mount the drivers because you're worried about the tiny vents on the back side of the drivers not breathing well enough, now the diaphrams are even closer to the center of the pipe where it would ring like a bell.
I'd use a pipe that was a bit larger diameter, maybe 8 inch diameter and use Vifa TC9 or Peerless TG9 three inch drivers instead. These drivers are relatively cheap, and have all the features that you want for a high end line array.
If the drivers are big enough and have a reasonable Xmax spec, such that you can take them down to at least 80 - 100HZ, you may get significantly less of the typical boominess due to room resonances. This is because of the fact that the lower mid energy will be emitted spread out not just left to right, but up and down as well. Reflection cancellations from each driver, will be largely filled in by the other drivers which will each have a slightly different cancellation frequency due to floor and ceiling bounce, relative to where the listener sits. Most rooms I've measured have problems in the 80HZ - 300HZ area, causing that irritating boominess. If you can fix that, you'll want to turn the bass way up and bath in it.
Roger Russel runs his $19000 line arrays (which I believe use the Peerless TG9 drivers) all the way down to 25HZ (with active EQ) and I asked him specifically in an email if there was any boominess and he said no. "no noticable boominess". I've been wanting to build some of those ever since, but I've already got too many speakers... :-/
Last edited:
Yep, I was thinking of borrowing some ideas from this:
Speaker Project - Dayton/Vifa Alu Pipe Speakers -- 11/28/2012
About the tube resonating or ringing, I thought of covering the inside with something I discovered when I waterproofed my balcony. It is a thick paint made of latex and cement. It has amazing dampening qualities. I first noticed when I dropped a couple of screws on the floor, the screws didn't make a sound, and didn't bounce either. Combine that with some bracing, and it should be pretty dampened against ringing.
Most people use polyfill inside, but I have extra rockwool from building bass traps here at home. Could I use rockwool (the heavy kind, 60kg/m) inside my speakers?
I'm designing and measuring to get my material... I'll also need to order an extra 16 drivers!
Speaker Project - Dayton/Vifa Alu Pipe Speakers -- 11/28/2012
About the tube resonating or ringing, I thought of covering the inside with something I discovered when I waterproofed my balcony. It is a thick paint made of latex and cement. It has amazing dampening qualities. I first noticed when I dropped a couple of screws on the floor, the screws didn't make a sound, and didn't bounce either. Combine that with some bracing, and it should be pretty dampened against ringing.
Most people use polyfill inside, but I have extra rockwool from building bass traps here at home. Could I use rockwool (the heavy kind, 60kg/m) inside my speakers?
I'm designing and measuring to get my material... I'll also need to order an extra 16 drivers!
Baffle size is normally a function of the XO point/slope and room gain curve, but this usually means a fairly wide one, so it usually boils down to how wide one can/will tolerate and/or how much EQ one is willing/able to use to shape its in room response.
So, your call; use as big as necessary for best overall performance to minimize the amount of EQ required or go with whatever looks 'right' in your room and/or needs to satisfy WAF considerations?
GM
So, your call; use as big as necessary for best overall performance to minimize the amount of EQ required or go with whatever looks 'right' in your room and/or needs to satisfy WAF considerations?
GM
For example, was looking at the low end losses occurring with a set of widerange planars and the lower limit is ~250 due to dipole cancellation losses. To make a baffle that remains OB but corrects for this loss would make for a baffle width of ~24". This can be achieved with two 1/4" thick plywood panels elliptically arced back leaving a flush mount front and an chamfered exit on the back side total thickness 1.5". Simple arced wedges form the structure that can then be filled with sand deadening this structure. If a regular box type enclosure is to be used same, shape can be employed, just enclose the rear space with a largely unseen box. Diffraction is minimized by adding edge treatment eg roundover, semicircle arc made of PVC run down the length, etc.
For acoustic dampening of the pipe, I wouldn't use any bracing. If the diaphrams are pretty near the edge of the pipe circle, the only resonant situation should be end to end. I'd break that up with a partial shelf maybe every 4 drivers along (Russel does this every two drivers along).
I would use "Liquid Nails" glue (or equivalent), to glue thick felt (or upholstering padding) to all the inside surfaces of the pipe (1/4 inch +). Then, I'd glue 1 inch thick foam rubber over that (if the pipe is 6-8 inches diameter). Then I'd put a bunch of the looser stuffing down the middle of the tube, but not so tight that it effectively blocks the vented pole piece of the drivers.
Adding side panels or wings lessen the requirements of active EQ, but may blur imaging significantly. Using a bigger pipe (8 inch) would provide a smoother diffraction rolloff, than a 4 inch pipe with the more abrupt edge to the driver frames. If you use a non-resonant driver mounting panel that's thin enough (1/4 to 3/8 inch), then you could front mount all the drivers, which would make it WAY easier to replace a bad or blown driver, should that ever be an issue down the road.
I would use "Liquid Nails" glue (or equivalent), to glue thick felt (or upholstering padding) to all the inside surfaces of the pipe (1/4 inch +). Then, I'd glue 1 inch thick foam rubber over that (if the pipe is 6-8 inches diameter). Then I'd put a bunch of the looser stuffing down the middle of the tube, but not so tight that it effectively blocks the vented pole piece of the drivers.
Adding side panels or wings lessen the requirements of active EQ, but may blur imaging significantly. Using a bigger pipe (8 inch) would provide a smoother diffraction rolloff, than a 4 inch pipe with the more abrupt edge to the driver frames. If you use a non-resonant driver mounting panel that's thin enough (1/4 to 3/8 inch), then you could front mount all the drivers, which would make it WAY easier to replace a bad or blown driver, should that ever be an issue down the road.
Thanks again for the feedback.
I can only hope that one day I will be able to share as much as you guys!
One more thing here. I'm starting to have second thoughts about the drivers I have. I've been on Youtube to try and listen to other 3" drivers and figure out if my 3" drivers can measure up, it's kinda hard.
Is there a way to test if my driver is suitable, other than building the array and finding out later? It would help to know before I shell out some more money for the extra 16 drivers on this project!
I can only hope that one day I will be able to share as much as you guys!
One more thing here. I'm starting to have second thoughts about the drivers I have. I've been on Youtube to try and listen to other 3" drivers and figure out if my 3" drivers can measure up, it's kinda hard.
Is there a way to test if my driver is suitable, other than building the array and finding out later? It would help to know before I shell out some more money for the extra 16 drivers on this project!
Perceval, I forget which driver you were planning to use. There's a lot to consider when you want one driver to run "wideband" (30 - 18kHZ).
Rubber suspension is a must, a vented spider and/or pole piece is a significant plus. Cavity effect can ruin an otherwise great driver when the half roll suspension sticks out too far (so they can have a higher Xmax). The shorted turn on the pole piece makes the high frequency response usable and the impedance curve flatter. A harder cone material has higher resolution, but also has uppermidrange and super sonic resonances that are very severe in many cases (always check the frequency response graphs - they should be pretty flat +/- 3dB, or at least very correctable with tone controls).
When using multiple drivers they should be mounted as close to each other as possible, so the nodal line of the interference pattern will be wider, so less annoying. You don't want drivers that have big wide mounting flanges that increase the distance between drivers. Xmax should be reasonable. The fancy new Dyminium (sp?) magnets are stronger, but the physical weight of the ceramic magnets help dampen any resonance in the driver frame better, and they're strong enough.
If an Engineer were to test a speaker driver, besides the usual frequency response test, they might use tone bursts inside a gaussian envelop, they might measure distortion products with a spectrum analyzer, they might look at a "waterfall graph" of the spectral decay.
Dispersion is important, but what's most important about it is you don't want an abrupt change in dispersion, especially in the upper-midrange, where the ear is most sensitive. You get an abrupt change in dispersion when a crossover switches you from a large diameter woofer or midrange driver to a 1 inch tweeter. The room reflections can work with that to color the sound in an unpleasant way. With small drivers you want to make sure the the vents on the back of the driver are pretty clear so that doesn't cause distortion.
Did I forget anything?
Rubber suspension is a must, a vented spider and/or pole piece is a significant plus. Cavity effect can ruin an otherwise great driver when the half roll suspension sticks out too far (so they can have a higher Xmax). The shorted turn on the pole piece makes the high frequency response usable and the impedance curve flatter. A harder cone material has higher resolution, but also has uppermidrange and super sonic resonances that are very severe in many cases (always check the frequency response graphs - they should be pretty flat +/- 3dB, or at least very correctable with tone controls).
When using multiple drivers they should be mounted as close to each other as possible, so the nodal line of the interference pattern will be wider, so less annoying. You don't want drivers that have big wide mounting flanges that increase the distance between drivers. Xmax should be reasonable. The fancy new Dyminium (sp?) magnets are stronger, but the physical weight of the ceramic magnets help dampen any resonance in the driver frame better, and they're strong enough.
If an Engineer were to test a speaker driver, besides the usual frequency response test, they might use tone bursts inside a gaussian envelop, they might measure distortion products with a spectrum analyzer, they might look at a "waterfall graph" of the spectral decay.
Dispersion is important, but what's most important about it is you don't want an abrupt change in dispersion, especially in the upper-midrange, where the ear is most sensitive. You get an abrupt change in dispersion when a crossover switches you from a large diameter woofer or midrange driver to a 1 inch tweeter. The room reflections can work with that to color the sound in an unpleasant way. With small drivers you want to make sure the the vents on the back of the driver are pretty clear so that doesn't cause distortion.
Did I forget anything?
Bob - could not a disservice be perpetrated by the implication that drivers not possessing the "significant plus" of vented pole pieces are at a disadvantage?
I can think of numerous widebanders from the currently well trafficked Vifa TC9FD & derivents , to every Fostex FE&FF I've seen, to the entire Mark Audio line, & numerous others that seem to stumble on quite nicely with solid back plate motor structures. Biga$$ subs (car audio & extreme HT ) were duty cycle could well create serious thermal loads on the VC assembly, yes - but more moderate residential systems, particularly a line array of which as I understand it one of the fundamental design criteria is the reduction of excursion/power consumption by each driver, I'm not so sure it's a major factor in the system design calculus.
I'd certainly agree about Neodymium - probably of more benefit in small dome tweeters for car installations, or headphones, earbuds, or micro speakers in computers, tablets, etc where overall dimensions are often highly constrained.
I can think of numerous widebanders from the currently well trafficked Vifa TC9FD & derivents , to every Fostex FE&FF I've seen, to the entire Mark Audio line, & numerous others that seem to stumble on quite nicely with solid back plate motor structures. Biga$$ subs (car audio & extreme HT ) were duty cycle could well create serious thermal loads on the VC assembly, yes - but more moderate residential systems, particularly a line array of which as I understand it one of the fundamental design criteria is the reduction of excursion/power consumption by each driver, I'm not so sure it's a major factor in the system design calculus.
I'd certainly agree about Neodymium - probably of more benefit in small dome tweeters for car installations, or headphones, earbuds, or micro speakers in computers, tablets, etc where overall dimensions are often highly constrained.
There are drivers that are as good as the Vifa TC9 or Peerless TG9 (identical except for cone material), but at many times the cost. I don't know of any that are significantly better, at any price, in the 3 inch category. No 4 inchers have as good a high end, and no 2 inchers can go low enough to retain the benefit of being able to go down to at least 100HZ (80HZ is better), to minimize boominess created by room acoustics.
I really like the Peerless woofers too. The 6.5 and 8 inch Nomex cone drivers would be hard to beat at any price. I've got the Nomex 6.5 inchers in one system and the 12 inch XXLS in another.
Put four 6.5 inch Peerless Nomex cone woofers vertically aligned, in a closed box with active EQ acoustically flattening them down to 30HZ, use a TC9 3 inch midrange in a subenclosure made out of 4 inch inner diameter plumber pipe full of acoustic stuffing, and a Fountek 1.5 inch ribbon tweeter, with X= 200HZ and 10kHZ, and you'd have a very nice speaker that worked better with room acoustics than most (because of the line array of woofers). A 4 pole active crossover would put frosting on the cake.
If your room is fairly large you might want to consider using 8 inch drivers for woofers instead. If the room's large and you like to turn it up hard-rock band loud sometimes, I'd go to the Peerless XXLS drivers. For really high volume, a 5.1 surround system might be a better approach.
I really like the Peerless woofers too. The 6.5 and 8 inch Nomex cone drivers would be hard to beat at any price. I've got the Nomex 6.5 inchers in one system and the 12 inch XXLS in another.
Put four 6.5 inch Peerless Nomex cone woofers vertically aligned, in a closed box with active EQ acoustically flattening them down to 30HZ, use a TC9 3 inch midrange in a subenclosure made out of 4 inch inner diameter plumber pipe full of acoustic stuffing, and a Fountek 1.5 inch ribbon tweeter, with X= 200HZ and 10kHZ, and you'd have a very nice speaker that worked better with room acoustics than most (because of the line array of woofers). A 4 pole active crossover would put frosting on the cake.
If your room is fairly large you might want to consider using 8 inch drivers for woofers instead. If the room's large and you like to turn it up hard-rock band loud sometimes, I'd go to the Peerless XXLS drivers. For really high volume, a 5.1 surround system might be a better approach.
Last edited:
I can do frequency tests and waterfalls with REW. I have some pretty good mics and a good soundcard. I'm not familiar with "tone bursts inside a gaussian envelop" though.
I would love to design something with 12" drivers, but I am not sure it would pass the WAF...
I am kinda limited in my choices of drivers. I tried a couple of online stores, and shipping 16 3" drivers, or a couple of 12/15 incher is quite prohibitive and was asked to forget about it.
Here is a link to a local maker and supplier of drivers:
Yahoo! ????
I also can get access to Usher drivers.
This is the 3" driver I already bought, and was planning to buy 16 more:
I would love to design something with 12" drivers, but I am not sure it would pass the WAF...
I am kinda limited in my choices of drivers. I tried a couple of online stores, and shipping 16 3" drivers, or a couple of 12/15 incher is quite prohibitive and was asked to forget about it.
Here is a link to a local maker and supplier of drivers:
Yahoo! ????
I also can get access to Usher drivers.
This is the 3" driver I already bought, and was planning to buy 16 more:
Attachments
Here's the actual page of the 3" driver's info:
http://goods.ruten.com.tw/item/show?21206258966085
And here is an updated list of the drivers available through that factory/store:
http://goods.ruten.com.tw/item/show?21206258966085
http://goods.ruten.com.tw/item/show?21206258966085
And here is an updated list of the drivers available through that factory/store:
http://goods.ruten.com.tw/item/show?21206258966085
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.