Hey guys, another board friday night with the wife at work,
seeing as I currently have 3 loudspeakers on the work bench a few considerations came up;
how do we define a good loudspeaker?
I know low distortion, short decay and flat FR sound 'ideal', but what else, and is flat FR really that important?
I have 2 3 ways and a 2 way I'm currently working on (a PMC IB1s clone, B&W XT4 rebuild and SB acoustics based mini monitors)
I guess this comes into the domains of what are your working practices and why.
For example, the B&W is very soft and nice to listen to, while the PMC clone is very dry and bland, but the PMC is far more 'accuratre'
can anyone share thoughts on how they do things and why?
let's get the ball rolling
(Side note, do certain drivers just 'play' together better than others? because I can't seem to get the seas 27tdfc to work with the SB15NRXC30-8, which is very frustrating as they're both superb)
happy fridays
seeing as I currently have 3 loudspeakers on the work bench a few considerations came up;
how do we define a good loudspeaker?
I know low distortion, short decay and flat FR sound 'ideal', but what else, and is flat FR really that important?
I have 2 3 ways and a 2 way I'm currently working on (a PMC IB1s clone, B&W XT4 rebuild and SB acoustics based mini monitors)
I guess this comes into the domains of what are your working practices and why.
For example, the B&W is very soft and nice to listen to, while the PMC clone is very dry and bland, but the PMC is far more 'accuratre'
can anyone share thoughts on how they do things and why?
let's get the ball rolling
(Side note, do certain drivers just 'play' together better than others? because I can't seem to get the seas 27tdfc to work with the SB15NRXC30-8, which is very frustrating as they're both superb)
happy fridays
Happy Friday. My name is Cal and I am a speakerholic. I have more speakers than most sane people but I have no favourites, that's why I own many. Like most addicts I don't have a reason, I do it because I like it. I know it will kill me but I do it anyways and while I don't often tell the world, I am happy with the decision(s) I have made. I am at peace with being a speakerholic.
I have a house full of speakers.
A pair of cheap disco speakers which are 400watts each.
A Fane 18 inch sub box.
A Fane 15 inch sub box.
A box with a pair of Fane 12-250TC's in it.
A box with a pair of Fane 12-200LC's in it.
A box with a pair of Maplin Big cat 12" speakers in it.
A box with 4 10" eminence speakers in it (cant remember what model of speaker)
A box with an old RS 100 watt speaker in it.
I first started in 1980 when I built a mobile disco.
It had 2 speaker boxes with 2 Fane 50WRMS speakers in each.
Sounded great despite drivers only going to about 4KHz.
A pair of cheap disco speakers which are 400watts each.
A Fane 18 inch sub box.
A Fane 15 inch sub box.
A box with a pair of Fane 12-250TC's in it.
A box with a pair of Fane 12-200LC's in it.
A box with a pair of Maplin Big cat 12" speakers in it.
A box with 4 10" eminence speakers in it (cant remember what model of speaker)
A box with an old RS 100 watt speaker in it.
I first started in 1980 when I built a mobile disco.
It had 2 speaker boxes with 2 Fane 50WRMS speakers in each.
Sounded great despite drivers only going to about 4KHz.
Listening tests are not only very subjective, but there are so many variables at play that it's really only a good approach when you are adjusting your multi-section tone controls to bring out the best in a completed speaker system. Otherwise you end up tayloring your speaker to the idiosyncracies of your particular ears, room acoustics, listening level and whatever the recording/mixing Engineers created. I design for flat as a default, largely so I can identify problem areas such as resonance or cancellations, but then I make it sound good with my own brew of tone controls (4 section Baxandall done right). Good tone controls are a must IMO.
In the speaker cabinet, I consider it very important to glue 1/4 - 1/2 inch thick felt (or upholstery padding - which is ugly but cheap) to all internal surfaces of the cabinet, except the driver panel. I double it up in any corners. I then put in soft foam rubber in certain strategic areas, and only then the fluffy stuffing, and it's all glued in place away from the rear of the drivers, with silicone rubber cement, or Liquid Nails, or similar glue that never gets brittle. Doing this right may be what most speaker builders fall short on. You want to create a "gaussian arrestor" to dissipate that back energy without it causing any kind of abrupt loading change on the drivers. Don Davis of the Sound System Engineering seminar demonstrated how important this approach is many decades ago. Fluffing that isn't tight to the internal surfaces of the cabinet has far less effectiveness.
Another place many miss the boat is crossover design. They don't all realize that the impedance of any driver may well be substantially different than the nominal rating, at the frequency they want to cross it over at. A 5 inch driver I was using, rated at 8 ohms, measured 15 ohms at about 3kHZ, where I wanted to cross it over to the tweeter. Not only will that necessarily throw off an 8 ohm crossover design amplitude response, but phase shift is then way off too, so you'd want to swap the phase on the various drivers in a multi way spkr system to see which way has less of a cancellation at the crossover frequency. So stock passive crossover products are a joke.
Preliminary testing of drivers should be for efficiency matching (you are likely to need to add R's to get that right, which changes the calculations for the reactive componenets of the crossover), and impedance at the frequencies you want to do the crossovers at. Measured impedance + efficiency matching resistor = actual impedance for calculating reactive components (coils and caps).
When done putting a spkr system together, you want to be able to turn off all but one driver at a time, so you can check with pink noise, a cal'd mic and RTA that each driver is rolling off as per your design, before turning them all on to check level matching. With an active crossover things are likely to be very accurate. With a passive crossover things are unlikely to be very accurate, but could be close enough if you did everything else right.
As for speaker drivers themselves, the harder the diaphram, the better the resolution, but hard diaphram drivers usually have severe resonance issues, often at the frequencies where the ear is most sensitive. I feel that hard diaphram drivers require 4th order crossover slopes (24dB/octave), which can really only be done well with active crossover electronics ahead of the poweramps.
With woofers, the Xmax spec is one of the more important things to consider. With midrange and tweeter drivers, off axis response and range of flatness are maybe the most important factors. Distortion too of coarse. I really like the Peerless TG9 or Vifa TC9 three inch drivers because they are almost ruler flat (in a good sub-enclosure) from 350HZ - 10kHZ (The 8 ohm Peerless TG driver is ruler flat to 15kHZ). This means I can keep crossover points away from where the ear is most sensitive (800HZ - 6kHZ), and where the majority of stereo effect is able to be generated (since inter-aural crosstalk screws up imaging in the lower midrange).
When a midrange driver can cover that wide of a range well, you're more likely to get away with using a harder cone woofer and a smaller tweeter. You may also be able to use 1 pole passive crossovers and have great results. Many tweeters generate significant distortion when run below about 5kHZ with a lower order crossover slope (coil tilt and I.M distortion). And that's right where the ear is most sensitive.
I could go on and on, but the one last thing I will say is that how a speaker interacts with the acoustics of the listening room is often the weakest link of the entire reproduction process no matter what else is going on. That should be one of the first things anyone thinks about when choosing a speaker design.
In the speaker cabinet, I consider it very important to glue 1/4 - 1/2 inch thick felt (or upholstery padding - which is ugly but cheap) to all internal surfaces of the cabinet, except the driver panel. I double it up in any corners. I then put in soft foam rubber in certain strategic areas, and only then the fluffy stuffing, and it's all glued in place away from the rear of the drivers, with silicone rubber cement, or Liquid Nails, or similar glue that never gets brittle. Doing this right may be what most speaker builders fall short on. You want to create a "gaussian arrestor" to dissipate that back energy without it causing any kind of abrupt loading change on the drivers. Don Davis of the Sound System Engineering seminar demonstrated how important this approach is many decades ago. Fluffing that isn't tight to the internal surfaces of the cabinet has far less effectiveness.
Another place many miss the boat is crossover design. They don't all realize that the impedance of any driver may well be substantially different than the nominal rating, at the frequency they want to cross it over at. A 5 inch driver I was using, rated at 8 ohms, measured 15 ohms at about 3kHZ, where I wanted to cross it over to the tweeter. Not only will that necessarily throw off an 8 ohm crossover design amplitude response, but phase shift is then way off too, so you'd want to swap the phase on the various drivers in a multi way spkr system to see which way has less of a cancellation at the crossover frequency. So stock passive crossover products are a joke.
Preliminary testing of drivers should be for efficiency matching (you are likely to need to add R's to get that right, which changes the calculations for the reactive componenets of the crossover), and impedance at the frequencies you want to do the crossovers at. Measured impedance + efficiency matching resistor = actual impedance for calculating reactive components (coils and caps).
When done putting a spkr system together, you want to be able to turn off all but one driver at a time, so you can check with pink noise, a cal'd mic and RTA that each driver is rolling off as per your design, before turning them all on to check level matching. With an active crossover things are likely to be very accurate. With a passive crossover things are unlikely to be very accurate, but could be close enough if you did everything else right.
As for speaker drivers themselves, the harder the diaphram, the better the resolution, but hard diaphram drivers usually have severe resonance issues, often at the frequencies where the ear is most sensitive. I feel that hard diaphram drivers require 4th order crossover slopes (24dB/octave), which can really only be done well with active crossover electronics ahead of the poweramps.
With woofers, the Xmax spec is one of the more important things to consider. With midrange and tweeter drivers, off axis response and range of flatness are maybe the most important factors. Distortion too of coarse. I really like the Peerless TG9 or Vifa TC9 three inch drivers because they are almost ruler flat (in a good sub-enclosure) from 350HZ - 10kHZ (The 8 ohm Peerless TG driver is ruler flat to 15kHZ). This means I can keep crossover points away from where the ear is most sensitive (800HZ - 6kHZ), and where the majority of stereo effect is able to be generated (since inter-aural crosstalk screws up imaging in the lower midrange).
When a midrange driver can cover that wide of a range well, you're more likely to get away with using a harder cone woofer and a smaller tweeter. You may also be able to use 1 pole passive crossovers and have great results. Many tweeters generate significant distortion when run below about 5kHZ with a lower order crossover slope (coil tilt and I.M distortion). And that's right where the ear is most sensitive.
I could go on and on, but the one last thing I will say is that how a speaker interacts with the acoustics of the listening room is often the weakest link of the entire reproduction process no matter what else is going on. That should be one of the first things anyone thinks about when choosing a speaker design.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.