About using "FLARED" ports (1 Viewer)

[Kerry Hackney]

As always, the more I learn about building speakers the more I realize I don't know. I am building an Adire alignment vented Tempest enclosure. I received the two 3" flared ports via UPS yesterday. Included in the box was the port instruction sheet that included the formula for calculating port length. As I read through it I discoverd that when using a flared port like the ones supplied by Adire or Madisound the finished port should have an overall length ONE INCH LONGER than calculated. I wonder if this would account for the tuning error in the 281 speakers tested by Joe DAppalito?

In the Tempest application sheets, Adire says that when using their 3" flared ports the center tube should be 11" long. I am fairly sure that they are taking this extra inch into account. However, in the original 281 instruction, I am thinking they listed the port length without regard to which port is being used. I know that there are other builders who have made ports in the past. Ultimately, I have observed that calculated port lengths are "RARELY" exact. Depending on driver variation from spec they can be off quite a bit.

Impedence sweeps or direct observation with a tone generator is a good sanity check if you have the equipment to do it. You may well be surprised at what you find.

 

[Jack Gilvey]

Ultimately, I have observed that calculated port lengths are "RARELY" exact. Depending on driver variation from spec they can be off quite a bit.

Impedence sweeps or direct observation with a tone generator is a good sanity check if you have the equipment to do it. You may well be surprised at what you find.

Yeah, this is much more of an issue with ported than sealed subs. It's probably best to start a little long and work your way down if you've got a way to do it. I don't think the sample variations among quality drivers like Adire's are enough to noticeably throw off an alignment, though, even if it may be detectable with a sweep. I'd assume this, anyway.

 

[Jon Hancock]

Hi guys,
The box Fb won't be affected by changes in the driver Fs and compliance; just the reponse and shape of the impedance curve will change (particularly the relative height of the upper and lower impedance peaks.
It's important to be sure how the calculated port length is derived. Many programs (LSPCAD for example) can be set based on using port flares or not, so if you follow the instructions packed with many of these port flare kits, you'll wind up with a port too long. In testing my little 8" two way design, even designs calculated with the Unibox default settings worked correctly using the total length *including* port flares, which even surprised me- since I'm pretty sure my box volume calculation (48L - 4 liter for driver, port, crossovers) was accurate. So, hitting 32 Hz with the 3" flared port required using the overall calculated length with no modifications.
The unforunate reality is that if you want to hit a certain number on the head (and that's also a debatable goal- can you hear the difference between an Fb of 32H and 33.5Hz?) you can only do so much by calculation- then you are left with the necessity to measure and modify as required.
Best regards,
Jon

 

[Kerry Hackney]

Jon is absolutely correct, box tuning is a mechanical calculation based on the volume of the enclosure and has nothing to do with the electrical/acoustical properties of the driver. I have at times had trouble finding the exact displacement for drivers and used the spec for a comparable known driver. As I think about using Bass Box, LSPCad, Winspeakerz they all come up with different answers for port length. Also, isn't there some consideration to height above sea level that will effect the outcome as well?

 

[Jack Gilvey]

The box Fb won't be affected by changes in the driver Fs and compliance; just the reponse and shape of the impedance curve will change (particularly the relative height of the upper and lower impedance peaks.

Oh, absolutely, Fb is fixed by the box and port. I was thinking more of the possibility of the response shape possibly not being what was hoped and that experimentation with Fb might be in order once the driver is actually in the box (if there's no way to test the individual driver sample beforehand). As I said, though, it probably wouldn't be enough to really throw things off (unless you wanted to hit a specific number like Jon mentioned). I would suspect that variations in Qts might be most likely to cause problems if run into, at least based on Vance Dickason's graphs.