Bandpass 4th

[Norcal Colby 0]

Hey guys. Looking to do my first bandpass 4th order. The reason for this enclosure would be to push all pressure into the cabin of a car through a ski port behind the center rear seat. I understand how vented and sealed enclosures are designed, however bandpass 4ths seem to confuse me a bit. I have searched and found many sources, but I get strange or mixed answers. Please help me clear up some things.

I am looking for a transient response with good output. Hoping to get the lows in my passed frequency without sacraficing too much of the 60-80hz range. I have heared that the sealed portion will mainly control the Low Pass and the ported portion will mainly control the High Pass. So if I have an SSA Icon 12 and use, lets just go simple here, a recomended 1.25cu' sealed enclosure and play into a, again simple, 2.25cu' vented enclosure tuned high, would that sound right? I have had no luck with my calculator because for me to design I feel like I first need to understand.

So please, what controls the Low Pass and High Pass and how should I design the volume of each chamber (I have been told to use recommended specs for the sealed and vented but just use a different tuning frequency).

if you know of good links i will appreciate those aswell

[ricksi30 923]

What do you drive? Is it the town car in your description?

[Norcal Colby 0]

sorry this enclosure is not for me. it is for my brothers lexus gs 300. i am going to instal it for him. i also won him the subwoofer through bryan on modified

[ricksi30 923]

Awesome! Have you considered a standard ported enclosure?

[Norcal Colby 0]

Awesome! Have you considered a standard ported enclosure?

i have, but my main goal is to keep the pressure in the cabin so the lows dont have to travel through the seats and also to keep trunk flex down a bit. if i dont go with this enclosure then i def will go with a low tuned vented box

[ricksi30 923]

Sounds good! For what it is worth I have had great success with the icon ported tuned to ~ 34 in my sedan.

[Norcal Colby 0]

thanks man. ill keep that that in mind

[dbsupra 69]

I don't know squat about designing it.

I just wanted to post to thank you for calling it a 4th order bandpass.

I see all the time people calling it just a 4th order and that drives me insane like nendo does mead.

[ssh 924]

Download Winisd, copy driver t/s specs into new file, then you can play around with the chamber volumes, and port tuning.

[Kangaroux 47]

Download Winisd, copy driver t/s specs into new file, then you can play around with the chamber volumes, and port tuning.

I wouldn't even bother with winisd. The second you put your enclosure in a vehicle that entire response changes. The only thing that winisd is useful for is seeing what variable effects what. With that being said, bandpasses can require a lot of tuning to get right, or sometimes you get lucky and they sound amazing on the first build. I suggest using an aero port or a sonotube or something that can easily be removed to adjust tuning. Sealed volume isn't too hard to get right, but be careful not to go too large on your ported side because it can be too peaky for your (or your brother's) liking. Stick with like a 2 - 2.5:1 ratio and play around with the tuning and you should be all set.

[ssh 924]

Download Winisd, copy driver t/s specs into new file, then you can play around with the chamber volumes, and port tuning.

I wouldn't even bother with winisd. The second you put your enclosure in a vehicle that entire response changes. The only thing that winisd is useful for is seeing what variable effects what. With that being said, bandpasses can require a lot of tuning to get right, or sometimes you get lucky and they sound amazing on the first build. I suggest using an aero port or a sonotube or something that can easily be removed to adjust tuning. Sealed volume isn't too hard to get right, but be careful not to go too large on your ported side because it can be too peaky for your (or your brother's) liking. Stick with like a 2 - 2.5:1 ratio and play around with the tuning and you should be all set.

So you would rather pick arbitrary numbers or ratios then at least see what the modeled response is out of car? That makes a lot of sense.

Edited August 25, 2012 by stefanhinote

[shizzzon 461]

In actuality, if you do model a 4th order in winisd, you will NOT want your response to be flat...

If you wanna get super crucial about it, it's still highly advised to model in winisd to get a base.

THEN.. with an already active setup in the vehicle(sealed, no influence with vent allowed), measure cabin gain in vehicle.

This requires an spl meter and patience. It takes a couple hrs of measuring and writing everything down.

Then based on info you obtained, make a graph of your cabin gain.

Then you can use that graph on top of what Winisd says and blend the two together to attempt a theoretical actual response curve where Winisd will then more accurate than just using it by itself.

[MikeMartel 64]

Fancy shit Sean.

[ssh 924]

In actuality, if you do model a 4th order in winisd, you will NOT want your response to be flat...

If you wanna get super crucial about it, it's still highly advised to model in winisd to get a base.

THEN.. with an already active setup in the vehicle(sealed, no influence with vent allowed), measure cabin gain in vehicle.

This requires an spl meter and patience. It takes a couple hrs of measuring and writing everything down.

Then based on info you obtained, make a graph of your cabin gain.

Then you can use that graph on top of what Winisd says and blend the two together to attempt a theoretical actual response curve where Winisd will then more accurate than just using it by itself.

Yep yep. Although I'd rather start off with a flatter response then peaky, then bank on cabin gain nulling it out. EQ could help, but the less the better.

[shizzzon 461]

This technique also comes in handy when you are burping for competition too.

Let me explain-

Let's say someone told you to find your car's peak by using a sealed box, etc.. bla bla bla.

So you do and it's 47hz.

So, you think for the rest of your life in this car, 47hz is the only frequency to aim for.... WRONG.

It was only the PEAK for 2 given reasons-

1- it was the peak found out of a whole array of frequencies.. But what were the DB values of all the other frequencies?

2- it was the peak found based on the displacement of the enclosure used and the remaining cabin volume at play...

So lets say you go and build a wall now..

Well, hehe, the remaining cabin volume will be a LOT smaller so the peak resonance in the vehicle will go UP.

So, let's say it went up to 51hz and you know it will go up..

So you measure with the wall sealed and sure enough, it went up.. in this example, it went up to 51hz.

So now you think every wall build needs to peak at 51hz... Lol.. no.

If you take a cabin gain graph of the vehicle....You may see other frequencies compared to 51hz only lower in amplitude by less than 0.1-3.0db.

What's interesting about this is this-

Even though 51hz is the most EFFICIENT peak, it can be overcome by forcing the enclosure to peak somewhere else and developing enough amplitude to overcome the efficient peak of 51hz in this example.

Why would one do this?

What if you didn't have the depth to tune low enough to get to your efficient tuning? You can only do large port area, but not depth?

Large port area = higher tuning = less depth of port.

Just some things for you to think about.

[///M5 2,833]

In actuality, if you do model a 4th order in winisd, you will NOT want your response to be flat...

Good luck getting it flat anyways. Huge rolloffs to be dealt with.

And to the OP, if you want 80Hz and the lows, you don't want a bandpass. By nature the passband is pretty small.

[Kangaroux 47]

Download Winisd, copy driver t/s specs into new file, then you can play around with the chamber volumes, and port tuning.

I wouldn't even bother with winisd. The second you put your enclosure in a vehicle that entire response changes. The only thing that winisd is useful for is seeing what variable effects what. With that being said, bandpasses can require a lot of tuning to get right, or sometimes you get lucky and they sound amazing on the first build. I suggest using an aero port or a sonotube or something that can easily be removed to adjust tuning. Sealed volume isn't too hard to get right, but be careful not to go too large on your ported side because it can be too peaky for your (or your brother's) liking. Stick with like a 2 - 2.5:1 ratio and play around with the tuning and you should be all set.

So you would rather pick arbitrary numbers or ratios then at least see what the modeled response is out of car? That makes a lot of sense.

The response changes so much when putting it in the vehicle that modeling it up isn't worth the time...I've modeled up stuff before that looked absolutely god awful on the graph but once it was in the vehicle it had a solid flat response. No I'm not saying pick arbitrary numbers but what I am saying is don't rely on the graph for what your setup is going to sound like...bandpasses are not easy to get right the first time which is why making them easy to tweak and tune is crucial if you're looking to get the most out of them

[///M5 2,833]

No I'm not saying pick arbitrary numbers

So then how and what would you pick?

[shizzzon 461]

You can't just use winisd and you can't just use what happens when you install it and turn it on.

If you truly want to be able to at least forsee or expect some sort of response, you must do both and overlap their results together.

98% of us on here would probably careless about taking the time to do this.. but i guarantee you if you did it just one time, you will enjoy learning what is happening by actually being able to mix experience with what's on a computer program.

[95Honda 486]

It would be wise to model 100% the time first, this way you know how the driver will behave in the enclosure you design. You know what the impedance, excursion, power handling and anechoic response should be. After you determain this, you can correlate what your overall system response will be inside your vehicle if you have an idea of the amount of cabin gain you will have in your vehicle.

Contrary to popular belief, little in the way of impedance, excursion and actaul output of the subwoofer system itself (minus cabin gain) changes at all inside or outside of the confinement of the car/truck. This has been measured and published in resources such as the Loudspeaker Design Cookbook.

Arbitrarily choosing rations, volumes and tuning is completely useless without modeling. Anyone who does this hasn't got a clue about what is actually going on and is purely lucky if they get favorable results...

[Norcal Colby 0]

wow, thanks for all the responses guys. i like to model so i can see how my subwoofer will like the enclosure. i know that once placed in the car, to the ear it will always sound diffent and in some bigger builds the car itself will act as almost part of the enclosure. i have heard of some pretty musical 4th order bandpasses before so ill keep seeing what i can do. as far as tweeking goes ill probably be doing a removable top using weather stripping and screws to seal the gap so i can test, pull off play with port and stuffing, putt back on, test... ect. i do not have access to any spl meters or term labs, but hopefully if i land a shop job soon i would have some equipment at my disposal

[Impious 1,476]

Arbitrarily choosing rations, volumes and tuning is completely useless without modeling. Anyone who does this hasn't got a clue about what is actually going on and is purely lucky if they get favorable results...

Or they waste an unnecessary amount of time "testing" in a vehicle when they could have eliminated a majority of the leg work by modeling first and applying cabin gain to the model to estimate in-vehicle response.