which WinISD do you have? get WinISD pro alpha. its free and the best ver to use.

to make a slot port, click the "vent" tab, then click on the "circle" icon to make it a "square" then change the dimensions for a rectangle.

 

That's what I was thinking, but my port needed to make two turns inside the enclosure, would that matter at all in the program?

 

Nope, all it cares about is length.

sent from my phone using digital farts

 

if the slot port uses three enclosure walls as it's own walls, then WinISD does not model the end correction correctly. In my personal experience, for a given pre-determined vent area, the slot port length needs to be reduced 40%, which is a good thing because you can fit in a very large port area for lower port compression and still achieve your target tuning frequecy without the slot getting ridiculous in length.

 

how so? It doesnt know anything about the enclosure shape. just enclosure volume and port volume. thats it! if you want to make an end correction, then change it! its right there in the vent tab.

what do you have to backup that you need 40% less port length?

 

x2^, but I think 40% shorter than square is a little too much......Thats why I allways use square aero port....at least you know what to do

 

You think wrong. I have found this through empirical, real world testing. It may not be 40% ON THE DOT, but it's right around there.

 

ok, I am not against the idea that a modeling program could be flawed. will be interesting

 

haven't forgotten, just long days at work. Will try to get the info/pics posted tonight.

 

Here is the ugly SOB

Port opening is 3" x 12". Obviously those are old pictures, and I realize I'm missing some pictures with more measurements, but rest assured, I WILL post it all---nothing to hide here.

 

click square. type in 3" by 12", enter in desired tuning, should give you length. then make sure youre in the 14-16sqin per cube range.

 

try reading this whole thread, thanks.

 

Alrighty, sorry for the delay, been busy working.

Here are some pics to validate my claims. 1.9 ft³ net, 3x12x33.25" port.

Granted, my initial claim of 40% reduction was off, but none the less, the end correction factor does get skewed.

Who wants to post up what WinISD says 1.9 ft³ with a 3"x12"x33.25" "should" be tuned to?

Then I'll post up the actual impedance plot.

Minibari, I'll be awaiting your response.

 

I will model it tommorow at work. My home pc died :/

sent from my phone using digital farts

 

want me to ruin the surprise?

 

Nah, i like suspence, lol

sent from my phone using digital farts

 

ok, so according to WinISD, FS for the box would be 36.5hz

Only caveat I see might be the subs FS, for using the impedance plot. but go ahead and show it and lets see how far off it shows.

 

I'll post up my screen shots after I get home from work. You might want to double check your WinIsD inputs for VB and Fb that correlate to my actual size port. Your 36.5Hz Fb is incorrect according to WinISD pro on my computer.

Also, I dont know what you mean by "caveat might be the subs Fs..."? Can you elaborate more on what you mean?

 

I did a lot of research on port length for rectangular ports a while back.Almost everything I found said to reduce the length by 1/2 the port size for ports where there was 1 wall of the box being used and 1 port size if 3 walls of the box where being used.
I built a few boxes using this theory and found that I came within 1 hz of the tuning frequency I was shooting for in all of them.
I believe that in very long ports it is less of an issue then very short ports because the longer ports where closer to 1/2 a hz and the shorter ones where closer to 1hz.All the ones I did where longer then 12" so if the port where just a few inches long I would think trial and error would be needed.

 

BBP6 claims that box should be tuned at 38.1hz

 

I just put in 1.9 cuft and the port you specified. (3 x 12 x 33.25)

I am using WinISD pro alpha .50a7

I get 36.5hz

If you tune your port right at FS of the woofer, it may skew imp plot a bit.

so what is the answer?

 

The impedance plot as a whole is irrelevant, tuning occurs at the local minimum between the two impedance peaks on the plot, period. That would be like saying the speaker "influences" the box tuning frequency, when in reality it is speaker independent.

Since you like suspense, I'll post the info later today.

 

subbed to see the results

 

Alrighty, I triple checked, and I get 36.9 Hz tuning using the specs I have provided proof of, using WinISD. (I get a vent length of 33.29", which is close enough to 33.25" since it is well within the margin of error).

Here are the simulation plots:

Here is the actual impedance plot with the Parts Express Woofer Tester/DATS system.

If you go back to WinISD, and set the Fb to get ~31.6Hz, you get a required vent length of 47.19". Yet my slot loaded vented box got there with only ~33.25".

For those who are not mathematically inclined, 33.25/47.19 * 100% ≈ 70 %.

Yeah, I was off initially because I said ~40% reduction----I'm getting old and some brain cells aren't quite as active as before, but none the less, a 30% reduction in required length is a big difference because it allows you to go BIGGER on the port area (without the penalty of a ridiculous vent length) to kill (lower) vent velocity and maintain port linearity to much higher levels.

Of course you have to factor in the larger volume a larger port takes up inside the box, but IMO the benefits of the lower vent mach air speed outweigh the negatives.

 

This is irritating, then, lol. Because bb6 and winisd both have the wrong answer. Both use the formulas that thiele and small came up with (and are regarded as correct)

So what is wrong here and how do model it?

Does hornresp or leap do better?

Also, does reality and modeling agree if the port is round or doesnt share enclosure walls?

sent from my phone using digital farts

 

it's not that WinISD and BassBox are wrong, it is WE that are wrong for assuming that the results hold true for ANY port configuration. The numbers those programs spit out are generally for a free-terminating end (not using any enclosure walls for the port-walls).

Also, does reality and modeling agree if the port is round or doesnt share enclosure walls?

Absolutely. When you specify the correct end correction factor, a round or even square/rectangular port models very accurately, so long as it is at least 1 equivalent-diameter away from any internal surfaces. It's once you get a circular port downright into the corner, or a square/rectangular port sharing one or more walls that requires modifying the length via the appropriate end-correction factor.

Horn Resp and Akabak lead to the same results. I believe they both use the lump-element model for the acoustic mass, so unless one personally accounts for the end correction factor from empirical data, slot ports like the ones in this discussion will be tuned lower than calculated.

Modeling with free-ware like WinISD/HornResp/Akabak/etc/etc is not a problem. Just specify the Fb, and chuck off 30% of the calculated port length to get in the ballpark. If you want the actual seat # inside the ball park, then chuck off about 25%, then see what the tuning is at via the actual impedance plot and modify accordingly by progressively making it shorter and shorter. If you're old-school like me, you can use an end correction factor in the vent length formula k=2.227 that I found a couple years ago, and correlates nicely.

 

I Will have to remember this when giving out advise, lol.

Most of the ported boxes i have made haf ports under 12", so have not run into a huge error. Either that or they were round ports down the center

sent from my phone using digital farts

 

design length has nothing to do with anything. It's all about proximity to the enclosure walls. You can use the formulas found on Brian Steele's DIY subwoofer page, and simply set K=2.227 in the port formula if using slot ports that use 3 enclosure walls. This will get tuning very close.

 

Any additional thoughts on the subject? I am building a sub with a Dayton RSS315HO-4 and was intending on using a slot port due to the length.

If I understood what I read here, when using WinISD, I subtract 30%, right?

 

Only if it uses 3 enclosure walls as port walls, I'd say

 

I realize this is a very old thread but it is the only one I've found in which someone other than myself is confirming that winisd calculates the port length for tuning slotted ports too long. so I apologize for reviving this thread however I have a question for Oscar.

I have been building (modeling with winisd pro alpha) and testing (with a dats3) subwoofer boxes for a few years and every single one that I build ends up being tuned around 4hz-8hz lower that what I had intended. even with subtracting half the port width from from the port length (I.E. 2" x 8" slot port, half the width would be 4", therefore subtract 4" from port length), I still end up tuned too low.

Oscar I understand you to say that winisd is approximately 30% over on port length calculations for desired tuning of a SLOT port. based on my own experience I can agree with that statement. and yes, I am using 3 inner walls for my slot ports.

however my question is this...

does that 30% error include subtracting half the port width or is it just a flat 30% from what winisd calculates for the length?

also, are you changing the settings for end factor in winisd whenever you use it to calculate port length or just leaving it in default setting?

again my apologies for reviving this thread and thank you in advance