Need help with Modeling Software

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hey Guys,

I have been trying to model up a subwoofer for my friend using the Tangband 8" W8-740c. The parameters for this woofer can be found here and here. We decided to use the plate amp from Partsexpress
here. We are looking to do a vented design with a long port to save money over pr's. I have tried to model the sub in "subwoofer simulater" and "winisd" but I'm not having much luck. I think I have all of the parameters entered in correctly but I wasn't sure about the "PE", "Z", or "dia". Is that the diameter of the voice coil or of the cone? Also for Le does the software want it at 1khz or 20 khz?
My goal in using these programs was to determine necessary box size and port length, and hopefully get an idea about how low it will go and how flat it will be. I was also confused with winisd about the power tab. I entered in 100 watts as opposed to 1 watt and the spl graph went off the chart, is that supposed to happen?
I appreciate the help, hopefully I can get up to speed with these programs.

George
 
Probably not a good idea to use that sub without a PR.

LE won't matter much so it doesn't matter what you enter.

To get a response shape that works OK with that bass amp - it has ~6dB boost at 30Hz - I tried a 12 liter box tuned to 28Hz with a 2" diameter by 24" long port. Port velocity reaches 40+m/s, which is high and you will likely have noise, even if flared. Using a 3" port 55" long, your port velocity goes down to 17m/s, which is better, but the port is ridiculously long...

Tuning small boxes to low frequencies is a job for a passive radiator.
 
The only parametes you need to model sub volume
are qts, fs, vas.

For tuning & vent velocity you will need power (pe) & Qes.

I get 10litres tuned to 40hz using winisd with a port 2 x 13"yields flattest response

This driver is very inefficient at 84db & low Re so will require high power & stable amp.

The parts express one looks up to the taks but hard to comment unless tested.
 
diyAudio Moderator Emeritus
Joined 2001
Wright said:
Hey Guys,

I have tried to model the sub in "subwoofer simulater" and "winisd" but I'm not having much luck. I think I have all of the parameters entered in correctly but I wasn't sure about the "PE", "Z", or "dia". Is that the diameter of the voice coil or of the cone? Also for Le does the software want it at 1khz or 20 khz?

George

Since I haven't modeled the TangBand, I cannot comment on the advisability of using a port or Drone Cone, (the original term for Passive Radiator, coined by it's inventor, Harry Olson), but I can give you the terms.

Pe = Max Continuous Power, (not instant peak power).

Z = Nominal impedance. Generally either 4 ohms, 8 ohms, or nowadays, even 6 ohms. Never 4.7 ohms or 6.8 ohms or whatever. It is the ohmage the manufacturer gives to fit into the 4 ohm/8ohm mold.

dia = The diameter of the circular frame around the cone. A speaker with a 10 inch circular frame and a 8½" cone will be called a 10 incher. It corresponds to the popular 15", 12", 10", 8", and 6½" that everyone uses anyway.

The actual cone itself is generally given as "Sd", and that is the area of the cone surface itself. It is a different number entirely from "dia".

Le is generally given as the Le at 1,000 Hz.
 
Applying boost below the tuning frequency is a no-no ;) Tuning to 40 Hz, no matter how flat it might appear in a simulator that doesn't account for amplifier response, is likely to cause overexcursion below resonance. A 2" port in a 10L box would have a velocity of ~47m/s when tuned to 40Hz, with ~35V input. Pretty cool if you want port noise and be able to blow out candles at 30 paces.

Without a PR, I'd probably use a small ~5-7L sealed box.

It'd make a nice (small) car woofer. There are better choices for use with a plate amp in a home.
 
Thanks!

Wow thanks for all the helpful replies! It is good to know what those t/s parameters mean. I wasn't having much luck with the other two programs, but I did try unibox, and found it to be more user friendly. I enjoyed entering some values and seeing it calculate the rest of them right on the money! (I am easily amused). The only thing I'm not fond of is that I can't figure out how to enter my own port length; seems the program does that for me?
Anyway about the subwoofer: I was talking to Al (RAW) before and he also encouraged me to use a passive radiator/drone clone, but since they run $50 for the high excursion version it is over our budget. He made it sound like I could go ported if I was desperate but perhaps it's a bad idea. Considering our target price range of $30-$50 for the driver, do you guys have any other driver suggestions that I could run in a ported box? The application is computer system/HT use to be coupled with a pair of SuperElf's and a pair of Elf1.0's in the rear. I would love to try and get down into the lower 30's (frequency, not price) if possible, but that might be asking to much of a small driver. Amp is as stated above, unless you have other suggestions at or below that price point. I really appreciate your suggestions, and I'm sure my buddy does too. ;)

George
 
diyAudio Moderator Emeritus
Joined 2001
How about a Peerless CC 260, or 831727 in a 28 liter, (1 cu ft) box, tuned to 30 Hz?

The box is only a little bigger than your planned box, the 10 inch has a pretty long throw. I will explain later, but it has a symmetrical magnetic system that in a ported box, will give much less distortion than a non-symmetrical system-and the TangBand does not mention a symmetrical system, which means it probably does not have one.

The Peerless is more efficient.

The simulation is by Subwoofer Simulator, written by our own F4ier.

The lower line is the response not counting the room, like most modeling programs do it.

The upper line includes standard room response-actual room response may vary.

The room response was an option added by our own member, Serow.

As you can see, at 30 Hz, even without room response, you are down to around 80 dB.

With room response, you are up to around 86 dB-more efficient than your TangBand even in the midband!

Price isn't much more than the TangBand.
http://www.partsexpress.com/pe/pshowdetl.cfm?&PartNumber=297-616&DID=7

You can always make the box bigger if you want, for even better response.

To tune it so low, you might get some port noise, but not too outrageous. This is a computer system, right? And a slight increase in box size leads to less port noise, if you want to get bigger.

Nice deal, I think.
 

Attachments

  • peerless cc 260 1 cu ft.gif
    peerless cc 260 1 cu ft.gif
    27.7 KB · Views: 360
The room response was an option added by our own member, Serow.

Hi kelticwizard, once again thanks for the vote of confidence on SubSim, but what did you mean by this? Room response plotting has always been a feature of Subwoofer Simulator from day one :)


Wright, with which part of SubSim did you have trouble while designing your sub? It would help me improve the user interface of SubSim if you could be more specific. I know a number of SubSim users find the interface to be odd at first. While they email me for feature suggestions, no one really comments on the user-interface :confused:


Cheers,

Isaac
 

Attachments

  • in-room.png
    in-room.png
    17.9 KB · Views: 332
Isaac,

I've been messing around with your program some more and now that I know the parameters it's easier to use. I did still have some questions though.
First, under vented box details, I didn't understand the terms ql, qa and vb; I'm assuming the latter is box volume?
Under vents, is Av vent area or diameter? And I am assuming lv is vent length...what is qp?
Also, I couldn't get the optimize tab to work at all, is it functional?
It was interesting to note that, using the sub sim software, I couldn't get the port velocity to rise above 4ms or so, so I must be doing something wrong.
I was also unsure about how to factor in the amp wattage.
All in all, I felt your interface was fine, I just don't really know what I'm doing. Thank you for devoting your time to providing such a free and useful resource.
Oh and since you mentioned feature suggestions, I think it would be cool to add a definable plate amp boost at a certain frequency and dB level.

All,
Concerning the peerless 10", I modeled it up and here is what I came up with.
An externally hosted image should be here but it was not working when we last tested it.


I'm trying to get this to work with the 5dB boost at 30hz, so it will be flat to 30. Since I couldn't figure out how to factor in power with this sim though, I'm not sure if my amp will still work with this driver. I appreciate your suggestions. Thanks,

George

PS. Any good way to add a little bit of boost to the 45hz region? I suspect my friend, who isn't yet into hi-fi, might not like the response as flat as I would.
 
Wright,

For Q values, have a look at this thread and yes, Vb is the net box volume.

Av is the vent radius while Lv is the vent length. The optimize routine is functional, but only in my personal prototype version -- not yet in the public version. The default input voltage is 2.83Vrms, which is probably/partly why port speed is so low. Increase this input voltage to see if the port is good enough to keep vent noise to a negligible level.

At the time of development, I did not include the option to input amplifier power for educational purposes. 99.999% of audio amplifiers are almost perfect voltage sources; meaning amplifier output voltage is constant regardless of load impedance. Since power = voltage^2 / impedance, and a speaker's impedance varries with frequency, actual applied power is not constant. Despite what a majority of hobbyists think, amplifier power is far from constant across the passband. Therefore, it would have been a bit misleading to let the user specify a constant amplifier power output. While other programs such as LspCAD let the user specify amplifier power, one can notice that the value is merely used to solve for amplifier voltage (prior to calculating system frequency response).

Thanks for the suggestion, you'll be pleased to know that amp EQ (as well as a lot of other features) is already on the features list. T-S parameters measurement is all that's left before I piece everything together.

Cheers,
Isaac
 
diyAudio Moderator Emeritus
Joined 2001
George, with all respect, what exactly is it that you are looking for from the forum with regards to this program and your project?

You asked for a suggestion from the forum for an alternative driver to the TangBand, which you proposed to put into a box slightly under one cubic foot. You wanted a driver in the $30-$50 range. You hoped for performance down to the high 30's range, though you doubted you could get it.

My suggestion was a well-known driver, slightly over $50, which , when put into a 1 cu ft box, will give performance superior to your proposed TangBand down to 30 Hz.

Next, I am reading that you feel that since Subwoofer Simulator does not presently have a capacity to chart filter performance, and you have an amp which you can add a +5 dB boost at 30 Hz, you do not know if this woofer will "work" with your amplifier. You also show a chart in which the woofer is put into a box of over twice the volume you were originally talking about!

Well, the woofer will indeed "work" with your amplifier. Hook the amp up to the finished speaker, and bass will indeed pour forth. If you add a +5 dB boost at 30 Hz, just imagine the chart 5 dB higher at 30 Hz than it presently appears. About the only thing I would add is to make sure any boost is above the tuning frequency.

People have built subwoofers on this forum ranging from half a cubic foot to 13 cubic feet. There are hundreds of drivers out there that could qualify as subwoofers of some sort.

In the beginning, I thought you were making it clear that you wanted a subwoofer box below 1 cu ft, with a driver between $30-$50, that could be hooked up to your specific amp to give good performance into the high 30's. There are darn few out there with the requisite price, parameters and air moving ability to do just that. Now I find out that because SubSim does not presently model amplifier boost, you don't know if the speaker will "work" with the amp.

Please let us know approximately what size you want this sub to be, (it doesn't have to be exact), give us a price range, and we will see what we can do. Moving box volumes around wildly and expressing doubt to the viability of a driver choice because the simulator program does not presently model a +5 dB amp boost serve only to confuse the issue.
 
KelticWizard,

My goal was originally to keep the enclosure size around one cubic foot. However, since you mentioned yesterday that I could increase the enclosure size for better response, I spoke with my friend and indicated it would not be a problem. So, last night I spent some time plugging in larger values for volume until I ended up with something that would give me a flattish response when the boost was added. My question about "working" with the amp was referring to being able to give the woofer enough power to use most of it's capability, and whether a 5db boost at 30hz would indeed give me the flat response I was looking for down to 30hz. I'm not sure on the nature of boosts and whether it would flatten out the whole curve down to 30 or decrease and go back up right at 30hz. So, more than anything, since this is my first time working with the software and I want to get the design right, I was looking for validation from the more experienced members of the forum. Here are the official goals of the design at this juncture:

- Flat down to ~30hz (disregard earlier comment about 45hz
peak)
- Size can range from less than 1 cubic feet up to 2.5 cubic feet.
- Amp will be as shown above.
- driver can be anything from 8-12" and price range from $30 to
$65.
- I basically have to conclude my design process by tomorrow
morning when I will get access to a woodshop and must begin
construction.

So far it seems that using the peerless 10" in this volume will work alright, but I would appreciate your thoughts on the overall design, given the criteria above.

I have not been able to test port velocity because I can't figure out which voltage this amp provides. I am assuming it is different from the 110 voltage supplied from the wall socket. Is there a general range for amps this size that I can use to model? I tried plugging in 15 volts and got ~29ms, which may still be too high...
What is considered a healthy value for port velocity?
Thanks for the help,

George
 
Ok, I noticed the 35v value Ron mentioned above and upon plugging it in, got a port velocity of over 60! :eek:
To get the velocity down I changed the port to a 2" radius 15" long and I got the velocity down to ~17ms. Is this low enough or will there still be port noise? Thanks,

George

Edit: Also I noticed that you mentioned boosting below tuning was a no-no. What about boosting right at tuning or just above?
 
diyAudio Moderator Emeritus
Joined 2001
Wright said:

Edit: Also I noticed that you mentioned boosting below tuning was a no-no. What about boosting right at tuning or just above?

Okay, now I know what is going on. You really had me guessing there for a little bit. :)

Actually, 2 Ft³ or a little above is a pretty good size. The old AR's and Advents-standard size in the seventies-were 2 Ft³. And everybody had two of them. So this is a good acceptable size.

Boosting at tuning is ok. If possible, boost a little above. Your speaker begins making large, wild excursions below tuning, so try to give the boost a chance to come down a bit before entering into the "below tuning" range. You can always tune just a bit lower than 30 Hz if you want.

A word on Peerless speakers. I have good luck with them. However, their Thiele-Small parameters tend to be different from what is listed. However, most of the time, the variations cancel out and when placed into a box, the performance is about the same as it would be with the listed parameters. Most of the time. Sometimes you have to send one back.

Just a warning. But I still recommend Peerless. Overall, I think their price/performance ratio is best.

Generally, I use winISD for port tuning velocity, because they give it in terms of Mach, (speed of sound). Generally, anything under .1 Mach is reasonably OK.

It should be noted that flared ports, on both ends of the port, are generally regarded as giving performance equal to a straight pipe of twice the cross section. Not twice the diameter-twice the cross section. see if you can fit them into your budget-they help a lot keeping your port size under control.

One more thing. About this "flat curve" thing. Under all circumstances, remember that any modelling program is at best an approximation. Your actual performance in the room will not only vary from room to room, it will vary from one spot in the room to another spot-and sometimes these spots are only a foot or two away. When we model, we are just shooting for a chance to get the performance in the room-nobody can guarantee anything. A speaker that in half-space, (against a single wall, no other walls, the way most modelling programs do it), which shows a 3 dB down point of 30 Hz has a better chance of giving good low performance than a speaker that models a 3 dB down point of 50 Hz. But these curves are all subject to change by the room.
 
diyAudio Moderator Emeritus
Joined 2001
According to WinISD, in a 60 Liter box, a 4" diameter port of 18.5 in length, (better put an elbow in there), will tune out to 27 Hz. A 4 dB boost at 30 should make it approximately flat all the way down to 30 Hz.

This is not counting average room gain.

The port itself should take up about 3.5 liters. Better make it a 65 liter box-I think that is still within your parameters.

Votage? Your amp is rated 170 watts into 8 ohms. The Peerless is rated 8 ohms. Under the formula Watts = V²¸R, we find that voltage should equal 37 Volts. That should be our drive voltage.

Below is a graph, at 37 Volts, of the Peerless CC 10", tuned to 27 Hz. The lower right line is the SPL without room gain, the upper white line is the SPL with room gain, and the purple line is the cone excursion.

As we can see, under full power we are living somewhat dangerously with cone excursion, (purple line), in the 42 Hz range. Max linear excursion is 9 mm. There should be extra in the non-linear range-but 15 mm might be coming fairly close to the limit, if it doesn't exceed it.

But take a look at the output, even without room gain. 108 dB @ 30 Hz. 112 dB with room gain. Pretty good for a single $54 woofer and a 2 Ft³ box.

You might want to forget the boost at 30 Hz. And if your friend's system has an equalizer, you might want him to back it off slightly at 40 Hz setting just to give the woofer a little more headroom at 42 Hz, where excursion is highest. You are already up to 112 dB without room gain in that range anyway, so you can afford to reduce things a couple of dB.
 

Attachments

  • peerless 10 cc line 60 liters excursion.gif
    peerless 10 cc line 60 liters excursion.gif
    21.5 KB · Views: 133
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.