What program to model a push-pull sub??

[monoiz15]

Hey there, I've got a few of these old 8" Dayton DVCs that were used in the Triska sub:
http://www.parts-express.com/pe/show...80&FTR=295-480

I want to design a push-pull type box for a pair of them. I have a 200W plate amp I'd like to use to drive it, and like most of my builds, size is no constraint, BASS is ever important (not that I'll get much out of these, but I'd like for the box to be optimal).

Anyway, I was kinda thinking having the front baffle with the one facing out of the box on top and the inverted one underneath, but on an angle like in this very crudely drawn in MS Paint picture. The reason for the angle is half aesthetic, half trying to keep it from protruding the entire depth of the driver out of the front of the box. I am curious as to what programs you all would use to design a push-pull type sub. All input greatly appreciated

[billfitzmaurice]

Use the same program you would for any box, it's just a dual driver cab with one driver reverse polarity.

[aarond]

WinISD will specifically allows you to model a box w/ multiple drivers but all it really does is multiply the volume by however many drivers you are using. In a sealed box it really is that simple, it's the same for a ported box as far as volume goes but the change in volume requires a change in port dimensions to maintain a speced tuning relative to a single driver. WinISD will model whatever you need, or you could just use seperate chambers for each driver and treat them as two single.

Are you wanting to go ported or sealed?

[dwigle]

What's the benefit of this alignment (one woofer mounted backwards) with reverse polarity? Isn't that just two woofer, moving in the same direction in a sealed box?

Alternatively, what's the benefit of reversing a driver (and not reversing polarity) so that the cone of one woofer pushes the backside of the other?

[monoiz15]

Wow, smart moment (sarcasm). I should have realized. Ok, so I can use WinISD. I have only done ported designs so far, and since low end extension is first priority, I'll be staying ported, but I'd like to try a sealed box one day. Two questions about WinISD:

I don't know why every time I enter driver info it says "Driver data fails integrity check." I enter all the available data.

And when your choosing the alignment, I don't really understand what the options mean. I just choose extended bass shelf -6dB because I figure that's automatically accounting for BSC?? Sorry if I'm a noob.

[Wolf]

Alternatively, what's the benefit of reversing a driver (and not reversing polarity) so that the cone of one woofer pushes the backside of the other?

There is in essence no benefit aka full cancellation.

Later,
Wolf

[Jeff B.]

What's the benefit of this alignment (one woofer mounted backwards) with reverse polarity? Isn't that just two woofer, moving in the same direction in a sealed box?

There's a theoretical cancellation of second order distortion.

[Jeff B.]

Wow, smart moment (sarcasm). I should have realized. Ok, so I can use WinISD. I have only done ported designs so far, and since low end extension is first priority, I'll be staying ported, but I'd like to try a sealed box one day. Two questions about WinISD:

I don't know why every time I enter driver info it says "Driver data fails integrity check." I enter all the available data.

And when your choosing the alignment, I don't really understand what the options mean. I just choose extended bass shelf -6dB because I figure that's automatically accounting for BSC?? Sorry if I'm a noob.

There is no baffle step in subwoofers. You should stick with B4 or QB3 alignments for starters.

[monoiz15]

Thanks Jeff!

It won't be a problem using one amp channel to power four voice coils as long as my wiring is solid, right?

dwigle, a speaker is obviously not symmetrical on either side of the cone. The basic idea of a push-pull sub is that by reversing the direction AND polarity of the second driver, the cones are still moving together, but since they are facing the opposite direction, it supposedly cancels out any distortion caused by irregularities in the suspension or magnetic field of the drivers. This page may help you understand different sub types if you're interested:
http://www.danmarx.org/audioinnovation/theories.html

If you didn't reverse the polarity, it would cancel out everything, basically silent. But the way you asked made me think isobaric, which is acoustically coupling two drivers either cone to cone, or both cones in the same direction. The key is that there is a constant pressure (isobaric) between the cones. This cuts Vas in half, allowing you to make a box much smaller, at the expense of having two woofers limited by the Xmax of one. Sonotubes would be useful for this purpose, I imagine.

[monoiz15]

I can't seem to find a way to even get an F3 around or below 30Hz without going way over excursion at 150W.
Anyone wanna give it a shot?? I'm very new to design software and I'm just getting used to the relationships between different sizes and bass response. I realized you can decrease the port velocity by making it wider, but that requires it to be longer as well.

The main issue I've been coming up against is overexcursion, at max power it's ALWAYS waayyy over and I'm not sure what to do about that :/

[Chris Roemer]

Enter Qes and Qms, NOT Qts (set ISD to autocalculate missing values).
For power handling, you have to either enter the cone dia. or it's surface area (Sd). Thes are usually not given in Dayton's T/S parms.

It's a significant figures thing. ISD should be smart enough to not worry about some calc. being off by one number, four decimal places over, but it isn't.

[duanebro]

I can't seem to find a way to even get an F3 around or below 30Hz without going way over excursion at 150W.
Anyone wanna give it a shot?? I'm very new to design software and I'm just getting used to the relationships between different sizes and bass response. I realized you can decrease the port velocity by making it wider, but that requires it to be longer as well.

The main issue I've been coming up against is overexcursion, at max power it's ALWAYS waayyy over and I'm not sure what to do about that :/

You are learning the limits of the drivers. As you increase the box size or lower the tuning you are also decreasing power handling. You are using 8" subs, trying to go below 30 Hz is not going to work very well with these. What are you going to be using these for?

[monoiz15]

I probably would have never realized it was a problem with the autocalculate. I tried with and without every combination of parameters I could think of, but I still wasn't able to make it pass the check. I'm using this spec sheet:
http://www.parts-express.com/pdf/295-480s.pdf

I went along anyway, hoping this would all be accurate and I came up with a 4.5 cu.ft. box tuned to 29Hz with an F3 of about 27Hz and should be able to produce 107dB with 50W. This was with the help of Ray Alden's Speaker Building 201 and WinISD. I had been planning on using the plate amp because I figured I would be able to put at least 120W RMS into a pair of them. At the moment I have one in a MLTL design by Paul K. I have even gone over excursion in that with my APA150, so up to 75W. This is making me curious as to what would happen if I built two of these and powered them with my APA150. They would have to be 8 ohms so it wouldn't get 75WPC, but with even 37W this is showing 106dB so two would be 109dB.

The purpose for this is an experiment to see how much (total output & sound quality) I can get out of two of these subs (even four in one box, if that's what I come up with).
If it goes well, it will probably end up supplementing bass (no hpf) for my D8s (Thanks Chris!). The D8s are incredible, I have a CleanBox Pro going to an iNuke 1000 to power them, and they can definitely take a pounding! They bass is very deep, but for some reason (originally suspected to be because of the rear port) it just doesn't project into the room. If you stand behind them, it's like an earthquake, but in front it's barely audible at times. I should mention, the music I listen to includes a lot of electronic dance music genres, so lower end bass is much more demanding. I know it's something crazy about my room, but I haven't been able to figure it out. I took them outside on Saturday while I was working on my IRON project, and the bass was great, they were about a foot away from my brick house. For some reason somethings getting cancelled out inside. It could more likely be the high frequencies just bouncing all over the posters on the walls and around the room that's the problem. I don't have a real EQ so I use the one on my computer or the iTunes one, which helps, but I could still use more bass. I plan on getting a 15" once I cant afford it.

Anyway, like I said this is an experiment, I don't mind if it doesn't end up with a good daily use, I'm just trying to learn and put something together for what I have available before I go spending money on more drivers, especially because I simply can't afford it right now. I have a job interview on Tuesday, and one just went well earlier this week, I need to be working before the end of the month.

[dwigle]

I found this winisd tutorial helpful. The order in which you enters the parameters matters as well. BTW, I built a couple isobarics in 1986. Didn't really see a lot of benefit for the costs and effort. It seems that drivers available today allow for flexibility that effectively mitigates potential benefits of isobarics. http://gonecatfishin.net/WINISD.htm

[Mayhem13]

For modeling a push pull, model it with one driver, double the VAS. Power handling won't be accurate but response will be. I've built a few with high impedance drivers as to purposefully drop the load. Without being able to back it up with measurements, these systems did sound noticeably smoother but the very large box size was counter productive IMO.

[billfitzmaurice]

I can't seem to find a way to even get an F3 around or below 30Hz without going way over excursion at 150W.

That's to be expected given the driver parameters. 6mm isn't a lot of xmax for a 30Hz capable sub, 12mm is more like it, so it will take two of them to do what one nice sub driver will. Then when you consider the Sd I'd be looking at using four of them for a respectable result.

[Howard Ferstler]

What's the benefit of this alignment (one woofer mounted backwards) with reverse polarity? Isn't that just two woofer, moving in the same direction in a sealed box?

Alternatively, what's the benefit of reversing a driver (and not reversing polarity) so that the cone of one woofer pushes the backside of the other?

The benefit is that it reduces even-order harmonic distortion, which will primarily be audible with the second-order complement. It has no impact on odd-order distortion products, and no significant impact on frequencies above 100 Hz. It reduces it, because distortion products caused by inward movement of a cone facing in one direction are nulled by the generation of out-of-phase distortion products by the reversed cone of the other driver. The artifacts from the drivers are out of phase with each other and cancel. On the other hand, the overall movement of the two cones remains the same, so the output is unchanged but the distortion products remain cancelled. Odd-order products are of a different kind and the cone reversal of one driver will not have an impact on them.

This approach can work. KEF used it with some of their models a while back and my Allison IC-20 full-range systems use it with their woofers, as did several other Allison models. However, the sonic effect is really going to be subtle if the drivers involved are good ones to begin with. Indeed, when I did a refurbishing of my IC systems a while back I eliminated the push-pull feature and returned the reversed driver to a forward-facing position. (I had to do this, because I rebuilt the crossover network and the larger mounting board made it impossible to use the reverse tube required to mount a reversed woofer.) I use outboard powered subs and so the advantages are just not there any more.

If you physically reverse the mounting of one of the drivers and do not reverse polarity, or simply reverse the polarity of one driver in a conventionally mounted pair, and the enclosure is a sealed one the baffling advantages of the sealed space are lost, and since the drivers will then be operating out of phase with each other the cancellation artifacts will be detrimental. You can stack drivers end to end in a sealed enclosure (with the rear driver hidden behind the outer one), but they still have to be wired in phase and the net result is that you basically have built up a reinforced driver with the same cone area.

Howard Ferstler

[monoiz15]

Mayhem, you're saying to model it with one driver, but in the parameters, to double the Vas? I don't understand why that is; why you shouldn't model it with two.

I'm now thinking all four of them in the box with that plate amp. That should work pretty nicely. Also it has a variable phase knob which is nice.

dwigle, What are the odds?! They are also using a Dayton 8" DVC! Thanks for the link, I'ma give it a shot right now

[monoiz15]

Sd, Re, Mms, Bl, Cms, Qms
I put them in and I'm still getting the error.. I don't know how much it is affecting my results.. I have the exact same screen as the example except a couple autocalculated values are the slightest bit off. Don't know what to make of it, but I'm pretty sure it's not making a big difference, I just don't like to see it.

[Mayhem13]

Mayhem, you're saying to model it with one driver, but in the parameters, to double the Vas? I don't understand why that is; why you shouldn't model it with two.

I'm now thinking all four of them in the box with that plate amp. That should work pretty nicely. Also it has a variable phase knob which is nice.

dwigle, What are the odds?! They are also using a Dayton 8" DVC! Thanks for the link, I'ma give it a shot right now

Because one the drivers is reverse phase of the other, so the compliance of the box stays the same. Another poster mentioned cancelling of even order HD.....that's been theorized but to my knowledge never proven.