AVS › AVS Forum › Audio › DIY Speakers and Subs › Andreas' Slow Rotary Sub build
New Posts  All Forums:Forum Nav:

Andreas' Slow Rotary Sub build

post #1 of 227
3/9/11 at 8:01am
Thread Starter 
So, I'm going to try and build one. I've been planning it for a while now. I still have a bit to do on my HT first, but I need some details worked out so I can do both.

I have a couple of questions:

1. Is 112db *that* impressive for 10hz - 20hz? It doesn't seem like it compared with a lot of traditional sub builds on the forum. I assume for 5hz it is much more impressive, but what the heck do I know? I want to order an SPL meter so I can measure what I have right now (a THT). I'm thinking I want to build my rotary with a bit more output if 112db isn't that big a deal. I want it to be worth my effort!

2. Can anyone point me in the right direction as to how big I need to make the enclosure/dead vent that connects the fan to the room? I have sound isolation in my room currently, and I don't want it to be for naught. I know I will lose isolation for the sub 25hz that the rotary puts out, but I'm ok with that. I'd like to keep the rest of the sound in the room. So, I think I need to compromise the enclosure/dead vent between output and sound isolation. Can I model it in WinISD perhaps?
Reply
Reply

AVS Top Picks

post #2 of 227
3/9/11 at 8:04am
112 db's is impressive from 5hz and up if it is clean. 10-25hz the DTS-10 is awesome. I would shoot for 3hz and up, that is the point of the rotary sub.
Reply
Reply
post #3 of 227
3/9/11 at 8:08am
Thread Starter 
Sure, I will take whatever I can get. The rotary is much less stressed at 1hz than 20hz. I'm just not sure if my amp (Crown XTI) and AVR (Denon 1911) will allow me <5hz. I'm hoping I might be able to overcome some of that with EQ like a MiniDSP.
Reply
Reply
post #4 of 227
3/9/11 at 8:10am
Quote:
Originally Posted by AndreasMergner View Post

Sure, I will take whatever I can get. The rotary is much less stressed at 1hz than 20hz. I'm just not sure if my amp (Crown XTI) and AVR (Denon 1911) will allow me <5hz. I'm hoping I might be able to overcome some of that with EQ like a MiniDSP.

With yout THT you don't need to go high with the rotary. 3-20hz would be perfect. If your amps to go to 3hz then I would sell them and invest in one that can, it will make a big difference and remove that from the equation if something goes wrong.
Reply
Reply
post #5 of 227
3/9/11 at 8:25am
Regarding the signal chain roll off you can measure this using REW if you have the right cables. You need to hook the last cable that would go to the amp to the input of your computer and run REW from 1hz up. You won't hear anything (obviously) but it will give you the FR of your signal chain (you can see any EQ filters this way too). This doesn't help with the amp but it will show you what you're equipment is capable of. Note you may want to run this first with just the computer soundcard out to soundcard in to see how much of the roll off is in the soundcard itself (you probably did this when configuring REW already though...).
Reply
Reply
post #6 of 227
3/9/11 at 8:56am
Thread Starter 
Sure, I'll get a new amp if need be. I'm just not sure if the amp or AVR will be problems at this point. I also am not sure I'd want to pay $2k for high end amps/AVRs for a 3-5hz either, so there is a balance there for sure.

lennon: Good idea....and no, I haven't used REW yet. I wanted to get a Radio Shack SPL meter (today?) and use it for testing the THT so I know where I stand with it. Then I would like to get a calibrated mic and mixer board for the rotary. Can I use an SPL meter to do the RTA distortion measurements?
Reply
Reply
post #7 of 227
3/9/11 at 9:31am
Quote:
Originally Posted by AndreasMergner View Post

Sure, I'll get a new amp if need be. I'm just not sure if the amp or AVR will be problems at this point. I also am not sure I'd want to pay $2k for high end amps/AVRs for a 3-5hz either, so there is a balance there for sure.

lennon: Good idea....and no, I haven't used REW yet. I wanted to get a Radio Shack SPL meter (today?) and use it for testing the THT so I know where I stand with it. Then I would like to get a calibrated mic and mixer board for the rotary. Can I use an SPL meter to do the RTA distortion measurements?

An SPL meter should work fine for that. The calibrated mic solution is nice for peace of mind (I feel much more comfortable posting graphs taken with my calibrated ECM8000 than I did posting RS SPL meter graphs) but it's really not necessary in the grand scheme of things. Note you can do the test I suggested above without needing an SPL meter, you're just looping the signal through your equipment to see where it cuts off on the low end.
Reply
Reply
post #8 of 227
3/9/11 at 10:02am
Quote:
Originally Posted by AndreasMergner View Post

I have a couple of questions:

1. Is 112db *that* impressive for 10hz - 20hz? It doesn't seem like it compared with a lot of traditional sub builds on the forum. I assume for 5hz it is much more impressive, but what the heck do I know? I want to order an SPL meter so I can measure what I have right now (a THT). I'm thinking I want to build my rotary with a bit more output if 112db isn't that big a deal. I want it to be worth my effort!

2. Can anyone point me in the right direction as to how big I need to make the enclosure/dead vent that connects the fan to the room? I have sound isolation in my room currently, and I don't want it to be for naught. I know I will lose isolation for the sub 25hz that the rotary puts out, but I'm ok with that. I'd like to keep the rest of the sound in the room.

1. If it is similar in size to the TRW (~19" wide fan) 120-125dB is pretty easy to get in room.

2. First off, you will not just loose isolation for <25Hz. You will loose all isolation from 0-20kHz, unless you also isolate the rear wave room. The rear room should be as close to the same size as the listening room as possible, or bigger. The manifold/vent should be square at ~4ft per side, and 4-8ft deep. I would go ~6ft myself.
Reply
Reply
post #9 of 227
3/9/11 at 10:19am
It LIVES! BWAHAHAHAHAHAHahahahaha! Good to see it rolling, Andreas.

IIRC, it was recommended to build a sound dampening box that has a travel length of 3-4m. The bigger it is, the less fan noise, but also the less efficient it is. A couple corners help dampen the sound, not to mention fit it in a given space .
Reply
Reply
post #10 of 227
3/9/11 at 11:06am
Andreas, The TC2000 driver has been shipped you should see it Monday or Tuesday.
Reply
Reply
post #11 of 227
3/9/11 at 11:08am
Quote:
Originally Posted by AndreasMergner View Post

So, I'm going to try and build one. I've been planning it for a while now. I still have a bit to do on my HT first, but I need some details worked out so I can do both.

I have a couple of questions:

1. Is 112db *that* impressive for 10hz - 20hz?

112dB @ 10Hz is just starting to be great...120dB is better 5Hz definitely needs 120dB for it to matter.
Reply
Reply
post #12 of 227
3/9/11 at 11:34am
Subscribed. Best of luck!
Reply
Reply
post #13 of 227
3/9/11 at 2:35pm
Thread Starter 
120-125db sounds much better. I got 112db from here: http://www.eminent-tech.com/woofercomparison.html

Soho: As far as loosing all sound isolation, I'm not sure why you say that. The purpose of the duct is to reduce noise from the fan, correct? This means that sound coming out will also be attenuated. Dead vents are designed to keep sound in while allowing airflow for occupants. This would just be a large dead vent. I want to find out how to model the frequency response of the rotary depending on the size of the dead vent/duct. Seems like if you had a 19" diameter fan, you would only need a 19" diameter duct...as you lengthen the duct, you'd need to increase the diameter. What if you didn't? How much output would you lose? What if I went to a 24" diameter fan? I'd have more output to begin with. So, I'd like to make a compromise between output, vent size, rotary diameter, rotary noise and sound isolation. Is there any way to model that? Why the 4'x4'x6' duct recommendation?

Penngray: Thanks! I look forward to receiving it and figuring out how I can use it best.

I picked up the RS SPL meter today. I'll see if I can hook it up with what I have here now so I can see my starting point.

Thanks for all of the encouragement.
Reply
Reply
post #14 of 227
3/9/11 at 10:46pm
Quote:
Originally Posted by AndreasMergner View Post

Sure, I will take whatever I can get. The rotary is much less stressed at 1hz than 20hz. I'm just not sure if my amp (Crown XTI) and AVR (Denon 1911) will allow me <5hz. I'm hoping I might be able to overcome some of that with EQ like a MiniDSP.

Check the threads for posts by Bruce Thigpen about using really cheap Audiosource amps - because they had no subsonic filters, and were essentially flat to DC! It is counter-intuitive but you might not need a massive and expensive amp for this project.
Reply
Reply
post #15 of 227
3/10/11 at 6:24am
Thread Starter 
Zora: Thanks! I know the rotary doesn't need much power....although that really depends on the actuating driver, the amount of excursion, the upper frequency you are trying to reach and the mass of the linkages. I have read as many of the old posts as I could find. I did not remember the inexpensive amp they had used!

I have REW working now. I need a longer RCA cable that I will pick up today so I can get the SPL meter past the doorway! I'm getting some high distortion levels (~15%) when I run the RTA tool in REW. I'll have to check into that. I think I read somewhere that a low battery in the SPL meter can cause that (and my battery is suspect), so I'll have to do a bit of detective work.
Reply
Reply
post #16 of 227
3/10/11 at 1:33pm
Thread Starter 
Ok, I have REW working better now. I'm getting <2% THD for all freqs, even down to 14hz....although 14hz is down 15 db from 25 hz. I had seen one THT REW THD graph showing 100% THD at 17 hz. I must have a better room or placement...or maybe I don't know what I'm doing with REW yet. In any case, I now have a starting point.

I'll attach it below. I have my AVR at -38db....not sure that matters. I'd like to crank it, but we are trying to get the baby asleep!
LL
Reply
Reply
post #17 of 227
3/10/11 at 3:38pm
Thread Starter 
I am messing around with WinISD. I'm using a 4th order bandpass with a very large rear chamber (IB), a very small (~0 L) front chamber, and a vent with a 66cm diameter and 555cm length. I basically made a speaker with 9999mm Xmax and enough watts (1111111) to get 120 db. Then I added a parametric filter with a 22db boost at 1hz. I get an SPL curve starting at 120db for 1hz and then slowly sloping down to 115 at 20hz. It is roughly equivalent to the FR graphs of the TRW17.

Then, I messed with the port diameter and length to see what effect it has on the FR. The port will be the dead vent from the fan to the HT. The rear chamber is the other room. It looks like the 555cm port length does not affect the FR since it has a resonance around 31 hz. If I go 10x longer, it will start to take some dbs from the higher freqs (20-25). Length of the port doesn't seem to be a problem. The 66cm diameter port (equivalent to a 23" square vent) looks like it has a port velocity of 17m/s at about 4.5 hz and 20m/s at 4 hz....so I might hear chuffing below those freqs?

Can anyone tell me if I'm on the right track here or am I way off base? I am not a speaker design expert by any means, but I catch on pretty quickly.
Reply
Reply
post #18 of 227
3/10/11 at 3:54pm
Quote:
Originally Posted by AndreasMergner View Post

So, I'm going to try and build one. I've been planning it for a while now. I still have a bit to do on my HT first, but I need some details worked out so I can do both.

I have a couple of questions:

1. Is 112db *that* impressive for 10hz - 20hz? It doesn't seem like it compared with a lot of traditional sub builds on the forum. I assume for 5hz it is much more impressive, but what the heck do I know? I want to order an SPL meter so I can measure what I have right now (a THT). I'm thinking I want to build my rotary with a bit more output if 112db isn't that big a deal. I want it to be worth my effort!

2. Can anyone point me in the right direction as to how big I need to make the enclosure/dead vent that connects the fan to the room? I have sound isolation in my room currently, and I don't want it to be for naught. I know I will lose isolation for the sub 25hz that the rotary puts out, but I'm ok with that. I'd like to keep the rest of the sound in the room. So, I think I need to compromise the enclosure/dead vent between output and sound isolation. Can I model it in WinISD perhaps?

I use a 8' long x 4' wide x 4' high enclosure for my TRW-17
Rotary Subwoofer. This works out well for me as I want
to keep the fan noise down to a very low level.
The vent opening into the room is 4' x 4'
The rotary woofer controller is by Marchand Electronics and
provides 200 watts to pitch blades and has a variable low
pass xover (15 hz to 25 hz), gain control, damping and
phase adjustment. The back venting should be much larger
than the listening room for full output (I vent into my
very large basement which is over 2000 sq. ft.
LL
LL
LL
Reply
Reply
post #19 of 227
3/10/11 at 6:10pm
Are you going to hack off the beefy frame of the TC-2000?




How will you attach (press fit) the bearing inside the voice coil without damaging it considering it is already installed and glued to the spider?

What solution have you decided to use to transfer the linear motion of the flange arms to rotate the blade arms?



The project is definitely do-able, but if you try to break down an existing driver, it gets complicated pretty quickly. Having the help of a driver fab shop makes it a lot easier.
Reply
Reply
post #20 of 227
3/10/11 at 6:37pm
Thread Starter 
rwaaudio: Your enclosure is similar to others that were built by Thigpen. I'm wondering why they are so large and what compromise I would be making if I went to a 2'x2'x15' enclosure with one bend. I would think I would get much better reduction in fan noise, but wonder what the performance hit would be.

As far as the back volume, I have read mentions of only 1/4 of the HT needed. It will affect the frequency response, however, similarly to an IB vs a sealed traditional sub. Maybe "similarly" is not the right word, but it will increase some freqs and decrease others.

Steve: I will have no problem hacking through that driver. I am no expert on driver construction, so I will need to see it to determine how I will attach the linkage. Hopefully, I will be able to get some help on how far I need to (or can) strip the driver to get something secure to mount to without destroying the driver. Considering it is a "free" driver thanks very much to Penngray, I don't mind putting in some labor and thought on how to do it.

I will be using an arm and clevis with the clevis pin going through a bearing....I don't mind experimenting a bit though. I want to keep it as light as possible to ensure good high freq response.
Reply
Reply
post #21 of 227
3/10/11 at 10:24pm
Marathon 5050 amp is not horribly expensive and plays down to 2 or 3hz I believe. I am sure the other amps from Marathon could do this that are not as powerful but I was just thinking of getting one of these soon.

I have a lot of questions but I will wait to hear more about this first.
Reply
Reply
post #22 of 227
3/11/11 at 6:11am
Thread Starter 
Now, I'm thinking the WinISD model is not entirely correct. A fan sub wouldn't have increasing port speeds as it went down in frequency, would it? I thought SPL was related to air velocity/cone velocity? A fan sub has a max air velocity without respect to frequency....the air velocity does not go up as the frequency goes down since it is just a fan.

Also, 20m/s (79 f/s) through a 2'x2' opening would be 19000cfm which is not going to be possible with a 24" fan.

Any thoughts? Maybe I should try modeling it in Hornresp?
Reply
Reply
post #23 of 227
3/11/11 at 6:43am
This is waaaay beyond my level of understanding, but I'll definitely be watching...good luck AM!
Reply
Reply
post #24 of 227
3/11/11 at 6:52am
If the TC2000 I shipped you does not work maybe we should contact a driver manufacturer and pay them to give us something to work with.

Im more then happy to fund some R&D on this!!!
Reply
Reply
post #25 of 227
3/11/11 at 7:37am
Quote:
Originally Posted by penngray View Post

If the TC2000 I shipped you does not work maybe we should contact a driver manufacturer and pay them to give us something to work with.

Im more then happy to fund some R&D on this!!!

Has anyone contacted John Janowitz? I know they've been having supply issues but I believe it's related to parts that wouldn't be required for something like this...
Reply
Reply
post #26 of 227
3/11/11 at 7:46am
Thread Starter 
Penngray: Thanks for the offer. Let's see what I can come up with for this driver. You have given me a good start already!

I did REW for my AVR. It looks pretty good, actually. 2hz is only 5 db down from 10hz's level.
Reply
Reply
post #27 of 227
3/11/11 at 7:58am
Quote:
Originally Posted by AndreasMergner View Post

rwaaudio: Your enclosure is similar to others that were built by Thigpen. I'm wondering why they are so large and what compromise I would be making if I went to a 2'x2'x15' enclosure with one bend. I would think I would get much better reduction in fan noise, but wonder what the performance hit would be.

As far as the back volume, I have read mentions of only 1/4 of the HT needed. It will affect the frequency response, however, similarly to an IB vs a sealed traditional sub. Maybe "similarly" is not the right word, but it will increase some freqs and decrease others.

Steve: I will have no problem hacking through that driver. I am no expert on driver construction, so I will need to see it to determine how I will attach the linkage. Hopefully, I will be able to get some help on how far I need to (or can) strip the driver to get something secure to mount to without destroying the driver. Considering it is a "free" driver thanks very much to Penngray, I don't mind putting in some labor and thought on how to do it.

I will be using an arm and clevis with the clevis pin going through a bearing....I don't mind experimenting a bit though. I want to keep it as light as possible to ensure good high freq response.


A minimum of 3' x 3' vent opening is required and 4' x 4'
is better, 5' x 3' would suffice. The opening should be
close to 1:1 ratio with the opening cross sectional area
carried through the entire enclosure. The length of
the enclosure should be 6' to 8', but not much longer as
this could cause turbulence. One bend is also required
with a 3' turn up to the floor.
For maximum output a larger volume than the room is required for back venting.

No enclosure is best for maximum output, but with fan noise that would be very high. A picture of my friend's rotary sub
shows no enclosure (testing only) and this gives more output,
but high noise.
LL
Reply
Reply
post #28 of 227
3/11/11 at 8:13am
Thread Starter 
rwaaudio: Thanks for the info. I'm assuming this is what Thigpen recommends?

I'm trying to quantify these types of recommendations. 3x3 is a minimum opening, but how much is the frequency response down from a 4x4 opening and which frequencies are affected? How does a 2x2 opening compare?

I am trying to minimize noise transmission -- not only from the fan in, but from the theater out. I can do that with more length and absorption. I would like to quantify changing the length on frequency response.

What I may end up doing is making the fan unit, measuring the frequency response while it is mounted in the doorway and then making the enclosure and measuring again. I wish someone could help me model the enclosure to predict frequency response!

I also wonder if designing the unit for more output would allow me to get away with a better sound isolated enclosure while still having the same performance.
Reply
Reply
post #29 of 227
3/11/11 at 9:08am
Andreas,
I am having a Movie night in Albany at the Imax(I have never been to an Imax) so when we go(we are waiting for a big summer movie) you should join us.
Reply
Reply
post #30 of 227
3/11/11 at 9:11am
Quote:
Originally Posted by AndreasMergner View Post

Now, I'm thinking the WinISD model is not entirely correct.

You can not model the rotary sub in any normal driver simulator.

Quote:
Originally Posted by AndreasMergner View Post

I am trying to minimize noise transmission -- not only from the fan in, but from the theater out. I can do that with more length and absorption. I would like to quantify changing the length on frequency response.

The length reduces all output across the subs bandwidth. If the diameter is made smaller (or the extra length is to great) you lose output on the low end, as the rotary sub blades are no longer able to create low frequency pressure waves, because the pressure in the manifold is to high compared to the rear side.

As rwaaudio has mentioned a bend is needed to help isolate the >100Hz sound, and it can work both ways. The only kicker is the manifold has to be larger to prevent any pressure build up in it, which kills LF creation. This is were the recommended 4x4ft diameter comes from. Also remember a 1ft of that diameter is taken up by the 6in of insulation on all sides.
Reply
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Speakers and Subs

AVS Top Picks

AVS › AVS Forum › Audio › DIY Speakers and Subs › Andreas' Slow Rotary Sub build