[burzel2]

First let me say that REW is fantastic. Since I discovered it last weekend I can't stop using it! I have a DIY sub with a 12" Dayton driver and a 300 watt amp in a 2 cubic foot sealed enclosure. F3 using WinISD is about 38 Hz. I'm using the older analog RS sound level meter. The first measurement below was taken with a 90 dB target with the preamp crossover set at 60 Hz. The second measurement was taken with a 95 dB target. There's always a spike at 45 Hz which is exactly the first order resonance of the width of the room. The second spike is around 90 which is perhaps the second order resonance. Here are my questions:

Why does the spike at 45 Hz get less pronounced at higher volumes? At 75 dB target the spike is huge.
Why do the measurements not look anything like WinISD? I don't see a distinct F3. Is that room gain?
How the do you minimize a first order resonance without putting the sub in the middle of the room or adding a second subwoofer?

Any help is greatly appreciated.

 

[burzel2]

Great suggestion. Is testing outside similar to an anechoic chamber? I'm not sure my laptop has a good enough sound card, but I'll look into it. To answer your question the plots have the 1/3 octave smoothing applied. The plots below do not have any smoothing. I've also included a third plot showing a 75 dB target from a previous measurement. I think the crossover is disabled in that plot (ignore the target curve). It shows the pronounced spike at 45 Hz and again at 90. I have no EQ in my system.

 

[burzel2]

My sub is a DIY 2.0 cu ft sealed enclosure using a Dayton RSS315HF-4 and Bash 300W amp. The room is 20'W X 12.5'L X 7.5'H. The couch is against the longer wall. The right side of the room has stairs and two open door ways. Behind the couch is a bay window. The sub is currently to the left of the front left floor standing speaker (Paradigm Studio 60s) and is near the back wall and about 5' from the left wall. I can reduce the 45 Hz spike by moving it to the center of the room, but there's a fireplace along the side wall. Experiments against the back wall were worse. The mic was located at the listening position in the middle of the couch.

I very much appreciate your help. I enjoy learning the technical aspects of this hobby.

 

[burzel2]

My cheap mic is on a tripod pointing horizontally right behind where your head would be. I did a little experimenting with having it vertical and in a few other locations, and the results were marginally different.

Btw, are you more curious about the source of the behavior you're seeing, or are you trying to tweak things to sound better? or both?

Definitely both.

First plot below is through my preamp (Rotel RSP-1066):

The next plot is directly to the sub amp. I would say the results are very similar to the previous plot:

This sweep has the source fixed and I'm changing the sub amp volume knob:

This sweep is through the preamp and includes the main speakers:

 

[burzel2]

I'm using the Radio Shack 33-2050 with the proper calibration file. Sorry for the confusion.

 

[burzel2]

I admit I had to Google subwoofer compression. I read about thermal compression and power compression. They seem to have the similar definitions. I'm not familiar with this so bear with me. My understanding is power compression occurs when the voice coil heats up, the impedance goes up. So as the power goes up the coil heats up, impedance rises, and less power is delivered to the driver. But my plots show compression at lower levels. Is my understanding of this correct?

 

[burzel2]

It appears to be a measurement issue (fantastic advice from everyone by the way!). The following plot has the mic right next to the sub. I set the meter to 80 and ran the sweep.

For the next test I adjusted the meter as it got louder. I re calibrated each time and verified by testing at the same volume as the previous run. The meter appears to be accurate even when the needle becomes pegged. However it becomes inaccurate beyond a certain point. For example when it's pegged during almost the entire sweep. Here's the plot with it adjusted and calibrated. I also set the crossover to 60 Hz on this run.

I'll re run the sweeps at the listening position this weekend. It's too loud to test with the family around. Some quick tests show the traces look identical at different volumes. Looks like there is no compression happening. My 45 Hz spike just goes up and up.

 

[burzel2]

Are you familiar with interpreting the Scope information?

Negative. I looked at it, but didn't dive into the help section. Any quick pointers?

I bet you can get that 70Hz dip to move around by changing the height of your meter

Unfortunately it's at the same height as my head in the main listening position. But I'll try some tests to see the effects of height. I did a few tests with the sub on it's side and at various heights. Some things looked better and some looked worse. I'm thinking of building another sealed enclosure that's a bit smaller with a front facing driver.

The 70 Hz dip is smoothed out a bit when driven with the mains, but both the sub and mains have that spike at 45 Hz. I can't get rid of it and have the sub in a wife approved location! I might have to bite the bullet and build two smaller subs.

 

[burzel2]

Wow. That was more than a quick pointer. Thanks for the visual aids. I definitely had an issue with the measuring tool. The plot below has the meter about 2" from the sub's cone. IMO it shows the subwoofer performing as it should. However my room needs some help. The big 60 Hz dip has something to do with the bay window. Only when the mic was not in front of it or the sub more to the side did it go away. I'd like to thank everyone again for all your help. REW is a great tool!

 

[burzel2]

You might try some outdoor measurements just to see what this looks like so that you can identify it in your room.

I missed winning an external USB sound card on Ebay by $1.00!! I need to get one soon for some outdoor testing. Also I'm considering an upgrade on measuring equipment. The RS meter is apparently only accurate down to 24 Hz.

Mike, if I get a good outdoor measurement, can I then compare it to the room measurement and draw some conclusions on what the sub is doing vs. the room modes? I've also learned that my Bash 300 has a 1 dB boost at around 28 Hz. This apparently is also a high pass filter at 17.7 Hz. It can be changed by replacing a few resistors. I assume this will change the response vs. WinISD as well. Do you guys ever send a high level signal to the sub to bypass the amp for comparison purposes? Below is my latest measurement with the SPL meter properly set and the sub in a better location (no more 60 Hz dip).

 

[burzel2]

I am using the calibration file.

One other question. The plot in post #21 was taken with the meter about 2" from the cone. Does this method minimize room gain? I'm asking because f3 is at about 25 Hz. Much lower frequency than predicted by WinISD. I need to measure this thing outside!

 

[burzel2]

Mike, your forcing me to learn about aspects of this hobby that are clearly more than just a hobby to many people in this forum. It's great! I'm a mechanical engineer, I enjoy learning about this stuff.

I had to read the first half of your post a few times, but the second half I understood. Transfer functions make sense to me and your explanation of

perceived "zero" as starting with the room's natural "frequency response"

makes perfect sense. Do you tune the simulation with equalization to represent the room's response?

Also so can you explain the difference between boundary gain and room gain? Thanks.