Question about multiple subs in REW

[pmandt]

Wel, I finally got my SB audigy 2ZS notebook card calibrated, checked levels, and made some preliminary measurements. These are attached along with a crude drawing of the room setup.

My setup is as follows. I have a lexicon MC1 MC1 processor feeding the sub output with an unbalanced cable and a RShack RCA to 1/4" plug adaptor to a BFD 1124 with Y-connectors splitting the sub out put signal to each of my 3 subs. Two of them ( one in each front corner of my 15'W x 35' long x 8' high room) are Rythnik audio 12" 350 watt servo kits in 2 cuft boxes. The third sub is a TC sounds 18" LMS 5400 with 2 Passive 18" radiators in a box slightly under 7 cuft. The VC's on the LMS5400 are wired inn series to a bridged QSC 4050HD power amp. I turned off the amps to all the mains to take my first 2s. measurement. The next one was done with the LMS 5400 only, which is sitting in the front left corner.

I have soom questions. What is the best sequence in whic to run REW? One sub at a time? All three together? The two Rythmiks at one time? Should I stack them together? Wher best to set the level controls and phase adjust ment on the LMS5400, since there are really not contols on the sub or amp, should i use the BFD filter. I cant currently use the lexicons output separately since it only has one sub output. I could use an ART cleanbox; I have one and also the caps recommended in the cleanbox mod article , but have not yet done the mod. Also, I don't have any high pass filter built into the system for the LMS5400, can I somehow use the BFD for this? What about using the Rhythmiks for midbass and the LMS5400 for the really low end, even though it seems to have really good extension. Thanks in advance for any advice with this.
Peter

[brucek]

What is the best sequence in which to run REW? One sub at a time? All three together?

Yeah, this is a tough one.

Let me say this first. Trying to mix a monster like the 5400 with a couple of 12" subs would be considered a poor idea. Others may certainly disagree.

The trouble is that you'll get reinforcement in the areas where the responses match (and that's good), but in the areas where the 5400 stands alone (at VLF), you don't get the reinforcement, so the overall response suffers. You'll find you're better off with the 5400 by itself. I would try placing the 5400 in the corner and see how it sounds solo.

This is a standard problem when someone steps up the food chain in the world of subs and feels there's just got to be a use for my old subs....................

That said, if you are trying to setup three subs, you really should EQ the 5400 by itself on one channel of the BFD and then co-locate the other two subs, and EQ them with the other channel of the BFD. Don't worry about level (except for REW setup levels) until you have EQ'd each channel separately and then played them together and tackle the tweaking that will be required to smooth it all out.

Then setup the levels with your receiver test tones and the sub amp volume controls to match the mains. Then you must also ensure the crossover area is clean by adjusting phase where possible.

I don't have any high pass filter built into the system for the LMS5400, can I somehow use the BFD for this?

No. What's the 5400 tuned to - you may not need a hi-pass. If it's tuned low enough (read LLT), then the electronics rolloff will become a reasonable substitute. , you can even play with the input filter of the ART box and tune your own hi-pass. Easy..

What about using the Rhythmiks for midbass

I'm sorry to say that I think the whole mid-bass theory is a bit wacky. Others may recount their anecdotes and sell you on the idea, but be ready for a lot of tweaking.......

Anyway, that's a great response you have posted that will EQ quite easily. With that sub, I would extend my horizontal scale down to 10Hz........... Interesting to see how low the 5400 goes.

Your the first person I know who has a TC Sounds LMS5400. Wow, what a driver... What size is the box etc, etc. Let's see a pic...

brucek

[pmandt]

Thanks for the quick reply. The last graph ( blue ) is one with only the LMS5400 on. Here are some pics. When you said I cdould use the ART cleanbox for a HP filter, did you mean by swapping out the caps for different values? I haven't quite figured out the tuning, yet. I think I have about 1400gms of mass added to each PR

[pmandt]

Sorry, I must have hit the post Key prematurely. The box volume is just under 7 cuft. Here are some photos. I haven't done any measurements with the mains on.

Peter

[pmandt]

I just finished running some sine waves through it. The tuning frequency looks to be about 17.5 HZ, based on movement of the active driver and the 2 passives.

Peter

[brucek]

You did a beautiful job. Looks great. I'll bet Ilkka would love to test that thing.....

I guess the answer of whether you need a hi-pass would be answered in a response test of your electronics. This is fairly easy to do with REW. It involves doing a frequency response of your processor and BFD in an REW loop. Often the electronics results in a 1st order rolloff that you would have designed into a high-pass filter anyway, so the high-pass isn't necessary.

With a tune of 17.5 Hz I doubt you'll need a high-pass (especially with the xmax and xlim of that driver).

You can certainly calculate a new value of coupling capacitor at the input to your ART box with the corner frequency being where you feel you require protection.

If I look at the ART box mod thread, I see that the claim is that the input pot is about 100K. So, if that's accurate, and if you wanted an fc around 10Hz for the filter, then the series capacitor would be about a 0.15uF (back of a napkin calculation). That's a fairly standard size you could try and test (with REW) and it would result in a 1st order ~10Hz high pass filter. That's if you think you need it.

I just finished running some sine waves through it.

Ignore anything below 10hz since the RS calibration file only has been tested to 10Hz.....

Have you tested it in a corner yet?

brucek

[pmandt]

brucek wrote:

You did a beautiful job. Looks great. I'll bet Ilkka would love to test that thing.....

I guess the answer of whether you need a hi-pass would be answered in a response test of your electronics. This is fairly easy to do with REW. It involves doing a frequency response of your processor and BFD in an REW loop. Often the electronics results in a 1st order rolloff that you would have designed into a high-pass filter anyway, so the high-pass isn't necessary.

With a tune of 17.5 Hz I doubt you'll need a high-pass (especially with the xmax and xlim of that driver).

You can certainly calculate a new value of coupling capacitor at the input to your ART box with the corner frequency being where you feel you require protection.

If I look at the ART box mod thread, I see that the claim is that the input pot is about 100K. So, if that's accurate, and if you wanted an fc around 10Hz for the filter, then the series capacitor would be about a 0.15uF (back of a napkin calculation). That's a fairly standard size you could try and test (with REW) and it would result in a 1st order ~10Hz high pass filter. That's if you think you need it.


Ignore anything below 10hz since the RS calibration file only has been tested to 10Hz.....

Have you tested it in a corner yet?

brucek

Thanks. No, I haven't put it all the way into a corner. I 'm going to need to move a few things around.
Peter