Testing theory : cascading crossovers

[Vaughan Odendaa]

Hi there,

I wanted to do a test that would confirm well known theory that adding two low pass filters in series results in gross cancellation.

What I did was I took a subwoofer from work (Jamo sub200), plugged it into my system with a 90 hz crossover as a start. I set the low pass filter on the subwoofer to 150 hz which is the upper limit.

Setting to 150 hz should bypass the low pass circuit altogether because the roll off is so steep at this point that it should be negligible at best. I also set the low pass filter to 90 hz to test what that would do.

In order for me to verify the results properly, I used my Velodyne SMS-1 real time analyser/EQ to confirm my results and provide objective data on the readings.

Interesting.

What I found is that by utilizing the same low pass filter on both AV amp and sub resulted in an increase in output over the crossover region. This goes against what I knew or what I thought I knew about electrical circuits unless I'm missing some serious points here.

"Bypassing" the low pass filter resulted in a big jump in output from 50 to 60 hz but a decrease in output at the crossover region of 90 hz. Not by a heck of a lot, mind you, but it was still surprising that it dropped by 2-3 dB's.

Now why is this ? Why would bypassing the low pass filter or more accurately, removing it's effect from the pass band result in a loss in output around the crossover point ? I thought that there should have been in increase not a loss.

As for the jump in output in between 50-60 hz, I can only imagine that because the roll off is not as steep once both filters aren't being used, there should be an increase over the range of the slope.

Everything else above 60 hz was pretty much identical to using both 90 hz crossovers in AV amp and sub. I will rerun the test again to make sure I haven't messed up but I just thought I would report this and let you guys know.

Graphs will follow.

Thoughts and opinions ?

--Regards,

 

[GTHill]

I can't wait until I understand what he said.

Gene

 

[Vaughan Odendaa]

All that I was trying to do was to confirm that one can't use the crossover in the AV receiver as well as the crossover (really just a low pass filter) in the subwoofer otherwise there will be cascading.

Cascading is when two filters overlap or combine. All crossovers (high and low pass filters make up a crossover) have slopes which define their rate of attentuation or roll off. By combining both low pass filters in AV receiver and subwoofer at the same frequency, the slopes should combine negatively.

I mean, I've demonstrated and anyone can do this too, that turning the low pass filter all the way up actually increases the perceived level of bass (when it's out of the pass band) while running the subwoofer warble tone. Not the best example to use but it's a start.

Not sure I understand what has happened either really. I just expected the graphs to show a huge difference when in reality there isn't a huge difference.

--Regards,

 

[Vaughan Odendaa]

Is anyone gonna bite this one ?

--Regards,

 

[mtrycrafts]

Is anyone gonna bite this one ?

--Regards,

WmAx is tho one to respond for more explanations on this
What happens if you place the crossover at 80Hz and the sub filter at say 90Hz?
By the way, how fine of a signals did you send, 1/3 octave? 1/6?

 

[WmAx]

I suggest you use the SMS and get a screenshot(if you have to use a camera -- please make sure it's in the exact same position for each image so that the images can be easily over-layed by me - use a tripod, or a table, or whatever) of it's graphs in each of these conditions:

1) Only subwoofer crossover - set at highest point (or bypass).
2) Only receiver crossover - set at highest point (or bypass)
3) Only subwoofer crossover, set at a fixed value of say 80 Hz.
4) Only receiver crossover - set at a fixed value of say 80 Hz.
5) Both receiver and crossover - set at fixed value of say 80 Hz - so that they cascade.

In every case, please specify if it is bypassed or set at highest frequency(and specify the frequency).

Of course, be certain the microphone is in an identical position for each test, and that no other variables other than the ones tested for change.

If you post the above graphs(please keep the images under 500 pixels wide), then someone may be able to help you.

-Chris

 

[Vaughan Odendaa]

Well. . cool. Way to make an introduction.

I'll be able to do that. Just give me a day to get my results up.

--Regards,

 

[Vaughan Odendaa]

Hi there,

Sorry I couldn't make the photo's smaller. I was in a hurry and didn't have time.

First pic :

Crossover set to 90 hz in AVR and low pass filter set to 150 hz (bypass) in subwoofer.

Second pic :

Crossover in AVR set to 90 hz. Low pass in subwoofer set to 90 hz.

Third pic :

I tried something. I set the low pass on sub to 150 hz and then set the crossover on my AV receiver to 40 hz.

Anyways, what do you think ?

--Regards,

 

[3db]

Hi there,

I wanted to do a test that would confirm well known theory that adding two low pass filters in series results in gross cancellation.

What I did was I took a subwoofer from work (Jamo sub200), plugged it into my system with a 90 hz crossover as a start. I set the low pass filter on the subwoofer to 150 hz which is the upper limit.

Setting to 150 hz should bypass the low pass circuit altogether because the roll off is so steep at this point that it should be negligible at best. I also set the low pass filter to 90 hz to test what that would do.

In order for me to verify the results properly, I used my Velodyne SMS-1 real time analyser/EQ to confirm my results and provide objective data on the readings.

Interesting.

What I found is that by utilizing the same low pass filter on both AV amp and sub resulted in an increase in output over the crossover region. This goes against what I knew or what I thought I knew about electrical circuits unless I'm missing some serious points here.

"Bypassing" the low pass filter resulted in a big jump in output from 50 to 60 hz but a decrease in output at the crossover region of 90 hz. Not by a heck of a lot, mind you, but it was still surprising that it dropped by 2-3 dB's.

Now why is this ? Why would bypassing the low pass filter or more accurately, removing it's effect from the pass band result in a loss in output around the crossover point ? I thought that there should have been in increase not a loss.

As for the jump in output in between 50-60 hz, I can only imagine that because the roll off is not as steep once both filters aren't being used, there should be an increase over the range of the slope.

Everything else above 60 hz was pretty much identical to using both 90 hz crossovers in AV amp and sub. I will rerun the test again to make sure I haven't messed up but I just thought I would report this and let you guys know.

Graphs will follow.

Thoughts and opinions ?

--Regards,

Let me understand this,

You connectd the LFE output of your receiver to the LFE input of your sub and set the low pass to 90 on your receiver and a 150 Hz on your sub?

If thats what you did, I would expect that your sub would only play frequencies from a 100 Hz and down because thats all the info the sub is getting. The only way your sub low pass filter would take effect is if you dailed it lower than 90Hz. As far as the sub's low pass filter is concerned, 90 Hz is allowd thru .

So what did I miss?

 

[annunaki]

When using two low pass filters together, is not recommended to set them both at the same frequency. Set your crossover on the receiver at 90 hz and then set the sub's crossover at 100hz or 110hz. This will help to roll off the upper end quicker. Setting them at the same frequency is not recommended.

If the sub's crossover dial is not completely accurate (as to what frequency it is actually at) it could be culprit to the issue at hand.

 

[3db]

When using two low pass filters together, is not recommended to set them both at the same frequency. Set your crossover on the receiver at 90 hz and then set the sub's crossover at 100hz or 110hz. This will help to roll off the upper end quicker. Setting them at the same frequency is not recommended.

If the sub's crossover dial is not completely accurate (as to what frequency it is actually at) it could be culprit to the issue at hand.

It also reaks have with phase. Butterworth filters if I remember my theorey correctly change the phase 90 degrees for each order.

Now my question is..in a set up like this, the filters are in series to one another?

 

[Vaughan Odendaa]

The LFE output is connected to the line in on the subwoofer (not LFE in). So low pass is in effect.

If thats what you did, I would expect that your sub would only play frequencies from a 100 Hz and down because thats all the info the sub is getting.

Yes. That is correct.

The only way your sub low pass filter would take effect is if you dailed it lower than 90Hz. As far as the sub's low pass filter is concerned, 90 Hz is allowd thru .

Again, correct. My test was to confirm that using two cascading filters results in gross cancellation. My test confirms that this will not happen (unless my testing was faulty). There seems to be more at work here.

--Regards,

 

[Thaedium]

When using two low pass filters together, is not recommended to set them both at the same frequency. Set your crossover on the receiver at 90 hz and then set the sub's crossover at 100hz or 110hz. This will help to roll off the upper end quicker. Setting them at the same frequency is not recommended.

If the sub's crossover dial is not completely accurate (as to what frequency it is actually at) it could be culprit to the issue at hand.

Just a quick question, I have my VSX-94TXH crossover set at 90hz and my DD-15 crossover set at 90hz. I've had to place the Sub in a non-optimal position due to room logistics and I'm getting a fairly significant dropout in the 55 to 60hz range that no amount of EQ is correcting. I've also ran the MCACC on the Pioneer and it has me set the DD-15 to a volume level of 20 in order for it to "match" the rest of the setup. 20 is just too low, and its lower then the factory presets for the sub. Could setting the crossover points at the same place be giving me some of this problem? Any recommendations would be greatly appreciated, thanks.

I find the Studio's to have pretty good bass on their own, but for some of my music tastes (IE: Rap - I know, not many audiophiles enjoy it, I'm ashamed, really ) I'd like the bass to kick in a little sooner to give the music its intended sound. However, when I watch TV or movies I'd like the bass to be managed a little bit more by the Studios to give that seamless feel and transition making the sub appear to be a natural extension of the speakers and not some crude overlay that I can pickout. Again, any suggestions or recommendations would be greatly appreciated.

Thanks!

 

[mike c]

Thaedium, I have the DD subwoofer and I like rap as well ... i think it's the wrong subwoofer for rap music. i highly doubt the rap songs were made with "low distortion" in mind. i bet those guys prefer high distortion bass. (as I do)

 

[Vaughan Odendaa]

Just a quick question, I have my VSX-94TXH crossover set at 90hz and my DD-15 crossover set at 90hz. I've had to place the Sub in a non-optimal position due to room logistics and I'm getting a fairly significant dropout in the 55 to 60hz range that no amount of EQ is correcting.

I would avoid using both 90 hz crossover in AV receiver and subwoofer. But besides that, you are most likely sitting in a null.

Could setting the crossover points at the same place be giving me some of this problem? Any recommendations would be greatly appreciated, thanks.

Cascading filters could be a problem. But it's most likely not the root issue here. Try moving your seating position, your subwoofers position, or where the speakers are placed.

I find the Studio's to have pretty good bass on their own, but for some of my music tastes (IE: Rap - I know, not many audiophiles enjoy it, I'm ashamed, really ) I'd like the bass to kick in a little sooner to give the music its intended sound.

Your subwoofer can produce far louder bass in the midbass region so I would personally set the speakers to small and leave it there.

However, when I watch TV or movies I'd like the bass to be managed a little bit more by the Studios to give that seamless feel and transition making the sub appear to be a natural extension of the speakers and not some crude overlay that I can pickout. Again, any suggestions or recommendations would be greatly appreciated.

Speakers set to small but perhaps choose a lower crossover frequency. Why don't you use 60 hz ? Although the lower the x-over, the less impact you'll get in that 40-80 hz range unless your speakers can outperform the subwoofer in that region.

Doubtful.

Try that, have a listen and then let us know.

Any thoughts on the graphs ?

--Regards,

 

[asdf]

/

1) Only subwoofer crossover - set at highest point (or bypass).

Hmmm

 

[annunaki]

Just a quick question, I have my VSX-94TXH crossover set at 90hz and my DD-15 crossover set at 90hz. I've had to place the Sub in a non-optimal position due to room logistics and I'm getting a fairly significant dropout in the 55 to 60hz range that no amount of EQ is correcting. I've also ran the MCACC on the Pioneer and it has me set the DD-15 to a volume level of 20 in order for it to "match" the rest of the setup. 20 is just too low, and its lower then the factory presets for the sub. Could setting the crossover points at the same place be giving me some of this problem? Any recommendations would be greatly appreciated, thanks.

I find the Studio's to have pretty good bass on their own, but for some of my music tastes (IE: Rap - I know, not many audiophiles enjoy it, I'm ashamed, really ) I'd like the bass to kick in a little sooner to give the music its intended sound. However, when I watch TV or movies I'd like the bass to be managed a little bit more by the Studios to give that seamless feel and transition making the sub appear to be a natural extension of the speakers and not some crude overlay that I can pickout. Again, any suggestions or recommendations would be greatly appreciated.

Thanks!

I would try what has been suggested so far. Lower the crossover point to 60hz or so. Bump the sub's crossover up though. You don't want them both at the same frequency. What is the highest -3db point of all of your speakers?

It does sound like your subwoofer is placed in a null. I would experiment with sub placement to fix the problem as well. Adjust the crossovers first and then check sub placement.

A good subwoofer can do well with all types of music. Rap tends to need the bass level upped a bit for ultimate "boom" factor.

 

[Halon451]

By convention, when selecting multiple crossover settings I have always believed that setting the AVR crossover to the -3dB point of your main speakers is ideal (or the next available setting upwards of that), then dialing in the sub x-over setting to a point just above that or higher. Thus to preclude omitting any of the frequencies within a certain range (you don't want any gaps for sure), and ideally marrying the low end response of your main speakers to a point where the sub picks up with no noticeable drops in output - and you have a smooth transition from high to mid to the lowest frequency your sub can output; all while not being able to audibly detect where and when it all happens.

One thing is for sure - you can't always go by what the specs tell you as far as a speaker's frequency response. The actual -3dB point could vary based on a number of other factors. Using proper test equipment like you are doing is the best way to hone in to the actual performance of your system.

 

[no. 5]

If thats what you did, I would expect that your sub would only play frequencies from a 100 Hz and down because thats all the info the sub is getting. The only way your sub low pass filter would take effect is if you dailed it lower than 90Hz.

If it was a brick wall filter. But because it's a slope, the sub will still be getting signal for frequencies above the crossover point (though at progressively lessening amplitude).

 

[Vaughan Odendaa]

Just to get the thread back on topic, can MDS, WmAx or Gene (or anyone with a techical background) offer their thoughts on the graphs and let me know what they think ?

I'm glad that at the same time this thread will help people configure their systems better, but I don't want this thread to be sidetracked.

Thank you.

--Regards,