Gestating large sensitive passive 15" augmenting subs

[afa]

 

Not wanting to nit pick 

But you said

On 28/05/2020 at 1:06 PM, davewantsmoore said:

That's how I know the impedance is at maximum at ~50Hz,

Whereas Brians graph shows the impedance peaking at  ~30Hz

 

On 29/05/2020 at 6:52 PM, BioBrian said:

 

 

cheers,

 

 

 

[Dirkgerman]

 

[BioBrian]

Maybe when my ears have recovered a bit? Feeling a bit shredded.

 

Too cold to go up there again today.

 

The neighbours'd probably agree .

[Dirkgerman]

Should find good, even attenuation up and down the octaves, is a wonderful thing !~! (yes it would be chilly on site, brr, the Mtn)

 

Edited June 2, 2020 by Dirkgerman

[BioBrian]

Why? Well I was expecting this to be a bit shocking - my first try at room measurement - I admit I didn't really want to find out, after reading of everyone else's ventures.

 

For a while now, I've stopped recording the volume level, and just bring it up so the loudest part of the graph is around 90 dB. But I found out that multiplying this by 4 large speakers was probably not a good idea. Those "Wheep, wheep" noises at full range and that output are pretty painful.

 

All were taken at 1.2m above floor - my listening chair/ear height. First, I got equidistant between mains and subs, and from that reference point, came forwards incrementally a couple of feet. That gave indication of how much variety to expect initially. Wow. No wonder I bob my head around a lot.

 

So I got some reliable traces, enough to show I'm not cheating here, then at the end I switched off the sub amp, and did another sweep.

 

 

Sorry it's a bit busy. I left the phase line in for the black trace - sure it can explain a few things.

 

Firstly RED - this is what us softies have been listening to, insisting on admiring the view, before the Gesties started jesting along.

BLUE - same thing - just the 4-ways, but with the Rolladoor closed. What a price, for a few more dB of the goods.

BLACK - door open, but Gesties filling in a few holes, just like the initial wish. Was it all worth it? Pretty flat from 12.5 to 50 Hz - enough to retire on??

GREEN - door shut, all still firing. Bass not as even, but I like how shutting the door heals some of the 215 - 267 Hz problem.

 

Lots to think about.

 

 

[davewantsmoore]

19 hours ago, afa said:

 

Not wanting to nit pick 

But you said

Whereas Brians graph shows the impedance peaking at  ~30Hz

HAH!   I also said "shorthand and sloppy".    No wonder Brian was confused.

 

I have "current drive" on the brain right now .... for my own design(s) that are cooking....  Not thinking right.... Said the opposite (up instead of down) and gave the explanation which fitted what Is said (not what was happening).

 

 

It's "due to the speaker/filter imedance" would have been a lot better.  HAH.

 

 

19 hours ago, afa said:

cheers

Nitpicking is always welcome.

[afa]

4 hours ago, davewantsmoore said:

Nitpicking is always welcome.

 

 

Hey Brian, nice bit of bass extension going on there,

try some of the recordings from this site, https://www.danleysoundlabs.com/tom-danleys-mic-recordings/

 

cheers

 

[BioBrian]

9 hours ago, davewantsmoore said:

No wonder Brian was confused.

I try to do things well .

[davewantsmoore]

5 hours ago, afa said:

try some of the recordings from this site, https://www.danleysoundlabs.com/tom-danleys-mic-recordings/

but be careful with the loud knob..... they large amounts of dynamic range and loud LF

Edited June 3, 2020 by davewantsmoore

[Auracle]

On 02/06/2020 at 6:41 PM, BioBrian said:

BLACK - door open, but Gesties filling in a few holes, just like the initial wish. Was it all worth it? Pretty flat from 12.5 to 50 Hz - enough to retire on??

GREEN - door shut, all still firing. Bass not as even, but I like how shutting the door heals some of the 215 - 267 Hz problem.

 

Lots to think about.

 

 

 

On 03/06/2020 at 1:05 PM, afa said:

 

Hey Brian, nice bit of bass extension going on there,

 

 

Coming out of the shadows to say this....

That is definitely a bass extension to be envious of. (Please, no phallic jokes). Yes, you can definitely hang your hat on this one Brian.

If you feel that there is any remaining fine tuning to be done, it would be what I mentioned privately. (I know how bad perfectionists can be, - I is one too!)

Other than that, consider yourself graduated (or is that gestated?).

Thanks for sharing your journey.

 

Next project:  - Extension down to DC? What do you say Brian?  ( - just kidding! However, 5Hz is possible if it can be recorded)

[davewantsmoore]

14 hours ago, Auracle said:

Coming out of the shadows to say this....

That is definitely a bass extension to be envious of. (Please, no phallic jokes).

I'd be very skepitcal of the data much below 20Hz..... meaning it is probbaly even more extended than the chart shows.

 

... but the real test is how the SPL changes with level.... ie. does the extension hold at 90dB, 100dB, 110dB  (or is there "compression").   Even a small number of dB is critical at these frequencies (due to increased ability to descriminate changes in level at LF).

 

Quote

If you feel that there is any remaining fine tuning to be done

From 20Hz to 150Hz where the drivers overlap significantly .....  should the SPL (green vs blue) be +3dB or +6dB? 

 

Is distortion within acceptable limits at ~20Hz and peak SPL?

[almikel]

On 02/06/2020 at 6:41 PM, BioBrian said:

Why? Well I was expecting this to be a bit shocking - my first try at room measurement - I admit I didn't really want to find out, after reading of everyone else's ventures.

 

For a while now, I've stopped recording the volume level, and just bring it up so the loudest part of the graph is around 90 dB. But I found out that multiplying this by 4 large speakers was probably not a good idea. Those "Wheep, wheep" noises at full range and that output are pretty painful.

 

All were taken at 1.2m above floor - my listening chair/ear height. First, I got equidistant between mains and subs, and from that reference point, came forwards incrementally a couple of feet. That gave indication of how much variety to expect initially. Wow. No wonder I bob my head around a lot.

 

So I got some reliable traces, enough to show I'm not cheating here, then at the end I switched off the sub amp, and did another sweep.

 

 

Sorry it's a bit busy. I left the phase line in for the black trace - sure it can explain a few things.

 

Firstly RED - this is what us softies have been listening to, insisting on admiring the view, before the Gesties started jesting along.

BLUE - same thing - just the 4-ways, but with the Rolladoor closed. What a price, for a few more dB of the goods.

BLACK - door open, but Gesties filling in a few holes, just like the initial wish. Was it all worth it? Pretty flat from 12.5 to 50 Hz - enough to retire on??

GREEN - door shut, all still firing. Bass not as even, but I like how shutting the door heals some of the 215 - 267 Hz problem.

 

Lots to think about.

 

 

nice work Brian! really great to see such an impressive bass extension with the Gesties!

Interesting the results you had door shut vs door open - I did some measurements in a thread years ago and found the bass smoother with door/windows open...

 

On 05/06/2020 at 11:33 AM, davewantsmoore said:

.. but the real test is how the SPL changes with level.... ie. does the extension hold at 90dB, 100dB, 110dB  (or is there "compression"). 

I've never done these type of measurements with my setup - something to try next time I'm in the mood for the woop woops...

 

Keep in mind one of @davewantsmoore's old adages - a wiggle in the FR will show a change in the phase response....you see your phase response is nice and smooth 15Hz- 50Hz - below 15Hz the FR is dropping off (so phase change increases) and above 50Hz the FR has peaks and dips - hence the phase also isn't smooth.

 

Mike

 

 

[davewantsmoore]

2 hours ago, almikel said:

Keep in mind one of @davewantsmoore's old adages - a wiggle in the FR will show a change in the phase response....you see your phase response is nice and smooth 15Hz- 50Hz - below 15Hz the FR is dropping off (so phase change increases) and above 50Hz the FR has peaks and dips - hence the phase also isn't smooth.

What we really need to be able to read the data..... main alone, gesties alone, both.

 

... but we're really in "the room" at these frequencies.    It isn't showing much (that we can be sure of) about the speaker themselves. (eg. there's a 6db droop from 50 to 20 Hz.... bad...... but I don't know what's responsible from this data)

 

It's ok, B will just tuck them under his arm, and take them outside.     

Edited June 7, 2020 by davewantsmoore

[BioBrian]

That's why they gave me 2 arms...

[Dirkgerman]

He also blessed you with with the brain to use a forklift or similar hydraulic device to lift a sub.  Mine I still have to take out the 50Kg of drivers before I can move the 80Kg double ply box, x 2. 

 

Fantastic Job Brian. Will visit again one day to have a chin wag and take it all in.

[BioBrian]

On 07/06/2020 at 5:26 PM, davewantsmoore said:

What we really need to be able to read the data..... main alone, gesties alone, both.

 

... but we're really in "the room" at these frequencies.    It isn't showing much (that we can be sure of) about the speaker themselves. (eg. there's a 6db droop from 50 to 20 Hz.... bad...... but I don't know what's responsible from this data)

 

It's ok, B will just tuck them under his arm, and take them outside.     

I can't believe I didn't think of getting the "Gesties alone" measurement - of course, that makes a lot of sense.

 

At some point, I found a correlation with some of the peaks and dips further up, and some older measurements of just the 4-ways, so I'll check that out some more.

 

I did get a message back from the Xkitz man today. He sort of agreed that my idea of using just the LP section of their line-level XO, to roll off the subs, would work. Next thing is to decide upon the (swappable module) XO frequency - it's not clear from my above graph whether to choose 60 Hz, 100 Hz, etc. I could get more than one, for a small fee, but feel I'm shooting a bit blind right now.

 

I'd love to get up there and do more measurements, but the last couple of days have been a bit shocking here, with over 12" of rain, and the river doing its "Cannonball Express" thing, hurtling boulders and logs past, and shaking the ground. The noise floor has gone through the roof, you might say. Not to mention wrecking my water system.

[Auracle]

Hi Brian,

 

"- it's not clear from my above graph whether to choose 60 Hz, 100 Hz, "

 

Don't wish to alter your thinking here, -as I see it, these "Gesties" were designed for augmentation, therefore I'd be using the lowest possible setting for the least amount of interaction and smoother transition with the main speakers at their lower extremities. 

1 hour ago, BioBrian said:

I'd love to get up there and do more measurements, but the last couple of days have been a bit shocking here, with over 12" of rain, and the river doing its "Cannonball Express" thing, hurtling boulders and logs past, and shaking the ground. The noise floor has gone through the roof, you might say. Not to mention wrecking my water system.

Mother nature is very impressive indeed, even on a small scale. If you could only harness a fraction of that energy. Do you have a hydro turbine system? It could help with the electricity bills , since a solar installation might be a bit challenging there.

 

Keep up the good work.

[davewantsmoore]

14 hours ago, BioBrian said:

I did get a message back from the Xkitz man today. He sort of agreed that my idea of using just the LP section of their line-level XO, to roll off the subs, would work

It will.

 

It will work fine in the frequency domain.

 

Having unmatched high pass and low pass responses is going to cause issue with phase/time.... and will mean that the woofers are further away from summing together.

 

This will mean at some frequencies the combined output will be the same (or even less) than the main speakers..... which is at best pointless .... and in reality a step backwards in SQ.

 

That being said, your chart indicate that you already have the problem now (ie. the frequency regions where adding in the gesties do not increase the SPL - ie. sound is being "cancelled")

[BioBrian]

13 hours ago, Auracle said:

Don't wish to alter your thinking here, -as I see it, these "Gesties" were designed for augmentation, therefore I'd be using the lowest possible setting for the least amount of interaction and smoother transition with the main speakers at their lower extremities. 

Not a problem at all - that's why I'm posting - my thinking does need help, to come to better conclusions.

 

The graph shows the subs filling in well from 22 Hz down, and most importantly, the 10 dB dip around 47 Hz.

 

It doesn't reveal to my brain what's going on around 60-70 Hz, but it could be interaction between the four 15" drivers as a group, compounded by the room.

 

Or, that the 4-ways have a 4th order LP filter, and the subs only a 2nd order LP, and this is doing something to the phase? More soldering needed to find that out.

 

Right now, I just want to spend some money and it'll all be fixed, yeah .

 

(Dave's post has just come through, while I was fishing round for OT photos, but I'll leave this as random pre-illumination thought process).

13 hours ago, Auracle said:

Mother nature is very impressive indeed, even on a small scale. If you could only harness a fraction of that energy.

Yes, it's now up to 13" of rain in the 3 days, finally subsiding. Someone has made a free firewood delivery; they just left it on a new island though, a bit difficult to access. I'll have to speak with them.

 

 

I've never been able to visualize a DIY hydro system that would survive 4' diameter logs flying past. Best to keep it a wild river and just wonder!

[davewantsmoore]

2 minutes ago, BioBrian said:

The graph shows the subs filling in well from 22 Hz down, and most importantly, the 10 dB dip around 47 Hz.

 

It doesn't reveal to my brain what's going on around 60-70 Hz

What is happening at both these frequencies is not a "10dB dip" .... it just looks like that on your chart as you only have a data point every 5 or so Hz.

 

What is actually happening is the "dip" goes all the way down to zero.   It's a complete "cancellation".

 

At ~47 Hz, you had a cancellation with just the mains ... its being "filld in" by the gesties, which don't have a cancellation at the frequncy, and at that location in the room.

 

At 50-70Hz .... It looks like you had a peak with just the mains.   I don't think so.   It's the dip at ~47Hz which makes it look like that.    There might be a slight bump (peak) here from the room, but it's hard to know/tell..... it doesn't really matter.     The issue is, that with the gesties added, it gets transformed into a cancellation.    Meaning one speaker is competely out of phase with the other - and the dip at ~57Hz, if it were displayed with much higher frequency resolution, would actually show a complete cancellation.

 

You can test this, by simply reversing the +/- leads on your subwoofers  (playing them out of phase).     You will see an extra peak at ~57Hz, when measured at the same location.

 

2 minutes ago, BioBrian said:

Or, that the 4-ways have a 4th order LP filter, and the subs only a 2nd order LP, and this is doing something to the phase? More soldering needed to find that out.

Everywhere that the speakers don't have the same frequency response .... or where they don't have a complientary highpass/lowpass slope (accoustic slope, not just filter slope) ..... they have unmatched phase.

 

So yes.... when the main woofer rolls off at -6dB = 100Hz  4th order ..... and the subwoofer rolls off a -6dB = 60Hz 2nd order  (I just made up some numbers, you know what they really are) ...... then they don't match phase, and they don't move together.

 

 

2 minutes ago, BioBrian said:

I've never been able to visualize a DIY hydro system that would survive 4' diameter logs flying past. Best to keep it a wild river and just wonder!

[BioBrian]

44 minutes ago, davewantsmoore said:

This will mean at some frequencies the combined output will be the same (or even less) than the main speakers..... which is at best pointless .... and in reality a step backwards in SQ.

 

That being said, your chart indicate that you already have the problem now (ie. the frequency regions where adding in the gesties do not increase the SPL - ie. sound is being "cancelled")

[Edit: you beat me again - I just clicked as your 2nd post came through, so I hadn't read it.]

 

Maybe it's just luck then, that the four 15" drivers aren't summing at 50 Hz, but filling in the bits I wanted.

 

I have to agree with the cancellation thing though - it's quite frustrating to walk around and have to feel the gestie drivers with a long rod, to even be sure they're earning their money.

 

What's happening around 60-70 Hz does need a look - ATM I think it's more the room, but I can try more mike positions.

 

I think at my level of understanding, it would be good to trot out some individual frequency tones and check out what's happening.

 

I have noticed that when playing organ music, there were notes missing in low scales before, and now it sounds like they are all there. I'd really like to get a recording of a slow chromatic scale, or notes from 20, to say 150 Hz. Sine wave tones are so out of tune!

Edited June 24, 2020 by BioBrian

[BioBrian]

At 60-70 Hz, if it's not the room, I suspect it's more to do with the interaction of the 8" and 15" drivers in the mains. I've had trouble getting meaningful measurements of this, and I'd not be surprised if measuring these from 3.2 m distance (as in listening position) shows up peaks/cancellations (including floor and ceiling bounce) there.

 

It all gets crazy complicated! Warning to prospective sub builders: "don't try this at home". Well unless you do it the proper way, haha.

 

Lots to try.

 

So with the Xkitz, @davewantsmoore, if I need to match slopes I need 2 lots of 24 dB/oct filters, and (groan) another stereo amp? Should save my money?

[davewantsmoore]

1 hour ago, BioBrian said:

At 60-70 Hz, if it's not the room, I suspect it's more to do with the interaction of the 8" and 15" drivers in the mains. I've had trouble getting meaningful measurements of this, and I'd not be surprised if measuring these from 3.2 m distance (as in listening position) shows up peaks/cancellations (including floor and ceiling bounce) there.

No.   Floor/ceiling bounce doesn't affect 70Hz..... only distances much larger than that do.

 

The "mains only" respons has a peak around 70Hz..... but I don't think it's as much a peak.....  rather, that you have a dip on either side (which is the room talking) .... and that around 70hz is also natural peak in the speaker response due to things we discussed a lot earlier in the thread  (tuning the cabinet very low -causes a slow rolloff between about 70Hz and 20Hz) ... and on the upper-side, the action of your low pass filter (24db at 100Hz, right!?)

 

 

Measurements of these frequencies are dominated by the "room".... by the location of the source and mic relative to the room.

 

1 hour ago, BioBrian said:

It all gets crazy complicated!

Having a good simulation is key.   Once you have a modelled response which you know is "correct" .... then whenever you take a measurement, you can ask "why is my measurement different from the model.... what caused it?!"

 

At these frequencies, it is the room modes, and the speakers and mic relation to them.

 

 

1 hour ago, BioBrian said:

@davewantsmoore, if I need to match slopes I need 2 lots of 24 dB/oct filters, and (groan) another stereo amp? Should save my money?

I don't really know what to tell you.

 

You need to have a design/plan for how you will make all the drivers in a multi-way system work together.    Incuding a subwoofer.

 

If you're going to "split" the mains and subs .... then, yes.   They both have wide bandwidth... so you could just pick a Hz, and get a highpass + lowpass filter there.

 

However, you might like to investigate the "gesties" response on their own in the room ..... to make sure, that you'd be happy with their response from 0 to 60Hz (or wheverever you decided).

 

 

-----

What I would have done/suggested..... Is to build a 3 way main speaker, with a response down to 60 to 80Hz... with a 24dB/octave high pass response   (so either ported.... or sealed with a 2nd order highpass added).

 

Then I would design 4 subwoofer, which play from 0 to 80Hz (or whaetever the XO frequency is) all with the same low pass response that matches up to the "3-way mains".

 

At 80Hz, the subwoofers will need to be within about a meter give/take of the same listening distance as the mains.... before you need to start thinking about delay....   BUT, what you could have done is use a mono subwoofer amplifier for each subwoofer, that had a delay knob on it ready to go.

 

 

 

[davewantsmoore]

2 hours ago, BioBrian said:

Maybe it's just luck then, that the four 15" drivers aren't summing at 50 Hz, but filling in the bits I wanted.

Yes...

 

Look at your chat for mains, and then mains+subs.

 

Below 20Hz.... the mains drop off, so of course the subs are filling in.  Also, that these frequencies everyting is in phase no matter what the disatcnes in room (as the wavelengths are too huge).

 

There's tew lumps (22, and 35, ish) where the gesties are reducing the output.... and in between they're increasing it a little bit (but not as much as they should be, ie. there's still some cancellation happening... on e driver working against the other).

 

From 40 to 50 odd.... the gesties are playing, where the mains are not (due to a null at the point int he room where you've put the mic)

 

Quote

 

What I would have done/suggested..... Is to build a 3 way main speaker, with a response down to 60 to 80Hz... with a 24dB/octave high pass response   (so either ported.... or sealed with a 2nd order highpass added).

 

Then I would design 4 subwoofer, which play from 0 to 80Hz (or whaetever the XO frequency is) all with the same low pass response that matches up to the "3-way mains".

 

At 80Hz, the subwoofers will need to be within about a meter give/take of the same listening distance as the mains.... before you need to start thinking about delay....   BUT, what you could have done is use a mono subwoofer amplifier for each subwoofer, that had a delay knob on it ready to go.

 

 

To go a step further towards great performance..... I would suggest you ran those 4 channels of woofer through some sort of "DSP" (just a simple miniDSP would be fine/great)    So you can individually EQ and delay each woofer so it is "flat" by it self ..... and then also EQ the sum of the 4 woofers so it matches to your subjective/preference target.

 

Big boxes and $ drivers ... aren't sufficient on their own.

[BioBrian]

All good, thanks. I'm digesting it all slowly, but some of it's a bit hard for my old teeth.

 

I went up to start measuring, but the roar of the river is still echoing around 'them thar hills' too much.

 

Can't get past this one for now:        IF I set the main LP filter at 100 Hz, 24 dB/oct, and the subs one at 60 Hz, 24 dB/oct, is it necessarily going to be "out of phase"? Say at 80 Hz, one would be simply outputting more volume than the other - could they not still be in phase, just different contribution levels to the whole?

 

Edited December 1, 2020 by BioBrian