DIY Dayton Audio UM18-22 sub

ATC & other makes of Hi-Fi equipment
User avatar
Posts: 668
Joined: Tue Jun 15, 2010 12:59 pm

Re: DIY Dayton Audio UM18-22 sub

Postby AlmaataKZ » Thu Jan 05, 2017 12:29 pm

Ah, great! big driver, two subs!

MattSPL wrote:The box design is roughly 4cuft, which is a good compact size for a driver like this, it's a 22 inch cube with curved sides, so about 18inch's at the back, then what I'd say is the maximum amount of acoustic wadding added to increase the size of the box the driver see's.

What I am trying to understand is how one arrives to these box parameters. What is the design process here? or is it just a guess or you put something (driver parameters etc) into a special calculator and it gives you the 'good' enclosure parameters?

the driver and amp choice looks sound to me (although I am in no way an expert).

MattSPL wrote:I've always been fascinated with bass, speakers/subs and as I listen to mostly electronic music, part of the hobby for me is finding music with 20hz and lower content, so I went for an 18 inch driver in order to get low with the least amount of distortion.
With the sealed cabinet, EQ and power is added to stretch the frequency response down to below 10hz.


another thing that I am trying to understand about designing and building speakers/subs is how these low Hz theoretical or measured ability translates into the user actually hearing the recorded bass notes. The measured response is via the microphone, which has the same sensitivity over the f range. The human ear is progressively less sensitive at lower f. So if, say, a 20Hz fundamental tone has a 40 Hz harmonic of a certain level it may feel several dB louder than the 20Hz fundamental (after including the ear sensitivity drop). This would mean the listener is listening to the distortion product and not to the original note. the lower the f, the higher the harmonic will be and the less sensitive the ear will be to the fundamental. So how does one know what is actually audible at these crazy low Hz?!
In order to understand the full picture the levels of distortion at various f and levels need to be understood. Did you have a chance to dive into this and if so, how does it look? When this sub is fed say a 15Hz tone at 90dB, what is the distortion picture? How loud are the harmonics and where does it put them after taking ear sensitivity (and room response) in mind?
having such a project and the equipment with the adjustments and measuring tools like you have gives a perfect opportunity to understand this.
would love to hear your thoughts and experience.
iMac/Bluray/PVR - Marantz SR7008 - 7.1 ATC SCM50SLAT fronts, C1C centre, SCM7 rears, MonitoAudio M3 rear surr, C4 sub
Airplay multiroom (aex to genelec 6010, b&w mm1, avi dm10)
System here: http://atcforums.co.uk/viewtopic.php?f=16&t=667

User avatar
Posts: 6434
Joined: Tue Feb 23, 2010 5:46 pm
Location: Ireland

Re: DIY Dayton Audio UM18-22 sub

Postby MattSPL » Thu Jan 05, 2017 8:53 pm

Hi Almaatakz,

The box volume is calculated using the winisd program. I didn't do this myself, I just took guidance from other people on AVSforums who've used this driver. They input the driver parameters and it gives the non eq'd response and then also the wattage required to reach the drivers xmax at 10hz, as this is the general goal for a diy sealed sub. so EQ is used to boost the low frequencies to compensate for a smaller box. Alternatively, a much larger box can be built which would require less or no EQ and therefore less wattage to reach xmax at 10hz.
So using winisd and your chosen cabinet size determines how much power you need.

The way I have tuned my subs crossover is a combination of using the microphone to plot a curve and ultimately by ear to get the desired result. I run a frequency sweep from 5hz to 120hz for example and aim for a smooth curve where the sweep sounds the same volume from top to bottom, or as close as I can get it.

I did take distortion measurements from the seating position over the Christmas, but have been told they are pointless as the room and it's contents are measured too. You need to measure close to the driver, preferably outside to get accurate results.
Image

You can see how much influence even closing the door has.
Everything above about 30hz is the speakers, so all those peaks are unavoidable. I will add room treatment and a minidsp 2x4 to address this.

Image

Image
Sony BDP-S790 > Matrix Quattro Dac > Digidesign RM2's with SB Acoustics drivers > DIY Dayton Audio UM18-22 Sub

User avatar
Posts: 668
Joined: Tue Jun 15, 2010 12:59 pm

Re: DIY Dayton Audio UM18-22 sub

Postby AlmaataKZ » Thu Jan 05, 2017 11:18 pm

thank you for explaining re box parameters and crossover setting

What distortion levels are you seeing with this design say at 30Hz, 20Hz and 10Hz at a useful spl level?Can you see the levels of individual harmonics?
iMac/Bluray/PVR - Marantz SR7008 - 7.1 ATC SCM50SLAT fronts, C1C centre, SCM7 rears, MonitoAudio M3 rear surr, C4 sub
Airplay multiroom (aex to genelec 6010, b&w mm1, avi dm10)
System here: http://atcforums.co.uk/viewtopic.php?f=16&t=667

User avatar
Posts: 6434
Joined: Tue Feb 23, 2010 5:46 pm
Location: Ireland

Re: DIY Dayton Audio UM18-22 sub

Postby MattSPL » Thu Jan 05, 2017 11:33 pm

Hi Almaatakz,

I've not really ventured that far into the software yet as I've not had much time, but if you look here http://www.data-bass.com/data?page=driver&id=58
This is a proper test done on my driver(plus many more) in a 4cuft enclosure like mine. He gives a very detailed write up of output and distortion.
Click the link at the bottom of the above linked page to see the sealed enclosure test.

Cheers
Matt
Sony BDP-S790 > Matrix Quattro Dac > Digidesign RM2's with SB Acoustics drivers > DIY Dayton Audio UM18-22 Sub

User avatar
Posts: 668
Joined: Tue Jun 15, 2010 12:59 pm

Re: DIY Dayton Audio UM18-22 sub

Postby AlmaataKZ » Fri Jan 06, 2017 10:59 am

I am trying to interpret the enclosed test graphs and translate the measurements into what sound can be heard when feeding different f and levels. I am not sure if I am dong it right and my figures are definitely inaccurate, but I get smth like this:

at 89.6V signal (giving 117db output at 50Hz), the sub has the following THD:
(note this is an extremely high level but this is the only level at which a curve is plotted at both the THD and the THD component graphs)

tone - component THD - 2nd harmonic SPL(-10dB) - 3rd harmonic level dB (-5dB)
20Hz - 40%--------------107dB-------------------------112dB
10Hz - off scale----------say115dB---------------------say115dB

Roughly corrected for ear sensitivity, the levels of the 2nd (15dB sensitivity difference 20Hz to 40Hz)) and 3rd harmonics (22dB sensitivity difference 20Hz to 60Hz) will be:
2nd, 40Hz = 122dB
3rd, 60Hz = 134dB

then, if the room has a resonance at say 40Hz (and most rooms will have it somewhere in the low region), that particular tone will be further boosted by 6-12dB.

this means that the original 20Hz, 10Hz or whatever extremely low tone will be completely inaudible as it will be masked by the sound of harmonics because we are more sensitive to them and also because some of them are further boosted by the room.

so, the question is - is it at all worth feeding the sub with below 20Hz frequencies? I think not. Because the system (sub-room-ear), in our perception, effectively replaces these tones with the dominant harmonics.

Is it perhaps better to only feed the sub with what it can handle within very low distortion levels (under 5%?) so then we know that what we hear is actually part of the recording?

I have to say that this is very unscientific and some figures (e.g. 134dB) are obviously unrealistic, but I think it still illustrates the point - if the system is theoretically capable of handling a low tone it is not necessarily what we will hear. So how does one understand what is one hearing?

Matt, this is not to criticise your build in any way. I would love to do something like this one day myself so you are part of the inspiration! I am just trying to understand how it works in principle and what is important. It may also help in your thinking while tuning the crossover and dsp settings, e.g. what to send and what not to send to the sub. I think yours is very powerful and is same or better than many many commercial subs in the 2-4K price range, plus the knowledge and experience that it will give to you.

adding the second sub will allow to decrease the level on each and therefore decrease the distortion and increase the chances of hearing that clean extremely low tone :-)
Last edited by AlmaataKZ on Sat Jan 07, 2017 1:16 am, edited 2 times in total.
iMac/Bluray/PVR - Marantz SR7008 - 7.1 ATC SCM50SLAT fronts, C1C centre, SCM7 rears, MonitoAudio M3 rear surr, C4 sub
Airplay multiroom (aex to genelec 6010, b&w mm1, avi dm10)
System here: http://atcforums.co.uk/viewtopic.php?f=16&t=667

User avatar
Posts: 6434
Joined: Tue Feb 23, 2010 5:46 pm
Location: Ireland

Re: DIY Dayton Audio UM18-22 sub

Postby MattSPL » Fri Jan 06, 2017 2:21 pm

Hi Almaatakz,

I don't think we need to worry about the higher frequency harmonics associated with the lower frequencies because my sub is only running upto 30hz or so. Even if running the sub upto 100hz for example, using either an auto or manual eq will allow a flat response to be achieved, so I imagine any upper harmonics can be flattened out.

I don't understand the science behind this, so can't explain anything other than what I hear and feel as pressure in the room, and to my ears it's doing a great job.
As you say, the performance of this sub will equal or better a lot of commercial subs, for example, comparing it on databass to the Velodyne DD18+ which costs £6K, a sub like mine on paper is equal or better performing. But as with everything Hifi, trust your ears ultimately.
Sony BDP-S790 > Matrix Quattro Dac > Digidesign RM2's with SB Acoustics drivers > DIY Dayton Audio UM18-22 Sub

User avatar
Posts: 668
Joined: Tue Jun 15, 2010 12:59 pm

Re: DIY Dayton Audio UM18-22 sub

Postby AlmaataKZ » Sat Jan 07, 2017 1:08 am

Matt
Higher harmonics will be there even if the crossover upper frequency is at 30hz. The harmonics do not come from the signal (which the crossover would have cut out), they come from the sub/driver itself so physically cannot be filtered out by any electronics or equalised out (but can be amplified by the room resonances, sadly). They will mask the sound once the distortion level is high enough. The question is how high is high enough. But it is relatively easy to answer (see the Linkwitz link for the logic). Most subs I ever heard of are incapable of producing audible 20Hz at medium to high spl levels, they hit the masking levels of distortion well before that. As can be seen on the excellent site that you referenced, only several huge subs can do this.

However, I agree that your sub will be as good or better than even very expensive subs!
iMac/Bluray/PVR - Marantz SR7008 - 7.1 ATC SCM50SLAT fronts, C1C centre, SCM7 rears, MonitoAudio M3 rear surr, C4 sub
Airplay multiroom (aex to genelec 6010, b&w mm1, avi dm10)
System here: http://atcforums.co.uk/viewtopic.php?f=16&t=667

User avatar
Posts: 6434
Joined: Tue Feb 23, 2010 5:46 pm
Location: Ireland

Re: DIY Dayton Audio UM18-22 sub

Postby MattSPL » Sat Jan 07, 2017 2:38 am

Hi Almaatakz,

I think once the driver remains well within it's rated xmax, these distortion levels shouldn't be a problem. So far, I've only seen about 10mm of travel from this sub, it's barely twitching most of the time. I'd be more concerned about the main speakers trying to reproduce the sub 20hz signals, because my speakers are giving their best at playing a 5hz test tone. This is severely limiting my maximum spl and testing of the sub. I intend to add the minidsp 2x4 in order to put a high pass filter on the speakers.

I agree though, most subs cannot play the sub 20hz material properly, but JL Audio and Velodyne do a great job and I've heard that Paradigm do too.
I've never heard driver noise/distortion from the JL Audio or this diy sub, but did hear some audible driver noise from the Velodyne DD12, even though the servo seriously restricted the subs spl on the low notes.
I was able to get usable 15hz from the Velodyne and strong 15hz output from the JL Audio. This diy sub is audible at 10hz, but it's mostly room pressure and a feeling in the ears.
Subtle things you get by producing these deep bass notes are for example in a film when somebody slams a door shut, you not only hear the door shut, but feel the pressure change in the room just like in real life.
Sony BDP-S790 > Matrix Quattro Dac > Digidesign RM2's with SB Acoustics drivers > DIY Dayton Audio UM18-22 Sub

User avatar
Posts: 668
Joined: Tue Jun 15, 2010 12:59 pm

Re: DIY Dayton Audio UM18-22 sub

Postby AlmaataKZ » Sat Jan 07, 2017 10:19 am

Great! One day I hope to do a similar build!
iMac/Bluray/PVR - Marantz SR7008 - 7.1 ATC SCM50SLAT fronts, C1C centre, SCM7 rears, MonitoAudio M3 rear surr, C4 sub
Airplay multiroom (aex to genelec 6010, b&w mm1, avi dm10)
System here: http://atcforums.co.uk/viewtopic.php?f=16&t=667

User avatar
Posts: 6434
Joined: Tue Feb 23, 2010 5:46 pm
Location: Ireland

Re: DIY Dayton Audio UM18-22 sub

Postby MattSPL » Sat Jan 07, 2017 3:10 pm

AlmaataKZ wrote:Great! One day I hope to do a similar build!


If you don't want to build the box yourself, you could get one made and then do the rest yourself.
Sony BDP-S790 > Matrix Quattro Dac > Digidesign RM2's with SB Acoustics drivers > DIY Dayton Audio UM18-22 Sub

Previous

Return to Hi-Fi

Who is online

Users browsing this forum: No registered users and 2 guests