The Bargain Hunter's subwoofer ...

Great thread MPW.. Have you decided on the Topology which you will Use. since you are going DIY and hence can actually make the cabinet furniture grade or even design it as a Corner table etc)
Do you want to consider the Push Pull Sub ? I believe it is larger in size but has one of the most satisfying -Musically -bass ( not an expert on design :) )

arj,

I am going for a 1 driver sealed subwoofer.

If you see pics of my setup http://www.hifivision.com/speakers/51650-merlin-tsm-bme-14.html

post no 139

i see the rack moved back in the centre..flanked by 2 subs and again flanked by the Merlins. In short the subs would be ideally at the feet of the Merlins.

The positioning of the subs may change depending on their interaction with the room.

Which is why i have started with 1 subwoofer and will ramp up to the 2nd one if required only. If should be fairly easy to mane the 2nd one once you make the first.

I am aware this is my first DiY and it may be possible my DiY ( the first time ) may need tuning.

I am open for mistakes and corrections.

Practicals is the best way of learning the theory :lol: ,, as long as you are enjoying it. For me personally the journey is more interesting than the destination

which is why i decided to get my feet wet in subwoofer territory and not to rely too much on...."reports "..

rgds
 
Last edited by a moderator:
For those who might be interested,The actual measured parameters of my driver are as follows:
LAB 12 -measured parameters

Fs = 22.64 Hz
Qms = 15.28
Vas = 113.3 liters
Cms = 0.34 mm/N
Mms = 146.8 g
Rms = 1.353 kg/s
P-Dia = 248.3 mm
Sd = 484.3 sq.cm
Qes = 0.38
Re = 4.7 ohms
Le = 0.97 mH
Z = 5.64 ohms
BL = 16.07 Tm
Qts = 0.371
no = 0.334 % ( calculated)
1-W SPL = 87.38 dB ( calculated)
2.83-V SPL = 89.69 dB (calculated)
-----------------------------------------
Box Properties
Type: Closed Box
Shape: Cube ( 39 cm cube ( ext dimen)-19mm MDF)
Vb = 40.08 liters
Qtc = 0.653
F3 = 48.06 Hz
No filling.

I did measure the actual in box parameters but I don't know where it is. I find a note against Lab12 in my notebook which says Qtc 0.65 Fr in box 43.5 Hz.
I'll have to do it again to confirm it but I do remember that it sounds fine. I like subs but not fanatical about them , so this unit has not been used for a few months now due to various reasons. I have just got the amp for it though the crossover isn't ready yet. Should finish this year !
This of course is slightly over damped . But you already know that as the unit goes closer to the wall or corner the LF gets lifted significantly and you will get quite a lot of extension.

Here are a couple of plots ( simulation of course !) of the box in the middle of the room and at one corner ( labeled - 'in room' ). You can see that the proximity of the wall increases the LF performance.
 
Last edited by a moderator:
Fantastic

Did u actually look at the math and graphs before you selected the Eminence 12 ?

The reason i asking; the subwoofer driver i selected initially ( Seas L26 ROY ) is not available in the Indian market anymore.

Its very heavy ( 10 kg driver alone ) and so shipping to India is a pipe dream on cabin baggage unless i could find a kind soul who would lug it for me. That was an improbable situation and put me off.

I am informed stocks would be back in June 2015 but i was unwilling to wait that long.

Aluminium woofer cone was recommended to me for fast agile movement and so with he Seas not available - we dug around a bit and came up with the Dayton RSS265HF-4

Google it for the spec sheet.

It got home last afternoon.

The selected amplifier is the Hypex DS4.0. I should have this with me in about 10 days.

Ideally i would have liked the DS 8.0 but i do not have the money for it and my target was to get a sealed subwoofer made at the INR 50000 range. The DS 4.0 fits the bill for the Dayton RS 10 HF just about right.

I used Win ISD and i am with the input values i get an internal volume of 15.8 liters for a sealed box.

Can someone re-check this for me please ?

Regards
mpw
 
Last edited by a moderator:
What is the size of your room?
What is the existing SPL capability of your mains?
How low can the mains go on their own? Are they ported or sealed?
 
You have a big room of 380 sqft and two sealed bass sources in you mains.

The problem with bass in home listening situations is that it is very modal in nature. What we hear is not only direct sound but direct sound and combination from multiple reflections. This gives rise to peaks and dips in bass. The bigger the room the lower in frequency the peaks and nulls are. This issue is of high significance in bass due to the fact that we dont have ability to differentiate between the direct and reflected sound in bass. What we hear is the combination. We hear bass in steady-state. The bass that one gets in listening position thus will be far from even what the alignment used suggests. Peaks and dips as high as 15dB are common in typical rooms. Rooms simulations dont help much either because the real room is very different due to windows, openings and furniture etc which hardly gets captured in simulation. Bass traps dont help much because the wavelengths concerned are large and lot of bass traps would be needed to really tame the reflections. Global EQ also does not work in the modal region

Bass in rooms is thus a spatial-steady state problem and needs a spatial-steady state solution. Much of what I am writing is from the below links which are good sources of recent work in this area
www.gedlee.com/downloads/OptimalBassPlaybackinSmallRooms.pptx
https://www.youtube.com/watch?v=SCWL-zusyqw


Good bass will need distributed bass sources. Even the mains can serve as bass sources if they go low enough and have shallow roll offs.

So, since you have mains that go low and have shallow roll offs, its a benefit that you can use. I would suggest an overlapping distributed bass solution, a 4th order bandpass subwoofer will perform very nice in 2 octave 45hz to 180hz bandwidth. The bandpass suppresses all higher order distortion harmonics and hence produces very clean sound. Below 40hz (ie below modal region) multiple subwoofers will combine in phase and so global EQ can be used to equalise the lowest octave.

Make stereo bandpass subwoofers and use just about any decent driver + amp. They are not really critical in the system. Some of the car subwoofers are infact very good. They are robust and have the right parameters like excursion and power handling. The amp can be a simple stereo amp that can go till 20hz and have sufficient power.

Use minidsp for crossover+EQ. One unit will suffice for a stereo sub.

Then take measurements to get a flatter response in the listening position. The resulting bass is tailored for the room and will be similar to the bass that we hear in headphones.
 
Last edited:
@mpw
Have you tried checking bass with and without the external zobel network? In my experience with zobles they introduce phase shifts by altering the filter phase which in turn can affect your low freq response.
If possible measure your speaker impedance with and without zobel using REW and check how the impedance phase varies. If the impedance phase varies more than 30 deg then it could possibly indicate issues.
 
You have a big room of 380 sqft and two sealed bass sources in you mains.

The problem with bass in home listening situations is that it is very modal in nature. What we hear is not only direct sound but direct sound and combination from multiple reflections. This gives rise to peaks and dips in bass. The bigger the room the lower in frequency the peaks and nulls are. This issue is of high significance in bass due to the fact that we dont have ability to differentiate between the direct and reflected sound in bass. What we hear is the combination. We hear bass in steady-state. The bass that one gets in listening position thus will be far from even what the alignment used suggests. Peaks and dips as high as 15dB are common in typical rooms. Rooms simulations dont help much either because the real room is very different due to windows, openings and furniture etc which hardly gets captured in simulation. Bass traps dont help much because the wavelengths concerned are large and lot of bass traps would be needed to really tame the reflections. Global EQ also does not work in the modal region

Bass in rooms is thus a spatial-steady state problem and needs a spatial-steady state solution. Much of what I am writing is from the below links which are good sources of recent work in this area
www.gedlee.com/downloads/OptimalBassPlaybackinSmallRooms.pptx


Good bass will need distributed bass sources. Even the mains can serve as bass sources if they go low enough and have shallow roll offs.

So, since you have mains that go low and have shallow roll offs, its a benefit that you can use. I would suggest an overlapping distributed bass solution, a 4th order bandpass subwoofer will perform very nice in 2 octave 45hz to 180hz bandwidth. The bandpass suppresses all higher order distortion harmonics and hence produces very clean sound. Below 40hz (ie below modal region) multiple subwoofers will combine in phase and so global EQ can be used to equalise the lowest octave.

Make stereo bandpass subwoofers and use just about any decent driver + amp. They are not really critical in the system. Some of the car subwoofers are infact very good. They are robust and have the right parameters like excursion and power handling. The amp can be a simple stereo amp that can go till 20hz and have sufficient power.

Use minidsp for crossover+EQ. One unit will suffice for a stereo sub.

Then take measurements to get a flatter response in the listening position. The resulting bass is tailored for the room and will be similar to the bass that we hear in headphones.

Hi goldy,

thanks for your post and also pointing out possible scenarios.

At this point i am simply trying to build a sealed subwoofer to handle everything below 55 Hz / 50 Hz.. The Merlins will continue to run full range.

Once i get the subwoofer done, then it would validate many things - practically in my room.

I am trying not to introduce too many electronics in the chain and to keep it as simple as i can.

But time will tell weather this endeavor leads me to " meet " my goals.

Like you correctly pointed out graphs and simulations are no substitute for real listening room experience.

Some of the graphs do are not even legible let alone inferable :lol:

@mpw
Have you tried checking bass with and without the external zobel network? In my experience with zobles they introduce phase shifts by altering the filter phase which in turn can affect your low freq response.
If possible measure your speaker impedance with and without zobel using REW and check how the impedance phase varies. If the impedance phase varies more than 30 deg then it could possibly indicate issues.

haribhai,

it would be best to visit the Merlin website and write to Bobby direct - whatever questions or doubts as to the efficacy of Zobel network.

He would possibly welcome intellectually stimulating questions which befuddle folks like me who play by the ear.

I am serious.

You could then share valuable information and add to your thread on the Zobel n/w.

Your intellectual quest would possibly be answered by an authority decades of speaker building experience and considered in very high regard.

As for me - i have tried it with / without.

I have also tried a Duelund cap zobel.

I prefer with.
 
Last edited by a moderator:
Fantastic

Did u actually look at the math and graphs before you selected the Eminence 12 ?

Mpw

Yes I did look at the simulations and graphs before buying the Lab12. Wouldn't plonk so much money down without the details. Driver is reasonably close to published spec. It's a good driver available right at our doorstep !
Linear Xmax is 13 mm if I am not mistaken. Mechanical max is more. I think it was 20mm.
 
Last edited:
If a zobel is added externally, then it most likely is before the x-over, right? Shouldn't a zobel be AFTER a x-over?
 
Yes I did look at the simulations and graphs before buying the Lab12. Wouldn't plonk so much money down without the details. Driver is reasonably close to published spec. It's a good driver available right at our doorstep !
Linear Xmax is 13 mm if I am not mistaken. Mechanical max is more. I think it was 20mm.

Out of all the technical data and the spec sheet we get with any driver we want to buy .... what are the 5 most important parameters for a subwoofer ?

I must admit :

a. i went on a recommendation ( The Dayton RC 10 HF is a close approximation to the original Seas L26ROY recommendation )

b. I looked at the weight of the subwoofer. Somehow this is a factor which turns me on. Heavy speakers are generally a good ( but not the sole ) indicator - for me atleast.

c. I looked at the price ( i did not want the price to be too low either )

regards
mpw
 
Received reply from Bobby-Merlins

The RC used in the merlins is not intented to be used as a zobel as thought. Its a RC network used in the HF region for preventing spurious signals from reaching the tweeter beyond its operating range. Depending upon the value of the capacitor it can be decided. Its mostly useful from 100KHz and above. The rc network also helps in reducing any RF & EMI noise and also prevents the amplifier from going into oscillations & ringing. This in turn as per Bobby helps to open up the sound stage in the HF region make its response more smoother and strain free.

This termination is useful if your power amplifier is not terminated with this rc network. I remember using them in my Beta 22 amplifier project giving similar explanations for using them. I am not sure if the Dared tubes uses them, but nevertheless using them will not harm in any way. I am planning to try them over this week using a 22ohms MOR and 0.047microfard across my speaker terminal. Though the simulation with the above did not in anyway impact any of my speaker parameters, i believe its more to do with the amplifier working than the speaker working. The speaker indirectly benefits with the amplifier termination.
 
hey guys..

i am building a subwoofer here :)

i wanted to know which ( say ) 5 parameters out all the tech spec would be most important in selecting the right subwoofer driver.

would the material of the cone be an important consideration ?

rgds
 
A robust build(which probably translates into weight), with good cone and basket, decent Xmax and power handling capability, good user feedback, decent after sales if you need replacement surrounds for example( I did), are some factors that come to my mind.
 
if that be the case then why are there so many published specs

Fantastic's post on the Eminence 12 is an example.

what quantitative determining points ( apart from budget ) would have led him to determine the suitability of the Eminence 12 or any other driver for his application ?

mpw
 
Let's take the case of a car-you get a whole raft of specs-length, width, height, ground clearance, engine displacement, transmission, fuel consumption, weight, fuel tank capacity-and so on-I could add to this list. Would all buyers consider all of them? I would guess not. Each buyer would consider the ones relevant to him/her.
 
Buy from India's official online dealer!
Back
Top