I am going to investigate in the directions you have pointed before fully committing to changes.
Sorry Joji, somehow I missed to see the pdf. Those
measurements look better than I would have thought.
measurements look better than I would have thought.
Rick,
The woofer XO and the mid/high XO are wired out of phase in the speakers.
Do you remember if that's as designed?
thanks
Joji
Jojip,
See I told ya to talk/email Rick. 😉 It is almost always that the original designer can shed most light on such issues.
Jojip, from what I understand the only access you have to the insides of the speakers are from the driver's mounting holes. Right?
if the cabinets are indeed so much oversized, then it will affect the midbass response and thereby the "attack" or "snap" of the speaker.
Rick,
I am not at liberty to discuss my speakers but given that you are the original designer, can I put you on the spot (forgive me) by asking what should be the target volumes for the W22, Peerless sub and W15 mid so JoJip can add the requisite number of "bricks".
Could he not roll up a thick rubber sheet (with glue) and introduce that inside the cabinet (the rubber sheet will unwrap itself and adhere to the side walls there by reducing volume)?
On a different but related topic, I am perfectly happy with the bass of the W22s in bass reflex driven by the 90W Yamaha S2000 amplifier. My room is 14' W, 9.5' ft high and 30' long. The speakers are along a short wall.
Newer versions of Dayton Audio WT3 actual is a LCR meter too, have a look specs and how usefull/helpful a tool at PartsExpress, Dayton Audio, or trueaudio websites. WT3 later got rebranded as DATS version 1 hardware and this year version 2 hardware inclusive version 2 software was released.
Last edited:
Thats correct Navin.
Removing the crossover boards was a battle, because they were glued to the cabinet + 4 screws with glue on the threads.
Had to break off part of one board without damaging any components.
One reason I suggest glued rubber sheet to walls is that the sheets wont move if you happen to move the speakers. Bricks may move and hit the speaker basket.
Will need the right volumes if Rick can reveal them. As he mentioned that's needed irrespective of the direction taken now.
Wooden blocks glued to the cabinet or rubber sheet you suggest are both options.
With Rick's comments, here is the direction i feel should be taken.
Please advice
1. Correct the cabinet volumes if Rick can provide the values
2. Adjust tweeter level with resistor changes.
3. Integrate the subs well with the 8" using ARTA measurements
4. Cleanup the mounting holes.
#2 is worth trying, though as suggested by Matt I am in favor of replacing the ribbon with a dome which can get the best out of the mid range and keep distortions and breakup out.
Matt can you comment on the Bass in light of Rick's inputs?
I can take more critical measurements to bottom out on understanding the bass response
FWIW,
5. ease up room issues. Placement experiments, carpet/rug, spike to tilt the speakers? (Note sure there are spikes to tilt these 160lb monsters 🙂
Please advice
1. Correct the cabinet volumes if Rick can provide the values
2. Adjust tweeter level with resistor changes.
3. Integrate the subs well with the 8" using ARTA measurements
4. Cleanup the mounting holes.
#2 is worth trying, though as suggested by Matt I am in favor of replacing the ribbon with a dome which can get the best out of the mid range and keep distortions and breakup out.
Matt can you comment on the Bass in light of Rick's inputs?
I can take more critical measurements to bottom out on understanding the bass response
FWIW,
5. ease up room issues. Placement experiments, carpet/rug, spike to tilt the speakers? (Note sure there are spikes to tilt these 160lb monsters 🙂
Last edited:
Why do you think your listening environment is poor? Big usually equates to good with one or two exceptions. A high roof is ideal if you opt for room treatment and can put much of it below a false ceiling possibly along with some of the subwoofers.
The reason you are down something like 6dB at 100 Hz at the listening position is because the speaker has no baffle step correction. Rick has confirmed this albeit implicitly.
Rick was suggesting cutting the mid and tweeter levels but you would also need to shape the bass response. 5th element was suggesting parallel woofers and using the output from the woofer and tweeter to help the inefficient midwoofer operating over a limited range but there is still significant design effort left to integrate with the baffle and the acoustic centres of the drivers. Adding another midwoofer does not look practical. Replacing the midwoofer with a more efficient midrange would make the crossover with parallel woofers easier and allow the midrange to operate in a conventional manner but this is becoming a major change.
If you changed to an active DSP all the issues would disappear except the gap between the operating ranges of the mid and tweeter. You would also have a better performing speaker and the flexibility to make a range of changes for zero cost except time and effort. As the costs rise to get an acceptable solution with a passive crossover switching to an active one becomes more attractive. I would suggest costing the full passive solution before spending anything significant on parts and test equipment.
If you are high passing a driver the loading provided by cabinet becomes less important. An overly large cabinet with appropriate stuffing might even be preferable to one sized to give the optimum compromise between bass extension and transient response. If you do decide to experiment with reducing the enclosure volumes I would suggest doing it in a reversible way.
The only driver that is not high passed is the subwoofer and I think you may need to look a bit closer at what it is doing. It doesn't simulate particularly well and may have been designed more to work with a passive radiator. At these frequencies one needs to equalise quite heavily to get a good in room response and so having a less than optimum unequalised response may not matter much so long as the equalised driver can move the air as required with the power available.
Since the subwoofer is being XOed at 90Hz can't we increase the gain on the plate amplifier and partially compensate for the baffle step. I know this is not exact but it may help besides it is reversible.
Baffle step would mean a about a 6db loss starting from about 1khz and flattening out at 100hz.
However I would also consider floor bounce. 100-120Hz is about the frequency Alison reported.
I would do the first 2 as these are reversible (don't glue the wood / rubber until you are sure).
Is there a gradual drop from 1k to 100hz as expected from baffle step, or is it just a narrow dip around 100hz?
Harbeth UK - High quality loudspeakers made in England
Oh and I found one more tweeter with a 110mm faceplate that may be able to be XOed at 2khz (but it is a ribbon).
http://www.lcy.com.hk/buy.php?kind=17
http://meniscusaudio.com/images/LCY-110.pdf
There is a LCY 130 too discussed here http://www.audiocircle.com/index.php?topic=107072.0
Last edited:
The mid and tweeter are reversed in polarity relative to the 8" drivers. This yields a in-phase acoustic response. You can reverse the polarity and measure to see the nulls at both crossover points.
99% chance it's a room mode. Placement can reduce the null, but not eliminate it. Fire up REW and run the room sim, define size, placement, etc. Do not try to eq it out, room modes cannot be overpowered.
My room creates 5 primary nulls, 3 belong to the room itself, the other two are the attatched kitchen and hallway acting as bass traps. One of these overlaps a primary and the other is close to another spredding it out a bit. The in room bass response looks like three mountain peaks from 25-140Hz and is fairly typical looking for an american living room.
My room creates 5 primary nulls, 3 belong to the room itself, the other two are the attatched kitchen and hallway acting as bass traps. One of these overlaps a primary and the other is close to another spredding it out a bit. The in room bass response looks like three mountain peaks from 25-140Hz and is fairly typical looking for an american living room.
My question to the OP earlier about looking at the subwoofer without filters was driven by thoughts along these lines. It seems likely the original builder did something to at least partially address the issues we are seeing.
I would expect it to be shifted a bit lower but sitting mostly in the woofer operating range. Padding the midrange and tweeter down a further 6 dB gives us something like an 82 dB sensitivity which does not look particularly sensible for a largish speaker of this type which ought to have a sensitivity more like 90 dB and be able to supply musical peaks in a medium sized room with a normal stereo amplifier rather than one rated in kWs.
I have my doubts whether there will be a positive benefit in making the volumes smaller because the drivers are high passed. The optimum cabinet size follows from the TS parameters, the amount of stuffing used and what the user believes is the optimum Q which typically varies from 0.5 to about 0.8 with 0.7 being the most commonly chosen. You can see all this for yourself by entering the TS parameters into a program like WinISD or equivalent.
Your in room plot shows that both the tweeter and midrange need to be padded down. This will lower the sensitivity of the speaker.
The objective is fine but I think you need to measure the sub and tell us what equalisation you have available.
Now I think everyone is in full agreement on this.
Bass is omni-directional which is what is causing your problem. Tilting the speakers is going to make no difference to the bass. However, measuring on the axis half way between the tweeter and midrange or even on the tweeter axis mmight bring a noticeable change to the tweeter response. When sitting down what lines up with your ear level on the speaker?
The room is 12x14' and 18-19' high with sloping roof. It opens into a smallish dining room with normal ceiling. Flooring is wood laminate on a concrete subfloor. I feel the room is playing its part because another speaker (wharfedale i mentioned earlier) seems to show similar issue but to a much lower extent when played in this room.
There is no doubt in my mind about existence of BSC issues. I can actually sense some of the missing bass if i crouch behind the speaker against its side walls and the rear room wall.
Replacing the midrange and tweeter and reworking both the crossovers makes this a completely new speaker design.
I was hoping to reuse the drivers (except tweeter).
Also no doubt in mind that something using such high end drivers and very good quality XO components should not be sounding as bad as these.
I am a bit lost right now, i want these to sound their absolute best, but i'm clear about the massive amount of work, time and money that may need to go in to achieve that. The speaker is fully dismantled and sitting in the room, with a very angry wife fuming about it 🙂
Agree that padding the mid and tweeter down bring the sensitivity down further, which is already fairly low for a speaker of this size.
Sounding their best, with modest effort and possibly saving money would be to sell the very expensive stereo amplifiers and DACs and replace then with sensibly priced multichannel hardware and DSP plus plus changing the tweeter.
If you want a passive solution I would suggest looking at changing the tweeter and all of the crossover. Connecting the woofer in parallel will give the bass boost required and return the sensitivity to "normal" but the midrange will be down on level. Exploring 5th elements idea of using the woofers and tweeter to help out may well be viable but it needs to be worked through for your baffle arrangement which will involve significant effort by someone. It may prove problematic but then again it may not.
I don't like the idea of padding down an inefficient midrange in order to provide baffle step correction. I think you would be better off simply leaving the woofers and midrange as they are and using DSP on the whole input signal. I think the tweeter needs changing whatever the option chosen to address the bass problem. This will mean a new mid/tweeter crossover but you might be able to use some of the existing parts.
- Status
- Not open for further replies.