What should we do about the values?
ideas? suggestions?
and please bear with the quick hand drawn schematic
thanks
Joji
ideas? suggestions?
and please bear with the quick hand drawn schematic
thanks
Joji
The MID and HIGH are eerily similar to what Dissi posted here, but they appeared to be wired out of phase in the speaker.
The circuit can be simulated and the values adjusted to match the measurements. An impedance plot would be helpful but is probably not essential and we already have an "intended" impedance plot from earlier.
Having established the circuit then tweaks can be examined to see how much they can move things in the right direction. If the tweaks cannot move things enough then a new circuit with the woofers in parallel along the lines of the one posted by 5th element could be simulated and developed. However, there will be reasons the designer opted to do things the way they are and not the more conventional/obvious way. If they were good reasons then expect a struggle.
Andy, Do you think the phase on the mid and tweeter is right and as likely intended?
Your in room response of the mid/high crossover region suggests the mid/high crossover is probably functioning as intended. If the mid or tweeter was incorrectly wired the drivers would cancel instead of sum in this region and there would be a cancellation dip like the ones you can see caused by the room at low frequencies.
At the listening position each drivers has an effective acoustic source. On a flat baffle the distance away of this acoustic source is not the same for each driver. This means that in the crossover region the two relevant drivers would not be moving together in phase and would not sum correctly unless something is done to adjust the relative phase. In an active crossover one of the signals would be delayed relative to the other to bring them into phase (5th element did this in his simulation). An alternative would be to step the baffle and physically move the tweeter and and midrange back to align with the woofers. But what is usually done is to fiddle with the filters which not only introduce changes in gain but also changes in phase. It is a bit like solving a puzzle and the challenge of getting a good behaviour can be satisfying.
On the topic of the subwoofer I am not sure that particular Peerless model was optimised for use in a sealed cabinet but the Tymphany website has been down for a day or two and I have been unable to check. Was the number read off the driver or obtained from somewhere else?
Jojip,
In the PDFs we sent I see a wide suck out (2k-8k) as deep as 3db in the mid-hf region (page 1 and 2). While I suspect this might be because the Mid's LP is around 2k while the tweeter's HP is around 4k and the crossover as Andy said "is performing as designed" it might also be due to a reverse in polarity although such dips are typically deeper and less wide.
Since it is not difficult to check that - reverse the polarity on the tweeter connections - why not try it. It costs nothing.
Exactly as Navin says, switching polarity costs nothing.
Joji, try recording the W15 response and then neo response
without moving mic and post the plot so we can see the overlap
of both of the graphs. Leave the W22's off. For instance, reference
mic point between the mid and tweeter, off to the right 40 cm,
away from baffle 1 m.
p.s. Solen caps have dimensions defined. Using the data provided
you could estimate unknown capacitance by measuring their
casings. The error would be +-10%, so the Solen says.
Joji, try recording the W15 response and then neo response
without moving mic and post the plot so we can see the overlap
of both of the graphs. Leave the W22's off. For instance, reference
mic point between the mid and tweeter, off to the right 40 cm,
away from baffle 1 m.
p.s. Solen caps have dimensions defined. Using the data provided
you could estimate unknown capacitance by measuring their
casings. The error would be +-10%, so the Solen says.
Last edited:
Andy/Navin,
I was actually referring to the mid/tweeter XO being out of phase with respect to the woofer XO. That also seems to be the case in the XO layout that Dissi had predicted.
I was wondering if that can explain the large bass dip around 100 Hz.
but both speakers seemed to have this, so it was likely as designed and not an error.
I was actually referring to the mid/tweeter XO being out of phase with respect to the woofer XO. That also seems to be the case in the XO layout that Dissi had predicted.
I was wondering if that can explain the large bass dip around 100 Hz.
but both speakers seemed to have this, so it was likely as designed and not an error.
Lojzek,
Most of these measurements are in the pdf i attached. I havent made any off axis. All on axi at 1 m at midrange height.
Good idea on the solen cap. Thanks.
Might get an LCR meter to do it the right way.
thanks
Joji
Matt,
Do you have any observations on the measurements and crossover diagrams?
Do these explain the issues we see.
Can you comment on the optimal direction to be taken at this time?
thanks
Joji
I am sorry that I haven't responded till now I've been quite busy with other things! The first thing that sticks out is, as everyone else has said, the lack of energy within the bass the lower mid region. Clearly something is not right with this loudspeaker and no wonder it sounds pretty terrible.
The mid to tweeter xover has taken a back seat imo as there are bigger problems that need solving. The first thing about the mid xover is the huge peak in the third order distortion at about 2.7kHz. This is one of the main reasons as to why you need to crossover around 2kHz, to eliminate this peak.
With the tweeters distortion I'd say that you weren't driving the speakers hard enough for the distortion to really be an issue. The tweeter is very sensitive, which is why I've said that it'd work reasonably well in a small two way with a lower xover. The test drive level on the audiogurman website is done at 2.83v, which is 93dB, this is loud, especially for a small two way, but in a three way, where you want to be able to hit 110dB with low distortion...the Neo3CD is gonna crumble. It'd be nice to see a distortion measurement of the midrange and tweeter done at a much higher drive level. This will be loud and when I do high volume sine wave testing I have to wear ear defenders as the sine waves are very impinging on the hearing.
Ignoring the distortion and the fact that the tweeter wont be happy at very high levels, the overall frequency response of the mid-tweeter section is fair. Yes there's a general lack of energy between 3 and 6kHz, but this wont make the speaker sound horrible, just a bit muted and possibly a little dull. It's probably a good thing next to the lack of bass and midrange energy.
This speaker really needs a completely new crossover and with it a new tweeter so that the crossover can be designed to fit the drivers appropriately.
The mid to tweeter xover has taken a back seat imo as there are bigger problems that need solving. The first thing about the mid xover is the huge peak in the third order distortion at about 2.7kHz. This is one of the main reasons as to why you need to crossover around 2kHz, to eliminate this peak.
With the tweeters distortion I'd say that you weren't driving the speakers hard enough for the distortion to really be an issue. The tweeter is very sensitive, which is why I've said that it'd work reasonably well in a small two way with a lower xover. The test drive level on the audiogurman website is done at 2.83v, which is 93dB, this is loud, especially for a small two way, but in a three way, where you want to be able to hit 110dB with low distortion...the Neo3CD is gonna crumble. It'd be nice to see a distortion measurement of the midrange and tweeter done at a much higher drive level. This will be loud and when I do high volume sine wave testing I have to wear ear defenders as the sine waves are very impinging on the hearing.
Ignoring the distortion and the fact that the tweeter wont be happy at very high levels, the overall frequency response of the mid-tweeter section is fair. Yes there's a general lack of energy between 3 and 6kHz, but this wont make the speaker sound horrible, just a bit muted and possibly a little dull. It's probably a good thing next to the lack of bass and midrange energy.
This speaker really needs a completely new crossover and with it a new tweeter so that the crossover can be designed to fit the drivers appropriately.
Thanks Matt.
I am willing to undertake a complete redesign. If you can guide me i am happy to follow along. In the end, i will very likely have a pair of wonderful sounding speakers doing justice to the drivers and beautiful cabinet and would have learnt something i have wanted to do for quite some time. Would also be several dollar poorer 🙂.
On a serious note, i will need to make some investments to get this project started. An LCR meter to dismantle the crossovers, measure and salvage what i can for the new XO. Also please advice if i need an impedance measurement unit like the Dayton WT3.
Which tweeter should we settle on from the options explored earlier?
Can you list out some steps i need to get going on ?
Would you be interested in taking the Neo CDs for your next bookshelf project as a token for your guidance?
Your help and the community's help is much appreciated.
thanks
Joji
Floor bounce? Roy Allison wrote some papers on this.
125 HZ Suckout
Speaker-room suckout and other tidbits. - Free Online Library
Big dip at 100hz - why? - AudioKarma.org Home Audio Stereo Discussion Forums
Cant you borrow one. I don't own one, but have access to one from a friend's workshop.
Matt, the lethargic midbass would have little relation to tweeters. Jojip and I have the same midbass drivers (dual 8" Excel) and midrange drivers (5" Excel). And my speakers rock. I live in an apartment and have rocked them so hard (when I was breaking them in and testing their limits in the initial days) that the neighbour below came up and requested me to turn the bass down as it was vibrating their ceiling and shaking their chandelier. The music I usually listen to for testing is Pink Floyd, Led Zeppelin, Jimi Hendrix, The Who, Black Sabbath, etc as I know this music well.
Another point to note, that I have not noticed the NeoX ribbon to "crumble" even when we put on less overdriven sounds like Miles Davis, Dizzy G, etc..but yes it does looks like the speaker needs a ground up re-design of the XO. At best JoJip can salvage some of the parts of the existing XO.
Jojip,
Matt suggested the SEAS dome. I have used the Crescendo version of his dome in a 2.5 way. It's nice and it rocks. I would consider the Millennium version for your use. Outside of this the ET338 from Morel is another option and the NeoX only because I have not had any serious issue with it.
Joji,
As I mentioned to you in my email I just ran across this thread and decided to respond here. Some of the posts have useful information while others fall short of giving good advice.
A little background on the design. I created the RC4 about ten years ago for customers who wanted a subwoofer but didn't have room for outboard cabinets. The sub is integral to the design and the dual 8" section was never intended to be run full-range. The high pass filter on the 8" drivers gives them more headroom by limiting excursion in the lower frequencies so that owners could enjoy the full benefits of a 4-way. The sealed box allows them to make a smooth transition to the subwoofer.
The 8" drivers are in series which provides the best level match to the midrange and results in a reasonable impedance across the entire range of the speaker. The extra surface area also adds impact in the range that they cover. Could they be ported and crossed lower? Sure, but the cabinet volume increases beyond what was practical and low bass isn't really the best attribute of the W22.
The plate amp I supplied for the kit was intended to be mounted on the rear of the cabinet. The customer who built this kit added the outboard Dayton amps since the curved rear didn't allow for a plate amp. The Dayton amp will suffice; however, I don't have the specific settings for that unit. You can use Arta to calibrate the amp as needed. The sealed sub uses equalization from the amp to extend the bass response.
I want to address the speculation on the mid /tweeter combination and crossover for them. The feedback on the sound has been positive from owners and I have no problem with the crossover point or slope. Given that your room has a high ceiling and wood floor it might present some problems, especially if you have mostly hard / reflective surfaces (as many homes do). The resistor values I gave you will help reduce the level of the tweeter and midrange as desired. Also be sure to have good reliable source material when changing the voicing so that you're not simply adjusting for problems in the recording.
I originally made a mistake when I gave you the 60hz figure. The crossover point to the subwoofer was 90hz and the 8" section without the high pass filter was -3db @ 60hz (maybe even a little lower than that - it's been ten years!). If the subwoofer isn't set up correctly there could be a dip in the response where it crosses to the 8" drivers. That can easily be tested.
Now for the bad news. Sometimes kit customers make changes without consenting me. In this case mistakes were made that could've been avoided. The overly thick baffle is going to add some coloration so chamfering the openings will help.
The estimated volumes you gave for each section are not correct. All three are oversized, especially the midrange and 8" sections. The subwoofer section could be changed with some added bracing or blocks to decrease the volume. The same is true for the midrange. The 8" section is a little different in that the volume is really closer to what a ported cabinet would be; however, the cabinet layout prohibits adding a port. The volume could be reduced in the same method as above to keep it as a sealed enclosure.
The volume issues with the midrange and tweeter could be part of the audible issues that you've noticed. Without being there to listen it's difficult to say exactly what you should do; however, if you make the changes I've suggested it will point you in the right direction. Work in small increments so that you can listen to the results as you make changes.
As far as going active I have no problem with that (I prefer DSP crossovers for my personal speakers); however, you'll need to be prepared to take the time to learn about making good measurements. After that comes knowing how to optimize crossover slopes / points and making changes to optimize the voicing for your listening preferences. Replacing the passive crossovers is really up to you but I think if you first make the changes I've suggested you may find that will take care of the issues that you have. The volumes need to be corrected regardless of what direction you choose. If you find changing the resistors doesn't alter the balance enough then you might consider a different tweeter and / or crossover change as a last resort. Any modifications really need to be made with a system approach that includes the subwoofers in the equation.
As I mentioned to you in my email I just ran across this thread and decided to respond here. Some of the posts have useful information while others fall short of giving good advice.
A little background on the design. I created the RC4 about ten years ago for customers who wanted a subwoofer but didn't have room for outboard cabinets. The sub is integral to the design and the dual 8" section was never intended to be run full-range. The high pass filter on the 8" drivers gives them more headroom by limiting excursion in the lower frequencies so that owners could enjoy the full benefits of a 4-way. The sealed box allows them to make a smooth transition to the subwoofer.
The 8" drivers are in series which provides the best level match to the midrange and results in a reasonable impedance across the entire range of the speaker. The extra surface area also adds impact in the range that they cover. Could they be ported and crossed lower? Sure, but the cabinet volume increases beyond what was practical and low bass isn't really the best attribute of the W22.
The plate amp I supplied for the kit was intended to be mounted on the rear of the cabinet. The customer who built this kit added the outboard Dayton amps since the curved rear didn't allow for a plate amp. The Dayton amp will suffice; however, I don't have the specific settings for that unit. You can use Arta to calibrate the amp as needed. The sealed sub uses equalization from the amp to extend the bass response.
I want to address the speculation on the mid /tweeter combination and crossover for them. The feedback on the sound has been positive from owners and I have no problem with the crossover point or slope. Given that your room has a high ceiling and wood floor it might present some problems, especially if you have mostly hard / reflective surfaces (as many homes do). The resistor values I gave you will help reduce the level of the tweeter and midrange as desired. Also be sure to have good reliable source material when changing the voicing so that you're not simply adjusting for problems in the recording.
I originally made a mistake when I gave you the 60hz figure. The crossover point to the subwoofer was 90hz and the 8" section without the high pass filter was -3db @ 60hz (maybe even a little lower than that - it's been ten years!). If the subwoofer isn't set up correctly there could be a dip in the response where it crosses to the 8" drivers. That can easily be tested.
Now for the bad news. Sometimes kit customers make changes without consenting me. In this case mistakes were made that could've been avoided. The overly thick baffle is going to add some coloration so chamfering the openings will help.
The estimated volumes you gave for each section are not correct. All three are oversized, especially the midrange and 8" sections. The subwoofer section could be changed with some added bracing or blocks to decrease the volume. The same is true for the midrange. The 8" section is a little different in that the volume is really closer to what a ported cabinet would be; however, the cabinet layout prohibits adding a port. The volume could be reduced in the same method as above to keep it as a sealed enclosure.
The volume issues with the midrange and tweeter could be part of the audible issues that you've noticed. Without being there to listen it's difficult to say exactly what you should do; however, if you make the changes I've suggested it will point you in the right direction. Work in small increments so that you can listen to the results as you make changes.
As far as going active I have no problem with that (I prefer DSP crossovers for my personal speakers); however, you'll need to be prepared to take the time to learn about making good measurements. After that comes knowing how to optimize crossover slopes / points and making changes to optimize the voicing for your listening preferences. Replacing the passive crossovers is really up to you but I think if you first make the changes I've suggested you may find that will take care of the issues that you have. The volumes need to be corrected regardless of what direction you choose. If you find changing the resistors doesn't alter the balance enough then you might consider a different tweeter and / or crossover change as a last resort. Any modifications really need to be made with a system approach that includes the subwoofers in the equation.
Last edited:
A few extra thoughts. I believe I used the center of the tweeter as the reference point for the measurement so you can check for the correct polarity at that position (the right polarity will have a null / dip in the response when the connections are reversed). The builder also added the facets on the front baffle which will change the response a little so some off-axis measurements will be helpful.
Last edited:
Thank You much for taking the time to post this detailed reply.
There are a lot of useful pointers here from the designer of the product and i am going to try them out.
For your benefit i am summarizing the issues i perceived, your insight into this would be most helpful
1. Lack of bass energy from the 8". They do balance out a bit better with the subs on.
2. poor imaging. The speakers just dont disappear in the music.
My reference experience here is B&W, high end Definitive and even Martin Logans
3. Mid, particularly vocals are the best i have ever heard. But i do not seem to enjoy the ribbon sound.
4. On complex , fast music tracks all resolution is lost, the sound is strained and muddy and not pleasant to hear. I cant seem to hear any form of rock music on this and find the experience enjoyable.
My reference experience is a pair of old Missions floor standers several years ago which played rock like nothing i have managed to hear since, within what is accessible to me.
from lots of information provided here by the community i tend to agree that the mid/tweeter pairing seem to not sound the best to my ears.
The Bass issue and mid/high level mismatch i attribute a lot to the very poor listening environment i am able to provide the speakers.
I completely agree that the baffle and mounting for the drivers are done very poorly in this implementation.
I love active DSP approach and is an option i am toying with except for the added cost and issues with return of some good gear i purchased to enjoy pure 2 ch stereo on these speakers.
Lots of learning is involved in the active option, but i think i am up for that.
Passive rework/tweaking I think is more involved, but seems more feasible at this point
Last edited:
- Status
- Not open for further replies.