Yes, err looser (subjectively more but less in slam) bass. Contradicts.
P.S. I see a weird in room FR chart in that link you gave about it in post#1, down in its page, with some strong in room bass cancellation. I hope its not about a difficult geometry for many rooms. I don't know if its for the 2 way or the 2.5 way.
EDIT. It says on the picture's tab legend. 2.5m in room for the 2.5 way.
Insignificant difference with the DCR variation and may be 0.2dB.
You can check to see if it's the port tuning by:
1. Change the port to half the length and listen.
2. If you can't do that, close the lower chamber by placing a piece of MDF over it and stick it on with Blu Tack or similar so there is no air leaks. This reduces the volume by around 10 litres and in turn raises the tuning frequency by around 4Hz.
Both these should lift the bottom end with #1 being more dramatic.
Ensure there is a clear path from the woofer to the port opening with no damping material obstructing this path.
I went back through my records and was tuning this size box to between 37Hz-39Hz according to the software. In actual fact the tuning frequency is about 4Hz lower as I've found I usually have to use a port that's around 70% of the modelled length to achieve the target bass performance. I've checked the Fb with test signals. This effect has been documented on the web and I've found it a lot in old Vifa driver datasheets.
You could also try removing some damping as too much can increase the Vb value so the box thinks and works as a larger size. This has an effect of lowering the Fb.
Last edited:
There is a clear path from the woofer to the port - in fact, i can reach my arm in through the port hole when the port is removed and touch the back of the woofer.
Cutting the port length in half changes the balance int he bass; there is a slight rise in the upper bass at the expense perhaps of some accuracy overall (it seems boomier). There is slightly more bass overall. However, the amount of low bass is still lower than it seems like it should be. There is much more bass when I remove the port entirely, but it is boomy and has a 'one note' sound.
Interestingly, bass instruments sound fairly good by themselves, when turned up; they are just buried too far in the mix when other instruments or voices are playing.
Doing a slow sweep with a test program, there is very little bass (relative to the main signal) below 60 Hz. There is dramatic rise around 60-70, a couple of peaks/dips where the bass gets louder-quieter-louder-quieter, and then it straightens out somewhere above 100 Hz.
I had a fair bit of stuffing in the bottom of the cabinet, and I removed half of it without much effect. I ended up using far less than Madisound sent me; I would be hard pressed to pack all of the stuffing into so small a space. How do i know how densely to stuff the bottom of the cabinet?
Cutting the port length in half changes the balance int he bass; there is a slight rise in the upper bass at the expense perhaps of some accuracy overall (it seems boomier). There is slightly more bass overall. However, the amount of low bass is still lower than it seems like it should be. There is much more bass when I remove the port entirely, but it is boomy and has a 'one note' sound.
Interestingly, bass instruments sound fairly good by themselves, when turned up; they are just buried too far in the mix when other instruments or voices are playing.
Doing a slow sweep with a test program, there is very little bass (relative to the main signal) below 60 Hz. There is dramatic rise around 60-70, a couple of peaks/dips where the bass gets louder-quieter-louder-quieter, and then it straightens out somewhere above 100 Hz.
I had a fair bit of stuffing in the bottom of the cabinet, and I removed half of it without much effect. I ended up using far less than Madisound sent me; I would be hard pressed to pack all of the stuffing into so small a space. How do i know how densely to stuff the bottom of the cabinet?
I've just completed a pair of ZRT's too. I had the same experience - lack of bass. This is what I think has happened:
It seems that the 2008 version of the tweeter is around 3db more sensitive than the previous version, and it doesn't seem to me that the crossover has been adjusted accordingly. The bass is there - but the tweeter sound is masking it, because it's too loud (and harsh). With the default XO, I found I needed 3-4 db of negative EQ above the crossing frequency to sound anything near reasonable. Of course, this approach (using EQ in itunes) causes problems in the crossover band, so I adjusted the level resistor in the tweeter XO circuit (R8 in the plans, from memory). At the moment I'm using an additional 3.3 ohm (in series), although I think my previous guess was better (an additional 2.8ohm in series).
Until making these changes the speakers were seriously poor and I thought I was taking crazy pills. I was telling my housemates that they should be awesome, but the cheapo yamaha 5.1 system definitely sounded better... Making the change above made a massive difference, and I think I'm getting close to an optimised setup. I don't have the facilities to measure freq response, so this has all be done by ear, best I can. Good luck, let us know if this makes sense!
It seems that the 2008 version of the tweeter is around 3db more sensitive than the previous version, and it doesn't seem to me that the crossover has been adjusted accordingly. The bass is there - but the tweeter sound is masking it, because it's too loud (and harsh). With the default XO, I found I needed 3-4 db of negative EQ above the crossing frequency to sound anything near reasonable. Of course, this approach (using EQ in itunes) causes problems in the crossover band, so I adjusted the level resistor in the tweeter XO circuit (R8 in the plans, from memory). At the moment I'm using an additional 3.3 ohm (in series), although I think my previous guess was better (an additional 2.8ohm in series).
Until making these changes the speakers were seriously poor and I thought I was taking crazy pills. I was telling my housemates that they should be awesome, but the cheapo yamaha 5.1 system definitely sounded better... Making the change above made a massive difference, and I think I'm getting close to an optimised setup. I don't have the facilities to measure freq response, so this has all be done by ear, best I can. Good luck, let us know if this makes sense!
Hi tomewing,
How do you know the the tweeter is 3db too sensitive? Is it a question of it sounding best to you when its 3db down, or is that in comparison with a particular earlier version of the tweeter?
Zaph writes on his website:
I ordered the kit from Madisound and had assumed that if there was a tweeter adjustment someone would made an adjustment per Zaph's comments.
I did some rudementary preliminary measurements using a laptop and its builtin mic. The measurements showed minimal low bass (probably both because there was roll off in the bass and also because the mic is not accurate at those frequencies). Above 200 Hz, the design (in room response is what is being measured, BTW) was fairly flat. A little bit flatter than my Odins, which had a bit of a dip around the crossover frequency.
The Odins do sound smoother and less trebly; perhaps because of the dip; or perhaps as you say, there is too much treble on the ZRT. Time for further testing when I can bring a proper mic into play!
How do you know the the tweeter is 3db too sensitive? Is it a question of it sounding best to you when its 3db down, or is that in comparison with a particular earlier version of the tweeter?
Zaph writes on his website:
I ordered the kit from Madisound and had assumed that if there was a tweeter adjustment someone would made an adjustment per Zaph's comments.
I did some rudementary preliminary measurements using a laptop and its builtin mic. The measurements showed minimal low bass (probably both because there was roll off in the bass and also because the mic is not accurate at those frequencies). Above 200 Hz, the design (in room response is what is being measured, BTW) was fairly flat. A little bit flatter than my Odins, which had a bit of a dip around the crossover frequency.
The Odins do sound smoother and less trebly; perhaps because of the dip; or perhaps as you say, there is too much treble on the ZRT. Time for further testing when I can bring a proper mic into play!
Last edited:
Monroe,
It's definitely a case of sounding better, to my ears, around 3db down - I don't know if it is a matter of taste, the nature of this room, or if there is genuinely some variability in the sensitivity of the drivers. It's difficult to describe, but the drivers didn't seem to 'integrate' well untill the changes were made.
I did read about another design using this tweeter and in that design they had modifications for the updated 2008 driver. That's where I got that figure from, although as you pointed out it seems like this has already been compensated for in the design.
What is the value of R8 on your crossover, out of interest? On the main ZRT page it says 6 ohms, but in the tweeter tweaking image linked down the page it says 7 ohms is the default...
I think a mic and some proper measurements are definitely the way to go. I look forward to those results. I'm going to eventually do the same here and will let you know the results (will take a while though).
Regards,
Tom
It's definitely a case of sounding better, to my ears, around 3db down - I don't know if it is a matter of taste, the nature of this room, or if there is genuinely some variability in the sensitivity of the drivers. It's difficult to describe, but the drivers didn't seem to 'integrate' well untill the changes were made.
I did read about another design using this tweeter and in that design they had modifications for the updated 2008 driver. That's where I got that figure from, although as you pointed out it seems like this has already been compensated for in the design.
What is the value of R8 on your crossover, out of interest? On the main ZRT page it says 6 ohms, but in the tweeter tweaking image linked down the page it says 7 ohms is the default...
I think a mic and some proper measurements are definitely the way to go. I look forward to those results. I'm going to eventually do the same here and will let you know the results (will take a while though).
Regards,
Tom
tomewing: I'm not sure what the R8 in my crossover is, and would have to remove the baffles to find out. I will let you know if i can figure it out.
rabbitz: thanks for the suggestion! I have posted a question here:
Madisound Audio Forum
rabbitz: thanks for the suggestion! I have posted a question here:
Madisound Audio Forum
Update:
Per the helpful suggestions made earlier on the thread and on the Madisound thread, I have made a couple of changes:
1.) I have increased the stuffing in the bottom of the line (very dense stuffing, as dense as possible) and have stuffed the penultimate section as heavily as possible without obstructing a path from the woofer to the port. I went back to the default port length. This seems to have improved the bass somewhat.
2.) Per tomewings's suggestion, I tried adding some series resistance to the tweeter. I started with 2 ohms, which was too much: it made the speakers sound a bit lifeless. Right now I am using an extra 0.66 ohms (i think) by adding a 1 and 2 ohm resistor in parallel. The extra resistance took a bit of the edge off the tweeters, although i have the feeling that the sounds is a touch muffled at times.
I have added resistance by adding extra resistors rather than chaning the R8 in the crossoever. Am i likely paying some kind of penalty for adding those extra resistors into the circuit, or is resistance resistance?
Per the helpful suggestions made earlier on the thread and on the Madisound thread, I have made a couple of changes:
1.) I have increased the stuffing in the bottom of the line (very dense stuffing, as dense as possible) and have stuffed the penultimate section as heavily as possible without obstructing a path from the woofer to the port. I went back to the default port length. This seems to have improved the bass somewhat.
2.) Per tomewings's suggestion, I tried adding some series resistance to the tweeter. I started with 2 ohms, which was too much: it made the speakers sound a bit lifeless. Right now I am using an extra 0.66 ohms (i think) by adding a 1 and 2 ohm resistor in parallel. The extra resistance took a bit of the edge off the tweeters, although i have the feeling that the sounds is a touch muffled at times.
I have added resistance by adding extra resistors rather than chaning the R8 in the crossoever. Am i likely paying some kind of penalty for adding those extra resistors into the circuit, or is resistance resistance?
Monroe,
Interesting update - by chance I've come to the same conclusion on point 1 - I've stuffed the lower chamber (below the port) to the limits of compression (while still getting my access panel on) and it had the same impact, a marginal improvement in the fullness and quality of the bass.
I've have tried something else interesting though, which may be of some interest. I have made a set of 2 way, 24db/octave active crossovers, with the same XO frequency as the Zaph passive crossovers. These are connected to 2 stereo class D amplifiers, then directly to the drivers. The level of each driver can be trimmed with this setup, which means a long process of adjustment to get set up, but plenty of control.
I'm very happy with this configuration. It still needs some tweaking, but to my ears it made a big difference. It has been a lot of work to set up. I need to listen more and upgrade a few of the components in the active XO and the amps before I make any final decisions, and remove the passive XOs from the speaker cabinets!
Tom
Interesting update - by chance I've come to the same conclusion on point 1 - I've stuffed the lower chamber (below the port) to the limits of compression (while still getting my access panel on) and it had the same impact, a marginal improvement in the fullness and quality of the bass.
I've have tried something else interesting though, which may be of some interest. I have made a set of 2 way, 24db/octave active crossovers, with the same XO frequency as the Zaph passive crossovers. These are connected to 2 stereo class D amplifiers, then directly to the drivers. The level of each driver can be trimmed with this setup, which means a long process of adjustment to get set up, but plenty of control.
I'm very happy with this configuration. It still needs some tweaking, but to my ears it made a big difference. It has been a lot of work to set up. I need to listen more and upgrade a few of the components in the active XO and the amps before I make any final decisions, and remove the passive XOs from the speaker cabinets!
Tom
I have just completed my ZRT's and the bass is certainly a problem. I have a strong resonance at 50Hz which dominates the sound. With the port open, the impedance drops nicely around 30Hz, with a peak at 20Hz and a much larger one at 50Hz. Blocking the port removes the 20Hz peak and the 50Hz one increases greatly. I therefore assume that the port tuning is OK but the driver resonance is too high. I know the Fs will drop with use, but probably not by that much.
My question is similar to Monroe's: could this be due to the shelf ventilation? I have three shelves in place and each has four 3" holes plus one 1" hole in it. Looking at the pictures of the Madisound cabinet, I would say that my clear area is no less than theirs, but circles rather than triangles.
At the moment I am running the speaker with stuffing in all internal sections, which reduces the resonance to an acceptable level. Conversion to a sealed system is on the cards, but I will be disappointed if that is the only remedy.
My question is similar to Monroe's: could this be due to the shelf ventilation? I have three shelves in place and each has four 3" holes plus one 1" hole in it. Looking at the pictures of the Madisound cabinet, I would say that my clear area is no less than theirs, but circles rather than triangles.
At the moment I am running the speaker with stuffing in all internal sections, which reduces the resonance to an acceptable level. Conversion to a sealed system is on the cards, but I will be disappointed if that is the only remedy.
Last edited:
My experience with this kit is difficult to account for. The bass started out weak and the sound in general was not what I had expected or hoped for.
Although increasing the stuffing and taking the tweeter level down slightly improved the overall sound, I do not think that those changes, alone or togeather, fully account for the improvement in sound over the past several weeks.
Is is possible that the speakers are breaking in? I am skeptical about the notion that a speaker could significantly break in over 50 hours of playing time, but I am at a loss to explain what appears to be a positive change to the overall sound of the speakers.
Although increasing the stuffing and taking the tweeter level down slightly improved the overall sound, I do not think that those changes, alone or togeather, fully account for the improvement in sound over the past several weeks.
Is is possible that the speakers are breaking in? I am skeptical about the notion that a speaker could significantly break in over 50 hours of playing time, but I am at a loss to explain what appears to be a positive change to the overall sound of the speakers.
Last edited:
Monroe,
I too found the tweeter level too high. For some reason Zaph's schematic shows the resistor as 6R although the text indicates 7R is the default. I was previously using a sealed and damar-treated version of the Proac 2.5 clone, which at the moment I still much prefer. I rather think the ZRT might need more development, as it measures reasonably well but does not sound quite right to me. I would like to take a few turns off the 2.7mH coil as suggested, but Madisound understandably glue the components in place.
I have not seen much discussion of the ZRT, but to me the Revelator has what I can only describe as a papery sound. Of course, it could be that the speaker works well with other systems and rooms, but at the moment I am siding with those who say this design is nothing special, despite the high cost!
I too found the tweeter level too high. For some reason Zaph's schematic shows the resistor as 6R although the text indicates 7R is the default. I was previously using a sealed and damar-treated version of the Proac 2.5 clone, which at the moment I still much prefer. I rather think the ZRT might need more development, as it measures reasonably well but does not sound quite right to me. I would like to take a few turns off the 2.7mH coil as suggested, but Madisound understandably glue the components in place.
I have not seen much discussion of the ZRT, but to me the Revelator has what I can only describe as a papery sound. Of course, it could be that the speaker works well with other systems and rooms, but at the moment I am siding with those who say this design is nothing special, despite the high cost!
I have finally had an opportunity to make some (crude) measurements. I used an uncalibrated Rode NT1 mic, which should be reasonably flat.
This is an in-room measurements, sum of both speakers, taken from the listening position.
I'm not familiar enough with measurements to know what is significant. It seems like the big drop at 300hz should be a concern; what could be causing that?
This is an in-room measurements, sum of both speakers, taken from the listening position.
I'm not familiar enough with measurements to know what is significant. It seems like the big drop at 300hz should be a concern; what could be causing that?
Attachments
Alternative xo
For those interested, here is an alternative crossover for the D3004-660000 and 18W8531G00. It's an AR series and was modelled when I was looking for a new tweeter for my mains but ended up using the Peerless 810921. It hasn't been tried but should be a good start and easy to tweak. I have used the 18W8531G00 with several tweeters from Scan Speak, Vifa, Peerless and SEAS so have an understanding of combining tweeters with this mid woofer. I designed a similar crossover on the SR71 for a local builder with good results.
This crossover only uses 2 caps, 2 inductors, 2 resistors and optionally another resistor and these are the values for the schematic posted.
L1 = 0.12mH (DCR<0R08)
R1 = 4R7 (10W)
C1 = 20uF (MKP)
L2 = 1.8mH (DCR<0R34)
C2 = 8.2uF (MKP)
R2 = 8R2 (10W)
R is for tweeter tweaking if required.
For tweaking:
To reduce tweeter level add R = 22R or 15R or 12R for taste. If not enough increase R1 to 5R6 and add R if required.
To increase mids reduce C2 in approx 2uF increments. If mids are still too low then decrease L2 to 1.5mH (with possibly decreasing L1 to 0.1mH).
To decrease mids increase C2 in approx 2uF increments.
Inductors should be 14AWG for a low DCR (especialy L1) and caps should be a good quality such as a Mundorf M-Cap (the white cheaper ones). R1 should not go below 3R3 with any tweaking as it lowers the impedance too much. The LCR from Zaph's crossover can be used across the tweeter if desired or warranted but can't be used if resistor R is in use.
As it's a series crossover all points are joined as shown and L2 can go on the leg between R2 and C1 if desired. The tweeter is -ve polarity as shown.
For small outlay it could be worth a try and may help with the sonic issues. I still think the enclosure is too large for this driver but that's another story (see Troel's site for comments on the 18W8531G00).
For those interested, here is an alternative crossover for the D3004-660000 and 18W8531G00. It's an AR series and was modelled when I was looking for a new tweeter for my mains but ended up using the Peerless 810921. It hasn't been tried but should be a good start and easy to tweak. I have used the 18W8531G00 with several tweeters from Scan Speak, Vifa, Peerless and SEAS so have an understanding of combining tweeters with this mid woofer. I designed a similar crossover on the SR71 for a local builder with good results.
This crossover only uses 2 caps, 2 inductors, 2 resistors and optionally another resistor and these are the values for the schematic posted.
L1 = 0.12mH (DCR<0R08)
R1 = 4R7 (10W)
C1 = 20uF (MKP)
L2 = 1.8mH (DCR<0R34)
C2 = 8.2uF (MKP)
R2 = 8R2 (10W)
R is for tweeter tweaking if required.
For tweaking:
To reduce tweeter level add R = 22R or 15R or 12R for taste. If not enough increase R1 to 5R6 and add R if required.
To increase mids reduce C2 in approx 2uF increments. If mids are still too low then decrease L2 to 1.5mH (with possibly decreasing L1 to 0.1mH).
To decrease mids increase C2 in approx 2uF increments.
Inductors should be 14AWG for a low DCR (especialy L1) and caps should be a good quality such as a Mundorf M-Cap (the white cheaper ones). R1 should not go below 3R3 with any tweaking as it lowers the impedance too much. The LCR from Zaph's crossover can be used across the tweeter if desired or warranted but can't be used if resistor R is in use.
As it's a series crossover all points are joined as shown and L2 can go on the leg between R2 and C1 if desired. The tweeter is -ve polarity as shown.
For small outlay it could be worth a try and may help with the sonic issues. I still think the enclosure is too large for this driver but that's another story (see Troel's site for comments on the 18W8531G00).
Attachments
ZRT vs Odin
I brought the ZRT downstairs to the same room/amp as my Odin to get a better measurement in a less coloured room, and to also record my subjective impressions regarding the sound.
I have posted both measurements.
My general impressions are:
1.) The ZRT's bass has improved over the last month (speaker break in?). Put side to side with the Odins, the ZRT's now seem to have bigger bass, although the bass is rather on the boomy side. The graph bears this out.
2.) The Odin's have more 'zing'. I don't know how to describe this difference: they sound like they are producing sound with no effort at all - there is a shimmering quality to the high notes. The ZRT's dont have that zing or apparent clarity. I don't see anything in the graph that would explain the difference.
Perhaps there is some kind of additional harmonic distortion on the Odins that is giving them this sound?
I think that I prefer the Odins' sound because I am used to listening to them. I can't defintivly say that they sound cleaner than the ZRTs.
The ZRT's subjectively sound darker than the Odins, but I dont see anything on the graphs that correlates with this impression.
I'm not sure where to go from here with respect to modifications. I think that I might leave the speakers as they are for a time and see if they grow on me.
I brought the ZRT downstairs to the same room/amp as my Odin to get a better measurement in a less coloured room, and to also record my subjective impressions regarding the sound.
I have posted both measurements.
My general impressions are:
1.) The ZRT's bass has improved over the last month (speaker break in?). Put side to side with the Odins, the ZRT's now seem to have bigger bass, although the bass is rather on the boomy side. The graph bears this out.
2.) The Odin's have more 'zing'. I don't know how to describe this difference: they sound like they are producing sound with no effort at all - there is a shimmering quality to the high notes. The ZRT's dont have that zing or apparent clarity. I don't see anything in the graph that would explain the difference.
Perhaps there is some kind of additional harmonic distortion on the Odins that is giving them this sound?
I think that I prefer the Odins' sound because I am used to listening to them. I can't defintivly say that they sound cleaner than the ZRTs.
The ZRT's subjectively sound darker than the Odins, but I dont see anything on the graphs that correlates with this impression.
I'm not sure where to go from here with respect to modifications. I think that I might leave the speakers as they are for a time and see if they grow on me.
Attachments
Monroe,
The shelf braces are definitely a problem. Even with the centres cut out I have found that the remaining skirt causes all kinds of internal reflections and resonances which really mess up the sound. removing the bracing completely has solved the basic problems and I now think these speakers will be a success. Which is a great relief.
The shelf braces are definitely a problem. Even with the centres cut out I have found that the remaining skirt causes all kinds of internal reflections and resonances which really mess up the sound. removing the bracing completely has solved the basic problems and I now think these speakers will be a success. Which is a great relief.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.