I will check the dc resistance. The woofers are marked as 8531 though.
I did have some type of open cell foam that I think I ordered from Madisound in my first boxes. In the second boxes I used that only on the sides right next to the drivers in a very small area because it was thin enough to not get in the way directly behind the driver cone. For the rest of the box I tried some fiberglass insulation. Perhaps not the number one choice but should not be causing this... Also, the problem persisted through both box builds with very different material choices.
I am hoping it's a mislabeled part or a problem similar to that. Otherwise I won't know where to go from here. 🙁
I did have some type of open cell foam that I think I ordered from Madisound in my first boxes. In the second boxes I used that only on the sides right next to the drivers in a very small area because it was thin enough to not get in the way directly behind the driver cone. For the rest of the box I tried some fiberglass insulation. Perhaps not the number one choice but should not be causing this... Also, the problem persisted through both box builds with very different material choices.
I am hoping it's a mislabeled part or a problem similar to that. Otherwise I won't know where to go from here. 🙁
I know of a case where the resistors were flipped and the bass tuning needed adjustments to the actual drivers installed. Reengineered, it sounded much better.
Wolf
Wolf
So one quick test I run is to measure the impedance of the entire speaker, and compare it to a known good graph, or a simulation.
Hi, merlinx76.
At a certain point in a fairly disastrous project you just have to abandon your dreams, hopes and ambitions.
I do this all the time.
What is wrong with your Zaph Revelator Project?
IMO, the basic design is woeful. Let me spell it out. Too low a crossover, a fairly difficult Scanspeak 18W/8531G00 woofer. And, WORST OF ALL, a soft-dome tweeter crossed too low at 1.7kHz. You'd do much better with the more musical ring-radiator version.
This all just adds up to a lousy speaker. Sorry.
There is scarcely a smidge of difference between Troels Gravesen's experienced Quick and Dirty and poor crossover and Zaph's effort here.
You thought that throwing big bucks at this would get you perfection. I think you are a sucker. Sorry.
IMO, the Scanspeak 18W/8531G00 sliced cone is a terrible woofer. For sure it needs a tank-notch around 5kHz here. The D3004-6600 soft-dome tweeter is not the fastest thing on the block either.
No wonder it sounds terrible. 🙄
At a certain point in a fairly disastrous project you just have to abandon your dreams, hopes and ambitions.
I do this all the time.
What is wrong with your Zaph Revelator Project?
IMO, the basic design is woeful. Let me spell it out. Too low a crossover, a fairly difficult Scanspeak 18W/8531G00 woofer. And, WORST OF ALL, a soft-dome tweeter crossed too low at 1.7kHz. You'd do much better with the more musical ring-radiator version.
This all just adds up to a lousy speaker. Sorry.
There is scarcely a smidge of difference between Troels Gravesen's experienced Quick and Dirty and poor crossover and Zaph's effort here.
You thought that throwing big bucks at this would get you perfection. I think you are a sucker. Sorry.
IMO, the Scanspeak 18W/8531G00 sliced cone is a terrible woofer. For sure it needs a tank-notch around 5kHz here. The D3004-6600 soft-dome tweeter is not the fastest thing on the block either.
No wonder it sounds terrible. 🙄
Attachments
I built the ZRT2.0 and the sound is very non -fatiguing. I chose it over the ZRT2.5 because the tweeter is less stressed so less likely to sound harsh. Amp is an Arcam FMJ19 with a mid level DAC.
Previously I built the SR71 which I did find fatiguing. I don't have test equipment so I mounted the woofer (ER18NRX) alone on a plywood baffle and ran a frequency sweep though it. There was a definite resonance at about 4.5khz which while reduced with the crossover was still somewhat annoying.
I suggest you do the same with the 18W8531. There should be no peaking with the crossover hooked up. Individual driver response is given on Zaph's site and the woofer is very smooth.
If the woofer sounds good then run the same test with the woofer mounted in your box to see if anything changes.
Previously I built the SR71 which I did find fatiguing. I don't have test equipment so I mounted the woofer (ER18NRX) alone on a plywood baffle and ran a frequency sweep though it. There was a definite resonance at about 4.5khz which while reduced with the crossover was still somewhat annoying.
I suggest you do the same with the 18W8531. There should be no peaking with the crossover hooked up. Individual driver response is given on Zaph's site and the woofer is very smooth.
If the woofer sounds good then run the same test with the woofer mounted in your box to see if anything changes.
Hi Steve,
So one thing is the FR measurements from SS did not match very well my 4531s. The shelved response at the top end was just non-existent. The driver is MUCH easier to integrate than the published measurements would leave one to believe.
Best,
E
So one thing is the FR measurements from SS did not match very well my 4531s. The shelved response at the top end was just non-existent. The driver is MUCH easier to integrate than the published measurements would leave one to believe.
Best,
E
The AirCirc's pretty good also...
Zaph's own measurements, widely used, widely respected, widely found accurate / representative.
The AirCirc could do with a bit more linear excursion, but it overloads reasonably gracefully, and LR4 @ 1.7KHz? Narry a problem, assuming you're not trying to outdo the SPLs generated by your mate down the road with the Altec VOTT A5s. 😉
Zaph's own measurements, widely used, widely respected, widely found accurate / representative.
The AirCirc could do with a bit more linear excursion, but it overloads reasonably gracefully, and LR4 @ 1.7KHz? Narry a problem, assuming you're not trying to outdo the SPLs generated by your mate down the road with the Altec VOTT A5s. 😉
Attachments
Zaph's measurements of the drivers in the box...
Based on this (and personal experience with the drivers), I suggest a more reasonable conclusion than the ZRT being 'a woeful design' (which it isn't) is that something is awry elsewhere, either with the components or the construction. Triple-checking the wiring might be an idea; it's also certainly possible that some drivers or components are out of spec., which unfortunately does happen from time to time. The AirCirc tweeter for e.g. did go through a period where consistency was an issue. This is all speculation though unfortunately; you really need to measure them to know for sure. It's good Madisound are assisting; they do try to help when they can.
Based on this (and personal experience with the drivers), I suggest a more reasonable conclusion than the ZRT being 'a woeful design' (which it isn't) is that something is awry elsewhere, either with the components or the construction. Triple-checking the wiring might be an idea; it's also certainly possible that some drivers or components are out of spec., which unfortunately does happen from time to time. The AirCirc tweeter for e.g. did go through a period where consistency was an issue. This is all speculation though unfortunately; you really need to measure them to know for sure. It's good Madisound are assisting; they do try to help when they can.
Attachments
Drivers are likely not the problem
Design is likely not the problem
Crossovers as shipped are probably not the problem (Madisound=reliable and both are the same). Would be nice to able to confirm.
Source may be a problem. Try your I-pod with your diy amp.
You tried two different boxes and different stuffing so they should not be the problem. You checked driver phase.
The only thing left really is crossover mounting. Is it near any significant metal objects or near the speaker magnets? This would throw off the inductors.
Other than that, as suggested, find someone to measure impedance and FR.
Design is likely not the problem
Crossovers as shipped are probably not the problem (Madisound=reliable and both are the same). Would be nice to able to confirm.
Source may be a problem. Try your I-pod with your diy amp.
You tried two different boxes and different stuffing so they should not be the problem. You checked driver phase.
The only thing left really is crossover mounting. Is it near any significant metal objects or near the speaker magnets? This would throw off the inductors.
Other than that, as suggested, find someone to measure impedance and FR.
That's the way forward (measuring the things).
Another set of data for the 8531G00 -FR & CSD track John's well & are consistent with other measurements I've seen (apart from the LspLab plot shown a few posts ago, which I suspect is a different driver, presumably mislabelled).
Another set of data for the 8531G00 -FR & CSD track John's well & are consistent with other measurements I've seen (apart from the LspLab plot shown a few posts ago, which I suspect is a different driver, presumably mislabelled).
Attachments
Well I am really discouraged in all of this. In swapping, sources, amp, and speakers with my very limited selection of equipment, it seems like every single part is contributing a little to the harshness. The more transparent the sound I get, the harsher everything sounds. It is hard to tell though because all of my other equipment is very cheap and crappy, and no matter what, it always sounded quite harsh. I still feel I am getting a pretty strong upper midrange peak or resonance that does not improve much at all by trying to EQ it out.
I have to just send the crossover to Madisound for the next step... it wont cost me much to do so.. to be sure everything is OK in the XO. I bought kits because I really didn't want to set up with measuring equipment and that is involved in designing a crossover.
The thing that really makes me feel that there has to be a flaw is that the Madisound tech said that he finds these to be a fairly warm sounding speaker which I am not getting at all AT ALL.
If they tell me the XO is fine to them, I might just have to give up on this altogether. I could mail the drivers to them but that will be pricier and is less likely to be the problem I would think.
I have to just send the crossover to Madisound for the next step... it wont cost me much to do so.. to be sure everything is OK in the XO. I bought kits because I really didn't want to set up with measuring equipment and that is involved in designing a crossover.
The thing that really makes me feel that there has to be a flaw is that the Madisound tech said that he finds these to be a fairly warm sounding speaker which I am not getting at all AT ALL.
If they tell me the XO is fine to them, I might just have to give up on this altogether. I could mail the drivers to them but that will be pricier and is less likely to be the problem I would think.
Go back and reconsider the room, regardless of it being different. Both could be poor.
Last edited:
merlin,
We haven't considered your problem may be room acoustics. Hang blankets, throw pillows around on the floor.
Best,
E
We haven't considered your problem may be room acoustics. Hang blankets, throw pillows around on the floor.
Best,
E
That's still a possibility i guess. For both rooms to have the same issue though? It sounds very much the same as it did at my previous living room. There is a large area rug between the speakers and my seat. They auto eq looks surprisingly similar in both rooms and the only difference between the left speaker and right speaker is mostly in the midbass+upper bass regions.
The measurement microphone tells me there is a sharp 6db peak at 1khz in both living rooms. The difference from 1khz to 10khz is 9db!!!
Here is a rough idea of what I am looking at. This is what the auto eq wants to do to achieve a flat response. 1khz is at least 6db louder than the midrange and 9db louder than the upper treble?
Attachments
It's WOEFUL! 😀
I had a little play with the 2.5 circuit today. Very low impedance around 3.5R, which might be part of the problem here. 2.5 way boomy bass just comes with the design.
Quite easy to modify for a 2.7kHz crossover. All worked very well. And actually, the cone breakup is no worse.
I had a little play with the 2.5 circuit today. Very low impedance around 3.5R, which might be part of the problem here. 2.5 way boomy bass just comes with the design.
Quite easy to modify for a 2.7kHz crossover. All worked very well. And actually, the cone breakup is no worse.
Attachments
Many people do find it has slightly elevated bass. It probably does. If you look at my auto eq, the bass needs no gain and the midrange + treble regions (other than the 1khz craziness) mostly needs a boost of 3db to 6db relative to the bass region. That is not something that bothers me since I prefer slightly elevated bass anyway.
You keep going on about how it has to be the low crossover point. If you go to Troels Gravessen's site (just for an example), he has designed probably 100 speakers and has many (including some 3 ways even in his latest designs) that use as low as 2nd order 1.7khz and I have even seen 1.5khz on his site. Everything is a compromise but 1.7khz tweeter compromise is not what I am hearing here. A low tweeter XO can be rather charming in its own way in my experience. At just a couple watts of output (or less), the tweeter is not straining by any stretch of the imagination. It also seems like the problem is still there when I disconnect the tweeter altogether anyway.
The low impedance I could see being a problem if I needed to push the amp loud to get problems but this sounds bad at even low volume.
I just tried listening to it nearfield which I realize it is not designed for, but still sounds peaky in that 1khz region to me. Damn I hope Madisound can find something incorrect in the crossover. Otherwise I might just need to amputate my ears.
The severe dip at ~ 1kHz means to me that you may have the tweeter connected IN PHASE, which is likely incorrect......
I see you made it with only changing a few parts. Some of which I probably have already. It will be something I give a try if Madisound comes up with nothing. Thank you for your time in setting that up.
- Status
- Not open for further replies.