Dave Bullet said:
Exactly. Consistency is the key. Kudos to Zaph for his efforts.
Yes, I traced the SPLs of Zaph's FR data and generated their phase using the FRC. I used impedance plots from manufacturers' sheets---the scale of Zaph's impedance plots is too coarse.
In all sims I set a -0.8 inch offset for the woofer.
Jay_WJ said:
I've done two crossover designs using 7" drivers and both have shown woofer acoustical offsets of about -2.25" relative to the tweeter on a flat baffle. The first was Seas W18's with Millenium, and the second was the Peerless 850122 with the Seas TBFC/G.
I determined this by experimenting with the crossover phase offsets until a reverse null was obtained though measurement at the crossover frequency.
jimangie1973 said:
This is strange. For all four designs I simulated---Jon Marsh's Modular MT, Zaph's L18/TBFC, BAMTM, and ZD5, the deepest reverse null at XO point was obtained at around -1.2" to -0.8" woofer offset relative to the tweeter. I think -2.25" is way too much for an acoustical offset of a 7" driver. In your case the in-phase listening axis would occur a feet above the tweeter level.
Jay_WJ said:
Well, in the case of my Peerless/Seas two-way, I am crossing over at about 1725 Hz. My crossover has 115 degrees phase lag in the tweeter (relative to woofer phase) to get a deep reverse null and excellent in-phase summing.
So, with the speed of sound 1130 ft/sec, a 1725 Hz tone has a wavelength of 7.86 inches. So 115 degrees phase shift is:
7.86 * 115/360 = 2.51 inches.
If I'm missing something LMK.
jimangie1973 said:
You don't miss anything. What I meant in my reply is that a 2.25 inch acoustic offset is way greater than a usual physical offset of a 7" woofer from a tweeter on a flat baffle. The result is that the optimal in-phase listening axis will be above the tweeter axis (I don't know how much above before calculating). If this is what you aimed for, that should work for you.
Which FRD tools did you use?
I've used SPL trace and Passive Crossover designer to sim Zaph's L18/ 27tbfcg results but I get this graph that looks like the drivers are connected with reverse polarity. Although it looks like a perfect reversed polarity graph, I dont know if it's supposed to be that way (meaning I need to somehow reverse the polarity) or Passive crossover designer just isn't cutting it. Forgive me I'm new and trying to learn.
Thanks
I've used SPL trace and Passive Crossover designer to sim Zaph's L18/ 27tbfcg results but I get this graph that looks like the drivers are connected with reverse polarity. Although it looks like a perfect reversed polarity graph, I dont know if it's supposed to be that way (meaning I need to somehow reverse the polarity) or Passive crossover designer just isn't cutting it. Forgive me I'm new and trying to learn.
Thanks
DefiantGSR said:
I used FRD tools exactly the way described here:
http://www.rjbaudio.com/Audiofiles/FRDtools.html
I used Zaph's data to trace the drivers' FR, but used manufacturer's technical sheets to trace their impedance (available at http://www.seas.no/ , different from those at madisound).
I used Speaker Workshop for the XO simulation.
Jay_WJ said:
I just did the measurement again. The deepest null is 15 dB down at 1600 Hz at about 1" below twetter axis. This is pretty good in room with reflections. Also, the real null is not exactly at one of the 62 RTA frequencies so energy around the null bleeds in, further reducing the null reading.
I did a vertical sweep and the null quickly fills in when the mic is raised or dropped below the nominal level 1" below tweeter.
jimangie1973 said:
This doesn't make sense to me. The mic is 1 meter from the speaker and 1" below tweeter. Then distances from the mic to the woofer's center and to the tweeter shouldn't differ at most by 1.5". But you said your XO made an acoustic offset of 2.5" between drivers. I don't think this is consistent with your measurement. How do you explain this? Do I miss something?
P.S. I just did some trigonometry. Assuming that a usual 7" midwoofer's center is located at 1" (at maximum) behind the baffle and that the tweeter and the midwoofer's center-to-center distance is 6", distances from the mic, which is 1m from baffle and 1" below tweeter, to the woofer's center and to the tweeter differ by 1.306".
Jay_WJ said:
That does make sense, but according to Dickason (Loudspeaker Cookbook 6th ed.), a good estimate is the center of the voice coil of each driver. I think what's going on is that there is a group delay between the electrical signal and the corresponding cone movement, and since woofers have greater mass than tweeters, the group delay is larger, meaning an acoustic center farther back.
According to the FRC simulation of several 7" drivers' group delay, the amounts of group delay are at most .3 msec, which is negligible, anywhere above 500Hz.
BTW, below are what I found about this issue, though I'm not sure if Roman JB is an expert by your standard.
"One important step that should be noted is that you have to assign a driver offset to each driver if their voice coils aren't acoustically aligned. For example, with a 2-way speaker with a flat baffle you should set the woofer offset to 0 and the tweeter offset somewhere around 0.5 to 1 which places the acoustic center of the tweeter closer to the listening position as is the case with a flat baffle speaker assuming the target listening axis is in line with the tweeter."
from http://www.rjbaudio.com/Audiofiles/FRDtools.html
"Another option allows you to define the driver's offset. This "offset" value refers to the offset of the driver's acoustic center relative to the baffle or a given reference point. A positive value indicates that the driver's acoustic center is sticking out beyond the baffle while a negative value indicates that the acoustic center is recessed into the baffle. Often when designing flat baffle speakers and using manufacturer's data you want to set the woofer's offset value to a negative number around 0.5 inches or more depending on the size of the woofer which indicates that the woofer's acoustic center is recessed into the baffle relative to the tweeter's acoustic center which is in line with the baffle. You can also change this offset value to get an idea what the frequency response will be at a different vertical listening height (listening at a taller height usually puts you closer to the tweeter relative to the woofer so you would make the tweeter offset more positive)."
from http://www.rjbaudio.com/Audiofiles/SWtutorial.html
So, according to him it seems that an offset for a usual 5" to 7" woofer relative to tweeter, which we need to assume to design an XO, is -.5" to -1".
BTW, below are what I found about this issue, though I'm not sure if Roman JB is an expert by your standard.
"One important step that should be noted is that you have to assign a driver offset to each driver if their voice coils aren't acoustically aligned. For example, with a 2-way speaker with a flat baffle you should set the woofer offset to 0 and the tweeter offset somewhere around 0.5 to 1 which places the acoustic center of the tweeter closer to the listening position as is the case with a flat baffle speaker assuming the target listening axis is in line with the tweeter."
from http://www.rjbaudio.com/Audiofiles/FRDtools.html
"Another option allows you to define the driver's offset. This "offset" value refers to the offset of the driver's acoustic center relative to the baffle or a given reference point. A positive value indicates that the driver's acoustic center is sticking out beyond the baffle while a negative value indicates that the acoustic center is recessed into the baffle. Often when designing flat baffle speakers and using manufacturer's data you want to set the woofer's offset value to a negative number around 0.5 inches or more depending on the size of the woofer which indicates that the woofer's acoustic center is recessed into the baffle relative to the tweeter's acoustic center which is in line with the baffle. You can also change this offset value to get an idea what the frequency response will be at a different vertical listening height (listening at a taller height usually puts you closer to the tweeter relative to the woofer so you would make the tweeter offset more positive)."
from http://www.rjbaudio.com/Audiofiles/SWtutorial.html
So, according to him it seems that an offset for a usual 5" to 7" woofer relative to tweeter, which we need to assume to design an XO, is -.5" to -1".
I think the following page of RJB Audio answers all of my questions about the driver offset:
http://www.rjbaudio.com/Audiofiles/driver offset calculator.html
jimangie,
What is the vertical center-to-center distance between woofer and tweeter of your design?
http://www.rjbaudio.com/Audiofiles/driver offset calculator.html
jimangie,
What is the vertical center-to-center distance between woofer and tweeter of your design?
I just got a reply to my email to RJB. I think he's right in saying,
"It is rare that you ever get perfect phase alignment between drivers and even if you do, a change in listening distance or height will change the phase alignment. I would focus more on the tweaking process through listening and adjusting the crossover in order to get a neutral sound. I think this is more important than struggling over perfect phase alignment between the drivers."
And reading Dickason's LDC 7th Edition I learned that a driver's ZDP (zero delay plane) is different from a driver to another and hard to know unless measured by sophisticated device.
jimangie,
Now I understand your case may happen and make sense. I learned a lot by having a discussion with you. THanks!
"It is rare that you ever get perfect phase alignment between drivers and even if you do, a change in listening distance or height will change the phase alignment. I would focus more on the tweaking process through listening and adjusting the crossover in order to get a neutral sound. I think this is more important than struggling over perfect phase alignment between the drivers."
And reading Dickason's LDC 7th Edition I learned that a driver's ZDP (zero delay plane) is different from a driver to another and hard to know unless measured by sophisticated device.
jimangie,
Now I understand your case may happen and make sense. I learned a lot by having a discussion with you. THanks!
I updated the crossover design posted here: http://www.diyaudio.com/forums/showthread.php?postid=1185885#post1185885
And also put it and its predicted FR below:
In the actual tweaking process, if I feel the sound a bit forward and want it to be more laidback I will shape the response in the upper midrange (or lower treble) by replacing C6 (currently 8.2 uF) with a 7.5 uF cap. The effect will be like:
And also put it and its predicted FR below:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
In the actual tweaking process, if I feel the sound a bit forward and want it to be more laidback I will shape the response in the upper midrange (or lower treble) by replacing C6 (currently 8.2 uF) with a 7.5 uF cap. The effect will be like:
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.