L2 and C2 look about right, but you may not need the Zobel on the woofer. Target should be 2nd order LR for the woofer, and selected should be 2nd order parallel. The tweeter should probably be 3rd order Butterworth (cap - coil to ground - cap) with a resistor before or after that. Do not parallel a 3N resistor with the tweeter....
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Ok....so, I am pretty sure what you are suggesting Don I had in a previous revision, but with bsc taken into account it looked like this one was better..... not doubting, just ignorant..... What's the reasoning for not wanting the 3N resistor there?
All my decisions were made on getting a relatively flat response (bsc considered) and getting the phase aligned properly....
I haven't been following this thread closely, as I am fortunate to remain gainfully employed during the apocalypse.
I assume that you are simulating using just traced manufacturer data? If so, you are likely to be using either an infinite baffle (aka 2pi) response, or an IEC baffle response which will be somewhat close to IB. While this is good for making direct comparisons between drivers on paper, it's not what we want to use for design work. Take that manufacturer data and apply the effects of the baffle diffraction in your intended cabinet, then to extract minimum phase from that, and enter the offsets, guesstimating the z axis offset (I notice you haven't entered any z offset in your screenshots) since you haven't the means to measure it yourself.
This is actually quite easily done with the "Blender" spreadsheet, and will provide much more useful data for simulation. Get the blender spreadsheet here:
This spreadsheet is made for splicing near and far field measurements, but is also useful for applying the diffraction to traced manufacturer data and extracting minimum phase.
1. Go to the "LF_FRD_Data" tab. Select import and load your manufacturer FRD data for either your woofer or your tweeter. Regardless of the driver, do this in the LF tab.
2. Review the imported data, observe the start and end frequency in column A. On the control tab, under Step 2, ensure that the start and end frequencies are set above and below the imported data. For example, if the data starts at 20Hz and ends at 19981Hz, enter 21hz and 19980Hz for the start and end frequencies.
3. Go to the "Diff_FRD_Data" tab, enter the baffle dimensional data and driver location and diameter, then click the "Use Diffraction Model Data" button. Piston diameter will not be 6" for a 6" driver, determine the diameter from the Sd spec for radiating surface area, usually around 4.5" for a 6" driver. Without math, you can get a good estimate by measuring edge to edge of the peak of the driver surround, so cone diameter + 1/2 of the surround.
4. On the Control tab, you will now observe the combination of the imported FRD with the baffle diffraction combined. You may now complete Step 4 on this page to "add tails" the response, this increases the accuracy of the phase extraction when done correctly. Yes, we are skipping step 3 for this process.
5. Now, select "Extract Minimum Phase", and then "Save Blended Result to FRD File".
Repeat for all drivers, then bring the resulting files into Passive Crossover Designer.
Ok....so, I am pretty sure what you are suggesting Don I had in a previous revision, but with bsc taken into account it looked like this one was better..... not doubting, just ignorant..... What's the reasoning for not wanting the 3N resistor there?
All my decisions were made on getting a relatively flat response (bsc considered) and getting the phase aligned properly....
The 3N resistor in parallel with the tweeter may drop the impedance too low. Hard to tell from your posted graph, but typically I put tweeter resistors in series unless I am trying to fix some sort of response problem, and then I'll use a 8N or similar. Not a pressing problem at this point. For some bizarre reason I never figured out, PCD usually gave me a tweeter response that was 2-3 dB off from my measurements, so I had to adjust later when I measured the drivers in the box. Your ears will catch that kind of imbalance pretty easily if that's the case for you.
But Chahly - Stahkist don't want speakers that look good, Stahkist wants speakers that sound good!
Please follow decibel's advice and take the time to learn either the Blender or Response Modeler. I know it seems like yet another complication, but IMO you are really just wasting your time (outside of learning PCD or XSim) trying to develop a crossover with IB files. All the posters have been gently suggesting this, but decibel took the time to write the long post.
If you have any questions or problems we're here to help. You could also post the cabinet dimensions, where the drivers are located on the cabinet (with respect to the lower left corner) and the cabinet volume for the woofers. That would allow us to do a cross check of your work.
I would also suggest, if you do want to take this beyond a one time shot is to get a USB mic. They are not that expensive, and if you look around someone might have a used one available - you don't have to spring for omni mic, a usb mic and a free w/s like Holm Response is all you need.
You need the mic not only to measure in box response, but also to validate your XO is working correctly. connect the XO and measure, the response should be pretty close to your modeled response.
I am not sure if you have measurement gear or not, but quite a few posts ask for XO help and they are ready to spend or have spent 100's if not in 1000's obtaining high end drivers, paying someone to cut/build/finish cabinets, but balk at 70$ for a usb mic.
The mic can also be used with REW to take care of room modes, EQ, DSP a lot of other applications that might find more frequent use.
But this is a great thread. What other tools (s/w) does everyone use and for what? I too need to revisit the basics... maybe this should be a new thread?
If you start the thread I'll add my .02. I assume you want to talk about tools to use if you are not measuring?
Wolf had a good blog that stepped through the whole procedure. Doesn't Paul C. have a procedure on his website as well? It's been so long since I've used traced graphs I've forgotten a lot of it.
I agree with Ani, USB mics are cheap enough compared to drivers/cabinets/xo-parts there isn't a strong argument for using manufacturer's graphs to design a speaker... unless you're trying to be an internet hero and spitball tons of crossover designs for people you've never met with drivers you've never heard. Just saying.
I agree 100%, but there will always be people who want to sim w/o measurement gear for whatever reason. And it is another level of complexity (far field, near field, port,merging, setting up the PC soundcard, etc.) for a newcomer. I think we tend to forget the trials and tribulations we went through (and help we received) before it became a standard operating procedure. And let's face it, not everyone wants to go "all in" like we have. Just saying, Craig
If you start the thread I'll add my .02. I assume you want to talk about tools to use if you are not measuring?
... unless you're trying to be an internet hero and spitball tons of crossover designs for people you've never met with drivers you've never heard. Just saying.
Haha, you might have me confused with someone else. I have done a few for people I have met
Ok. So. I was planning on buying a mic. It's looking like that should be step one.
I am willing to do this the right way now that life affords me the time and opportunity to have a quiet space to do the proper work. I just thought this was the beginning step to ball park and then use the mic afterwards, yada yada yada....
To be honest, for this project I wasn't too concerned about the outcome being amazing but the educational aspect got me so may as well do it to the max.
Stop, hit the reboot. What do I need to do now. What mic, what else.... I have a DATS that I need to try with v2 (I think that's it). I can't thnak you guys enough for this crash course and hand holding.
Also, here is the baffle.....I guess it doesn't matter much now....
The umm6 is the only mic I have and it leaves me wondering if the umik1 might be a better choice. While measuring it becomes apparent the umm6 is rather directional and the umik is more omnidirectional. The umm6 cal file is suspect and the umik more of a known, plus they're within $1 in price.
The umm6 is the only mic I have and it leaves me wondering if the umik1 might be a better choice. While measuring it becomes apparent the umm6 is rather directional and the umik is more omnidirectional. The umm6 cal file is suspect and the umik more of a known, plus they're within $1 in price.
As far as calibration goes they are all suspect - you need to go for the cross spectrum labs to get a decent one that works.
There was a thread on TT about the calibration files the umm6 is supplied with being generic although the cal files and mic are numbered as if they are meant for each other.
Comments
Anyone here have a minute to fire up RM and demonstrate?
Shit. How can you tell?
What's the reasoning for not wanting the 3N resistor there?
All my decisions were made on getting a relatively flat response (bsc considered) and getting the phase aligned properly....
Please follow decibel's advice and take the time to learn either the Blender or Response Modeler. I know it seems like yet another complication, but IMO you are really just wasting your time (outside of learning PCD or XSim) trying to develop a crossover with IB files. All the posters have been gently suggesting this, but decibel took the time to write the long post.
If you have any questions or problems we're here to help. You could also post the cabinet dimensions, where the drivers are located on the cabinet (with respect to the lower left corner) and the cabinet volume for the woofers. That would allow us to do a cross check of your work.
You need the mic not only to measure in box response, but also to validate your XO is working correctly. connect the XO and measure, the response should be pretty close to your modeled response.
I am not sure if you have measurement gear or not, but quite a few posts ask for XO help and they are ready to spend or have spent 100's if not in 1000's obtaining high end drivers, paying someone to cut/build/finish cabinets, but balk at 70$ for a usb mic.
The mic can also be used with REW to take care of room modes, EQ, DSP a lot of other applications that might find more frequent use.
I agree with Ani, USB mics are cheap enough compared to drivers/cabinets/xo-parts there isn't a strong argument for using manufacturer's graphs to design a speaker... unless you're trying to be an internet hero and spitball tons of crossover designs for people you've never met with drivers you've never heard. Just saying.
InDIYana Event Website
Ben that really stinks. I bet you must have put a ton of work into that.
Ok. So. I was planning on buying a mic. It's looking like that should be step one.
I am willing to do this the right way now that life affords me the time and opportunity to have a quiet space to do the proper work. I just thought this was the beginning step to ball park and then use the mic afterwards, yada yada yada....
To be honest, for this project I wasn't too concerned about the outcome being amazing but the educational aspect got me so may as well do it to the max.
Stop, hit the reboot. What do I need to do now. What mic, what else.... I have a DATS that I need to try with v2 (I think that's it). I can't thnak you guys enough for this crash course and hand holding.
Also, here is the baffle.....I guess it doesn't matter much now....
And there will be another right over the red line with a tweeter in the top right corner.
herb at cross spectrum labs actually does calibration files for mics they sell. And hes a big supporter of the diy community