This is just one of those have fun building speakers thread with some good pointers that is supposed to help people. Calling it insane is hmm... simply wrong.
Sharing suspicious FRD and ZMA files on the web is indeed an insane way of having fun with building speakers...
Hello,
so I was very fortunate to get a nice e-mail from SEAS today 🙂, they sent me the original datas for the A26 woofer and the T35
:
Enclosed were the .txt files that I provide here in a ZIP file. These were exactly frd and zma files, I just duplicated them to get the usable files.
There is indeed a substantial discrepancy with the datas that circulated on this forum so far, so please take care of using only these original datas in the future.
I played a little bit in XSim and it appears that a 1mH inductor in // on the tweeter makes a big improvement.
Without the inductor :
and with the inductor :
Strange that SEAS didn't think of adding it in the original filter!
so I was very fortunate to get a nice e-mail from SEAS today 🙂, they sent me the original datas for the A26 woofer and the T35
: Enclosed were the .txt files that I provide here in a ZIP file. These were exactly frd and zma files, I just duplicated them to get the usable files.
There is indeed a substantial discrepancy with the datas that circulated on this forum so far, so please take care of using only these original datas in the future.
I played a little bit in XSim and it appears that a 1mH inductor in // on the tweeter makes a big improvement.
Without the inductor :
An externally hosted image should be here but it was not working when we last tested it.
and with the inductor :
An externally hosted image should be here but it was not working when we last tested it.
Strange that SEAS didn't think of adding it in the original filter!
Last edited:
Hello,
so I was very fortunate to get a nice e-mail from SEAS today 🙂, they sent me the original datas for the A26 woofer and the T35
Enclosed were the .txt files that I provide here in a ZIP file. These were exactly frd and zma files, I just duplicated them to get the usable files.
Thanks for sharing! SEAS customer support has proven pretty awesome. I'll consider adding an inductor to my build and report back when I have.
The woofer of the A26 speaker actually is a bit too loud in the range from 600 to 1100 Hz. The tweeter is of opposite phase there and brings the level down by 2 dB. That's how the original works.
With an additional inductor in the tweeter filter this doesn't happen anymore and the hump at 900 Hz might become a problem.
Attachments
True, but wouldn't you agree that the flatter response from 1 - 2,5 kHz is better overall? As well as higher sensitivity.
Forgot to attach the files. 🙂
Just for fun, here's the FR plot. The complete XSim file is zipped and attached.
Best,
Erik
The bump at the ~1.8KHz probably will make the sound a little forward. The overall level at 2K - 4K level a little too high compared to the 200 - 300Hz so the overall sound probably could use some warmness.
Last edited:
I really don't know. It's a simple modification you can easily try. I only expect too much mids from the modification and think the original version is more balanced. 🙂
EaZy, many thanks for sharing. I'm keen to see the Xsim response with 0.82mH & 10R on the tweeter (the values I have right now) but don't know how I can turn the .txt files into .frd & .zma. Please can you give me a steer?
A frd or zma is just a text file with a different extension on the filename. If each row of numbers is set up right in the txt file, just rename the file from FILE.txt to "FILE.frd" or "FILE.zma". Make sure you aren't changing it to FILE.frd.txt (if using Windows Notepad, save it as type "*.* all files".
EaZy -- does the SEAS data include correct delay relationships in the FRD file? Most manufacturer data doesn't, it usually is with delay set that gives the flattest phase plot for each driver alone, not accounting for distance of source-to-microphone. Without phase set for actual, as-used in the mounting of stuff, relative delays, the simulation in the crossover region won't be correct and can even be horribly wrong.
By the way everyone, I'm not hurt others are improving the idea I started with, but do apologize for any errors. I honestly don't remember how much I did for the original. 🙂
I'm learning quite a bit watching how others approach the topic of simulation with limited evidence. While I personally find it frustrating to talk to hobbyists who want all sorts of advice about how to do things the hardest possible way, I think this entire thread is a pretty fun and enlightening exercise, so thank you to everyone who has contributed constructively and with a light touch.
Best,
E
I'm learning quite a bit watching how others approach the topic of simulation with limited evidence. While I personally find it frustrating to talk to hobbyists who want all sorts of advice about how to do things the hardest possible way, I think this entire thread is a pretty fun and enlightening exercise, so thank you to everyone who has contributed constructively and with a light touch.
Best,
E
Just as said bwaslo, you just need to rename the files extensions. The ZIP package contains the files you need BTW.
The files contain the lab measurements coming from SEAS, I guess phase datas are correct!
Boxsim's crossover optimizer with balanced settings comes back with a 1 mH 0,41 Ohm inductor. Just ordered two 1 mH Jantzen air coils 18AWG. DCR is 0,48 Ohm, just north of 10% of the T35's Re of 4,6 Ohm.
- Status
- Not open for further replies.