having major issues with my crossover
Thank you, i really appreciate it, im going to change my crossover tomorrow after work and re test the speaker, ill take the video of my set up like i mentioned earlier and send it over to you.
I definitely struggle with the software side of things, i tried to upload measurements into xsim to create the crossover when i first started these speakers, i got the drivers loaded but for whatever reason the components i added hardly changed the response graph so i know i must have uploaded it wrong.
So i just took the crossover out of the enclosure to check how i wired it, it turns out i had wired it to correct side of the cap when i built it. The mistake id made was only on the quick mock up i did for the picture to go on here. Im actually so pissed off with this now, i thought i had the answer as to why it doesnt match my xsim design but im back to having no idea whats wrong. So im gonna start fresh. Im deleting all my old files from omnimic and doing them again, then im gonna try to add my measured files to xsim and disregard the manufacturers measurements, its just not working for me.
and thanks to you both for all the help you've given me this far with all of my builds
Just finished taking the new measurements, and actually managed to get them into xsim following your awesome youtube video on the subject, i added the frd file of both woofers playing together onto each separate woofer in the diagram, i then added all the components for my crossover and it gave me a graph perfectly matching the one i got when i measured my speaker. It had the same dip at around 4000hz. So ive been messing around with the crossover but i cant get it anything like flat, the impedance is at like 16ohms too. im beginning to think that these drivers dont match well, which is a pain since my boxes are all made to suit that tweeter. Also the very top end is giving me issues, i think thats down to the pc83, every time ive measured those drivers ive had a weird thing going on after 16000hz. Anyways here are the new measurements in case you would be good enough to have a mess around with them for me, a good crossover at 4500 hz seems impossible to me.
First of all that is excellent news with regards to the response. Accurate measurements make all the difference. I'm so excited that we got that part out of the way. I'll go ahead and take a look at these this weekend and see what I can come up with. Feel free to post what you have already, it'll give me a good starting point I might be able to help out explain what you could have done a little differently.
Thankyou i would really appreciate that. I took more frd and zma measurements today of just one of the pc83 drivers connected in the box, then did the tweeter again then did both playing together. So now i have the measurements posted above of the woofers in series, and some on my laptop of one woofer and the tweeter, i took that measurement just so i could try it in xsim as i wasnt sure if i was to upload the frd of both woofers or just one to each simulated driver. When i first tried the crossover by adding frd of the woofers in series to each sim driver i was getting an impedance over 16ohms so i wasnt sure that was right.
I tried tonight to get the response where i wanted it but my impedance wasnt good, and i didnt know how to fix it, it dropped to 1.5ohms at 12khz so i put it to one side to try again tomorrow. So once ive got as far as i can with it ill post my results and maybe you might have a fix for my impedance problem. These two drivers, to me, seem very hard to knit together, its as though the pc83 doesnt like that im trying to quiet down its higher end, and give its job to the tweeter.. the son of a bitch 😄
Yeah i was planning to post it tomorrow but i may as well post it now, itl take me around 15 minutes though as i have to get my computer on and it takes forever to load
Sorry, i cant find where i saved it, currently trying to reproduce it, ill be as fast as i can
So ive got it something like how i had it before, i want the slight arc kind of dip around 4000hz area, but im finding it hard to find the right components/circuit to properly adjust the middle area, i think the crossover point is in the right place but i realise that this is probably a useless overall design that will have many flaws, but messing around with it is the the only way i can see what changes what. Like i said though, still having trouble knowing how to adjust the middle.
I also have a quick question, since these frd files are from the actual speaker, does that mean no baffle step has to be worked out as the graph once finished will be how the speaker will sound? So instead of needing a 6db decrease in the treble i can just lower it by however many i want in relation to how i want to colour the sound?
As you can see, the impedance is well and truly "fubar"
Also, the woofers in that design were done with the frd and zma file of a single woofer in the box, i couldnt keep the impedance below 16ohms when i used the woofers in series files. I know there is probably a way to sort that, i just dont know it lol
These are the files for the single woofer tested, just in case they are any use to you. The z-offset i added to each woofer was 0.8
Something is wrong with your measurements or the files I am receiving. Why don't you just save your xsim file and zip it. All the data you have will transfer.
Just to be clear on how you measure your z-offset.
- Place your microphone approximately 16" away from your speaker, centered on the tweeter.
- Take a measurement of your tweeter
- Measure both of the PC83-4 however they will be hooked up (in this case in series). These are now acting as one driver.
- Hook up the tweeter in parallel to the to the PC83-4's. These will stay in series. So you will take the +of the tweeter and hook it up to the positive side of the new combined woofer. And the negative to the tweeter the same way. There should only be a total of 4 wires connected to each other.
Something was off with the first files you sent over, I could not find the z-offset with them. The second files you did not include the zma files. You really should be measuring the PC83-s as one driver though.
Yeah this is what has been bothering me, im convinced my microphone is faulty because everything you described above is exactly how i measured them. Ive even changed the amplifier in case that was causing the issues with it, im not sure what else to do, im wondering if the microphone calibration files could be wrong? Or my laptop could be changing the sweep somehow?
Ill upload the zma for the second set of files, im sorry that i forgot to do that, it was 4.15am when i uploaded them. I think the reason you cant use the first set of files is because i forgot to measure them with the "show phase" button clicked. Again, sorry about that.
If i should be making the crossover with the woofers in series rather than single then thats how i will do it, i was just unsure how to add the files to the simulated woofers because it was seeing each one as an 8ohm driver instead of a 4ohm, is that what should happen?
And yeah ill save the xsim file next time, i didnt realise that's how you did it 🙂
Ive deleted and reloaded the original files for you, i think they should work now as i took new ones with phase included yesterday, if they still dont work there has to be something wrong with my set up. Thankyou for your patience with me, i know im probably causing you a headache by now
Just a rundown if how i take a measurement
I set the speaker about half distance between roof and floor,
Set the mic up 16" away from the centre of the tweeter
I have two wires from the tweeter through the port hole, and two wires from the series woofer through the port hole too.
I have the laptop headphone jack connected into the amp aux in, then i connect the pos and neg wires from the amp to the tweeter
I press play and turn the soundcard on in omnimic and adjust the amp volume until the graph is around 80db then i click show phase and wait until the graph settles down. Then i pause and save the frd curve.
I do the same without changing the amp volume or moving the speaker for the other measurements.
If that method is fine then it must be either a setting in omnimic or on the laptop that isnt producing the correct response. I dont know what else to change or try. Hopefully you see an issue in my method because thats the easiest fix
It sounds correct. Like I said, if you want to upload your xsim information, I'll take a look at what you found and see if I can help. It sounded like you found the z-offset with some measurements you had. I think they were the ones, that you didn't attach the zma files with. Either way, if you want to send that info over, I would be happy to look it over.
Thank you, ive replaced all the files i had already uploaded with new ones that contain the phase information, or so i believe. I have also uploaded the xsim file in the post above that has my set of single woofer measurements. Im not sure how much use it will be to you though as i modelled the xsim design from the one driver frd file rather than two in series.
I want to do a knew xsim design using your method of two drivers in series but i still haven't figured out how to upload the frd without making the software believe it has 4 drivers and one tweeter in the impedance chart.
I hope the new uploaded files are something you can work with now though, please let me know if i have to take new measurements or do something else to get it to work for you . And if thats the case i will record the omnimic screen from my phone during the measurement so you can see whats going on