having major issues with my crossover
A few months ago i started a thread about building some surround speakers, the cabinets are built and working using just the pc83 drivers, i made them in an mtm style, 2 pc83 and one td20f-4 for tweeter, i just didnt connect the tweeter or the crossover. ive finally done that with one of the speakers now with plans to get that right before doing the rest. The only problem is that the crossover i designed in xsim is absolutely nothing like it should be in regards to frequency response.
This was my xsim crossover
And after wiring together all the parts and installing it in the speaker i got this after testing at 16" away on tweeter level. I wired the tweeter inverted as the diagram says
As you can see its absolutely nothing like it, so i tried wiring the tweeter normally rather than inverted and actually got a better graph, slightly that is
So given the above, ive lost all confidence in the method of using xsim to design the crossover, please can someone help me figure out if its the initial design thats the problem or something else? Whats bothering me the most though is that the speaker with the crossover and tweeter installed isnt much different from the graph i get when testing just the pc83 on their own. This graph below is with just the pc83 playing (i tested one of my incomplete speakers as a reference)
I apologise for the massive first post, im just trying to give as much detail as i can to start off with, one last thing to add is that i know the crossover point is where it should be, at 4800hz. I played a test tone from 20-20000khz through the speaker and put my ear to the woofers, they stopped playing a tone at 4700khz and the tweeter took over from there.
I seriously appreciate any advice on this as im losing all hope of ever being good at crossover design
@dameo182 First, where are you taking these measurements from? How far away and are these on the Tweeter axis, center on the baffle?
Next, after you answer those questions it would be best to take three measurements with the crossover unhooked. The first is the Tweeter centered on the Tweeter at Tweeter level 16 inches away. Then unhook your Tweeter and play just the PC 83 with the microphone in the exact same position. Finally take one last measurement of both of the drivers in parallel. Once you have those measurements, upload them here and I'll help with the design.
Hi, thanks for taking the time, i took the measurement 16" away on tweeter level, and yes centre on the baffle. I dont know if you remember but you helped me to figure out the z offset a few months ago on these speakers. I took all those measurements and uploaded them to my surround sound speaker thread, its the last post i think, but i figured i would build the crossover and see how it measured. I copied the diagram and i was expecting to at least see a similar response to the xsim design. So when it turned out to be nothing like it i was so disappointed, because im back to square one. I dont know if its my measurement thats faulty or the build of my crossover. Which do you think i messed up on? I was simply trying to get the xsim result on the speaker then adjust from there, but im completely lost with it all now since nothing i have tried has worked.
Should the real life crossover have been closer to the xsim design than it is or is it normally quite different due to my box and room etc?
Sorry it took me so long, with the time difference here in the uk i only see your messages while at work, i have to wait until lunch time to reply.
Erm tbh im not sure which file is the right one, i plan to start from scratch and re-do all the measurements either tonight or tomorrow anyway, if you dont mind waiting that is, as ive got to rearrange the insides of one of my speakers ready for testing again.
Im thinking theres only 2 things i could have messed up with, as i think the xsim design was ok. Id have been happy to get that response in real life. So ive either made a mistake on tge measurements, or the way i built the crossover was wrong. I thought i might record the next measurements i do and send you the video so you can see if anything im doing is wrong? That way if it all seems ok then i can turn to the crossover and see if my mistake was with that?
No worries. I'll definitely look it over when you get it. it'd be a good idea to take pictures of the crossover as well next time. now one of the things that I do is I use alligator clips to clip together the different crossover components before I solder on a board. That helps me make sure that there were no issues with it. And if there are issues I should catch it there before I solder it.
Yeah thats a good idea, i think ill just twist the components together next time so its easier to mess with when i inevitably do it wrong 😄
I dont own alligator clips yet, but i will get some for future crossovers.
And thank you for again taking the time to try and help me find the reason none of this is working for me
Tbh i would prefer to be able to measure future drivers in my enclosures, and then upload the measured files to xsim and create the crossover to adjust the actual response i get, rather than using frd files from the website, is that something you could teach me how to do? Ive treid to learn it from other posts but they started to talk about putting tails on the driver responses and extracting minimum phase n stuff, i didnt really understand it
Just quickly wired the xover parts together in the way i did it in the speaker, hope you can see clearly enough from the pic to see whether its right or not
At first glance, I notice that the initial connection of the woofer circuit is in the wrong place. You have it at the junction of C1, L3, and R1. It should be on the other side of C1, ie, the positive binding post where you connect the amp. Other than that, it's correct, AFAICS.
*Note that the other side of L3 connects to the ground bus (-), as well as the ground terminals of the drivers as depicted in the schematic.
I think i know what you mean but im not 100%, do you mean that the baffle step should have its own positive wire from the amp and then go into L2 and then on into the woofer pos? I thought i could connect it to the 3 other parts, i didnt realise it mattered which side it was connected to so thankyou for that. Could that be causing the dip i have in the response?
And just to check, when you say ground bus you mean the common ground? I connected the nuetrals from the speakers and the remaining inductor together, then joined them to a single wire and sent that to the negative binding post on the back of the speaker, is that right?
No, ive just realised what you meant, the power for the inductor is going through the cap first isnt it, im such an idjit!
Not an idjit if you can learn from these setbacks. Give it a try and see how it sounds.
Yeah i will, hopefully it fixes my issue, if theres still some problems with the response after changing it no doubt ill be back on here begging for help again lol thank you for noticing that, i hadnt thought about posting a pic of the crossover until @123toid had mentioned it, now im glad i did, it may have saved me weeks of headache if the response improves
It sure will. You had basically started with a 1st order high pass feeding all of the drivers. With a 1.33uF @ 4ohm, that's pretty high up there.
Let us know the outcome.