- Thread Starter
- #61
I did switch the speakers from left to right just to see if position made much of a difference. It did not.
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
My first REW Measurement - DIY Speakers - Did I get speaker polarity wrong?
- Thread starter nikb
- Start date
I did switch the speakers from left to right just to see if position made much of a difference. It did not.
Interesting. So. I assume the RCL for the woofer is a series inductor with the CL in parallel. If you had made a mistake here, it would affect the lowpass the of the woofer, but not the performance in the bass. I see 2 ways to debug this:
- Remove the woofers from the enclosure and measure them without crossover in short distance. If one of the chassis has a defect, it should show up I hope.
- Check the enclosure damping and the bass-reflex port. Any damping material (like Sonofil) should ideally be in the middle of the enclosure and the reflex port should be free.
Aside from the woofer, there is also a significant different in the high frequency response. Do you have any explanation for this? It's very strange. Maybe another idea would be to switch the crossover boards between speakers, just to rule them out.
- Thread Starter
- #63
OK, I was thinking the same.
The port is free and the damping material is along the side walls and back wall of the speaker. Looks fine to me.
I think the quickest check will be to swap woofers between the speakers (only switched speaker position before).
Originally the wireing was such that you could not easily remove the crossover. I have since made the crossover in the bad sounding speaker pluggable/disconnectable. Still have to do this with the other speaker.
- Thread Starter
- #65
Interesting. I swapped the woofer chassis between the two speakers. Compared to my measurement earlier, the bad woofer response has now switched from right side to the left side. The treble issue can of course still be seen on the right side.
It looks like I'll need to order a new woofer. But I better look at the treble issue before ordering any parts.
Last edited:
When you finished your research, before you buy anything I would try contacting Troels. He is usually supportive if you ask smart questions especially after you already bought one of his kits. Maybe he has an idea.
- Thread Starter
- #67
I will certainly do that. First, I will probably pull both crossovers and compare the tweeter sections.
Today I really hate my old self for making super dirty crossover builds and then cramming them into the boxes and even soldering on some of the connecting wires making sure that you absolutely cannot easily remove them if you have to debug an issue
To any new speaker builder who sees this: dont do it like that! Take your time and do it properly.
Last edited:
JayGilb
Major Contributor
If you have a voltmeter play pink noise and measure the voltage on the low/mid/high outputs and see if there a significant difference between the left and right crossovers. Also measure the signal at the input to the crossover to verify channel balance.I will certainly do that. First, I will probably pull both crossovers and compare the tweeter sections.
Today I really hate my old self for making super dirty crossover builds and then cramming them into the boxes and even soldering on some of the connecting wires making sure that you absolutely cannot easily remove them if you have to debug an issue
To any new speaker builder who sees this: dont do it like that! Take your time and do it properly.
Pareto Pragmatic
Addicted to Fun and Learning
I'm glad to see you got your measurement issues worked out.
I am not sure how much you have in parts in the crossovers, but I would likely just start over. But then I would have used cheap parts for my first version, and then better parts after confirming it worked. Or not, I'm pretty frugal.
I do suggest you confirm your cross over design before messing with things like checking individual components. If the bass filter is correct, then you likely have a parts issue. So at that point troubleshooting components will be the way to go.
I am certain there are many here who can tell you what's what with your design, I am not one of them.
I can't see how. In every crossover I have seen, the signal is the same to the point where it splits to the drivers, so the same signal is going to the drivers, then the crossover filters some of that signal out before it hits the drivers.
Ok, I can think of one, very very low probability way it could happen. If there were different gauge wires in the crossover that could do it. But why would anyone do that? By mistake, maybe? Usually people are just using the same spool of wire for the whole thing, so that might happen if someone was using scraps of wire I guess.
I have used scrap wire, but not on a project, just for patching and repair of non-audio things. So I get the impulse for sure. I very much doubt this is the case here though.
Today I really hate my old self for making super dirty crossover builds and then cramming them into the boxes and even soldering on some of the connecting wires making sure that you absolutely cannot easily remove them if you have to debug an issue
I am not sure how much you have in parts in the crossovers, but I would likely just start over. But then I would have used cheap parts for my first version, and then better parts after confirming it worked. Or not, I'm pretty frugal.
I do suggest you confirm your cross over design before messing with things like checking individual components. If the bass filter is correct, then you likely have a parts issue. So at that point troubleshooting components will be the way to go.
I am certain there are many here who can tell you what's what with your design, I am not one of them.
I can't see how. In every crossover I have seen, the signal is the same to the point where it splits to the drivers, so the same signal is going to the drivers, then the crossover filters some of that signal out before it hits the drivers.
Ok, I can think of one, very very low probability way it could happen. If there were different gauge wires in the crossover that could do it. But why would anyone do that? By mistake, maybe? Usually people are just using the same spool of wire for the whole thing, so that might happen if someone was using scraps of wire I guess.
I have used scrap wire, but not on a project, just for patching and repair of non-audio things. So I get the impulse for sure. I very much doubt this is the case here though.
- Thread Starter
- #70
Hi,
After two weeks of absence I want to give you ashort update on my diy speakers:
I have double checked the crossovers, but did not find anything. My crossovers are really messy, but look correct. I made sure that they follow the design now (I had a slightly different value for one of the resistors in the tweeter sections. Now its back to the original value).
I have been removing and changing crossovers and chassis, so its hard to tell the original connection, but I believe that both mids were connected the wrong way.
Yesterday my new woofer arrived and I exchanged it for the damaged one.
Even though both chassis came with the same kit and both speakers have been stored together, the woofers are very different. While the "good" woofer has a flexible and compliant sorround, the surround of the bad woofer is very brittle and gets damaged by even slightly touching it. Both speakers probably always performed slightly different, even when they were new.
Even now the surround of tthe new woofer feels much more flexible and compliant than the "good" old woofer. Maybe I should change the second original woofer, too? Anyway, I will adress the issue with the tweeter on the right side first.
Here's a new REW measurement:
As you can see, the right side still exhibits the drop in the high frequencies. It seems that I missed something there.
I will have to (again) make sure that both drivers are performing alike and then remove the crossover and check again.
Despite the remaining issues I can now hear that they are very good and transparent speakers, and the time it takes to fix them will be worth it.
After two weeks of absence I want to give you a
I have double checked the crossovers, but did not find anything. My crossovers are really messy, but look correct. I made sure that they follow the design now (I had a slightly different value for one of the resistors in the tweeter sections. Now its back to the original value).
I have been removing and changing crossovers and chassis, so its hard to tell the original connection, but I believe that both mids were connected the wrong way.
Yesterday my new woofer arrived and I exchanged it for the damaged one.
Even though both chassis came with the same kit and both speakers have been stored together, the woofers are very different. While the "good" woofer has a flexible and compliant sorround, the surround of the bad woofer is very brittle and gets damaged by even slightly touching it. Both speakers probably always performed slightly different, even when they were new.
Even now the surround of tthe new woofer feels much more flexible and compliant than the "good" old woofer. Maybe I should change the second original woofer, too? Anyway, I will adress the issue with the tweeter on the right side first.
Here's a new REW measurement:
As you can see, the right side still exhibits the drop in the high frequencies. It seems that I missed something there.
I will have to (again) make sure that both drivers are performing alike and then remove the crossover and check again.
Despite the remaining issues I can now hear that they are very good and transparent speakers, and the time it takes to fix them will be worth it.
Keith_W
Master Contributor
I suggest you read Book 2 of the REW eBook so that you can learn to interpret these measurements yourself. You have shown the frequency response too zoomed out. If you look a bit closer, you will find the right speaker is -20dB at 20kHz compared to the left speaker. This is intolerable. It's going to be either your driver or your crossover that is at fault. The easiest way to diagnose it is to swap the tweeter left to right and re-measure. If the problem follows the tweeter, then you need a new tweeter. If the problem remains with the speaker, you need to check your crossover.
I also want to show you something. This is the step response of your left speaker (red) compared to the right (green). I have zoomed out to 450ms.
The "head" of the step response shows how well integrated your speakers are, and the "tail" of the step response shows ringing. I don't know if it's room ringing or driver ringing, but you can see a few unpleasant features:
- the amplitude of the ringing is higher than the main impulse itself
- the left speaker (red) rings for roughly the same duration, but the amplitude of the ringing is significantly lower.
For comparison, this is my system with the same horizontal scale. Notice that the main impulse is narrow and sharp, and all subsequent ringing is lower in amplitude than the main impulse. It also decays much faster than yours.
You will need to find out why there is so much ringing. Like I said, it could be the speaker, it could be the room. The easiest way to find out is to remove the room from the equation. Take the speaker outside and repeat the measurement. If the ringing stays the same ... I would be looking at whether your driver is underdamped (high Qts), or your cabinet is vibrating, or something like that.
I also want to show you something. This is the step response of your left speaker (red) compared to the right (green). I have zoomed out to 450ms.
The "head" of the step response shows how well integrated your speakers are, and the "tail" of the step response shows ringing. I don't know if it's room ringing or driver ringing, but you can see a few unpleasant features:
- the amplitude of the ringing is higher than the main impulse itself
- the left speaker (red) rings for roughly the same duration, but the amplitude of the ringing is significantly lower.
For comparison, this is my system with the same horizontal scale. Notice that the main impulse is narrow and sharp, and all subsequent ringing is lower in amplitude than the main impulse. It also decays much faster than yours.
You will need to find out why there is so much ringing. Like I said, it could be the speaker, it could be the room. The easiest way to find out is to remove the room from the equation. Take the speaker outside and repeat the measurement. If the ringing stays the same ... I would be looking at whether your driver is underdamped (high Qts), or your cabinet is vibrating, or something like that.
- Thread Starter
- #72
Thanks for pointing me to the ebook - I will read it.
- Thread Starter
- #73
I have since swapped the tweeter chassis between the left and right speakers and the problem does NOT follow the chassis.
I have also re-measured the left speaker in the position of the right speaker to see how much of the drop is dependent on speaker position. There is no fundamental change caused by the position in the room (I will not post these measurements here, because essentially they do not show any change).
Thus the culprit seems to be the crossover. I have removed it and I will check all the values again.
- Thread Starter
- #74
This is very interesting. I am convinced (but measurements need to show it first) that my room plays a significant part in this. The room is very bare with lots of hard surfaces. Based on my previous measurements I believe that both speakers should show the same response in my room. Also, both cabinets are very sturdy.
Problem is, I do not see how I can measure the speakers outside: I live in an apartment without a back yard and it is amost winter here in Switzerland. Is there any alternative to measuring outside?
Keith_W
Master Contributor
If your room looks like an echo chamber, it probably is. What struck me is that you said that one woofer was stiffer than the other one, and as you can see, the ringing characteristics between L and R are not the same. It could be speaker, it could be the room - no way to tell from this measurement.
Maybe you have access to a sports hall somewhere? https://www.freifeld-akustik.de/2019/01/24/lautsprechermessung-wo-soll-ich-messen
- Thread Starter
- #77
Good idea. I have a school nearby - although this will have to wait until the crossovers work properly.
- Thread Starter
- #78
This is getting frustrating. I have checked all the component values of the crossover. They seem to be OK. Only one resistor is unreadable - I will desolder it and measure it outside the circuit. I will reflow all solder joints just to be sure.
Keith_W
Master Contributor
How did you check the component values? By looking at what is printed on the components? Or measuring the response with a digital multimeter?
- Thread Starter
- #80
Only by looking at the print. Its all stil in circuit atm, and I only have a multimeter and cannot measure capacitance or inductance.
Most of the inductors have hard to read labels by now, but I still have the kit instructions and they contain things like wire sizes, helping me distinguish them.
Similar threads
