To make a long story short I discovered a loss of output in a speaker using Audiolense and I wanted to share how to go about diagnosing the problem. What audiophiles don't know is how incredibly useful the impedance plots are to diagnosing speaker issues, if any.
95% of the time, if the speaker's impedance chart is correct the speaker is functioning as designed, and when it's not working correctly the impedance plot can accelerate diagnosis.
Audiolense
Audiolense is a suite of tools to make it easy to do room and speaker correction. It's similar in some ways to Dirac or rePhase. I've never used it before so wanted to see how easy it would be to "fix" my desktop speaker's Roon playback. The reasons are complicated but I don't run Windows on my desktop, but I do run Roon here and in the living room so I wanted to see how easy or hard it would be to apply the full suite of corrections Audiolense has to my existing speakers via Roon.
The Speakers
My desktop speakers are
DIY models I call LM-1. They are small 2-ways with a 5" woofer. Not high output or deep bass. Perfect speakers to show you how incredibly useful impedance measurement tools like the
Dayton Audio Test System (DATS) is. Any DIYer or audiophile who needs to get into their speakers to fix or update them should have one in their arsenal, which is the point of this post.
The Problem
The issue showed up immediately after running my first test suites with Audiolense. Mind you, at this point it's no different than running OmniMic or Room EQ Wizard. We aren't really taking advantage of any feature other than frequency response measurement.
Unfortunately I lost the very first measurement I took so I'm having to simulate it for you, below.
So, what happened? I had no idea yet. Clearly the tweeter is at least under performing. I moved the speaker around to take impedance measurements and the tweeter got magically better. I strongly suspect a loose connection. I can show you the second FR measusmrents from Audiolense here:
Without an impedance measurement we'd probably suspect a bad tweeter here, but truth is anything between the jack and the tweeter can be bad. Here's where DATS comes in to rescue us from uncertainty, and why having a frequency plot alone alone is not good enough.
Before
I even open the speaker I run an impedance sweep of both of them.
Lucky for me, being the original builder I also have the reference
impedance plots to help guide me.
The plots above are the left and right speaker. The bottom is the left. The left speaker tracks my design simulations, and is about 2 Ohms lower across the board than the right speaker, but it's even worse in the tweeter section. .
What does this tell me? Lets start with some basics. The hump at about 1,700 Hz is where the low and high pass filters meet. The impedance above that is mostly the tweeter circuitry. The impedance below that is mostly the woofer and it's circuits.
If the tweeter was shorted the tweeter impedance would be LOWER than reference. If it was open the impedance would shoot up. Neither is happening. We also note that the impedance problem is wide-band. The whole speaker is suffering a problem, not any one driver. Take a look for instance around 500 Hz. That is an area completely up to the woofer. It's an Ohm higher than it should be.
This tells me the problem is unlikely to be the crossover or the driver. I probably have a high resistance (1 Ohm is high in this case) problem somewhere between the jacks and the crossover. These were some of the first speakers I ever built, so it's quite possible things have gotten a little loose over time. Plus I know I used to use cheap crimp on spades and tools compared to how I build today.
Worth pointing out that if all we had was noticed the tweeter "wasn't working" we'd be doing a lot of physical labor to diagnose the problem. We'd probably attempt to take the tweeter out, measure the DC resistance (useless here) and swap with the other speaker. A whole lot of effort that would tell us the tweeter wasn't the problem. I hope you already see how DATS has changed our workflow.
Mind you, at this point the tweeter may be a problem, but it's clearly not the only problem.
Disassembly
Having some idea of what to look for I took the front panel off. The leads to the tweeter came off almost immediately. Inspecting the connection to the rear jacks I discovered the ground wiring was loose internally. The first issue would explain why the first time I ran the frequency plots (which I lost) the tweeter was so far down. The second would explain the high impedance. So, 3-4 hours of reworking the connectors and finding and asjusting the tweeter spades and I get to this before and after pic:
Your eye should first notice that these images have only 1 hump in the bass section. That's because the before and after measurements were taken with the front baffle off, but the drivers still attached. Notice the after picture shows about an Ohm drop everywhere. The tweeter impedance is also looking a lot better. Great!! Perhaps the problem is solved?
I reassemble the speaker, retest the final impedance. It looks good. Now lets do Audiolense again (but REW or OmniMic would also be fine):
Badabing, badaboom. Problems solved. Now that both speakers are measuring identically I know the rolloff is now due to speaker toe-in. These ring radiators are directional.
The moral to this story is that an impedance measurement tool like DATS V3 or rigging up your own version for Room EQ Wizard is a tool you should use liberally. It can save you a lot of time and effort and get you to the right answers faster than trying to do this without them.
