|
|
| Re: 4 pi speakers + 3 pi subwoofers, pics & measurements [message #81367 is a reply to message #81366] |
Sun, 31 May 2015 09:48   |
 |
Wayne Parham
Messages: 18783
Registered: January 2001
|
Illuminati (33rd Degree) |
|
|
The waveguide is ours, I knew that. We're the only ones that make it. But I wasn't sure about the driver or crossover. Now that I know the driver is the DE250, in my mind, that leaves only crossover and measurement system. And I doubt the crossover is doing it. It looks like you have it right. You can always double-check it, but I'm sure you've already done that.
My guess is the measurement system is at fault. Not the microphones, but the system, itself. Seeing two microphones show the same trend tells me it's probably not the microphones. The only thing that's really left is the system. The input circuitry could cause this, or any number of other things internal to the system.
How does it sound? If there is as much rolloff as would seem to be shown in the measurements, it would sound very muddy and dull, almost like no tweeter was connected. Does it sound that way?
|
|
|
|
|
|
| Re: 4 pi speakers + 3 pi subwoofers, pics & measurements [message #81374 is a reply to message #81371] |
Thu, 04 June 2015 10:19 |
|
Wayne Parham
Messages: 18783
Registered: January 2001
|
Illuminati (33rd Degree) |
|
|
You're right, there's more mutual inductance when they're positioned that way, close together and in the same plane. They are coupled together, and act sort of like a transformer. This causes the HF crossover to shift some, and even causes some of the energy from the woofer circuit to bleed into the tweeter circuit, and vice versa. But this woud be most evident in the crossover range, and even worset case, it's a fairly small effect. I wouldn't say it's trivial, but the coupling isn't so strong that it could create a 10dB boost or drop.
What we're seeing is HF rolloff, and by a pretty significant amount. I think it's probably the measurement system, and that the charts aren't giving an accurate picture. If they are, if there is something wrong with the crossover or drivers, then the HF is so low that these speakers would sound very dull, almost like no tweeter is connected at all.
|
|
|
|
| Re: 4 pi speakers + 3 pi subwoofers, pics & measurements [message #81376 is a reply to message #81374] |
Thu, 04 June 2015 10:53 |
vandevoordekoen
Messages: 18
Registered: January 2013
|
Chancellor |
|
|
Hello,
I can't say the speakers sound like there is no tweeter at all. Neither can I say that the HF are "sparkling", but maybe that's not necessary.
To test the measurement system, I measured the HF signal 3 times: with and without USB audio card + using 2 different mics (ECM8000 and TASCAM DR-1 (TASCAM is not great but no garbage either)). The following overlap gives an idea:
Red line = TASCAM mic directly connected with laptop (mic-in)
Green line = TASCAM mic connected to the USB audio card (M-audio Fast Track)
Blue line = EMC8000 mic connected to the USB audio card (M-audio Fast Track) WITHOUT .cal-file
Red and Green line are approximately the same. The Fast track isn't the weak link. 
The ECM8000 measurement without .cal file gives approximately the same result. So the .cal file is corrupted IMHO.
Just to give an idea, these are the same measurements, WITH .cal file applied for the ECM8000 mic
Looking at the first overlap, I see a 5db to 10db roll-off. But this might be normal? Aaaargh, it's driving me a little crazy, honestly. 
Would it be the amp?
|
|
|
|
|
|
| Re: 4 pi speakers + 3 pi subwoofers, pics & measurements [message #81383 is a reply to message #81352] |
Fri, 05 June 2015 22:09 |
tom-m
Messages: 56
Registered: December 2009
Location: Texas
|
Baron |
|
|
To help narrow down where the problem is, I would ask if you have a set of speakers other than the Pi4. If so, measure another set of speakers. If they show the same HF roll off, then it would be in your measurement gear, or audio gear. If no roll off, then it is in the Pi4 speakers.
Tom
|
|
|
|
Posting Rules
Members
Search
Help
Register
Login
Get a T-Shirt!
