And Now For Something REALLY BIG

Status
Not open for further replies.
Jump to Latest
#401
buzzforb said:
......but should have good transient attack and clarity in midbass, which most people are talking about when they talk about bass
Click to expand...

Yup! Thats what I think of as accuracy

buzzforb said:
For the horns, i am choosing Iwata, becaus ei can diy one failry easily and will basically amount to cost of wood and time. Jmlc vs tractric is just a decision. Kohlbreck, the guy who measured the jmlc for lynn, is now listening to another horn. That tells me that the jmlc is not the undisputed end of th eline. Also, i fyou look at the ETF article i linked, while the JMLC and Kulwegen horns measured well, they were not chosen first in terms of sound. Go figure.
Click to expand...

'just a dicision'?????? Its driving me loopy 😀 As for someone listening to something else........I wouldn't read that much into it. Going on the general populace of DIYAudio changing 'is just something we do' - sometimes changing up, sometimes down, and sometimes sideways!!!

As for the results/measurements they are interesting but as the competition was of different CD & horn combinations, none of which I'll probably ever own, it goes to prove nothing is, or should be, decided by measurement alone. As you said - Go figure!!

buzzforb said:
Truth is, we won tknow till we try
Click to expand...

Amen brother, amen
 
#403
It goes without saying that to align the voice-coils sounds like the right idea as far as time alignment is concerned, but that is only part of the story. While that approach seems simple enough in the end you have to look at what you have created by doing that. Think about that horn lens or waveguide sticking out in front of a bass driver directly below the waveguide. You have now created a large physical surface that is quite irregular in shape usually on top of the bass driver that the mids and lows are now going to reflect off of. The reflective properties of a waveguide hanging out there like that are going to be a real problem if you look at this as a reflective surface and not just as two completely disconnected devices. You won't have problem with the high frequency response as it is unaffected but you surly will with the lower frequency dispersion. If things were only simple things could be done so easy. In today's active electronic arena it would seem that it would actually be smarter to use a simple time delay to align the two acoustical centers rather than to physically move the waveguide forward to align the voicecoils which could be anywhere from a few inches on a short waveguide to almost a couple of feet with a long horn.
 
#404
Thanks guys

Pano said:
That will get you close. Then align the phase.
Click to expand...

How would one go about ascertaining the phase difference?

Kindhornman said:
In today's active electronic arena it would seem that it would actually be smarter to use a simple time delay to align the two acoustical centers rather than to physically move the waveguide......
Click to expand...

That all makes sense. So with the CD behind the mid/woof, how do you go about ascertaining the delay needed to time align the drivers? I assume (again!) that you time align the CD to the mid/bass - don't know why, I just do 🙄
 
#408
Here is an example of how I do it with HOLMImpulse.

Let's suppose that you've built a crossover where the acoustic slopes (important, acoustic!) end up as 4th order Bessel at 800 Hz. What would the response, phase and impulse of an ideal Bessel 4th order look like? You can see them in the graph below.

You will notice that LP impulse peak lags the HP peak by 20cm. If your drivers don't measure like that, they are out of alignment/phase. On this graph the impulse is in dB mode, otherwise you'd hardly see the LP peak.

When I first started doing this I made the mistake of trying the line up the LP and HP impulses. Phase was never right doing that. One day it occurred to me to see what an ideal filter would look like - and that's what you see below. Of course real speakers are rarely that ideal, so you have to fudge, tweak and use your judgment. It's not all that difficult.
 

Attachments

  • bessel-4th-800.png
    bessel-4th-800.png
    39.5 KB · Views: 467
#409
thanks Pano.

So i'll need a mic and HOLMImpulse.

My crossover understanding is limited as I've only used 1st order.

I know that 1st order = 6db/octave - phase unchanged? 2nd order = 12db/octave - phase 180º? 3rd order = 18db/octave - no idea! 4th order = 24db/octave - no idea! I understand low-pass, band-pass & high pass.

I've also been reading up on active vs passive & the 3rd way - PLLXO

Pano, you mention 'acoustic slopes (important, acoustic!)'. Dumb question I'm sure, but why do you highlight acoustic slope?

tnx for the help
 
#412
JRKO said:
A penny for your thoughts ....
Click to expand...

impedance interfere with passive xo, and changes its function and resulting response, phase etc

and compression driver/horn combo is even worse, with multiple impedance ressonance peak

active is at least free of those speaker impedance issues
still have to deal with acoustic response tho

I guess those problems might have made the CD's bad reputation of sounding very harsh much worse than it deserves
and no doubt it will sound unbareably harsh if the xo is not functioning properly

on the good side, I don't think you need any fancy amplification for a CD
probably a simple low power chip amp will do fine, and be very simple

the remaining other drivers might still work with relatively simple passive xo
 
#413
tnx tinitus.

So the passive crossover is passive in both directions - subject to the drivers impedance anomalies - just because it is designed to give the correct crossover freq. & slope , it may not due to the driver attached.

An active crossover is essentially line level (before amplifier) and not subject to driver feedback, stamping its exact filter upon the signal before amplification with no speaker level interaction.

This should also be true of a PLLXO?
 
#414
JRKO said:
tnx tinitus.

So the passive crossover is passive in both directions - subject to the drivers impedance anomalies - just because it is designed to give the correct crossover freq. & slope , it may not due to the driver attached.

An active crossover is essentially line level (before amplifier) and not subject to driver feedback, stamping its exact filter upon the signal before amplification with no speaker level interaction.

This should also be true of a PLLXO?
Click to expand...

Yep, this is the case AFAICT. I've tried both PLLXO and active via mini-DSP and it's much better than passive speaker-level Xover, at least at my design skill level. 🙂

IG
 
#416
Little update on my project. Decided to get some Altec 416's. Got them used for a good price. Also got some Altec 288 compression drivers. That leaves just the horn profile. I will begin my build in a couple of weeks, so hope to have some more extensive impressions soon:up:
 
Status
Not open for further replies.
Top Bottom