Impedance Issue

ani_101

I am having an issue with the impedance dropping down to 2 ohms. Is there anything I can do to raise the impedance? I am planning to re-do the crossover in any case, but this is the one I could get close to a flattish response, but can't get the impedance up.

ani_101

The PCD file - Please give any suggestions I can try to get the impedance up or anything else regarding the overall crossover.

https://drive.google.com/open?id=0B2bbY-YB6ikpMmlvVTcydG9PRU0

kennyk

I'm using a pad, what drivers are you using? Throw in zobel. That may help.

ani_101

Sb19 tweeter, rs100-4 mid and m8n woofer. The in box driver responses are pretty wonky, but this I what I got to work with.

ani_101

The box is also a bit oversized at 60 liters for the m8n, but it looks decent with a low tune, but can't turn it up too much. Should be decent enough though for apartment living.

kennyk

What does the tweeter look like with 2nd order and do you need a filter?

Mzisserson

Hi Ani,
   The mid and the tweeter may have a bit too much overlap. The soft roll off makes it look like they are in parallel and the Z will drop.

Even though the mid has a peak, in its own impedance at thise frequencies, the total Z will be lower than the lowest (tweeter). 

I would re-think whats going on. I wonder, too what the Z phase is doing. 

Or, leave it. If its a 4 ohm nom speaker, 2 ohm dips are common. The Wilson Alexx drops to 1.6R at 2800... Not that that is a great example, but commercial companies will not worry about dipping too low on a 4 ohm speaker. 

One last note: Despite it beibg the tweeter, that load will still draw quite a bit of power at moderate levels and cause power compression. Mind the resostors in both the tweeter and midrange circuits. They may get toasty at suprisingly low levels. 
  

ani_101

I would think it is a 4 ohm speaker. The and tweet are 4 ohm nominal and the woofer too dips close to 4 ohm and there is no scope for padding the mid. 

I'll beef up the resistors

ani_101

What is z phase and how do I find it?

Mzisserson

ani_101 said:

I would think it is a 4 ohm speaker. The and tweet are 4 ohm nominal and the woofer too dips close to 4 ohm and there is no scope for padding the mid. 

I'll beef up the resistors

That works. In my 4-way between the 5FE120 and ND28 there is a drop to about 3.2 at 3K. I was gobsmacked when doing some high power testing and burned out a 5W resistor.  

Lol listening to SRV at Blues Club levels (I typically listen fairly soft, but like to push the designs to see how they fare) when I see magic smoke start to come from the back of one speaker (the side he was soloing on...X-overs were mounted to thw back). I dived for the amp power button, too late... the resistor was burned black and when I cut it out measured 1.2K insted of 5.6R. 

So went 21 years of never blowing a speaker or circuit. 

A 12 watter works dandy, though it still can get too hot too touch at those levels. 

greywarden

Yeah the first thing I noticed was that the knee on the top of the mid response is probably  causing the dip in impedance, maybe a tank would work?

ani_101

Sorry, didn't understand the knee on the mid. The tweeter impedance is dipping to 2 ohms which is taking the minimum system response to 2 ohms. Theid might be contributing, but looking at the tweeter

ani_101

The red tweeter impedance on the tweeter filter section is going down to 2 ohms

Mzisserson

From the pics, the tweeter is closer to 3 than 2. The total impedence is just about 2. If you look at the midrange, it will still have output in that 8K-is range, which means BOTH drivers are still in-circuit. I would wager the Midrange is contributing to the lower impedence. 

It seems like splitting hairs, but 2.7R vs 2.1R is quite a bit if difference (30% ish) when that low to begin with. 

Look at the Z data in the tabs of PCD, you will get a better idra of real numbers there, too. 

jhollander

Here are few ideas.  Not optimized

ani_101

Thanks John. I will give this a try.

Question: during a few of the earlier iterations, I could get the woofer and mid to blend, but the tweeter / mid had quite a ragged response - I could also get the phase to overlap or get very close between the mid and tweeter around the crossover point and the impedance above 4ohm, but could not get rid of the wriggles in the FR. This crossover which I posted was the smoothest I could get the FR and reduce the ripples in the FR, but the impedance fell. How critical are the wriggles in the FR (5bd+) in the tweeter region?

I am not looking for a super clean response, but how much should I obsess over it? I think the ripples are due to the box / baffle effect

jhollander

If you are confident in your measurement set up, then equalize the peaks, do an x-o, and listen.

I would have liked to use XSim on this one as the wiggles make it hard to use PCD.  I don't like using a lot of parts so I listen before trying to flatten with more parts