Danny, regarding your posts on adding a ribbon tweeter to a full range driver.

Dracule1 · « **on:** 19 Jun 2010, 08:28 pm »

I think the Von Schweikert Unifield 3 - touted as an augmented full range - uses a 1st order crossover for the ribbon tweeter to compliment the Fostek driver, but frequency response does not have the roller coaster peaks and valleys. There is a gentle broad dip going down to ~5dB centered around 5kHz. How do you think VS managed this?

http://www.avguide.com/review/von-schweikert-audio-unifield-model-three-loudspeaker-tas-198?page=2

Danny Richie · « **Reply #1 on:** 19 Jun 2010, 09:49 pm »

It is easy to find a measuring height where everything is somewhat in phase and with minimal cancellation like the small 5db dropped area at 5kHz. No doubt about it though, if you move the mic (or your ear) up or down a few inches then you'll find much greater levels of cancellation.

Dracule1 · « **Reply #2 on:** 20 Jun 2010, 03:23 am »

The measurement was actually done by the reviewer, not by VS, which is may be more credible than the manufacturers measurement as long as he knows what he is doing. The examples you showed on prior thread showed much worse FR near the xover point, despite your attempts to get the best FR. Also the Unifield 3s midrange is 5'' compared to the 3" driver you used and the ribbons center is quite a ways from the center of the 5" driver. So I'm surprised VS was able to get such smooth response based on what you were saying.

BTW, do you know which Fostex driver he is using?

Danny Richie · « **Reply #3 on:** 20 Jun 2010, 09:50 pm »

What the reviewer did was not accurate at all. It was an RTA measurement. So it included all room reflections. It then takes an average. It could have had a 10db swing each way averaged out to get that line. A lot of smoothing was used also.

A real MLS measurement (gated) with no room reflections would have been just as bad or worse than my example that I measured.

Off axis responses are needed to really get the true story.

*Scotty* · « **Reply #4 on:** 21 Jun 2010, 12:34 am »

Here is a quote from the Unifield 3 product description regarding the crossover design

Quote

Crossover: Global Axis Integration Network (our proprietary design that eliminates â€œbeamingâ€) set at 100 Hz and 8 kHz (trailing slope @ -6dB, 300Hz/4kHz overlap). Optimized for flat off-axis response and phase consistency. Very high quality parts are used from France, Germany, and the U.S., including proprietary film-foil capacitors, low distortion inductors, film-foil resistors, and high purity copper wiring.

This might mean that the crossover slopes are steep enough minimize the off axis response problems that a 6dB/oct slope at 8kHz would bring to the design.
Scotty

Danny Richie · « **Reply #5 on:** 21 Jun 2010, 01:14 am »

Quote

This might mean that the crossover slopes are steep enough minimize the off axis response problems that a 6dB/oct slope at 8kHz would bring to the design.

Nope, the first order slope will not be steep at all. Vertical off axis cancellation will clearly be there.

Quote

our proprietary design that eliminates â€œbeamingâ€) set at 100 Hz and 8 kHz

That's funny. 100Hz is omni-direction but there proprietary crossover will eliminate beaming.

It's pretty easy to eliminate beaming at 8kHz too. That wavelength is about 1.7" long. So use a driver with a diaphragm that is smaller than that.

*Scotty* · « **Reply #6 on:** 21 Jun 2010, 01:52 am »

The way I see it beaming is the least of their worries. I always thought that beaming was another word for controlled vertical dispersion.
Scotty

planet10 · « **Reply #7 on:** 21 Jun 2010, 05:10 am »

Quote from: Dracule1 on 20 Jun 2010, 03:23 am

BTW, do you know which Fostex driver he is using?

FX120.

Has anyone listened to this loudspeaker?

dave

Dracule1 · « **Reply #8 on:** 21 Jun 2010, 08:18 am »

Dave, I have listened to them. The midrange is exceptional - transparent, holographic, full, detailed, and natural. How much is the FX120?

Dracule1 · « **Reply #9 on:** 21 Jun 2010, 08:30 am »

Danny, thanks for clearing that up. I'm a novice at speaker measurements. I know RTA stands for real time analyzer, but I didn't realized FR obtained by RTA includes room reflections and averaged over multiple samples. I really don't know what MLS (maximum length sequence?) means. Would appreciated an explanation.

DFaulds · « **Reply #10 on:** 21 Jun 2010, 11:29 am »

Quote from: Dracule1 on 21 Jun 2010, 08:18 am

Dave, I have listened to them. The midrange is exceptional - transparent, holographic, full, detailed, and natural. How much is the FX120?

$113 @ Madisound

http://www.madisound.com/store/product_info.php?manufacturers_id=131&products_id=282

Danny Richie · « **Reply #11 on:** 21 Jun 2010, 03:24 pm »

And RTA can work several ways depending on the software or settings. It is a continuous signal that allows the user to take a snap shot at any given time or set a number of averages. You can typically set it to take 20 or 40 samples and average them.

A MLS response is a short duration signal that allows the signal to be viewed over time. Looking at the impulse response, the user can gate the time window and start just as the signal reaches the mic and stop it soon after and before any reflections can arrive.

Taking gated measurements on and off axis will tell you exactly what the speaker itself is doing.

Oddly, I often see RTA measurements made with both speakers playing at the same time. So you not only get room interaction, coupling, and cancellation effects, but the a compounded effect of having two sources playing the same signal in that same room.

Dracule1 · « **Reply #12 on:** 22 Jun 2010, 03:01 pm »

Danny, have you ever tried to make a multiway with a fullrange driver, similar to the Unifield 3?

Danny Richie · « **Reply #13 on:** 22 Jun 2010, 06:37 pm »

Quote

Danny, have you ever tried to make a multiway with a fullrange driver, similar to the Unifield 3?

Yeah, I have actually.

My experience with it showed that the best place for any additional tweeter was on the back side of the speaker. It added that little increase in the sense of air and space without adversely effecting the response. I have also integrated passive and powered low frequency drivers.

I also did the crossover design work on this speaker: http://www.azzolinaaudio.com/page3/page3.html

And this one: http://www.azzolinaaudio.com/page2/page2.html

Dracule1 · « **Reply #14 on:** 23 Jun 2010, 02:37 am »

Where is the full range driver in the horn speakers? So tell us Danny, any opinion on the sonics mating a full range to a high crossover point tweeter and low crossover bass driver.

Danny Richie · « **Reply #15 on:** 23 Jun 2010, 03:06 am »

Quote

Where is the full range driver in the horn speakers?

Yes, and it is open in the back.

Quote

So tell us Danny, any opinion on the sonics mating a full range to a high crossover point tweeter and low crossover bass driver.

On the top end there is no real "matching". A rear firing tweeter just adds a little ambience. Crossing it the bass driver was not a real problem either. It would have been real anemic without it.

HT cOz · « **Reply #16 on:** 23 Jun 2010, 03:07 am »

Ill take a stab at this,,, see if I'm learning anything.

Take a look at this pic

Now look at the Neo2X

Ok so here is a speaker that covers much of the important frequences with one driver and then crosses over to the fantastic Neo3 for a great top end. If you employed a super wide band driver and super tweeter that wide driver is likely to start beeming before the crossover and will kill your off axis room response which provides a great deal of the speakers sound. Add a pair of servo subs and you have your 3 way speaker.

So I would think what is the point? What is to be gained and what is lost?

Danny Richie · « **Reply #17 on:** 23 Jun 2010, 03:16 am »

Nice points HT cOz.

I have yet to hear a full range (or wide band driver) have detail levels and resolution levels as good as an average dome tweeter. And the smaller diaphragm dome tweeter will always have better off axis response.