Siegfried discovers added dimension

johnk... · « **Reply #20 on:** 1 Feb 2007, 03:41 pm »

This is a measurement of a NeO3 PDR I just took on an 11" x 17" flat baffle, dipole baffle configuration. NO REAR CUP. Theses samples were purchased several years ago when I was first developing the NaO because I thought they might be just the ticket. I went through 3 pairs of them because the response was so off the manufactures data. I finally quit returning them and just wrote them off. The blue trace is the tweeter, the red an LR4 target 2K Hz at 85 dB, about as good as might be obtained with the samples I had. Now maybe things have changed but this is what I experienced and how the samples I still have from a few years ago measure today, Feb 1, 2007.

Anyone think this is acceptable? I'm looking for reasonable flat response and 90 to 92 dB sensitivity. I don't see it with these samples of the Neo3 PDR. I'd be happy to see some measurement on newer samples.

Danny Richie · « **Reply #21 on:** 1 Feb 2007, 04:25 pm »

Dang John,

I have never seen any look like that. Were you using one of their face plates?

I stock a lot of these tweeters and I have been a dealer for them since they were released. I have never seen responses like that.

I recently used them in the OB-7 and OB-5 kits.

See response on 10db scale:

Same response on a 5db scale:

It is within +/-1.5db once you are about 10 degrees off axis as the peak in the top end smooths out making the overall response even smoother.

Crossover points:

The tweeter only has a simple 3rd order network on it. It also does have the rear back cup.

I also used it in my O-3 omni surround speaker. Below is the on axis response.

These were designed to be listened to slightly off axis (10 to 20 degrees). In this design it uses a simple second order network on the tweeter.

Here it is on axis and at 10, 20, 30, and 40 degrees off axis. 20 degrees off axis is pretty damn smooth.

More info on those designs can be seen on our web site.

http://www.gr-research.com/kits/ob-7.shtm

http://www.gr-research.com/kits/o3.shtm

johnk... · « **Reply #22 on:** 2 Feb 2007, 12:53 am »

(JPK) Not using nothing but the metal slotted grils and diaphram. No face plates, not rear cups, not nothing. Looks just like the picture on page 201 of the PE 2007 catalog mounted on a flat, 1.4" think baffle (no caviety behid the tweeter either). If your measurements are as good as mine are

then something has changed over the past few years. These are samples 5 and 6. the other 4 looked the same.

Davey · « **Reply #23 on:** 2 Feb 2007, 01:23 am »

I measured one of those tweeters about a year ago....with the back off. John K's measurements are almost identical to what I saw. There is nothing unexpected about this response. When you remove the back the damping on the diaphragm changes quite a bit and the under-damped response is the result.

You'll see much the same thing if removing the rear chamber from a variety of conventional dome tweeters also.

Davey.

Danny Richie · « **Reply #24 on:** 2 Feb 2007, 05:35 pm »

The rise in the lower end is due to the back cup being off, but the peak at 14kHz and the fall there after is likely do to not having it mounted in a face plate.

ctviggen · « **Reply #25 on:** 2 Feb 2007, 06:04 pm »

This is why frequency responses can be misleading. You've used two different techniques for taking the same measurements, and based on these measurements, you've reached two different conclusions. (Where "you" refers to the posters who have examined the tweeter.)

johnk... · « **Reply #26 on:** 2 Feb 2007, 07:19 pm »

(JPK) Nothing confusing about it from my end. The suggestion was why not uses the Neo3 for a dipole tweeter. The siimple answer for me is, and remains until I see differently, it doesn't work well in dipole mode. What it does with a rear cup isn't relavent to the dipole problem. No confusion at all.

konut · « **Reply #27 on:** 2 Feb 2007, 08:34 pm »

Quote from: johnk... on 2 Feb 2007, 07:19 pm

(JPK) Nothing confusing about it from my end. The suggestion was why not uses the Neo3 for a dipole tweeter. The siimple answer for me is, and remains until I see differently, it doesn't work well in dipole mode. What it does with a rear cup isn't relavent to the dipole problem. No confusion at all.

What if I were to use 2, with the cups on, back to back?

Danny Richie · « **Reply #28 on:** 2 Feb 2007, 08:51 pm »

Quote

What if I were to use 2, with the cups on, back to back?

I think it would work great and be really tough to beat.

scorpion · « **Reply #29 on:** 2 Feb 2007, 11:07 pm »

This is a debate over the Pope's beard. I can't for instance see that the NAO II has any tweeter operating in dipole mode, that is open back. Davey has made the point. In a recent Danish comparison only the Seas Excel Millenium tweeter was considered the equal of BG Neo3 and there it was compared to units from ScanSpeak, Focal, Fountek, AMT and Seas all up to more than six times as expensive as BG Neo 3.

/Erling

Kevin Haskins · « **Reply #30 on:** 2 Feb 2007, 11:59 pm »

I've measured a bunch of these also. I'd say there was something wrong with the production run you where measuring John. The ones I've measured are easy to work with, back cup on or off.

wikin · « **Reply #31 on:** 3 Feb 2007, 02:28 am »

Hmmm, I can only think of 2 possibilities for these sets of conflicting data.
JohnK's samples were from much earlier production run where the manufacturer could be on their steep learning curve, hence the inconsistent and quality issues in terms of response flatness. Later on the manufacturer honed their production processes and obtained higher Cpk (process capability).

Davey · « **Reply #32 on:** 3 Feb 2007, 02:34 am »

I don't think you guys are fully understanding the critique and performance standard John K. is applying to this tweeter.

His measurement is absolutely correct. There is nothing wrong with the "production run" or anything else. The ones I tested are fairly current ones and also the ones that Zaph tested.

Here's a quote from Zaph's page that confirms the response:

"This is a very clean and natural sounding tweeter. Note: these were tested with the rear chamber in place. Later, with a different pair, I also tested with the chamber removed. The tweeter was far too overdamped in that case, and I recommend leaving the rear chamber installed."

John meant to say underdamped vice "overdamped" but the result is clear.

Davey.

Al Garay · « **Reply #33 on:** 14 Feb 2007, 04:50 am »

Hi Davey,

Curiosity got the best of me, I ordered a new pair of BG NEO3s to build the Adire Dakini (new name for Dan Wiggin's DDR). I already have plenty of Extremis woofers from my LCR home theater.

Once I'm done, I will be glad to have you, Mac and Andy over. Kevin, you will be invited as well, perhaps I will have the 2641 ready to play. It's only taken me a year or has it been two...

Al

chrismercurio · « **Reply #34 on:** 25 Feb 2007, 07:53 pm »

I have been lurking here a while and wanted to bump this thread a bit.

http://www.troelsgravesen.dk/Tweeters_dist-test.htm

Troels measured the NEO3 with and without the back cup and found that the distortion levels actually went down with the cup removed. His curves also don't agree with the published curves by JohnK, and fall more in line with what BG publishes.

I'm wondering if Danny or others have gone back and listened/measured with the rear cup removed?

C

johnk... · « **Reply #35 on:** 26 Feb 2007, 07:09 pm »

(JPK) I'm still ready to accept that the samples I have are different. They are several years old and are what they are. I just stick a mic in fornt of them and what come out comes out. Since everything else I measure looks pretty much like what every body else get for other drivers I see no reason to doubt the measurements of the samples I looked at or curently have.

I'd be pleased if newer samples measured better.

Zero One · « **Reply #36 on:** 27 Feb 2007, 12:40 pm »

I just found this thread and thought I might a my two cents worth. I have experimented with rear firing tweets on both bass reflex and open baffle set-ups, in both cases the drivers were heavily modded full range 4 inch drivers with whizzer cones with 10 inch sub augmentation below 50hz.

In both cases the tweeters are just small 1 inch car tweeters and are aimed slightly upwards and towards the room corners, they are crossed at about 6000hz, but the front drivers are not filtered at all.

The results are wonderful, very easy to listen too with wonderful sounstaging and presence, they can be tuned via an L pad, though I have them turned right up.
I figure that much of the info above say 6000 is really more ambient rather than primary sounds and since harmonics are a mixture of relections etc having the tweeters rear firing and using the room boundaries give a more natural effect, well thats my thought anyway. Doesn't really matter why, in my setup it just works a treat, well worth a try.

pixelpusher · « **Reply #37 on:** 3 Mar 2007, 06:27 pm »

Hi, My first post. Hope this works! Just stumbled on to this discussion. I have just measured Neo3pdr with and without back cup from 1 meter on a 5'X7' baffle. The red is with the cup. I used the new surface mount plate. Next I will measure on my dipole baffle which is 18"X48". I was discouraged to read some of the comments regarding Neo3pdr use as a dipole. But this measurement doesn't look too bad.
Mike

johnk... · « **Reply #38 on:** 4 Mar 2007, 11:30 am »

pixelpusher,

When did you purchase your samples of the Neo3? They certainly look better than my samples with regrads to the resonance peak. But they still seem to require considerable equalization due to the rising response. And if the level at 2 to 3 k with the cup is indicative of the quoted sensitivity then without the cup it's looking like the sensitivity is down in the 83 to 84 dB range.

pixelpusher · « **Reply #39 on:** 4 Mar 2007, 07:09 pm »

They were purchased from GR Research just a few weeks ago. Yes there is a price to pay in sensitivity. I am intrigued by their use as dipoles and hope that they will impart some of this rear tweeter effect.