BG Radia Neo drivers vs conventional tweeters and mids

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,
Since it looks like Radia's Neo drivers are not going to be offered to us diy'ers any more, I'm starting to think I should pick up a pair of Neo3's and Neo8's or 10's before it's too late. I've never heard any of these planar drivers before. I've heard Magnepan's but those are so different in concept I'm not sure if they're really comparable. I'm wondering how the Radia drivers compare to conventional cone and dome mids and tweeters, and also the HiVi planars.

I've seen Zaph's measurements on the Neo3's and Neo10's so I know they have low distortion and good frequency response. I'm not sure about thermal compression though. I read Linkwitz's article on sine burst testing with the Neo3 but it appeared that it wouldn't apply to the Neo3 crossed over at a higher frequency. I'd also be looking at trying these drivers in their own enclosures instead of open baffle. How well do they work "boxed" compared to conventional drivers? Also, while the frequency response looks usable, do they suffer from sympathetic resonance compared to conventional drivers?

thanks,
Dan
 
I called BG and one of the employees happened to pick up the phone even though it was already a done deal. The person I spoke with said there is a company in Europe that basically makes clones of the BG drivers. He said they were very good. He also said there is a Chinese company that makes knock offs that are junk.
 
Meniscus still has a few in stock and so does GR Research. I just order a pair of the Neo3's with the deeper back cup and the Neo8's from Danny at GR Research. The Neo8's sound interesting. They are an 8 ohm version, but not the Neo8S. They have thinner traces and what look to be more traces running between the gaps in between each set of magnets. So the sensitivity is the same as the 4 ohm version but with the higher impedance so apparently a bit better efficiency and hopefully better for thermal compression as well. We'll see how they work out.

I'm still interested in any subjective impressions of these drivers compared to domes/cones though, especially experiences running them in an enclosure.

Dan
 
I think the main reason the ribbons sound the way they do is because of their directivity.
IE, if you use a waveguide to achieve a similar response pattern, it will sound similar.

I have BG planars here, ribbons, domes, etc.

One thing I like a lot about the NEO3 and NEO8 is the depth; you can put it in spots that other speakers will not fit.

I'd be curious to see if an array of three or four of the cheap neo tweeters that PE sells would sound similar to a NEO3. Four of these in an array would have a sensitivity of about 95dB, in a package that's a similar size as a NEO3, for under sixty bucks:

Dayton Audio ND20FA-6 3/4" Neodymium Dome Tweeter
 
I have looked at all the sites when google lists them and so far no one has any. There are some Neo10's and some Neo8's like the links posted earlier. I might just try and get the Neo10. But they are more expensive compared to the Neo8. And the Neo*S does everything I want as a mid and does a little better than the Neo10 from what I have read.

BUT I might have to settle.
 
Maybe they also have a better time behaviour?

Agreed. Particularly with "true ribbons."

There's a way to prove this mathematically:

First off, we have to ask ourselves, "what do we want in the time domain?"
I think that ideally we would want a radiator that plays the signal, and then returns to it's starting point in a timely fashion.

IE, if my radiator is playing a signal, I want it to play that signal then return to it's original state.

Next up, we have to figure out what is "a timely fashion."

I think that timeframe is defined by frequency. For instance, if you're playing a signal that's 10khz, "a timely fashion" would be 1/10,000th of a second, or 0.1ms.

You can see that this time period will get longer as you go lower in frequency. By the time you get down to 1khz, that time period is one millisecond, by 100hz it's 10 milliseconds.

If all of that makes sense, you'll start to see why ribbons, particularly "real" ribbons have an advantage. Their CSD is superior to conventional drivers.

On the upside, the news isn't "all bad" for conventional drivers, because one simple way to get better CSD is to simply use a smaller driver, or an array of small drivers.
Dayton_RS28A4-CSD.gif

Here's the CSD for a well made conventional tweeter, a Dayton RS28. We can see that it meets our criteria on the low end, but as frequency goes higher it's not capable of "returning to zero" as quickly as we need it to. I think that's part of the reason that ribbons sound "airy." If you really want to wince, take a look at the CSD for a compression driver on a horn.

By the way, this sounds like an argument for using small drivers for bass, but I don't feel that way. If you look at it mathematically, tweeters need a clean CSD way more than woofers do.

And, obviously, a lot of this will depend on what you're listening to. For instance, a well mic'd classical recording of an orchestra will have complex high frequency information. The type of music I tend to listen to doesn't (EDM.)

Oddly enough, heavy metal has a lot of this info in the music too, because there's a lot of complex guitar work in heavy metal.
 
Last edited:
I have looked at all the sites when google lists them and so far no one has any. There are some Neo10's and some Neo8's like the links posted earlier. I might just try and get the Neo10. But they are more expensive compared to the Neo8. And the Neo*S does everything I want as a mid and does a little better than the Neo10 from what I have read.

BUT I might have to settle.

Danny at GR Research did have more Neo8's left after I had ordered mine although that was more than a week ago. The Neo8's he had were custom though, they were 8 ohm rather than 4 ohm but same sensitivity according to him. I'm still trying to wrap my head around what the real world difference is. To achieve 8 ohms I think they use thinner traces but loop them through the gaps between the columns of magnets more times. That could give more drive force for the same current yet the thinner traces might lead to the onset of thermal compression sooner. We'll see how they match up to the Neo3's.

It looks like Meniscus still lists the 4 ohm Neo8 as in stock too.
Dan
 
Here's the CSD for a well made conventional tweeter, a Dayton RS28. We can see that it meets our criteria on the low end, but as frequency goes higher it's not capable of "returning to zero" as quickly as we need it to. I think that's part of the reason that ribbons sound "airy." If you really want to wince, take a look at the CSD for a compression driver on a horn.

If you look at the CSD's of a few true ribbons on Zaph's site, they don't look any better and typically worse. He didnt test some of the higher end ribbons like the Raal's though. I haven't heard enough true ribbons to have much of an opinion of the sound and I've never heard a double ended planar magnetic like the Neo's. BG's buyout kind of forced me to finally buy the Neo drivers to start a project with them before it's too late!
Dan
 
If you look at the CSD's of a few true ribbons on Zaph's site, they don't look any better and typically worse. He didnt test some of the higher end ribbons like the Raal's though. I haven't heard enough true ribbons to have much of an opinion of the sound and I've never heard a double ended planar magnetic like the Neo's. BG's buyout kind of forced me to finally buy the Neo drivers to start a project with them before it's too late!
Dan

One "neat" thing about good CSD is that I haven't seen any evidence that you DON'T want it.

IE, I think there are some reasons that one might *want* a speaker with moderate amounts of distortion. I find that distortion makes things sound louder and euphonic.

But bad CSD? I can't see any upside.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.