Pe Dotw: Rs28a!

[cjd]

Wow... PE is too nice. And allll weekend too!

my wife is gonna be gruuumpy. ahh well.

C

[dyazdani]

That is a VERY good price...

[JonMarsh]

will check it out- due to lab move, no network access!

[Scott Simonian]

I saw the price and it looks about the same. How much cheaper is it this week? Doesn't look like more than $10.

[cjd]

$10 is about 21%.

That's a pretty good deal in my book!

C

[dyazdani]

$10 is about 21%.

That's a pretty good deal in my book!

C

Yep, mine too - I may have to buy a couple.

[Bent]

... ordered 3 of em.

[Drew]

This is certianly a nice discount, but I have to ask: what do these get you that you don't get with the Seas 27TBFC/G that's $32 all the time?

I'm not trying to rain on anybody's parade here, this is an honest question. I've never heard or used the Dayton, but I do quite like the Seas.

[cotdt]

the RS28A sounds smoother than the seas and if you get a good sample has even lower distortion figures

[Jim Holtz]

the RS28A sounds smoother than the seas and if you get a good sample has even lower distortion figures

I've used the RS28 and liked it but my ears aren't agreeing with yours. I prefer the sound of the H1212 to the RS28 and I'm not so sure that I don't like the TDFC better than both even though it has ever so slightly higher distortion measurements. The TFDC has just a bit more air that I think sounds more accurate, even though the distortion measurements won't back up what I hear.

BTW, the TBFC-G is the budget low distortion king based on every test I've seen. Here's one. http://www.zaphaudio.com/tweetermishmash/ The RS28's also have consistency issues that would make me choose a H1212 for the same or similar money. I believe Mark K's tests also agreed with John Krutke's in that regard. The RS28 is a nice tweeter but not the giant killer the RS woofers are, IMHO.

My $.02 worth...

Jim

[oneoldude]

The TFDC has just a bit more air that I think sounds more accurate, even though the distortion measurements won't back up what I hear.

Jim

The distortion measurements probably do back up what you hear. Increasing distortion in a reproduced signal cannot reproduce that signal more accurately. That "air" you hear is likely greater distortion. So even though the signal is less accurate, you nevertheless "like" it more.

It is like adding salt to your poached egg. Adding salt does not make the egg a more accurate egg, but you like it more with salt.

Just my .02

[cjd]

To my ears, the biggest advantage of the RS28A over the H1212 is that it doesn't have that nasty huge peak just beyond where most response graphs stop and our ears "officially" stop hearing.

Plus, the grill looks better! And it's a metal faceplate.

At its normal $47 price, I have to think long and hard about using it. At $37, it becomes a simple matter of preference.

C

[Jim Holtz]

The distortion measurements probably do back up what you hear. Increasing distortion in a reproduced signal cannot reproduce that signal more accurately. That "air" you hear is likely greater distortion. So even though the signal is less accurate, you nevertheless "like" it more.

It is like adding salt to your poached egg. Adding salt does not make the egg a more accurate egg, but you like it more with salt.

Just my .02

Yes, this is the everlasting debate of objectivity Vs. subjectivity. I think I fall somewhere in the middle. You start with measurements but you "always" have to end up with your ears because measurements are not what you listen to in the end. Measurements give you an idea of what you can expect but few speaker designs reach their potential without tweaking by ear after the crossover has been summed flat. Different drivers/topologys will sound different with the same frequency response. That's about all thats left of the "magic" of speaker design with todays sofisticated software. I might add it is IMHO what sets speaker designs apart.

I'd also submit that live music, which should be our reference, contains distortion. So how much of that distortion is removed/modified in the recording process? If a driver adds some of it back, it might be deemed most accurate subjectively even though it might not be the measurement champ.

Of course, I also readly admit to being in love with line arrays and ribbon tweeters. :T YMMV

Jim

[oneoldude]

I think that arguing classic "objectivist" or "subjectivist" positions is of little benefit to our hobby. Polarization occurs and little useful information is passed.

This often occurs when those arguing use concepts and words in a thoughtless way. Like the objectivist who essentially says, "This sounds good because it meets my measurement criteria." Or the subjectivist who says, "This is accurate because I like the way it sounds." Both are usually wrong.

I believe that improvement in communication comes when the objectivist admits that a system that meets his measurement criteria may not sound good and when the subjectivist admits that what he likes may not be accurate reproduction. Once that happens, real communication can begin.

Otherwise, we end up with:

"When I use a word," Humpty Dumpty said in rather a scornful tone. "It means just what I choose it to mean - neither more or less."
"The question is," said Alice, "whether you can make words mean so many different things."
"The question is," said Humpty Dumpty, "which is to be master - that's all."
Lewis Carroll

[Dennis H]

Oneoldude, lighten up a little. You're preaching to the choir about distortion and objective vs. subjective. The TDFC is a VERY low distortion tweeter, among the very best at any price. I doubt very much that Jim's preference for it is because he likes the sound of distortion so your initial premise is flawed.

[noah katz]

"I'd also submit that live music, which should be our reference, contains distortion."

Only if people playing music through speakers is "live".

[cotdt]

"I'd also submit that live music, which should be our reference, contains distortion."

Only if people playing music through speakers is "live".

instruments have a lot of linear and nonlinear distortion of their own is what he means

[joecarrow]

To my ears, the biggest advantage of the RS28A over the H1212 is that it doesn't have that nasty huge peak just beyond where most response graphs stop and our ears "officially" stop hearing.

Plus, the grill looks better! And it's a metal faceplate.

At its normal $47 price, I have to think long and hard about using it. At $37, it becomes a simple matter of preference.

C

Does this metal faceplate eliminate the need for a tweeter isolation chamber? Right now I'm getting ready to buy the parts for a Modula MT, and the price difference looks like about $25/pair more for the RS28a version (with the sale price), and I'm unsure if I should take advantage of the sale and get the Dayton tweeters.

If it would simplify cabinet construction and there is some sonic benefit, I guess it's worth the extra money.

I really can't decide which way to go!

Dayton RS28a Pros:
-metal faceplate (no need for isolation?)
-no high frequency breakup
-slightly more accurate?
Dayton RS28a Cons:
-$25/pair more expensive (budget is kind of tight, but $25 is reasonable)
-unit to unit variability (even if they are both OK, will it be a good match?)

Seas H1212 Pros:
-Good reputation for quality control
-More extended top end
-cheaper
Seas H1212 Cons:
-high frequency breakup (possibly audible?)
-needs internal enclosure due to plastic faceplate

Seas H1189 (27TDFC) Pros:
-Good reputation for quality control
-Most high frequency extension
-no metal cone breakup
Seas H1189 Cons:
-Slightly higher distortion
-No grille (I hate when tweeters get pushed in, and I move a lot)

So, any thoughts? Anybody with a lot of experience with any two who can make first-hand comparisons?

[cotdt]

Does this metal faceplate eliminate the need for a tweeter isolation chamber? Right now I'm getting ready to buy the parts for a Modula MT, and the price difference looks like about $25/pair more for the RS28a version (with the sale price), and I'm unsure if I should take advantage of the sale and get the Dayton tweeters.

If it would simplify cabinet construction and there is some sonic benefit, I guess it's worth the extra money.

I really can't decide which way to go!

Dayton RS28a Pros:
-metal faceplate (no need for isolation?)
-no high frequency breakup
-slightly more accurate?
Dayton RS28a Cons:
-$25/pair more expensive (budget is kind of tight, but $25 is reasonable)
-unit to unit variability (even if they are both OK, will it be a good match?)

Seas H1212 Pros:
-Good reputation for quality control
-More extended top end
-cheaper
Seas H1212 Cons:
-high frequency breakup (possibly audible?)
-needs internal enclosure due to plastic faceplate

Seas H1189 (27TDFC) Pros:
-Good reputation for quality control
-Most high frequency extension
-no metal cone breakup
Seas H1189 Cons:
-Slightly higher distortion
-No grille (I hate when tweeters get pushed in, and I move a lot)

So, any thoughts? Anybody with a lot of experience with any two who can make first-hand comparisons?

RS28A looks much better.

[joecarrow]

RS28A looks much better.

Do you mean it's more visually pleasing?

[Mazeroth]

Yes, I think it's more visually pleasing as well. I have a set of the RS28s and absolutely love them. Granted, I've never used the two aforementioned tweeters before. :P

[JonMarsh]

Does this metal faceplate eliminate the need for a tweeter isolation chamber? Right now I'm getting ready to buy the parts for a Modula MT, and the price difference looks like about $25/pair more for the RS28a version (with the sale price), and I'm unsure if I should take advantage of the sale and get the Dayton tweeters.

If it would simplify cabinet construction and there is some sonic benefit, I guess it's worth the extra money.

I really can't decide which way to go!

Dayton RS28a Pros:
-metal faceplate (no need for isolation?)
-no high frequency breakup
-slightly more accurate?
Dayton RS28a Cons:
-$25/pair more expensive (budget is kind of tight, but $25 is reasonable)
-unit to unit variability (even if they are both OK, will it be a good match?)

Seas H1212 Pros:
-Good reputation for quality control
-More extended top end
-cheaper
Seas H1212 Cons:
-high frequency breakup (possibly audible?)
-needs internal enclosure due to plastic faceplate

Seas H1189 (27TDFC) Pros:
-Good reputation for quality control
-Most high frequency extension
-no metal cone breakup
Seas H1189 Cons:
-Slightly higher distortion
-No grille (I hate when tweeters get pushed in, and I move a lot)

So, any thoughts? Anybody with a lot of experience with any two who can make first-hand comparisons?

you know, 10 or more years ago we would have killed to have tweeters that worked as well as any of these units. Some have been close, but they cost a small fortune in comparison (Focal TC120dx2, for example).

Interesting comparison you're listing above.

My comments:

RS28a- very neutral sound for the most part to my ears, like a slightly more dynamic version of the SS98000 with more dynamics in the lower part of the range, but a little less air on top. The FR measurements show why. I don't hear anything that sounds like upper range breakup modes being excited and causing IM down low. I still use a backing enclsoure with this, as I do even with the Hales Trancendence, which has a much massive aluminum front plate. I can see how folks might figure it's not necessary, and perhaps it isn't. Another reason I like this part is that it is compatible enough with the XT25 and SS98000 in response profile and impedance curve to be dropped into existing designs with little or now modification.

H1212- no critical listening experience, only a little measuring experience to date. Have dropped one into a Modula MTM in place of 27TDFC with good results, but only have auditioned with my Sony HT reciever (haven't had big two channel system up in a year, due to having setup that part of living room as construction area for crossovers and test fixtures) (and then having no time to construct!) I'm wondering if the upper range breakup does make any problems or not with IM into the base band. just speculation.


Seas H1189 - I'm not the major fan boy of soft dome tweeters, but this one and the Millenium Excel are the one's I'd use if there were no good hard dome tweeters. It's not as flat as a Milleinum Excel, and takes a little work to get a neutral voicing. Still, it may be arguably one of the best deals in the DIY world, in terms of cost performance ratio that's pretty unbelievable. Wish I had these back in the 70's... wish I had these back in the 80's. Wish I had these back in the 90s....
Cheap to repair, too.

Not on this list, and not usuable in most two way configurations, but very interesting to me for new three way designs, is the

Seas H1283- only usable from 2 kHz up, it's response is very smooth to 30 kHz, has a fairly well controlled upper resonance at 35 kHz, and excellent off axis performance. This is the one I want to try in a waveguide. I'm hoping it's possible to further reduce the distortion in the critical 2-6 kHz area, and keep most of the pluses of this driver. If I'm daring, or just to annoy folks, I'll try a bandpass design as Avalon often implements, and see if we can kill the 6 dB 35 kHz peak without messing anything else up. Could be a reall sweetheart with the right midrange or midwoofer.

[jkrutke]

To my ears, the biggest advantage of the RS28A over the H1212 is that it doesn't have that nasty huge peak just beyond where most response graphs stop and our ears "officially" stop hearing.

Plus, the grill looks better! And it's a metal faceplate.

At its normal $47 price, I have to think long and hard about using it. At $37, it becomes a simple matter of preference.

C

IMHO, ultrasonic breakups are not audible. There's been "studies" and "white papers" by a few companies that say they are, but the ones I've seen are nothing more than marketing BS to sell a product.

But take note of this: I've tested lots of RS28's. Half of them DO have a large, high intensity breakup node and half of them don't. A couple had two breakup nodes, one at 20 and one at 28kHz. Some have a 7dB dip between 10 and 20kHz and some only have a 1 or 2dB dip there. For a while, I had lots of people requesting that I test their tweeters to see if they got a "good pair". I've stopped doing that since I got bored with it, and I got tired of giving people the news that their $100 pair of tweeters didn't perform well.

The RS28 is a tweeter constructed of $50 worth of excellent internal parts, but put together with $0.25 worth of quality control. There's a lot of potential with this tweeter however, so I intend to revisit it in 6 months to a year to see if things have gotten under control.

[Dennis H]

The RS28 is a tweeter constructed of $50 worth of excellent internal parts, but put together with $0.25 worth of quality control. There's a lot of potential with this tweeter however, so I intend to revisit it in 6 months to a year to see if things have gotten under control.

Yup, there's the rub (oops, bad pun. ) Where's Dave Ralph when we need him? He has more experience tearing these apart than anyone else. Can the voice-coil centering be tweaked, like Jon does with the Hales, or do the alignment posts fit so tightly that what you get is what you get?

[cotdt]

Yup, there's the rub (oops, bad pun. ) Where's Dave Ralph when we need him? He has more experience tearing these apart than anyone else. Can the voice-coil centering be tweaked, like Jon does with the Hales, or do the alignment posts fit so tightly that what you get is what you get?

... and has anyone experimented with removing the grills and diffusor?

I've also went through around 8 RS28A's and ended up with the two best and returned the rest. Some of them had large chips in the magnet! But the two I have now doesn't have the big dip in the frequency response from 12-20kHz that the other RS28A's have and are very flat. Soon I will see just how low these good ones can play.

[cjd]

IMHO, ultrasonic breakups are not audible.

Well, I accidentally erased my post ... the short of it is: I'm weird, I know. But I am sensitive to some ultrasonic frequencies. And the TBFC is one that in my experience fatigues me more than most. My RS28A's don't. Nor do the 27TDFC's.

And, I really wish the RS28A were more consistent than it seems like it is. What can you do.

C

[TacoD]

And, I really wish the RS28A were more consistent than it seems like it is. What can you do.

C

It is a good thing that the RS28A exists, but there is no free lunch imho. Maybe the Seas tweeters are not as good, but you get where you paid for.

If you have to buy 8 RS28A to get a good pair, I think there is something wrong with the product .

[Mark K]

Well,

In general, I agree with the premise that the RS28a seem to have more variability. So far though, I've only tested two out of ~15 that were not, what I would consider, usable. At any frequency.

However, in terms of reaching down low, it's only 50-50 that a particular unit would get to 1.3-1.4k

I've only tested one pair of 27tbfc/g's so I don't have overwhelming data for that. One the other hand, the other 27 series have been very good as well.

The design of the RS28 is outstanding; but I agree that it's very murky where the low end limit is for any particular unit.

Can you hear stuff above 20k? Well, there's not really convincing evidence, although it might be possible in some very limited circumstances. As Jon mentioned, IM distortion. Try playing a 20 and 21k signal out of two separate amp/freq generator combo's and you will hear a clear 1k tone. So it's certainly possible.


arty:

:evil:

:rx

sorry, was letting my sons pick smileys...

[jkrutke]

the short of it is: I'm weird, I know. But I am sensitive to some ultrasonic frequencies.

Well, as long as you don't hang upside down during the day and go around town looking for blood at night, I'll let it slide.

In general, I agree with the premise that the RS28a seem to have more variability. So far though, I've only tested two out of ~15 that were not, what I would consider, usable. At any frequency.

You're the only person I know who has tested more RS28's than I have, so I think I'd trust your consistency numbers more than mine.

Have you noticed that the better performing ones have slightly pointier impedance peaks? Not higher in amplitude, just less rounded over. I'm sure this isn't coincidence.

Yup, there's the rub (oops, bad pun.)

I've actuall only had one rub, and it was a pre-QC version fresh off the pallet from Usher. I promply sent it to Mark K to validate some really wacked-out looking harmonic distortion and spectrum noise plots.

I think the bigger problem was crooked voice coils. Not as in concentricity, but axis angle. You could tell by removing the coil and looking at it. The ventilation holes in the former were closer to the dome on one side. Every RS28 I opened up had this problem to some extent.

I've heard some other guys say the coil was pinched between the top plate and the pole piece on one side, a serious concentricity issue in that case. I haven't seen that myself though.

[Bent]

Does anybody know a quick and dirty way to ensure they have been sent decent RS28's?
a fast bench test of some sort?

[cotdt]

Does anybody know a quick and dirty way to ensure they have been sent decent RS28's?
a fast bench test of some sort?

Have you noticed that the better performing ones have slightly pointier impedance peaks? Not higher in amplitude, just less rounded over. I'm sure this isn't coincidence.

Maybe the pointiness of the impedance peaks will do it! Impedence testing I guess...

Never tried it myself but you can probably open up the tweeter and rebuild it better, if you have the skills.

[joecarrow]

Maybe the pointiness of the impedance peaks will do it! Impedence testing I guess...

Never tried it myself but you can probably open up the tweeter and rebuild it better, if you have the skills.

Maybe not- not if the voicecoil is crooked like John described.

I think a REALLY quick and dirty way to tell if one of these is any good would be to run a sine wave sweep at moderate volume from 1 khz to 10 or 20 khz. It's fairly easy to tell the difference between a clean, low distortion, sine sweep apart from one that contains a lot of harmonic and IM distortion. With a clean one, you'll hear a single note slowly increasing in pitch. The distorted signal will sound like a few notes at once, with the pitch increasing from low to high a few times during the single sweep.

I know that this works for sound cards on computers (it was a real eye opener when I heard this the first time), and I assume that it's audible for any component that adds distortion.

To the gurus- is that reasonable?

To conclude my story, I think I've settled on the Seas H1212 because it seems to fit my needs best. This is going to be a tight budget project, and any gamble with needing to ship tweeters for replacement just isn't going to help the budget.

Thanks for all the commentary

[JonMarsh]

Most of the "bad" RS28a aren't probably "that" bad. For example, an average LPG 26T is NO where near properly aligned (In my experience), and on a sine test tone you can audibly align them for clean sound sans any rubbing. OTOH, maybe most of what I've gotten are what people would call "good" drivers. The ones that don't measure that hot have higher high order THD and IM below 1400 Hz, but they're also being pushed pretty hard. An RS28a playing at 100 dB with a 25/50/25 mix of 800 Hz, 1 kHz, and 1200 Hz will set your ears on edge, and has pretty high peak levels becuase of the beat components adding at peaks.

~Jon

[Dotay]

Thanks for the heads up. :T I just bought five of them to build my planned Natalie P, Modula MT, and yet to be determined center. Mark K had said in the past that he'd be willing to test out the tweeters if one would pay for the shipping costs. Hopefully he's still willing to do this so I/we can get an idea if anything has changed with a more recent sampling.

[JonMarsh]

Well, I have a bucket full of them I'm going to be scrutinizing closely also, so we shall see what we shall see. Too bad there isn't someone else closer than Mark or I to you- both of us being in Northern CA.

[kgveteran]

Picked up two today.This will round out my LCR.saved 24.00 total over the shielded version.

So, now tell me there is a sonic differance between shielded and non...... :M

[cjd]

Sonic differences should be none. Response profile will be slightly different due to impedance differences. Don't remember how much different.

C

[Dotay]

Too bad there isn't someone else closer than Mark or I to you- both of us being in Northern CA.

cjd - What kind of measurement equipment do you have at your disposal? I used to make it to Chicago quite frequently but I haven't been in a while...might be time for another mini road trip.

[rumatt]

BUMP.

It's now 9 months later. Have there been any updates on the RS28a's quality control?

I'm having a hard time starting out a complete DIY HT set of speakers based around the RS28 if it has QC issues... that is unless I can easily test for the problem and return the bad ones.

But a lot of folks are using them with the stickied designs and seem happy with them, so I guess they can't be that bad.

[cjd]

QC issues were on the low end only I think, largely affecting crossovers below 1.6k.

I'm not aware that any recent ones have been as heavily tested, but I know PE did step up their efforts.

[ThomasW]

Each tweeter is now individually tested and there's a sticker indicating it's tolerance pasted on the magnet.

[Scott Simonian]

So now lets say Im doing a 'Modula' project and I get a pair that are ~.5-1dB. Would I be able to compinsate for that in the project xover or would I just "know" that one side is maybe a .5dB too hot?

Just curious.

[ThomasW]

I wouldn't get all that concerned. There are variations in other components drivers, caps, inductors, etc, etc.

And I'd bet your ears aren't matched with-in 0.5dB...:wink:

If you can hear an in-room difference of 0.5dB with the completed design, then solve the problem.

[fjhuerta]

That's good to know, Thomas. I wasn't sure what the sticker was for.

I built my Natalies a month ago, but have been listening to them pretty closely, just to be sure of what I hear. I'll post a review shortly.

[Scott Simonian]

And I'd bet your ears aren't matched with-in 0.5dB...

Ah...you got me there, Thomas.