I'm seriously rethinking Dayton RS drivers..

[Jim Holtz]

I've been excited about trying the RS drivers in a lower budget design but after reading the various threads and Jon's comments about super low crossovers for the RS180, I'm beginning to wonder where the RS150 should be crossed at to keep it in its optimium range.

I'd planned on using it as a mid in a 3-way but the tweets I have on hand all require a 2.5K or higher crossover. Also one of my design goals is to end up with a SPL capability in the 105 DB range to handle peaks easily. These will be speakers for my kids, BTW. The exceptionally low distortion, excellent build quality and low price are certainly attractive but not at the expense of compromising the final design. I may just consider the RS225's and RS270's for bass duty and forget the RS line as mid ranges. :cry:

Anyone have any experience with the RS150's?

Thanks!


Jim

[Sdiver2489]

Same situation here. I wll be interested to hear the replies

[tktran]

I've been following MarkK's tests in detail, and I think selecting the MOST suitable driver is not as easy as it seems.

https://web.archive.org/web/20041010171518/http://206.13.113.199/ncdiyaudio/mark/Testing_page.htm

The recent 7" shootout suggests that the RS180 is great value, and ok in a 2way as a midwoofer, but in 3 way as a pure mid, where we can let an 10" or 8" woofer handle bass all the way up to lower mids, 300Hz, and let the mid handle just ~300Hz to 3-4Khz, a dedicated midrange, like the Seas MCA15 is probably a better choice.

https://web.archive.org/web/20050326030519/http://206.13.113.199/ncdiyaudio/mark/Midrange%20test%20group/midrange_test_data.htm

[JonMarsh]

I agree regarding a three way, if you want a more conventional crossover and higher crossover frequency.

You've gotta put this in some perspective- the W18 has much better ETC plots, but then you can buy 5 RS180's for the price of one W18. The 850439 is easier to work with, and has OK ETC plots to 1800Hz, but it's over 2X the price.

The RS150 holds up better on the top than the RS180, BUT, it's still no W18- it's pretty fair up to 1600-1800 Hz, but's starting to really come unglued (IMO) by 2200 -






And is seriously spazzing out by 2400...




Let's not even think about 3K....





Now, a couple more points to keep in perspective...

If you're going for a relatively uniform power response, a 3 k crossover point makes no sense to a 6" driver or 7" driver- the dispersion is already narrowing down by 1500-1800Hz.

To keep the driver center to center distance under control for good dispersion in the crossover lobe, we want to be under 1 wavelength at the crossover frequency, but preferably around 1/2 wavelength. One wavelength at 3 kHz is about 4.5". See the problem?

Folks think I'm a little crazy with my 1200 Hz and 1400 Hz crossovers, but I have my reasons... and they translate into how the system sounds.

But I shudder to think of what some of the RS MTM's with 3 kHz crossover points must sound like. :roll:

Good engineering is figuring out how to work with the limitations and get around them- how to make a bucks worth of parts sound like $10. OK, I'm not really doing that, as I spend some of the money saved on drivers on crossovers. It's a matter of tradeoffs...

But the RS drivers do have some strong points, nonlinear distortion and bang for the buck certainly being the main ones. And they look pretty cool, too. :B

[Captain Cojo]

Would there be any advantage to empoly a silk-domed mid range, such as the Dynaudio MD 140/2 or the Morel to fill in between-say,1400 to 4k? The sensitivity is certainly there, but a 4-way is probably getting out of hand....

[cjd]

I have an MTM crossed just shy of 2200Hz (to the 27TDFC).

It sounds darned good. That is not to say it can't sound better, just that it sounds good.

I'm in-progress mostly on a revision of the crossover to drop that to ~1800Hz, in part because the design is also intended to be CC suitable, and 2200Hz has obvious issues there. In fact, the issues stem from the fact that a driver of that size really is starting to get out of its ideal range at 2200Hz. All that stuff Jon spouts, you know? It's not just stuff he makes up I think. But, I think for most folks 2200Hz is within the realistic good sound range of the RS150. Look at the results from the recent warm-up event in WI where they messed with a digital crossover to settle on their ideals - they arrived at about 2200Hz.

I have a 3-way sketched out with 225's and the 150 as mid. Not to mention the RS270/RS180/RS28A 3-way I'm working on with Mr. Bunge and occasional tips from the guru's. And some RS125/RS180/tweet 3-ways. Lots of sketching.

If you limit your design by picking a tweeter and refusing to budge on it, yeah - the RS150 probably isn't the right driver. If you look at the total package and pick your components, I think you can achieve your goals with these drivers and be exceptionally happy with the results. Given their value, I suspect you would end up being better financially off picking up new tweeters such as the 27TDFC or the RS28A when it becomes available, than finding mids. Though, there are some mid options out there I'm ignoring.

Pick your compromises and run.

C

[JonMarsh]

Would there be any advantage to empoly a silk-domed mid range, such as the Dynaudio MD 140/2 or the Morel to fill in between-say,1400 to 4k? The sensitivity is certainly there, but a 4-way is probably getting out of hand....

Well, as you mention, a four way IS getting out of hand.. (BT, DT- just ask ThomasW about some of the things we used to do...)

Some folks are using the Morel dome mid with the Extremis, to stay away from it's problematic areas in the upper midrange, and crossing to a tweeter higher up. More crossover parts, and to deal with BSC cleanly, you don't want to run the dome mid lower than ~800 Hz. If you go for two octaves, that's 3200Hz upper crossover. So, how do you get the dome mid and tweeter domes close enough together? (less than 4" Ctr to Ctr). Plus, there's the matter of getting consistent power response in the crossover region... 3 kHz is a very sensitive part of human hearing, so paying attention to these details is pretty mandatory.


I think it would be easier to get the system sounding good over a wide window using the more robust tweeter (RS28A) and stay with the lower crossover frequency (1200-1400 Hz). And skip the extra mid driver and expense. Where you might gain more is converting the design to a three way by taking the load off the midwoofers below 200-250 Hz.

But everyone knows I'm a wild man... at least, wild eyed. :B

~Jon

[ThomasW]

(BT, DT- just ask ThomasW about some of the things we used to do...)

Ah yes the 'good ol days' ..... :roll:

Now everyone's going to want to build a pair of 350 lb fullrange, faceted front, black monoliths, with removable baffle inserts. Or maybe monster 4 and 5 way TL loaded designs with supertweeters. Then they'll wonder why their 700 watt amp is brought to it's knees driving a single 10" woofer..... :E.

Oh yea and let's make them design it using a manual drafting table, and slide rule, we couldn't afford a $700 calculator in 1969. Those certainly were the 'good' ol days ...:wink:

[JonMarsh]

Ah yes the 'good ol days' ..... :roll:

Now everyone's going to want to build a pair of 350 lb fullrange, faceted front, black monoliths, with removable baffle inserts. Or maybe monster 4 and 5 way TL loaded designs with supertweeters. Then they'll wonder why their 700 watt amp is brought to it's knees driving a single 10" woofer..... :E.

Yeah, but has anyone in the early 70's ever heard a single 10" woofer go that low and that loud in the 20-40 Hz region? Course, they were almost 6 feet tall... that was a very special long Xmax CTS 10...

[Jim Holtz]

Jon,

Since I've been involved with a couple mid dome designs with Rick, I'll toss in my $.02 worth. You'll never get to the 1/2 wave length C-T-C spacing using a mid dome but it isn't that hard to get to one wave length using a smaller faceplate 3/4" dome and trimming it slightly. Better yet use a ribbon which eliminates the C-T-C spacing issue since they radiate from the complete length rather than just the center as a dome does. As far as the potential for problems crossing in the 800 Hz and 3K Hz range, careful crossover design takes care of that quite well.

Here are the results of the Iowa 2003 DIY event which I brought the JH3 Reference design that Rick and I put together. http://www.kaiaudio.com

I like mid domes a lot but I'm a bit of an image and detail freak so they fit my tastes nicely. The line arrays I have now fit even better. :lol:

Jim

Well, as you mention, a four way IS getting out of hand.. (BT, DT- just ask ThomasW about some of the things we used to do...)

Some folks are using the Morel dome mid with the Extremis, to stay away from it's problematic areas in the upper midrange, and crossing to a tweeter higher up. More crossover parts, and to deal with BSC cleanly, you don't want to run the dome mid lower than ~800 Hz. If you go for two octaves, that's 3200Hz upper crossover. So, how do you get the dome mid and tweeter domes close enough together? (less than 4" Ctr to Ctr). Plus, there's the matter of getting consistent power response in the crossover region... 3 kHz is a very sensitive part of human hearing, so paying attention to these details is pretty mandatory.


I think it would be easier to get the system sounding good over a wide window using the more robust tweeter (RS28A) and stay with the lower crossover frequency (1200-1400 Hz). And skip the extra mid driver and expense. Where you might gain more is converting the design to a three way by taking the load off the midwoofers below 200-250 Hz.

But everyone knows I'm a wild man... at least, wild eyed. :B

~Jon

[Jim Holtz]

Actually, the reason I'm limiting my choices is because I already own a pair of DX19's and Neo3's which both require a crossover in the mid 2K range or higher. This is a budget project so I'm not interested in more expensive tweets. That said, I'm not a big fan of 1" tweeters. They just don't have the off axis dispersion of a 3/4" dome or a ribbon/planar tweet.

Also the designs I do for myself require high SPL capability because of my fetish for home theater and the budget projects end up in kids hands so high SPL's are again the requirement. Even at a normal listening volume of 80 - 85 DB, the peaks will hit 20 DB higher so 105 DB isn't unrealistic for a requirement. It doesn't fit with low crossover points very well though.

Jim

I have an MTM crossed just shy of 2200Hz (to the 27TDFC).

It sounds darned good. That is not to say it can't sound better, just that it sounds good.

I'm in-progress mostly on a revision of the crossover to drop that to ~1800Hz, in part because the design is also intended to be CC suitable, and 2200Hz has obvious issues there. In fact, the issues stem from the fact that a driver of that size really is starting to get out of its ideal range at 2200Hz. All that stuff Jon spouts, you know? It's not just stuff he makes up I think. But, I think for most folks 2200Hz is within the realistic good sound range of the RS150. Look at the results from the recent warm-up event in WI where they messed with a digital crossover to settle on their ideals - they arrived at about 2200Hz.

I have a 3-way sketched out with 225's and the 150 as mid. Not to mention the RS270/RS180/RS28A 3-way I'm working on with Mr. Bunge and occasional tips from the guru's. And some RS125/RS180/tweet 3-ways. Lots of sketching.

If you limit your design by picking a tweeter and refusing to budge on it, yeah - the RS150 probably isn't the right driver. If you look at the total package and pick your components, I think you can achieve your goals with these drivers and be exceptionally happy with the results. Given their value, I suspect you would end up being better financially off picking up new tweeters such as the 27TDFC or the RS28A when it becomes available, than finding mids. Though, there are some mid options out there I'm ignoring.

Pick your compromises and run.

C

[cjd]

Jim: I'll reiterate my last comment.

Pick your compromises, and run.

Budget design? Go ahead use an RS150 and don't mind that it may be outside its ideal range. Or pick a different driver (T-B full range 4" driver?). Or something. I was merely speaking the various aspects as I saw them.

Given that I spent the entire morning in a requirements gathering/business practice grasping/whatever meeting, I was just in the mindset of analyzing and re-stating it with the hopes it might consolidate thinking.

C

[Mark K]

Well,

I think you have to be careful how you interpret the linear distortion data above 2k. I think that you might be able to use an RS150 above 2k, maybe to 2.5k or higher. This depends on youre effort in the xover. Linear distortion can be equalized out.

This 1-2k range is an area where there is probably some measurement artifact going on. If I was really interested in using an RS150 above 2k I'd put it in a test box and try to do a very high resolution FR curve in the traditional way, at 1-2m, before making up my mind on this.

While I think there are practical reasons to use something like an MCA15, you still have issues.

It depends on what compromise you want.

I tend to favor Jon's view. With some of the very robust tweeters available, pushing the tweeter xover down is a better idea.

[JoshK]

Well, as you mention, a four way IS getting out of hand.. (BT, DT- just ask ThomasW about some of the things we used to do...)

Some folks are using the Morel dome mid with the Extremis, to stay away from it's problematic areas in the upper midrange, and crossing to a tweeter higher up. More crossover parts, and to deal with BSC cleanly, you don't want to run the dome mid lower than ~800 Hz. If you go for two octaves, that's 3200Hz upper crossover. So, how do you get the dome mid and tweeter domes close enough together? (less than 4" Ctr to Ctr). Plus, there's the matter of getting consistent power response in the crossover region... 3 kHz is a very sensitive part of human hearing, so paying attention to these details is pretty mandatory.


I think it would be easier to get the system sounding good over a wide window using the more robust tweeter (RS28A) and stay with the lower crossover frequency (1200-1400 Hz). And skip the extra mid driver and expense. Where you might gain more is converting the design to a three way by taking the load off the midwoofers below 200-250 Hz.

But everyone knows I'm a wild man... at least, wild eyed. :B

~Jon

Thanks for this concise and informative explanation. I have long been googly-eyed over some three way designs employing the morel mid, but as you say it is tough to get the dome and the tweets centers close enough.

Just another thought. If you went with say a robust tweeter (RS28a or other) and then a mid like the MCA15 would there be any advantages over another midbass woofer, of course xo'ing to a woofer?

[Feyz]

The RS150 holds up better on the top than the RS180, BUT, it's still no W18- it's pretty fair up to 1600-1800 Hz, but's starting to really come unglued (IMO) by 2200 -

This may be a side point, but I got curious to see what would the ETC of SL style tone burst of an elliptic-cauer filter would look like. And simulated this filter



found at https://web.archive.org/web/20060815.../elliptic.php3

The input signal I used for the simulation was a 3-cycle 1Khz sine wave with a hamming window shaping applied to it. The resultant ETC of the filter's sum (LP -HP sum) is in this picture:



The red curve is the ETC of the shaped 1Khz tone burst used, and the yellow curve is the ETC of the output of the filter's sum. The filter's xover frequency is 900Hz. (A CSD shows the same thing also, I actually find CSD's to have more frequency resolution power on the mid and high frequencies vs shaped tone burst ETCs)

The stored energy whether on the driver's mechanical parts or the xover's electrical parts should be considered equally I think. There is the danger of falling into a steep filter's stored energy and time smearing while trying to avoid the driver's high storage regions.

Just playing devil's :twisted: advocate here....

[Dennis H]

Interesting, Feyz. Personally, since you and others on the Mad board showed how the driver's burst ETC is directly related to the frequency response - flatten the frequency response and the ETC cleans up - I haven't paid a whole lot of attention to them. Everything you need to know is on the frequency response curve -- driver not flat = not good in that region.

A question - how well does the circuit you tested sum to flat frequency response? If it's not flat, the ETC will be distorted more than you would expect from the allpass behavior of the XO.

[Mark K]

The stored energy whether on the driver's mechanical parts or the xover's electrical parts should be considered equally I think. There is the danger of falling into a steep filter's stored energy and time smearing while trying to avoid the driver's high storage regions.

Just playing devil's :twisted: advocate here....

I agree that these are less useful in the stopband, where the xover starts to dominate.

But I'm not sure if they are the same or not. In the case of a driver only, the FR changes are the cause of the ringing. In the other, the resultant FR is flat, but the phase is changing. An interesting question. Does the ear perceive ringing from a phase aberration the same way it does an inherent driver aberration, even if the time domain ringing is the same? Could the phase shift from a higher order xover be about as audible as a modest FR aberration. And maybe that's why it's very hard to hear.

[Feyz]

A question - how well does the circuit you tested sum to flat frequency response? If it's not flat, the ETC will be distorted more than you would expect from the allpass behavior of the XO.

Yes it is not very flat at the xover region, 0.9db dip to the left of the xover frequency and 1.2db peak to the right of it.

[Feyz]

Does the ear perceive ringing from a phase aberration the same way it does an inherent driver aberration, even if the time domain ringing is the same? Could the phase shift from a higher order xover be about as audible as a modest FR aberration. And maybe that's why it's very hard to hear.

That's a good point. I don't know much about how the ear and brain process received sound.

[Feyz]

A correction

found at https://web.archive.org/web/20060815.../elliptic.php3

The input signal I used for the simulation was a 3-cycle 1Khz sine wave with a hamming window shaping applied to it. The resultant ETC of the filter's sum (LP -HP sum) is in this picture:

It turns out the reason that ETC looked so bad was because that particular cauer xover circuit I had found on the 'Net wasn't designed very good. The reality turns out not to be so with well designed cauers. Please see this post of mine that I did on another thread:

https://www.htguide.com/forum/showthread.php?p=161206#post161206

[Mark K]

A correction


It turns out the reason that ETC looked so bad was because that particular cauer xover circuit I had found on the 'Net wasn't designed very good. The reality turns out not to be so with well designed cauers. Please see this post of mine that I did on another thread:

https://www.htguide.com/forum/showth...206#post161206

Well, that makes sense. A while back I tried to do the ETC plots for an actual built 2 way LR4 and the plots looked relatively good. So I was suprised to see them so bad.

Well, I guess you can forget my comment on the MAD board then...