Living Room Bookshelves - sim #1

**JonMarsh** · 13 July 2016, 16:09 Wednesday

Cabinet size? Baffle width? Where did the driver data come from? Is it showing the normal baffle step in response due to the specific cabinet mounting? Planning to do anything about that bump in the mid woofer response at 1 kHz? Why not use a conjugate network on the tweeter Fs so you don't have the bump up in the transition band response at Fs?

Just some thoughts...

**scottvalentin** · 13 July 2016, 17:20 Wednesday

Jon I have simmed the baffle response and box tuning in an 8.5 wide by 14 high baffle box of 11 litres. For the tweeter, I traced the SB Acoustics data and then simmed the baffle response in Response Modeler and exported the minimum phase files for PCD. For the DS135, I used the Parts Express FRD files, simmed the baffle step in Response Modeler and exported the minimum phase for PCD.

I could do a notch filter to tame that bump at 1k which seems to be both a baffle item and a particular response of the driver. I was trying to use as few parts as possible in the crossover.

I don't know what a conjugate network is otherwise I'd have tried it! Thanks for looking though, I'm trying to learn a bit so any explanation of what I've done (since I don't really know, I clicked buttons til the graph looked good), where it's wrong, where it works, would be helpful.

Thanks!

**JonMarsh** · 13 July 2016, 18:52 Wednesday

A conjugate network is an LCR network used to flatten the impedance rise at a driver Fs; so the L, C, and r are tuned so that the net impedance curve seen by the filter is at the nominal driver value as measured in the main part of it's operating range (for example, typically 6 ohm for a so-called 8 ohm woofer).

A bump like that midrange response issue can be tamed by an LCR shunt network, too. At some point, though, you have to ask yourself, "Do I want to fix these issues with more crossover parts, or spend the money on better drivers?

Though the filter transfer function isn't plotted, what I suspect, based on the Dayton supplied curves and the obvious dip in the 1400-1600 Hz area, its that you've compensated for that in the network filter, with a rising response in that area, but then that bumps up the 1kHz region stuff, which can include slight peaking due to the baffle diffraction effects.

I'm not a huge fan of paper cone drivers; overall, I think the RS150P would be flatter and easier to work with, and the extra Sd would give a bit more bottom end weight and mid bass definition. OTOH, I'd be inclined in that price range to go with the Dayton RS150-4, which has pretty flat response to 2500Hz, no dip, and a fairly flat impedance curve, which is the sign of a good motor with low inductivity modulation (low IM at higher frequencies).

I could make a lovely paper cone driver recommendation, too, the Motus Audio UH130PW1, with a very low distortion underhung motor....

This is actually about a 6-1/4" driver, I have a pair here, even thinking about them for a "budget" high end 3 way as the midrange, or an MTM.

Expensive parts (Meniscus and Solen have them) but with some care you could probably do a very nice LR2 with not many parts.

**scottvalentin** · 20 July 2016, 12:45 Wednesday

Alright Jon,
I did make some changes, and have since modelled the RS150-4 with the Dayton ND25. I know, what about the RS28, there's about 100 designs with those 2 drivers. Well in Canada, the RS28's are about $75 each from Solen, and the ND25 is $22. The RS150-4 is $47, (or the difference in cost from the RS28 to the ND, so $135 for drivers seems reasonable.

Quick summary, these are simulated in an 8w X 14h X 10d (outside dim) giving 10 or so litres of internal volume. Tweeter is offset 1.5" from centre.

Here are the system graphs and responses and crossover schematic. Lots of parts on the tweeter to drop the hump at 1500-1800 while not rolling it off too fast... this seems to be a problem no matter what tweeter I use since the diffraction models all seem to get a large 2-3 db bump around here (or just below).

These were done without referencing any similar projects only because I wanted to see what I came up with first on my own. Then hopefully by comparing and putting in values for completed designs I can learn alternative methods to what I did for manipulating the response.

Also, at $200 apiece for the Motus drivers...I'm hoping to do a pair for that much, but they do look like nice drivers!

**scottvalentin** · 20 July 2016, 13:56 Wednesday

I looked at a few designs using the RS150 and cleaned up the woofer section, and also significantly cleaned up the tweeter section with fewer parts and probably a better reverse null. Even started putting a BOM together and put some parts in a cart! This will come to around $200 a pair actually including wood... BUT my project points are gone for this year especially as I still have two large subs to veneer... I should probably finish those off!

**scottvalentin** · 20 July 2016, 15:39 Wednesday

Here is the updated simulation with fewer parts:

**JonMarsh** · 20 July 2016, 16:39 Wednesday

Quick question- are you accounting for driver time origin offset? (the acoustic of the woofer is acoustically behind the tweeter).

Also, that little ND tweeter doesn't go very low- it's a part better suited (when running it naked) to a three way. How much would an SB Acoustics SB26STAC-C000-4 cost you up there? That's another possibility in a relatively budget tweeter that has the range and Xmax needed for a two way. That ND25FN is really more intended for a three way, given that the range needs to include the transition region in the crossover- it should probably be used from around 3500 and up. The Fs is about 1500 Hz. Most tweeters for two ways have an Fs around 600 to 800 Hz.

People have put these into waveguides to increase the loading capability and reduce distortion and run them down to 2500Hz crossover points.

**scottvalentin** · 20 July 2016, 17:33 Wednesday

Jon to your first question, I haven't set up the acoustic offsets in PCD. I understand that will affect the phase response, but my thought was that getting the frequency response was more important. Writing this, seems to me that they go hand in hand, so I should probably set the offset up first!

I guess I should have looked at the Fs of the tweeter first - but this is why I did this so that I can start getting a practical understanding of all the things I've been reading about for the past year.

Jon the SB26STAC-C004 are running at $51 each from Solen. I'd say that is do-able. Again, this will be next year's project, so hopefully if I quit working on speaker sims at work and we get some wins here, my bonus will be good and I will have some funds to spare!

Thanks for taking the time to look and give some feedback - really appreciate it.

**JonMarsh** · 20 July 2016, 18:51 Wednesday

You're welcome. I have measured data on the SB26STAC-C004 and they look like pretty good parts for the money.

**scottvalentin** · 22 July 2016, 11:59 Friday

John,
Any experience with the Creative Sound Solutions LD22 or comments based on its datasheet?
Here is the PDF at this link with measurements on a 7.5" baffle and a 10.5" baffle (PE cab sizes I think).

Creative Sound

http://creativesound.ca/pdf/LD22.pdf

These would run me $35 each or $20 each (CAD!) if I order factory seconds that have a slight crease in the dome.

Living Room Bookshelves - sim #1

Living Room Bookshelves - sim #1

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment