Pondering an IDS-25 Clone (sort of) for HT

[JonMarsh]

This is a spin off thread, to avoid thread pollution.

Here I'm going to discuss and present ideas for creating a basic clone of the IDS-25, with a few twists, to use for GF's HT front mains.

First, when considering clones of the IDS-25 system, it's a good idea to be familiar with the original! This is the main page, and much more info can be found from other linked pages:

http://www.roger-russell.com/ids/ids.htm

Russell either had a custom driver built, or was very careful in selecting the driver which he uses. It's not a simple challenge, as one is looking for wide frequency extension, low distortion, and freedom from any high Q break up modes in the top end. I like his approach for the equalizer, but then, like Russel, I'm kind of old school myself, and would do it pretty much the same say - including balanced inputs and outputs available, low noise, low distortion op amps, film caps, etc. :T


Most of the IDS clones I've read about, including Javier's, are somewhat flawed due to their choice of driver, usually made on the basis of cutting cost.



BTW, this Vifa TG9FD-10-04 is purported to be the driver used in the IDS-25. Verified on several sites.

http://www.parts-express.com/vifa-tg9fd-10-04-3-1-2-glass-fiber-cone-full-range--264-1064

It does use a copper cap to limit impedance rise and upper end distortion, and the cone resonances are well damped.





With the shown driver Fs, they will need some significant LF EQ to perform below 100 Hz. BUT, the sensitivity gain should be high; not really all that much power. Xmax is rated at 2.6mm- not very high.


Still, I'm tempted as a giggle to build some new mains for my GF's family room HT with these, see how they sound. Its a fair sized room, about 18' by 30'.


It's certainly worth running some numbers and modeling. But I would recommend using these like the company below, with a some extra LF reinforcement. I like the physical design because it appears well optimized for achieving good Sd relative to the driver area, and of course, you can pack these together. Front panel design and routing will be an issue- might have to go to phenolic panels, maybe even aluminum.


Interesting... let's see, 50 at $19 each is still pretty close to $1K. Still not an impulse purchase.


Good first step would be to buy a few and see how the distortion measures at low drive levels, and how that would scale with an array to 90 dB level baseline.


*****************************


Geez, another distraction from current projects. Just ordered 4 for testing.

I ran the numbers, and using a configuration of 25 on each side, with five sets of drivers wired in series and then paralleled (a 5 x 5 matrix) with the net efficiency gain and the starting sensitivity of 84 dB/W (W, not 2.83VRMS; remember, these are 4 ohm drivers) the net total system sensitivity calculated out to 97.98 dB.

Now, looking at the intrinsic frequency response, as well as the low Xmax, I don't personally like the idea of EQ'ing out the bottom end... but crossing these at 100 Hz to an LF Line source (which only need be one LF/sub at the floor, and one at the ceiling), things should be pretty hunky dory using a 12 dB electrical high pass and a 4th order low pass. I can imagine all kinds of configurations that might work well for that, such as those Wavecor woofers, or the 10" RS265, combined with the SS W26/00-0 PR.

now, OF COURSE, the top end is not going to sound like a ribbon tweeter or Beryllium dome top end... but I bet there will be some things the system does overall in a very interesting manner. Interesting enough to try out and if it works dedicate these to GF's front HT mains? I think so...

So, for analysis purposes, what we'd be interested in is, with 4 or 5 wired in series, with a 2.83VRMS drive (4V Pk, 8V Pk-Pk), what does the distortion look like? Because that's about what we'll see at nearly 98 dB output from the finished array. So, that copper cap and other things may be pretty important here.

I'll build a 4 driver test box and see how this looks.

BTW, as usual, I'm thinking that if I eliminate the LF EQ network, then everything else (BSC, HF EQ) may likely be accomplished passively. Possibly. Who knows...

[CraigJ]

Hi Jon,

For the less informed (me), how is the IDS clone different from something like the Murphy Corner Line Array; https://www.trueaudio.com/array/index.htm, or even a CBT array (aside from price)?

From the IDS-25 brochure; "Eighty one feet of high grade Cardas chassis wire is used in each system to interconnect all of the drivers." :thud:

[augerpro]

I'm liking Huerta's LA too. Jon, if bass response isn't a priority, have you considered the TC9FD-18-08?

[oneplustwo]

The Murphy Corner line array was really interesting to me for a while. It takes up very little practical space and much of the theory made sense to me. But when I dug deeper, I didn't get the sense that the sound quality was all that great. The CBT is really cool (I heard them in a less than optimal room at burning amp here in SF a couple years ago). Not only the look, but they were really dynamic, very detailed, and delivered on the huge sweet spot. Even in the weird room. One of the subs was acting up if I remember correctly that day, so the low end wasn't great, but I would expect either a dedicated sub or infinite baffle sub to supplement down there anyway.

CBT has been out of stock for a while though and although cool looking, isn't super practical from a space utilization perspective. Then there's the bi-amping part of them that isn't terribly problematic but does add cost/complexity.

And so... waiting with bated breath to see a Jon Marsh line array!

[---k---]

I have a problem with the "Murphy Corner Array". It is a line array placed in a corner.

[JonMarsh]

Hi Jon,

For the less informed (me), how is the IDS clone different from something like the Murphy Corner Line Array; https://www.trueaudio.com/array/index.htm, or even a CBT array (aside from price)?

From the IDS-25 brochure; "Eighty one feet of high grade Cardas chassis wire is used in each system to interconnect all of the drivers." :thud:

Well, fundamentally, it's not. I have only heard tangentially once or twice about the Murphy Corner line Array- I checked that out yesterday evening, too, and it's very similar to the idea I've been considering for GF's mains line array, as her current speakers are in the far corners up high.

This corner placement at the boundary gives you some advantages in bass performance, not just the boundary reinforcement, but greatly reducing secondary paths and nulls. This is something the IDS-25 will have some issues with, as I've seen pictures of it setup online. The thing is, you could build a cabinet optimized for the corner placement, and still use it elsewhere, with changes in the EQ. It's something to ponder.

Where I suspect the Murphy corner array may not be optimized (or is optimized in a skewed way) is the choice of driver- the ND-90 is clearly a direct clone of the Aurasound NS3-193- so it will have the long throw capability of the underhung motor, and as that driver is tuned, more extended bass response than the other small drivers. BUT, the motor doesn't appear to use copper in the gap to control inductivity modulation, so I expect the upper midrange and high frequency distortion to be poorer, even though it is no where near running out of excursion- they haven't implemented copper in the gap in this version of the NRT motor the way the NS12 and NS10 and most of the other NRT woofers do.


Witness the impedance curve of the ND-90 (which is essentially identical to the NS3-193 posted curves):




Versus the Vifa used in the IDS-25:




I suppose I should still buy a few samples of the NS3-193 and see how it actually measures- I expect the top end to be a bit more rough with the metal cone, too. The Vifa has something of a peak on axis, but is pretty nice at 30 degrees- if it's well controlled between 15 and 30 degrees, then that will dominate the room sound.

[JonMarsh]

I'm liking Huerta's LA too. Jon, if bass response isn't a priority, have you considered the TC9FD-18-08?

Good question. It has some of the same characteristics, apparently having the same chassis and motor.



And the price is almost half the TG9FD.

But then look at the response curve- ignoring cost, which would you want in your system? And below a certain threshold of pain ($1K roughly) cost was not a factor for me.



Not everyone would make the same decision, and you might come close to the same subjective sound quality after putting more EQ in the mid and treble area of the TC9FD. But those are physical cone issues, as MarkK and Sigfried point out, creating energy storage and linearity performance issues, and I'd rather just avoid them to begin with, as much as possible. But I can see the attraction for someone wanting to steer more towards the budget side of things (like Javier) and with a multi-band EQ of some kind. I think this would be a good driver choice compared with the comparably prided Founteks.

[fjhuerta]

Hello Jon,

I'm a bit worried about this thread, since I've done the IDS Clone with the Founteks, and I've done many of the things you are about to try - but how could I try to tell you the things I encountered without sounding a bit... silly? I mean - YOU are the expert on these things!

Nevertheless, I'll share my findings, and I hope they help you with what will surely become THE IDS-25 killer.

(For those of you who don't know about my build - here it is: http://techtalk.parts-express.com/sh...n-a-line-array)

First: My inspiration comes directly from Roger Russell's 25 driver IDS-25, and the Murphy Corner Line Array (http://www.trueaudio.com/array/downl...%20Project.pdf), a 25 driver array Dayton ND-90. I credit both designs with the idea of actually going ahead and building the array.

Several comments:

On why I picked the Fountek FE-83. Price, and luck. I was going to build a small Bluetooth bar (!!) with 2 of them - but ordered 8, just for the sake of it. Then I decided a small array with 8 drivers would be fun, got 8 extra. Then I thought that maybe 12 per side would be a better idea. And so on, and so forth. Getting to 51 drivers (50 + 1 extra) was somewhat easy. I was aware of the lack of copper in the motor, and the somewhat not even FR, but since the initial idea was to build a cheap and cheeky speaker, I didn't care. Also, FR didn't look that bad, so for $6.90 I could probably not do wrong.

With the shown driver Fs, they will need some significant LF EQ to perform below 100 Hz. BUT, the sensitivity gain should be high; not really all that much power. Xmax is rated at 2.6mm- not very high.

Yes. There's sensitivity gain. The thing is, there's not that much gain below 180 Hz in my design. I don't know if it's related to baffle or driver size or both.

Roger did use quite a bit of EQ, and I did, too! Here's my EQ graph.



This is the resulting FR, at my listening seat.



As I mentioned before, I'm using the room correction curve suggested by Dr. Sean Olive here.

Without active EQ, I'd find it terribly difficult for these speakers to work, if at all. A line array comb filtering effect kills treble and boosts bass somewhat not evenly. You can see this effect on Murphy's PDF regarding his array.

Good first step would be to buy a few and see how the distortion measures at low drive levels, and how that would scale with an array to 90 dB level baseline.

Good thing there are 25 drivers sharing the load... distortion really became a non-issue with my design, even at high levels! Still, I'll need a subwoofer because around 40 Hz I have quite a lot of it... at high levels.

Roger Russell mentions distortion as very, very low, and the Murphy array mentions it can hit 110 dB at 50 Hz with low distortion levels, so I don' know if this is exclusive to my design. Since the Murphy array is installed on a corner, maybe it has quite a lot of bass boost and EQ levels aren't as high as mine.

Now, looking at the intrinsic frequency response, as well as the low Xmax, I don't personally like the idea of EQ'ing out the bottom end... but crossing these at 100 Hz to an LF Line source (which only need be one LF/sub at the floor, and one at the ceiling), things should be pretty hunky dory using a 12 dB electrical high pass and a 4th order low pass.

I'll be integrating my arrays with 2 woofers per side, RS-270s (actually, I'll use my 3 way design's bass module) at 100 Hz. Yes, EQ does add distortion in my design.

now, OF COURSE, the top end is not going to sound like a ribbon tweeter or Beryllium dome top end...

It has been my experience that the top end, when properly EQ'd, sounds better than at least my SB-29 tweeters.

Comb filtering should destroy highs. But it doesn't, at least when listening 1m or more away. Active EQ does wonders here, and I assume, sharing highs between 25 drivers lowers distortion quite a bit.

I'm thinking that if I eliminate the LF EQ network, then everything else (BSC, HF EQ) may likely be accomplished passively. Possibly. Who knows...

I'm sure you could... but I'd be downright painful... I would decidedly not recommend it, especially considering the losses involved.

But I can see the attraction for someone wanting to steer more towards the budget side of things (like Javier) and with a multi-band EQ of some kind. I think this would be a good driver choice compared with the comparably prided Founteks.

The multi-band EQ is IMHO a must, a deal-breaker. I must stress this: IMHO. But if I did everything again, I'd definitely get the Vifas.

[JonMarsh]

Thanks for all the info you've shared- I haven't gone through all of it yet, but should during the weekend.

I'm thinking to use my Metric Halo LIO-8 that's been updated to ULN-8 for developing a test EQ for the system- then developing an analog EQ box that will deliver the same curve, so it can be used in any signal path at line level. I have the full DSP and analysis setup, so coming up with something workable shouldn't be hard.

Another quick comparison I put together is highlighting some data differences between candidate drivers... one wouldn't make final decisions based on this, and one might need to double check some parameters, but some patterns emerge.




I highlighted parameters that are sub par in yellow, and above "average" in green. The impedance curves for the Fountek problems basically look nice, clearly have copper in the gap, but a detailed look shows lots of bumps and squiggles which usually are evidence of mechanical resonances. Would be nice NOT to have those.


A few factors I specifically see in favor of the Vifa part:

• Largest effect Sd (cone area) of the contenders, by as much as 30%
• Good frame form factor for stacking
• Published impedance data suggests farly low cone resonances-
• Qts is suited to some extension with a sealed enclosure.

It would be nice if it had more Xmax, but the real trick would be seeing a Kippel test on all of these candidates, and I don't see that happening. I could just do distortion comps, but how OCD do I want to be about this? I'm thinking that if Roger settled on these, it's probably worth just going with them for this experiment, eh? As I expect to be using these if they work well in an HT setup crossed at 100Hz, deeper bass response isn't really the big criteria- but I'll test it full range for fun, too... and get some data analogous to yours.

[fjhuerta]

I can provide a very detailed free air impedance curve for a single FE-83 and for the entire array. I'm wondering if you think there's some copper on the FE-83. The impedance doesn't rise as much as I expected it to.

I'd definitely avoid the FE-85, based on the Xmax alone...

It's a great table! One can see why Roger settled on the Vifa.

[oneplustwo]

Can I ask what may be a silly question to the gurus? Is there an inherent problem with the corner loaded concept? It seemed to make a lot of sense to me and like I mentioned earlier, has definite practical benefits from a placement perspective (assuming your room works for it). So perhaps it hasn't gotten a great reception because of the execution of the EQ? Or is it purely the driver choice as Jon mentioned previously? Considering using bits and pieces from all the discussion and rolling my own for the fun of it.

EDIT: Also, Javier (and others besides Jon as his POV is fairly clear) curious about your perspective on the TC9FD. It is quite a bit cheaper and appears to have the same frame/motor. The FR/impedance curve doesn't look that terribly bad to me. And if it was in the same table as above, it would compare pretty well to all except perhaps the TG9FD. Again, asking because if I end up rolling my own, I don't need to be dependent on what the Great Marsh ends up doing.

EDIT AGAIN: Did some more reading and found this comparison of the paper and glass fiber drivers. The appear to be essentially identical practically speaking. And this post has results of those driver from Zaph.

[augerpro]

Jon posted the response of both Vifa's above. I had also thought the poly one was better behaved in the treble, but looking a those plots the FG does indeed look better. Of course there are always those immeasurables, the poly cone has high praise for subjective sound quality.

[fjhuerta]

Here's a 500 point impedance sweep of a single FE-83.


Here's the entire array impedance. 500 points.

[Juhazi]

Javier,
Can you plese show us step response from the line? At various distance say 1m/2m/3m. I have never seen such before. I am wondering if combing causes "ringing" in impulse and step response, I think it should show up.

[---k---]

Jon,
Did you consider the new Dayton PS95? Even from a theoretical perspective? It was designed to be a full range driver. Since it is the newest kid on the block, it might have some advantages in the motor design. We know it has copper. Looks darn sexy! ET would love them. It costs more, but if we're just talking theory or testing.... The response graph looks a little a little more ragged, but I think some of that is scale and some baffle. Xmax is average. Probably all of the parameters you're looking at fall in the middle of the considered drivers. But, I'm still expecting it to have the lowest distortion.

[JonMarsh]

I can provide a very detailed free air impedance curve for a single FE-83 and for the entire array. I'm wondering if you think there's some copper on the FE-83. The impedance doesn't rise as much as I expected it to.

I'd definitely avoid the FE-85, based on the Xmax alone...

It's a great table! One can see why Roger settled on the Vifa.

The factory curve with it's flat slow rise is definitely indicative of copper in the gap- that helps keep down the midrange and HF distortion, which otherwise would rise due to inductivity modulation. You see this with most good tweeters, and the better midrange and mid woofers, say, some ScanSpeak models, as well as Aurasound, Wavecor, and some others.

Yeah, that Xmax spec has me scratching my head- everything else is similar enough to the FE83 that I wouldn't expect a 2:1 difference. Maybe it's just spec'd more accurately? Could be the FE83 is on the low side, and that would explain the results you get in LF distortion compared with what Roger Russel claims for the IDS-25.

[fjhuerta]

Yeah, that Xmax spec has me scratching my head- everything else is similar enough to the FE83 that I wouldn't expect a 2:1 difference. Maybe it's just spec'd more accurately? Could be the FE83 is on the low side, and that would explain the results you get in LF distortion compared with what Roger Russel claims for the IDS-25.

On the build thread at TechTalk I actually posted that I believed that XMax was closer to 1.5 mm than 3 mm. I have a minispeaker based on the ND-91 and the Founteks can't even come close to the displacement of the Dayton. I agree - it's quite possible that it's just that the 85's specs are more accurate.

[JonMarsh]

Jon,
Did you consider the new Dayton PS95? Even from a theoretical perspective? It was designed to be a full range driver. Since it is the newest kid on the block, it might have some advantages in the motor design. We know it has copper. Looks darn sexy! ET would love them. It costs more, but if we're just talking theory or testing.... The response graph looks a little a little more ragged, but I think some of that is scale and some baffle. Xmax is average. Probably all of the parameters you're looking at fall in the middle of the considered drivers. But, I'm still expecting it to have the lowest distortion.

Hadn't even heard of that one (small full range drivers are NOT something that's been on my ongoing radar...)

Checked it out at the DaytonAudio site, and the data sheet link is broken. Sd is 28cm2, similar to the Fountek parts. Xmax is 2.5mm. Unfortunately, it is round...

Found the data sheet on PE; response plot below:



I dunno, what do you think? Looks kind of rough above 1 kHz to me...


Given all the interest in the TC9FD, I'll order some of those this AM before the EMI/EMC seminar if I can squeeze it in.

Several people have done lines on DIYAUDIO using that driver; it's weird (to me) that I don't see any projects using the original driver... and some rather elaborate versions have been built with the paper cone driver, like the Two Towers project in the Netherlands.



This is an elaborate build, using a well thought out and tooled translam construction - cutting corners on the drivers seems odd, to me.. but then I may seem odd to a lot of people, too!

[JonMarsh]

On the build thread at TechTalk I actually posted that I believed that XMax was closer to 1.5 mm than 3 mm. I have a minispeaker based on the ND-91 and the Founteks can't even come close to the displacement of the Dayton. I agree - it's quite possible that it's just that the 85's specs are more accurate.

That explains a lot... :T

[JonMarsh]

Here's a 500 point impedance sweep of a single FE-83.
[ATTACH=CONFIG]23809[/ATTACH]

Here's the entire array impedance. 500 points.[ATTACH=CONFIG]23810[/ATTACH]

Looks like some possible broad low level resonance in the 2500-3000 area, but I could be reading too much into that broad bump. I'll measure and do similar plots to post for the two Vifa's.

[---k---]

I dunno, what do you think? Looks kind of rough above 1 kHz to me...

Agreed. I said the FR looked a little more ragged. So, that leads to the question of why this driver exists? PE wanted to round there line up out with a new driver, but couldn't do better than a driver several years old? Even at several dollars more? I guess that was what was kicking around in my brain when I looked at the ragged response.

[Steve Manning]

Hadn't even heard of that one (small full range drivers are NOT something that's been on my ongoing radar...)

Checked it out at the DaytonAudio site, and the data sheet link is broken. Sd is 28cm2, similar to the Fountek parts. Xmax is 2.5mm. Unfortunately, it is round...

Found the data sheet on PE; response plot below:



I dunno, what do you think? Looks kind of rough above 1 kHz to me...


Given all the interest in the TC9FD, I'll order some of those this AM before the EMI/EMC seminar if I can squeeze it in.

Several people have done lines on DIYAUDIO using that driver; it's weird (to me) that I don't see any projects using the original driver... and some rather elaborate versions have been built with the paper cone driver, like the Two Towers project in the Netherlands.



This is an elaborate build, using a well thought out and tooled translam construction - cutting corners on the drivers seems odd, to me.. but then I may seem odd to a lot of people, too!

Thanks for posting this Jon ...... I went and found this build since I could not imagine trying to keep all of those layers lined up for a speaker that tall. Very nice build up to the point one of the cabinets split apart. I had exactly the same thing happen to me on my translam build. He also came to the same conclusion I did ....... don't constrain the layers, mother nature will win. 8O I also think LBL would be a better, though more expensive, option than BB, since it's more stable. Your baffle also needs to float so the wood can expand. I wonder how long it took Magico to figure that out with their original Mini, or they just had somebody that knew what they were doing from the get go.

[BOBinGA]

Jon,
I have a set of the PE PS95s and they are what I would consider good, but not great. They are made for full range use and are fine for that, but they really need a separate tweeter. Their response if flat enough with a simple 0.2 mH coil in series to tame the rising top end, but the ragged response makes them sound, well, a little dull in comparison to a similar sized 3" plus tweeter design I built. YMMV.

I also ran across this test site that has Soundeasy tests of many of the drivers on your list.

http://feleppa.com.au/speakermeasmid.html

It would seem to confirm your choice of the TG9FD10 unless you want to up the budget and spring for the Fountek FR88EXs.

[augerpro]

It would seem to confirm your choice of the TG9FD10 unless you want to up the budget and spring for the Fountek FR88EXs.

In my experience (YMMV) with the FR88EX the upper treble is not pleasant. I wouldn't use it fullrange. Although Javier likes his smaller Founteks in the treble...maybe the line's combing covers up the weaknesses there? Still, if I had to pick I'd start with a driver that subjectively sounds good as a single full range driver.

[fjhuerta]

In my experience (YMMV) with the FR88EX the upper treble is not pleasant. I wouldn't use it fullrange. Although Javier likes his smaller Founteks in the treble...maybe the line's combing covers up the weaknesses there? Still, if I had to pick I'd start with a driver that subjectively sounds good as a single full range driver.

You have to remember I'm using heavy EQ on the speaker. It's not as if I'm listening to its "natural" FR at all - I'd probably dislike it; I'm not a fan of full range single driver speaker designs.

Listening to the array without EQ is pretty awful. I'll try to post an RTA measurement later for reference.

[5th element]

The TG9FD-10-04 is a nice little driver, but in its stock form it suffers from significant resonance issues beneath the dust cap and through the pole vent, this caused dramatically increased levels of third and fifth order distortion products, way in excess of what Zaph measured for the TC9 paper coned version. To circumvent this, quite literally, I removed the dustcap and installed a wooden phase plug and the 3rd and 5th dropped almost by an order of magnitude. Even with the modification it doesn't perform quite as well as Zaph shows the TC9 performing.

These are cheap drivers with a bargain basement price, but the TG9FD really needs a tweeter. I tried listening to it full range and it didn't work well, but then that's not a surprise, its response drops like a stone over 12k. I used mine with an infinity tweeter in a small waveguide that I bought from ebay. These are used near field where I do all my sparks and wires soldering and sound sublime.



And now the distortion at 2.83vrms, the crossover is 4th order acoustic at about 3.2kHz. Note the rising 5th order below around 700Hz, something the TC9, in Zaph's measurements do not show. This is with the modification, without it the performance was quite terrible and run naked the pole vent resonated like an organ pipe, it sounded quite terrible.



These are used with a DSP to equalise the response and to apply some bass EQ.

If money wasn't an issue at all I'd be interested in building a large line array of the SB65WBAC from SB acoustics. These little drivers measure far better than the TG9 does and really DO have extension out to 20k.

[JonMarsh]

Thanks for the inputs and feedback, Matt!

The SPL response looks pretty good-



The main disappointment is the fairly small cone and low Sd value, 20cm2.

Xmax about 2.6mm.

Worth putting on the test list. Solen and Madisound carry these, and they are pricey, but with the Neo magnet that have a sort of decent rear window area- I'm curious how the cone break up and distortion profile look. Ordering some this morning. Would probably need at least 30 per side. Ouch!

One of the things I've been thinking about doing is building a small test line (four drivers) and measure and EQ using my Metric Halo LIO-8, then do listening tests. I can do this swapping out baffles in the 0.75 cu ft PE MTM cab.

[JonMarsh]

Ordered four of the SB parts AND four of the 8 ohm Vifa - (have the four ohm on the way) Peerless TG9FD10-08 3.5" Fiberglass Full Range 8 ohm.

[JonMarsh]

The published response for the 8 ohm version seems to be somewhat more different than I would expect.



I think the midrange issues COULD be related both to rear window issues and to pressure under the dust cap as Matt relates. But I would expect them to be more similar... well, we'll see what we shall see. I plan to use Phenolic test plates for mounting the drivers under test to try to minimize rear window issues, as that's how I will build this if we get to that step.

Also, the 8 ohm version is available from Madisound but not from Parts Epxress- their buyout well on that part has run dry. These may not be in production anymore, per se- this may all be new old stock or NOS, as we call it.

[5th element]

Jon your memory really is failing you these days, or maybe it's just a case of Homer Simpson style stuff, your learn more and all it does is push other stuff out!

Here are distortion measurements for the SB65

2.83vrms



2.83vrms +10dB



2.83vrms -10dB (probably more representative of a line array)



Very nice.

[JonMarsh]

I have serious Homer Simpson problems, but then, I also like to see the data as measured on my own setup, for correlation. That's for sharing! :T

[fjhuerta]

Fun with Arrays

So, I have this monster next to the arrays (http://www.htguide.com/forum/showthr...-speaker/page2) and I started wondering why on earth I couldn't use the woofer section (2 RS-270s, ported, 100L each) as a sub for the arrays.

Answer: no answer. So I configured the MiniDSP for 4 channel use, crossed the filters over at 100 Hz, 24 db/Oct L-R, optimized everything with the for the -1 dB/Oct target curve, and..


110 dB... my entire house was rattling at 25 Hz...





Now, THIS IS BETTER!

[---k---]

Yeah. That sounds like a better plan than trying to eq the line array to boost the base. It may not be theoretically correct, and the sound will drop off quicker than the line array, so you're equing the base level to one spot. But, I think that is what everyone, but Jon :P , is typically after.

[oneplustwo]

Very nice. So are you effectively bi-amping then, correct? What do you think about using the 8 channel to do parametric EQ and sub management for a HT application? Or have you looked at other minidsp products (Dirac/OpenDRC)? I'm thinking it may be simpler to just have the miniDSP take care of the mains and then let the receiver take care of the surrounds and sub. May need to wait and see how the bass response is if (when?) I try the corner loaded array with the TC9. Anyone have any thoughts as to whether that concept is fundamentally flawed?

[fjhuerta]

Yeah. That sounds like a better plan than trying to eq the line array to boost the base. It may not be theoretically correct, and the sound will drop off quicker than the line array, so you're equing the base level to one spot. But, I think that is what everyone, but Jon :P , is typically after.

Funny. I mentioned this exact same thing on TechTalk, and at least a couple of people told me since subs were basically omnipolar in their response, they would blend perfectly with the array. I thought this was incorrect, since point sources lose power twice as fast as line arrays for a given distance.

The sub modules work nicely, but I'm debating whether to keep them. Something bothers me about them. I can't put my finger on it. It's as if the soundstaging magic trick these speakers did was gone, or as if I lost some top to bottom coherence. Perhaps it's psychological. Also, there's the thing about rock - subs helped immensely... I need to listen to them far more to make a decision.

Anyway, regarding the Juhazi's request - showing the array + sub response at different distances can also help to see if the subs are properly integrating to the array. So I took 3 very informal measurements using SynRTA. Here they are.



By looking at the graph I can see that at least from 1 to 3 M, the subs are working fine. Having said that, you can clearly see the destructive interference in the treble from the 25 drivers. The comb filter effect in full force. The farther away you stand from the array, the less evident the comb filter effect is.

Then again, I have close to +12dB at 20 KHz. I cannot emphasize this enough - without EQ, these speakers sound weird.

Very nice. So are you effectively bi-amping then, correct? What do you think about using the 8 channel to do parametric EQ and sub management for a HT application? Or have you looked at other minidsp products (Dirac/OpenDRC)? I'm thinking it may be simpler to just have the miniDSP take care of the mains and then let the receiver take care of the surrounds and sub. May need to wait and see how the bass response is if (when?) I try the corner loaded array with the TC9. Anyone have any thoughts as to whether that concept is fundamentally flawed?

I have never used anything other than the MiniDSP, although I don't see any reason why the 8 channel unit wouldn't work for HT. I'd use it as you say - MiniDSP on the fronts, EQ for the sub on the receiver.

Bass response without an EQ will probably be around -10 dB from 150 Hz or so and down. Then you'll have a bit of a flat response from 150 Hz to 2 KHz, and treble response will dro very fast from there to 20 KHz. Think -12 dB down or more...

[Paul W]

The sub modules work nicely, but I'm debating whether to keep them. Something bothers me about them. I can't put my finger on it. It's as if the soundstaging magic trick these speakers did was gone, or as if I lost some top to bottom coherence. Perhaps it's psychological. Also, there's the thing about rock - subs helped immensely... I need to listen to them far more to make a decision.

With the stronger bass, consider trying a couple db less tilt toward the top-end to see if that restores coherence.

[oneplustwo]

Not sure if it's ok to post detailed info, but parts express did offer $9.65/ea pricing for 80 TC9FD drivers. (I'm considering using this many to cover the left and right as well as a center channel.)

[JonMarsh]

I don't think that's a problem to post.

My samples from Madisound and parts express shipped today.

Most of today was back to work prep (unpacking and setting up new LG 31" 4K monitor, doing some camera and lens tests with the new Pentax 645z just arrived, and working on the outdoor apperal list and ordering stuff for the Antarctica trip. GF has a lot of snow gear that can be adapted but I'm nearly starting from scratch, and stores are running out of winter season stuff already! Got several pieces on the way, though. :W

[oneplustwo]

Nice! I love the idea of the 645. Can't see myself carrying it around though. Considering going the other way actually with one of the Sony Alpha A7 full frame mirror-less bodies. But in reality, I'll probably get the next canon 5D. My Mark II is definitely getting a bit long in the tooth.

Back on topic, if I did a short line array for the center channel, I assume I could EQ it similar to the L/R channels? I'm less worried about the bass response given its a center channel and even the highs to some extent. But I assume the EQ could get me to a satisfactory performance?

[JonMarsh]

As long as your listening distance is fairly near field, and representative of the distance you listen at, I think it ought to work.

I have the original 645D, but the new Z is such a capable camera with the improvements to the dynamic range and ISO capabilities of the sensor, and the much faster processing electronics. Still, it's very familiar feeling, and I'll be taking the 645D body, too, as backup and reassembled into a specific lens and hardware setup. I think I can get by with just 3-4 lenses, as I'm find the pentax 67 55-100mm covers bases well when I need a medium zoom, and it's almost as sharp as the 120mm macro. I have several 67 series lenses that of course only shoot manual; that works ok on the D, but even better on the z with the live view with magnification.

[fjhuerta]

Here's one thing that I honestly don't understand, and I mentioned it on the other thread.

Roger Russell claims his array can reach 20 Hz. He mentions a couple of times on his FAQ about the bass capabilities of his speaker.

I could get 20 Hz at ver low listening levels (90dB? I can't remember) and could only listen to them at OK levels while reaching 40 Hz.

The thing is, I simulated the output of 25 FE-83s on an array... and the simulation pointed out that the maximum output lower than 40 Hz would be severely limited.

Hmm.

So, I thought that maybe the Vifa drivers Roger used could get lower than my Founteks, and louder.

But - I ran the same simulation today of 25 TC9s, and they run neck and neck with the Founteks - and bass output is about equal between both. Actually, they are basically identical.

So, I have no clue whether Roger Russell's claims are a bit of a stretch (VERY doubtful - everything I've read about him says he DOES know what he's doing) or maybe the music he listens to doesn't strain the speakers that much. I do know I could live with my array, without a sub, if all I listened to was jazz or classical. I'll try to post some graphs later.

[oneplustwo]

Out of curiosity, have you tried putting your speakers as close to a corner as you can to see if you get any boost in bass from the corner loading? Or perhaps see if there are bad things that start happening. Not sure if your room lends itself to that or if you even wanted to take the effort of re-EQ-ing with that configuration. Just a thought.

Back to the center channel array, how many drivers do you think I should plan for? I had considered doing a line above the screen and one below to try to get an image in the center. The other thought was to put a single longer line in wall behind the screen. But that would require an acoustically transparent screen.

[fjhuerta]

Out of curiosity, have you tried putting your speakers as close to a corner as you can to see if you get any boost in bass from the corner loading? Or perhaps see if there are bad things that start happening. Not sure if your room lends itself to that or if you even wanted to take the effort of re-EQ-ing with that configuration. Just a thought.

I'm ashamed to say, that when I measured my room, I only did so from one side. So, yeah, the speakers can fit on...one corner of my room. They are very far away from the corners for that reason.

I'd guess that the Murphy Line Array has impressive 50 Hz response because it's corner loaded.

[JonMarsh]

I'm ashamed to say, that when I measured my room, I only did so from one side. So, yeah, the speakers can fit on...one corner of my room. They are very far away from the corners for that reason.

I'd guess that the Murphy Line Array has impressive 50 Hz response because it's corner loaded.

I'm sure that doesn't hurt. One reason studio monitors are usually flush mounted and in the corners is both the boundary reinforcement, and the fact that in this position you minimize time delayed boundary reflections. With speakers, you need to either have the designed for and mounted at a plane or corner boundary, OR, well out from the boundaries (think Cardas speaker guides) and spaced appropriately, with the listening position in something close to near field, for best results.

Today both of the Vifa/Tymphany driver samples arrived, but the fiberglass coned ones are much prettier (of course, as they are more expensive!




Hopefully I can get setup on the weekend for some measurements, but truthfully, we're still in a mode where dealing with preparation logistics for the trip to Antarctica the end of February is taking precedence. Cool mitten gloves (Outdoor Research Meteor Mitts) and a new camera backpack (ThinkTank Airport Accelerator) arrived yesterday, along with GF's SealLine water proof pack shell.



These are used to transport your "real" pack which is carried inside; they are water proof to the degree that they will also float just find (absent the use of bricks) for quite a while...

Now I've got to double check that the camera pack will fit in my SealLine 75.

[JonMarsh]

More drivers arrived yesterday, including the TG9F10-08 8 ohm version from Madisound,




and those itsy-bitsy 2.5" full range SB acoustics parts...



They look like a candidate for a phenolic front baffle, too. Not much rear window.