Corner FTC line array with ~2.5" drivers

[onni]

Inspired by ra7 (link) and oneplustwo (link) I've come to the conclusion that I'd like to try a corner floor-to-ceiling line array. The difference is I'll be sitting closer to the line arrays due to a smaller room (3.9 m x 4.7 m = 12.8' x 15.4'). This leads me to think that a slightly smaller driver is appropriate (Combing worse when close to array and when driver ctc distance is large).

The SB acoustics SB65WBAC25-4 gets a lot of praise, but putting 40 of them in each speaker is a little expensive for me. So I've looked around and found 3 alternatives:

Peerless TA6FC00
$8.9 each.
Aluminum cone.
Neo motor.

An externally hosted image should be here but it was not working when we last tested it.

Peerless TC6FC00/TC6FD00
$7.1 each (neo)/$5.1 each (ferrite).
Paper cone.

Visaton FR 58
$6 each.
Paper cone.
Ferrite motor.

An externally hosted image should be here but it was not working when we last tested it.

They all need EQ to reach 80 Hz (where subs will take over), but the Peerless ones have about twice as much Xmax which should make things easier.


I found this thread interesting: link. The use of waveguides to reduce combing problems is interesting. I'll try to simulate this.

I've just started construction of the house where the speakers would be used, so this is in a very early stage...

/Anton

 

[onni]

Forgot a picture of the FR 58:

All models are suitable for rear mounting, which I think is a good idea in this case. The plan is to place the drivers in the throat of a 90 degree horn, as if the speakers were wall integrated. I'll upload a few sketches later.

/Anton

 

[wesayso]

Another array!

I'd recommend getting one of each model driver and listen and measure. Keep the cleanest, most neutral one that has reasonable x-max.

I must say I did not regret not being able to do the corner array. I don't thing I'm loosing much. Except for missing some low end re-enforcement on one side perhaps. Yeah and a dip on the other side. Ok, ok, I get it .

Keep an eye on each vertical plane parallel to the array of drivers.

If that Visaton graph is real I like it. its the only one with a smooth rising response, quite handy for array use. Too bad it has lower x-max numbers.

 

[fluid]

Onni,

How close do you think you will sit?

If you can be 2m away from the array then I don't think you need to worry too much about the combing with music signals.

The room I will have mine in the short term is 4.1m x 4.7m.

The smaller drivers also tend to introduce different compromises and I haven't seen one that competes that well with the TC9 unless price is no object.

40 to 50 of the Aurasound NSW could be interesting as they have enough x-max to change the frequency response. Not cheap though.

 

[perceval]

I second the advice of getting one of each driver you like, and measure for yourself.

Most T/S parameters and freq graphs are happily tweaked by the marketing division.

Also, most places will offer discount when buying large quantities... so don't be afraid to ask for a discount if buying 50+ drivers!

 

[wesayso]

Onni,

How close do you think you will sit?

If you can be 2m away from the array then I don't think you need to worry too much about the combing with music signals.

The room I will have mine in the short term is 4.1m x 4.7m.

The smaller drivers also tend to introduce different compromises and I haven't seen one that competes that well with the TC9 unless price is no object.

40 to 50 of the Aurasound NSW could be interesting as they have enough x-max to change the frequency response. Not cheap though.

Comparing to these above mentioned drivers I'd also go with the TC9 FD18-08.
It's tough to beat, as long as you're able to sit just over ~2 to 2.5 M away.
Onni, you'd have them in a corner, so this should work in your size room.

 

[koldby]

The problem with the smaller drivers are their x-max.
1mm for the TA6FC compared to 2,6 mm for the TC9FC types.
You can get the TA7FC a lot cheaper than $8,9... For 50 pcs they are $6.

But you can get the TC9FD15 (a very close cusine to TC9FC00) for $8,15 a pcs for 50 pcs.

And you can get the original IDS25 driver TG9FD10-04 (4 ohm) for $11,83 min 48 pcs...

And here comes a really bargain.
How about the TG9FD10-08 @ $5,5 ???
Problem we need to be a lot that goes into that deal as min order is 2016 pcs... And lead time is 18 weeks.
At that price I would take 100 pcs myself.
Anybody for some group buy?
Wesayso?
Fluid?
Perceval?

 

[koldby]

An all out line array HT room could be 50 (48?)pcs for front speakers 24 for center (12 above 12 below) 32 for rear and 16 for side omni arrays (4+4 in a triangular array).

Total 122 drivers @ lets say $6... $732 for the ultimate HT system and then a subwoofer.
Not that bad or..?

 

[koldby]

Comparing to these above mentioned drivers I'd also go with the TC9 FD18-08.
It's tough to beat, as long as you're able to sit just over ~2 to 2.5 M away.
Onni, you'd have them in a corner, so this should work in your size room.

And TC9FD18-08 is $ 8,15 @ 50 pcs...

 

[onni]

Another array!

I'd recommend getting one of each model driver and listen and measure. Keep the cleanest, most neutral one that has reasonable x-max.

I must say I did not regret not being able to do the corner array. I don't thing I'm loosing much. Except for missing some low end re-enforcement on one side perhaps. Yeah and a dip on the other side. Ok, ok, I get it .

Keep an eye on each vertical plane parallel to the array of drivers.

If that Visaton graph is real I like it. its the only one with a smooth rising response, quite handy for array use. Too bad it has lower x-max numbers.

Testing them is a great idea, especially when they are so cheap!

I have a possibility to make an "invisible" speaker that avoids several problems:
1. Floor and ceiling reflections.
2. Front wall reflections.
3. First side wall reflection.
4. Handing over between drivers.

And a speaker that might fill a lot of aspects that are needed for a "good" speaker:
5. Close to perfect impulse response.
6. Good headroom (>100 dB at 1 m from 100 Hz).
7. High sensitivity.

I would also like to achieve good (constant) directivity, to get a wide sweet spot. Here the size of the driver plays in. With a 2"-driver I can get the desired behavior to a higher frequency than with a 3.5"-driver.

I have specifically designed the room so that it is symmetrical and with strategically placed windows:

This makes line arrays possible (they are not in my current house).

Onni,

How close do you think you will sit?

If you can be 2m away from the array then I don't think you need to worry too much about the combing with music signals.

The room I will have mine in the short term is 4.1m x 4.7m.

The smaller drivers also tend to introduce different compromises and I haven't seen one that competes that well with the TC9 unless price is no object.

40 to 50 of the Aurasound NSW could be interesting as they have enough x-max to change the frequency response. Not cheap though.

I'll be around 2.5 m from the closest speaker.

All drivers introduce compromises
Pros and cons for the TC9FD18:
+ Larger Sd and Xmax = reaches deeper.
+ Cheap ("only" need 30 per side).
+ High power rating.
- Slightly uneven response above ~7 kHz.
- Difficult to control directivity due to size.

Pros and cons for the TA6FC00:
+ Smooth response all the way to 20 kHz.
+ Easier to control directivity (small).
- Expensive (at least 40 per side).
- Low power rating.

Pros and cons for the FR 58:
+ Smooth response all the way to 20 kHz.
+ Rising response.
+ Easier to control directivity (small).
+ Fairly cheap
- Low power rating.
- Small Xmax.
- Low sensitivity.

Pros and cons for the NSW2:
+ Large Xmax.
+ Easier to control directivity (small).
+ Low Qts
- Falling response above 6 kHz.
- Expensive (at least 50 per side).
- High Fs

I second the advice of getting one of each driver you like, and measure for yourself.

Most T/S parameters and freq graphs are happily tweaked by the marketing division.

Also, most places will offer discount when buying large quantities... so don't be afraid to ask for a discount if buying 50+ drivers!

I've been looking at mainly digikey, especially for their discounts for large quantities

 

[onni]

The problem with the smaller drivers are their x-max.
1mm for the TA6FC compared to 2,6 mm for the TC9FC types.
You can get the TA7FC a lot cheaper than $8,9... For 50 pcs they are $6.

But you can get the TC9FD15 (a very close cusine to TC9FC00) for $8,15 a pcs for 50 pcs.

And you can get the original IDS25 driver TG9FD10-04 (4 ohm) for $11,83 min 48 pcs...

And here comes a really bargain.
How about the TG9FD10-08 @ $5,5 ???
Problem we need to be a lot that goes into that deal as min order is 2016 pcs... And lead time is 18 weeks.
At that price I would take 100 pcs myself.
Anybody for some group buy?
Wesayso?
Fluid?
Perceval?

I can't find a TA7FC, did you mean the TC7FD00? That looks good as well. Slightly less smooth response and same Xmax, so not really better than the TC6/TA6.

I'm only going for a stereo set (for now at least).

/Anton

 

[onni]

Enclosure idea

So the idea is to
1. Make an "invisible" speaker, i.e. in the corner.
2. Make a waveguide to control directivity (wide sweet spot).

This is my first idea for the enclosure:

The thickness of the enclosure is about 10-12 cm. The volume can be made big if necessary, the rest of the waveguide can also be made into a bass trap.

Looks like this in room:

The blue lines that extend the speaker enclosure (waveguide) are absorbers with the main purpose to remove the reflection from the opposite speaker. The idea is to cover the entire thing (speaker + absorber) with the same fabric.

/Anton

 

[wesayso]

Ambitious, I like it!

Seems like a real enough solution to me, can't wait to see that happen.

 

[koldby]

I can't find a TA7FC, did you mean the TC7FD00? That looks good as well. Slightly less smooth response and same Xmax, so not really better than the TC6/TA6.

I'm only going for a stereo set (for now at least).

/Anton

Sorry I misspelled.. TA6FC

 

[fluid]

I would also like to achieve good (constant) directivity, to get a wide sweet spot. Here the size of the driver plays in. With a 2"-driver I can get the desired behavior to a higher frequency than with a 3.5"-driver.

I'm not sure I understand this part. When I look at the graphs all the Peerless ones and the TC9 start to roll off, off axis at about 2K. The TC9 actually looks better to me in this regard. Maybe it is different scales on the graphs.

I understand that the smaller the driver the higher it will begin to beam but the geometry of the actual drivers has an effect too as seen in the graphs.

All drivers introduce compromises

They certainly do Pick your poison.

Here is a quick comparison of a TC6FC00-04 (Orange) vs a TC9FD18-08 (Yellow) vs Aura NS2 (Green) The other's aren't in the database.

I have used 3 watts as that is where the TC6 reaches x-max.

SPL at 3W


Excursion


Maximum Output


Box parameters used in model


I think this sort of graph is relevant as it gives you a better idea of the sort of spl the system will be capable of.

I still think the TC9 has the best set of compromises but I can see why you would want a smaller driver. The tricky part is finding a good one that doesn't cost too much.

I really like the integrated waveguide, seems like the best way to do a corner array to me.

Looking forward to see what you come up with

 

[onni]

Ambitious, I like it!

Seems like a real enough solution to me, can't wait to see that happen.

Hopefully But to quote Top Gear "Ambitious, but rubbish".

Sorry I misspelled.. TA6FC

Ah, I see. The 8 ohm model I suppose. 18 week lead time is a lot, but probably worth it

I'm not sure I understand this part. When I look at the graphs all the Peerless ones and the TC9 start to roll off, off axis at about 2K. The TC9 actually looks better to me in this regard. Maybe it is different scales on the graphs.

I understand that the smaller the driver the higher it will begin to beam but the geometry of the actual drivers has an effect too as seen in the graphs.



They certainly do Pick your poison.

Here is a quick comparison of a TC6FC00-04 (Orange) vs a TC9FD18-08 (Yellow) vs Aura NS2 (Green) The other's aren't in the database.

I have used 3 watts as that is where the TC6 reaches x-max.

SPL at 3W
View attachment 601701

Excursion
View attachment 601702

Maximum Output
View attachment 601703

Box parameters used in model
View attachment 601704

I think this sort of graph is relevant as it gives you a better idea of the sort of spl the system will be capable of.

I still think the TC9 has the best set of compromises but I can see why you would want a smaller driver. The tricky part is finding a good one that doesn't cost too much.

I really like the integrated waveguide, seems like the best way to do a corner array to me.

Looking forward to see what you come up with

A smaller driver should (at least in theory) be easier to control. The directivity should be really wide up to a higher frequency and thereby controllable by a waveguide. When the cone diameter is in the same size as the wavelength it starts beaming and a waveguide doesn't really work.

The TC9 really has a nice behavior in the spec sheet, there is however a quite apparent widening in the response around 3 kHz that I don't really like:

Measured 0-90 degrees. Figure from mehlau.net here: link.

My earlier work with waveguides combined with a 1" ringradiator (ScanSpeak R2604) resulted in a speaker with this response:

If I could come even close to the directional behavior of that I'm happy I'll have a really large waveguide this time, so I have my hopes up

EDIT: Are you using the corner in the sims (1xPi)? Or are they assuming open air conditions (4xPi)?

I tried simming 40 TA6FC in WinISD and found that I could reach about 100 dB with a Linkwitz transform down to 70 Hz (Q = 0.65) at 1 m. This is also assuming free air. That's plenty for me.

/Anton

 

[onni]

Simulations!

Alright, so I've started doing some sims. The first ones I've put in ra7's thread, as they are for the TC9FD18. Here is the latest one with a flat waveguide 1 mm from surrounds (approximately as in ra7's line arrays):

I have the TC9FD18 and Peerless 830970 at home, so I've started with them (as I can measure them fairly well):

The 830970 is well suited to rear mount in a waveguide with its inverted surround. Here are simulation results from a simple rear mounting in a 7 mm thick baffle and chamfering the hole. Line 55 mm ctc distance, floor to ceiling array, result at 2 m distance.

And here is with an approximated geometry for the TA6FC00:

The surround makes it more difficult to make a smooth transition from cone to waveguide which is probably the reason for the dip at 16 kHz.

They are both a first iteration, I'm pretty sure I can improve the response a little.

/Anton

 

[onni]

A few drivers ordered

As digi-key suddenly had the TA6FC00-08 in stock I ordered a few interesting drivers:
TC5FB00
TC6FC00
TA6FC00
TC7FD00

I'm kinda hoping that the TA6FC00 is best, I'll make simple enclosures to compare them.

For comparison of the results in my last post is the TA6FC when front mounted in a 59 mm baffle:

Rear mounting seems like a really good idea

/Anton

 

[onni]

Drivers have arrived!

The drivers I ordered Thursday evening arrived today, that's impressive! Here they are:

Here with a comparison to the TC9 and a ScanSpeak R2604 (that I have from earlier). From left to right: TC5FB00, TC6FC00, TA6FC00, TC7FD00, TC9FD18.

I didn't know that the TC6FC was smaller than the TA6FC (55 vs 59 mm width).

/Anton

 

[geotone]

Hey there, I was silently following this with great interest! Did you make any progress? I find these little peerless drivers super interesting and would be curious to hear of your impressions!

Regards, G