TangBand 75-1558SE on a Waveguide with Measurements

[augerpro]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Sorry for the late reply, I'm not on forums too much anymore. I have not made any more progress on this, though I still have the last baffle sitting in the garage awaiting measurements. Pete Schumacher seemed interested in carrying on with using this driver in a waveguide, but I don't know if he has.

[Pete Schumacher]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Hi,
Thank you for your hard work. I'm considering to use the Tangband-75-1558SE in my speaker mainly based on the good distortion measurements done by Zaph. Did you found a way to solve the acoustic center issues with this driver?

Best regards
Adam
Denmark

It's pretty simple actually. if you mount the driver to the back of a .75" - 1" thick baffle, you can get decent alignment with most flush mount tweeters. The hole through the baffle can then be rounded over, and chamfered, to get a smooth transition to the baffle. You can fill any gaps using gasket material if you're worried about cavities. But because the guide is so shallow, the only benefit will be acoustic centering, leaving the same low and high frequency limits in place. In lower sensitivity designs, perhaps 500Hz on the low end; maybe 700Hz in higher sensitivity efforts. upper limit is pretty much 2500-3000.

Here's the 80884, 75-1558SE, and SEAS 22TFF on the baffle. I've even modeled up a crossover . . . just never got around to building it yet.

[augerpro]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Looks good Pete! Funny *when* I get around to building mine it will also use the 830884. I'm relieved to see the AC is a nice match for a flush mount tweeter, I was thinking it may about the same as a 4" cone woofer. Do you have the impulse or phase data to let you estimate the actual offset?

[Pete Schumacher]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Looks good Pete! Funny *when* I get around to building mine it will also use the 830884. I'm relieved to see the AC is a nice match for a flush mount tweeter, I was thinking it may about the same as a 4" cone woofer. Do you have the impulse or phase data to let you estimate the actual offset?

I ran a sum of the 22TFF and TB along with their individual responses, and used PCD to determine that mounted on the back of the 1" thick PE baffle they way we did, the mid was 10mm behind the tweeter.

[augerpro]

Re: TangBand 75-1558SE on a Waveguide with Measurements

I'm not familiar with PCD, are you saying that is completely a sim, no actual measurements used a starting point for later calculating (HBT or whatever)?

[Pete Schumacher]

Re: TangBand 75-1558SE on a Waveguide with Measurements

I'm not familiar with PCD, are you saying that is completely a sim, no actual measurements used a starting point for later calculating (HBT or whatever)?

The measurements are actual, all taken 1m from a point mid way between the mid and tweeter. You take a summed response measurement, along with the individual measurements, extract minimum phase from the individual responses, put them into PCD, adding in the vertical and horizontal offset information into X and Y, and then adjusting Z until the summed response in PCD matches what you measured. You then have the offset. It's a very accurate way to determine offset using Holm, since it doesn't measure actual phase (time of flight).

I think there's a way to determine it directly with Holm, but the PCD method works very well.

[augerpro]

Re: TangBand 75-1558SE on a Waveguide with Measurements

I see, thanks Pete.

[augerpro]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Tried Dan's idea of using some foam strip to fill the gap between the driver flange and underside of the baffle. This is with the third iteration of the waveguide, green is with the foam:



I think we're almost there

So, a couple years later I'm finally coming back to this and it is kicking my a$$. So in the green plot above, I would call usable. I labeled that WG3 and I also had a measurement from a further WG4:

Unfortunately I have no idea what the dimensions are of those waveguides, and I threw them away a long time ago :( So today I measure a new waveguide (WG5) with what I intended to try back then, and I got this: So this damn mid is like Goldilocks' porridge, and I'll have to go back and figure out what I built before and see if I can get the WG4 response again.

Pete S. do you remember the exact construction of your speaker? Do you have the mid response sans crossover?

[Pete Schumacher]

Re: TangBand 75-1558SE on a Waveguide with Measurements

So, a couple years later I'm finally coming back to this and it is kicking my a$$. So in the green plot above, I would call usable. I labeled that WG3 and I also had a measurement from a further WG4: [ATTACH=CONFIG]44223[/ATTACH]

Unfortunately I have no idea what the dimensions are of those waveguides, and I threw them away a long time ago :( So today I measure a new waveguide (WG5) with what I intended to try back then, and I got this: [ATTACH=CONFIG]44224[/ATTACH] So this damn mid is like Goldilocks' porridge, and I'll have to go back and figure out what I built before and see if I can get the WG4 response again.

Pete S. do you remember the exact construction of your speaker? Do you have the mid response sans crossover?

It's pretty simple actually. if you mount the driver to the back of a .75" - 1" thick baffle, you can get decent alignment with most flush mount tweeters. The hole through the baffle can then be rounded over, and chamfered, to get a smooth transition to the baffle. You can fill any gaps using gasket material if you're worried about cavities. But because the guide is so shallow, the only benefit will be acoustic centering, leaving the same low and high frequency limits in place. In lower sensitivity designs, perhaps 500Hz on the low end; maybe 700Hz in higher sensitivity efforts. upper limit is pretty much 2500-3000.

Here's the 80884, 75-1558SE, and SEAS 22TFF on the baffle. I've even modeled up a crossover . . . just never got around to building it yet.

The construction of the mid recess was a simple chamfer for about half the depth into the PE baffle (1" thick) and topped off with a roundover to the baffle edge and then hand sanded to smooth transition overall.

The raw mid response on this baffle is the plot in red which looks very much like some of yours. Seems to be quite adequate for a 3000Hz LR4.

[augerpro]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Missed the red plot when I looked at it, thanks. So the driver was mounted at a depth of 1/2", or is it 1" deep with only the top 1/2" chamfered? And the cutout (throat) was probably 3, 7/8" or so?

[Pete Schumacher]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Missed the red plot when I looked at it, thanks. So the driver was mounted at a depth of 1/2", or is it 1" deep with only the top 1/2" chamfered? And the cutout (throat) was probably 3, 7/8" or so?

It's 1" deep if I remember correctly. The first half inch from the opening was chamfered with a 60 degree bevel. The top half inch was with a larger radius roundover tangential to the baffle. I think it was a 1" roundover. The transition point was then hand sanded to smooth. Again, if I remember correctly, I had the baffle secured to a waste piece that served as the shelf for the bit bearing.

[critofur]

Re: TangBand 75-1558SE on a Waveguide with Measurements

Do you think you get audibly lower distortion vs the CAM-558 going this route? Some of the best liked speakers at DIY events have used the Morel mid (Tzu-jan is one I can remember at the moment).

I wonder if the fish-tank-filter foam plug could help with this type of waveguide? Perhaps a this might be a candidate for a large baffle speaker - both deeper and shallower waveguide on big broad baffle?

[davepellegrene]

Re: TangBand 75-1558SE on a Waveguide with Measurements

I've been working with this mid for a client who wants to use it in a waveguide from 400Hz to 3000Hz. I have it in an 11" wide by 6" tall guide 2" deep to be able to get down to 500 Hz with good directivity. He is crossing it to the XT25 which I am also working on an over lapping guide that will fit them both. That's why i went with 2" on the guide depth so I can keep the acoustical centers matched up. I just molded the first prototype last night and it went well. I need to do some fine tuning on the mold but it's fine for prototyping. Once I get mounting plates on it I'll post my measurements.

Here is the Mid in its own guide. With a 3" throat boost should be up to 4.5 kHz. Which looks to be pretty close to the SPL of the flat baffle.

https://picasaweb.google.com/m/zoom?...portHeight=268.

Distortion looks real good at 90 dbs at 1 meter.

Dave

[augerpro]

Re: TangBand 75-1558SE on a Waveguide with Measurements

I'm really not trying to load the mid at all, just trying to move it back some so it isn't a PITA to cross to, and isn't sticking out from the baffle 1". Regarding the Morel, I've worked with it in what would become the inspiration for this new speeker. It is very good, but I was hoping for something that can cross lower this time around.

[Pete Schumacher]

Re: TangBand 75-1558SE on a Waveguide with Measurements

I've been working with this mid for a client who wants to use it in a waveguide from 400Hz to 3000Hz. I have it in an 11" wide by 6" tall guide 2" deep to be able to get down to 500 Hz with good directivity. He is crossing it to the XT25 which I am also working on an over lapping guide that will fit them both. That's why i went with 2" on the guide depth so I can keep the acoustical centers matched up. I just molded the first prototype last night and it went well. I need to do some fine tuning on the mold but it's fine for prototyping. Once I get mounting plates on it I'll post my measurements.

Here is the Mid in its own guide. With a 3" throat boost should be up to 4.5 kHz. Which looks to be pretty close to the SPL of the flat baffle.

https://picasaweb.google.com/m/zoom?...portHeight=268.

Distortion looks real good at 90 dbs at 1 meter.

Dave

Link doesn't work for me . . . :(