Post by: Tamas Tako on June 02, 2005, 07:03:36 AM
Just in case if anyone is interrested for frequency response and Xcursion tests of the LABsub (both V1 and V3). I have some data.
If there is interrest for it, I may can upload the results.
Cheers,
Tamas
Post by: TonyPybus on June 02, 2005, 07:44:18 AM
Post by: fernand on June 02, 2005, 08:31:08 AM
Post by: Dave Rickard on June 02, 2005, 09:03:51 AM
Thanks,
Dave
Post by: Mike MacWillie on June 05, 2005, 08:19:59 PM
Post by: Too Tall (Curtis H. List) on June 06, 2005, 07:47:58 AM
| TT wrote on Thu, 02 June 2005 07:03 |
Hi, Just in case if anyone is interrested for frequency response and Xcursion tests of the LABsub (both V1 and V3). I have some data. If there is interrest for it, I may can upload the results. Cheers, Tamas |
Hey Tamas,
Sure thing, send some my way.
clist@airadv.net
Also a breif explanation on how you did the tests.
I'd be very interested how you did the excursion test.
PS-If I can help displaying multiple curves with weighting, etc in Praxis let me know.
Post by: jeffhtg (Jeff Kenney) on June 06, 2005, 03:16:25 PM
Post by: Michael_Elliston¶ on June 06, 2005, 08:40:06 PM
Post by: Johan Rademakers on June 07, 2005, 03:32:04 AM
| Quote: |
I'd be very interested how you did the excursion test |
And yes, what they say.
Mvg Johan
Post by: Tamas Tako on June 07, 2005, 05:10:16 AM
I will do that today evening...
Tamas
Post by: Tamas Tako on June 07, 2005, 01:57:03 PM
I try to upload a zipped folder where you can find some pictures and excelsheets with the results and the measurement arrangement.
I also will attach some jpg screenshots of the measured responses.
There are some LAB12 VoiceCoil photos as well, showing the possibly most common failure reason and a solution for it.
The glue used for this purpose is a two component Araldite epoxy.
Type:Araldite Resin AW106 with Araldite Hardener HV953U (1:1 mix ratio)
I hope this is usefull information for all here.
Best Regards,
Tamas
Post by: Brad Litz on June 07, 2005, 06:30:59 PM
Post by: Tamas Tako on June 07, 2005, 06:44:15 PM
Yes. I also mailed with Dave McBean about the results.
For very High power usage (and without your mouth extension) however the excursion starts to be too big around (and below)30Hz for the V3..
All measurements where done in halfspace using 4 enclosures to get a reasonable loading.
It would be great to have some 2x 15" version of the V3 to get less excursion down there.
(well the BT7 was a 2x15" box....)
Best Regards,
Tamas
Post by: Brad Litz on June 07, 2005, 07:16:39 PM
I like the 12" drivers because the cone can be made very stiff and does not break-up into higher resonant modes.
Post by: Michael_Elliston¶ on June 07, 2005, 10:32:31 PM
Post by: Bogdan Popescu on June 08, 2005, 12:29:44 AM
You have some nice tools there.
What is actualy the limit excursion where one would have to worry? Is it the 44mm mechanical limit or the Xmax ?
All the best,
Bogdan
Post by: Tamas Tako on June 08, 2005, 05:05:38 AM
1. Machanical demage:
You would demage the surround and/or the spider and/or the Voice coil when you exceed the mechanical limit of movement, what is 44mm If I remember correct.
However to reach such a large excursion one would need a big amount of power (current) which would burn the VoiceCoil.
2. Acoustical output:
The Magnetic Gap is ca. 10mm deep and the VoiceCoil is ca. 37mm long.
This gives a linear excursion of ca. 26mm (+/-13mm)
When I don't count with the magnet structure's (thermal)power compression (however I should!!) then exceeding this excursion gives me a decreased BL product and THEN all TS parameters changes...
This means for example that at 31mm(+/-15.5mm) excursion I just have the -3dB points of the BL product. While F= BxIL, this means that for reaching this excursion, you would need ca. 3dB more power than you would need normaly (if you where in the linear excursion range)
This also means, that above 27Hz you surely can't exceed the
-6dB BL points (+/-18mm) with an INPUT power that wouldn't burn the VoiceCoil. This point would be maybe the absolute non repeated peak.
So if you use a HPF at 30Hz with a slope of -24dB/oct, you surely will not reach the mechanical Xmax.
Brad,
I agree, the 12" cone is a great winner in this application.
BTW I think, one could get ca. 3-4dB more output @30Hz with a pair of good 15" drivers, where the cones should be very stiff...
Cheers,
Tamas
Post by: Michael_Elliston¶ on June 08, 2005, 04:45:54 PM
Voice coil heating seems a worse problem than over excursion,when appropriate subsonic filtering is applied.
What about XBL2 - Good for high horn SPL?
http://www.adireaudio.com/TechInfo.htm
Post by: Phillip_Graham on June 08, 2005, 09:38:09 PM
| mike_nz wrote on Wed, 08 June 2005 16:45 |
Hi Tamas Voice coil heating seems a worse problem than over excursion,when appropriate subsonic filtering is applied. What about XBL2 - Good for high horn SPL? http://www.adireaudio.com/TechInfo.htm |
XBL is a cleverly designed gap structure that results in very linear BL product over a broad range. Outside of this, it has the same basic other types of limitations (heat dissipation, cone stiffness, etc.) as any other driver.
One of the principles of Adire is Dr. David Hyre, formerly a researcher in NMR (nuclear magnetic resonance) at the University of Washington, and active back on the HiFi design lists in the mid 90's.
He was one of the first people I ran across who REALLY knew what was going on with designing loudspeakers, and I learned a lot from email conversations with him. I am sure future products from adire will continue to be most impressive.
Post by: Tamas Tako on June 09, 2005, 04:50:25 AM
I knew this Motor structure, and Yes this is a real good solution to have long excursion.
However there is a small disadvantage of this design:
While you have two air gaps (below each other), you need twice the B*A in the magnetic path to get the same air gap Flux density as you would get in a conventional Motor structure design.
This means double the B in the magnet (expensive magnet material) with the same surface, or the same Magnetic material with same B, but twice the surface (2*B)*A= B*(2*A)...
This means heavier and more expensive Assembly...
(this may be only expensive but not heavy by using high temp. NDy. magnets.)
Anyway, having such a 15" driver with good TS parameters and a quite stiff cone would make it possible to design "2x15" LABhorn" a la BT7. (Mike, you have the File on your website where Tom Danley was writting about the possibly TS parameters for the LABhorn if they were 15" drivers instead of the 12"...)
Cheers,
Tamas
Post by: Michael_Elliston¶ on June 09, 2005, 04:51:08 PM
| TT wrote on Thu, 09 June 2005 20:50 |
Hi, Anyway, having such a 15" driver with good TS parameters and a quite stiff cone would make it possible to design "2x15" LABhorn" a la BT7. (Mike, you have the File on your website where Tom Danley was writting about the possibly TS parameters for the LABhorn if they were 15" drivers instead of the 12"...) Cheers, Tamas |
Tamas-The interesting thing about the 'ideal' Driver,is that changing BL only seems to only change the excursion peak below cutoff,so if one used a good subsonic filter one could use a low bl driver rather than go to high expense with large powerful motors. The other small effect is seemingly worse ripple in HF response from undersized horns with low BL drivers.
Post by: Tamas Tako on June 09, 2005, 05:05:57 PM
You are right, however what is in your view "lo BL" 15" driver?
I think around 20-25Tm... But reach some 2x 20-25Tm with two air gaps is not so easy, but possible...
Anyway the ADIRE's motor is quite clever...
If you know some concrete driver, please let me know!
In the meantime take a look to the BMS 15N840 15" Nd driver...
(1200W AES, +/-14mm Xmax ca. 30Hz Fs etc...)
Post by: Michael_Elliston¶ on June 10, 2005, 08:13:54 PM
| TT wrote on Fri, 10 June 2005 09:05 |
Mike, You are right, however what is in your view "lo BL" 15" driver? I think around 20-25Tm... But reach some 2x 20-25Tm with two air gaps is not so easy, but possible... Anyway the ADIRE's motor is quite clever... If you know some concrete driver, please let me know! In the meantime take a look to the BMS 15N840 15" Nd driver... (1200W AES, +/-14mm Xmax ca. 30Hz Fs etc...) |
Low BL for a 12" would be near 12TM
Locally all that is possible here is PAUDIO or EMINENCE.
Im not sure if a high Xmax HT type woofer will provide reliability in prosound-sudden destruction.
Post by: Too Tall (Curtis H. List) on June 12, 2005, 09:26:03 AM
Post by: Gareth James on July 02, 2005, 09:57:47 AM
I have been playing around with some figures for a possible prototype using the BMS 15N840 drivers, as mentioned earlier in the thread.
Sorry if sound stupid here but im having trouble interpreting the excursion data that Tamas measured and am trying to compare the benefits of the 15's.
Tamas, in your spreadsheet it states a 16mm p-p or +/- 8mm movement at 30hz for the V1 at 40v. Is this correct?
Am i missing something vital, the displacement i get in McBean is +/- 16mm which is 32mm overall?!? I read and re-read the help on McBean but it clearly states one way displacement. If it was peak-peak displacement this would make sense, (i think).
Just want to clear this up in my mind as its hampering my progress on a few projects im designing for possible prototyping in McBean.
Please someone help!
Gareth James
Post by: Tamas Tako on July 06, 2005, 03:55:04 AM
I was waitin for that question.....
I did the tests 100% precise, so I am sure the test results are good.
If you can remember TOM DANLEY posted predictions for 4 and 6 LABSUBS halfspace @30Hz and at excursion limit (143dB rms, 149dB peak for 6 LABSUBS)
Well I measured 142-143 dB @30Hz for 4 LABsubs, which is very close to TOM's predictions.
However If you calculate with HORNRESP, at this level (143dB@30Hz for 4x LABSUBS in HALFSPACE), you would get an excursion, which is not real, becouse the driver units are not capable to move as big!!!
I also had a fellow, who made an AKABAK modell for 4x LABSUBS in HALFSPACE to see what Excursion this would predict, but this also was approx. the same as HORNRESP predicted.
So all I can say is, that in reality my results are true, but nor HORNRESP's nor AKABAK's prediction is true in this case!!!!
However TOM's prediction was TRUE!!!!!
This man knows something, you can belive me.........
Post by: Gareth James on July 06, 2005, 10:45:32 AM
Do you think the reason the actual excursion is so much lower than the predicted results (hornresp etc.) could be that these programs do not take into account the coupling effect of the cabinets accurately?
Your tests were all done for 4xLABs in an array, do you think the predicted would be closer for a single unit?
Cheers,
Gareth James
Post by: Brad Litz on July 06, 2005, 09:06:06 PM
Look at p17 of this article:
http://www.klippel.de/pubs/assesslargsign/assessing_large_si gnal_performance.pdf
This site has a lot of info on non-linear large signal (high power) driver performance.
http://www.klippel.de/background/default.asp
Post by: Gareth James on July 07, 2005, 04:32:23 AM
I didn't realise about the other factors limiting displacement, its pretty complicated stuff!
Guess ill just wait til someone incorporates these klippel prediction parameters into a simulation..mr mcbean?
Gareth James
Post by: Walt de Jong on July 08, 2005, 07:52:09 AM
Please notice the theoretical difference between halfspace radiation and putting a cabinet on the floor. Normally we call the later one also halfspace but that ignores the fact that for a large stack the frontal area of the cabinets also starts to work as a boundary. A true halfspace measurement is a driver or cabinet flush mounted in a very large wall (infinite baffle)
By the way, using AJ-horn with the option 'floor' for placement gives +/- 8mm at 40V at 30Hz.
Best regards,
Walt
Post by: Tamas Tako on July 08, 2005, 09:23:54 AM
Could you send me screenshots like all modelling parameter, SPL vs Frequency and Xcursion vs Frequency plots made with AJhorn for 4 cabinets on the floor?
It would be very interresting for me....
Thanks,
Tamas
Post by: Walt de Jong on July 08, 2005, 09:31:05 AM
http://img24.imageshack.us/img24/3108/tt4dy.jpg
The modelling of the frontchamber is critical. If you use a large value excursion at 30Hz goes up. I think the horns starts right in between the two drivers so I used a small volume of 6 litres in my simulation. If I use 24 litres I also get around 15-16mm.
Best regards,
Walt
Post by: Wayne Parham on July 09, 2005, 06:24:15 PM
Post by: Wayne Parham on July 18, 2005, 09:57:08 AM
Excellent data, Tamas, thanks!