Speakerplans.com Homepage
Forum Home Forum Home > General > General Forum
  New Posts New Posts RSS Feed - Peak limiting for driver protection
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Peak limiting for driver protection

 Post Reply Post Reply Page  <1234 6>
Author
Message
toastyghost View Drop Down
The 10,000 Points Club
The 10,000 Points Club
Avatar

Joined: 09 January 2007
Location: Manchester
Status: Offline
Points: 10861
Post Options Post Options   Thanks (0) Thanks(0)   Quote toastyghost Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 5:12am
Originally posted by Augusts Augusts wrote:

But I still don't get where the big guys get 1500W AES and 6000W peak from a speaker enclosure.


Like almost everyone else, the spec sheets are written by marketers not engineers. They use the AES power rating for the transducers mounted in the box on the spec sheet, but rarely drive them at this level on their DAP amp platform unless they’re keen to make even more cash selling the operators of their systems spare parts

As I said before, the AES rating for a driver is not a lie. People just seem to rarely check the what the test standard entails and how that applies to their use case.

Originally posted by Conanski Conanski wrote:

Originally posted by Augusts Augusts wrote:

But I still don't get where the big guys get 1500W AES and 6000W peak from a speaker enclosure.


It's theoretically possible with program material that has 12db crest factor or more, that would result in 375w continuous with 6000w peaks.


As the vast majority of music does, for the majority of the 20 Hz to 20 kHz that we are interested in.

Even if you're playing modern electronic bass-heavy music which can have a 6 dB crest factor in the lower notes when analysing the full track, the resulting crest factor of the signal the speaker terminals will typically be closer to 9-12 dB due to the use of a high pass and low pass filter on the DSP.

It's quite easy to test this out for yourself. Just analyse the electrical signal at the output stage of your DSP's sub channel while playing your heaviest tracks.
Either in real-time using an RTA tool such as the one in REW, or by making a recording of it using a direct balanced connection to the input of an audio interface and then analyse that with something like Audacity or Reaper.

The M-Noise study that led to the AES75-2022 standard did essentially this process for hundreds of thousands of music tracks across all genres, and of course found that crest factor is actually far higher for the upper HF region, at 16-20 dB.

A 20 dB increase in output level to handle those peaks without clipping requires 10 times the voltage and 100 times the power in comparison to the average signal level.

If you're not shattering HF diaphragms regularly when using an 8 ohm nominal compression driver rated at 60 Watt AES, and we use the common spec sheet peak power handling of 4 times the AES test value, that means the RMS voltage applied to your compression driver should be just 4.38 Volts, or 2.4 Watts average to ensure you are reproducing the music as it's recorded.

That's not such a problem when the driver and waveguide has a voltage sensitivity of 114-116 dB with 2.83 V applied to it. You'll be hitting 135 dB on the peaks with the above values.

An arguably bigger problem is that a whole heap of premium loudspeakers are well into harmonic and modulation distortion levels above 5% by that point, which almost every study shows is clearly audible and unpleasant.

Some of the peak or max SPL values listed on manufacturer spec sheets are at 10% distortion for full-range speakers, which is almost unlistenable.

Look up some of the Production Partner “im test” reviews via Google Translate. They’re quite thorough, and show the average and maximum SPL thresholds achieved for a suitable distortion level when using various test signals that represent music well. There are several subwoofers and full-range speakers of various sizes in their archive.
Back to Top
Augusts View Drop Down
Registered User
Registered User
Avatar

Joined: 07 January 2017
Status: Offline
Points: 150
Post Options Post Options   Thanks (0) Thanks(0)   Quote Augusts Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 11:17am
I might have posed the question inaccurately. I have no questions about MR and HF drivers because they are kinda in their own acoustical environment and I usually trust the manufacturers given power ratings.

My concern was with subwoofers where excursion comes into consideration about how to limit everything. It does not matter wether the system goes into limiting or does not, I want my drivers to be safe.

For example for a Xilica processor, it has just one limiter, the standart RMS one or some might say that it is an AES value that we input into the limiter calculator (Funktion BTW has updated their calculators section.)

So the question would be, if I am designing my own speakers and I know my way around hornresp, I can see the actual driver power at every frequency but I want to choose the right amp and the right treshold for a Xilica limiter - Hornresp power input is 1000W RMS 4 ohms and the driver side of the simulation is under control, everything is fine.
Then should I set the limiters treshold to 1000W RMS 4ohms or 500W RMS 4ohms? And would I have to have and amp of 2000W RMS 4ohm or 1000W 4ohm? Because I know that I need at least 3db headroom for the peaks that the limiter lets through for the signal to stay dynamic. All of this that I am talking about is theoretically, what would be the right approach of setting the treshold for a subwoofer to stay in the Xmax zone.
If a peak is shorter in time that the most upper frequency in time that the upper LP frequency of the woofer (100hz) then I can safely set the amps Vpeak limiter +3db than the Xilicas dsp treshold.
Back to Top
Augusts View Drop Down
Registered User
Registered User
Avatar

Joined: 07 January 2017
Status: Offline
Points: 150
Post Options Post Options   Thanks (0) Thanks(0)   Quote Augusts Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 11:24am
I have never had problems with overheating. But I have however destroyed quite a few coils by hitting them at the backplate due to feeding excessive power to them.

I know I have to have more boxes but that does not mean that they should be limited incorrectly.

In my case, I think I am going to go with 1000W treshold for Program limiter and +3db for peaks
Back to Top
toastyghost View Drop Down
The 10,000 Points Club
The 10,000 Points Club
Avatar

Joined: 09 January 2007
Location: Manchester
Status: Offline
Points: 10861
Post Options Post Options   Thanks (0) Thanks(0)   Quote toastyghost Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 11:48am
Originally posted by Augusts Augusts wrote:


In my case, I think I am going to go with 1000W treshold for Program limiter and +3db for peaks


I’m not sure why you still think a 3 dB overshoot is sufficient, in any pass band. It suggests you either haven’t read or haven’t understood everything that’s been posted in this thread.

For subs, a 6 dB ratio between RMS and Peak voltage limiter values is common for a reason.

Also, as I mentioned on the previous page, Hornresp is calculating the Apparent power from the drive RMS voltage and electrical impedance curve. The Wattage power value cannot be a single number for all frequencies; this is why sensitivity and such must be measured by applying a fixed voltage.

The vast majority of DSP limiters are also based on an RMS voltage value. This is because you cannot have RMS Power, since it is calculated from the RMS voltage.

Only modern “smart amps” which directly integrate the DSP into active monitoring of the amplification stages and load that can use limiters based on the Real Power or current delivered to the drivers. Powersoft have white papers on their TruePower limiter which explain this, and the limitations of a single-value RMS voltage limiters to control thermal rise.

If you check the Xilica manual, I believe there is a peak limiter in the signal chain but it’s values are fixed in relation to the pass band’s electrical HPF. Similar to the “Auto” option on a XTA DP series.

Back to Top
Augusts View Drop Down
Registered User
Registered User
Avatar

Joined: 07 January 2017
Status: Offline
Points: 150
Post Options Post Options   Thanks (0) Thanks(0)   Quote Augusts Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 12:47pm
Have not understood.



This is by applying 1000W@4ohm or 63Volts from the amp, and this is what the driver receives at given frequencies, this is with HPF and LPF applied.

But I still dont understand does setting the limiters treshold at a 1000w would protect the driver from over excursion.
Just a simple question limit at a 1000W or limit at 1/4of a 1000w to leave space for +3db program material and +3db peaks.
I dont care about overheating because this driver is rated 1200w AES and they have been playing in a club with limiters engaged for about 5 years and still are working fine.


Edited by Augusts - 27 December 2022 at 12:49pm
Back to Top
Elliot Thompson View Drop Down
Old Croc
Old Croc
Avatar

Joined: 02 April 2004
Location: United States
Status: Offline
Points: 5055
Post Options Post Options   Thanks (0) Thanks(0)   Quote Elliot Thompson Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 12:50pm
Originally posted by Augusts Augusts wrote:

I have never had problems with overheating. But I have however destroyed quite a few coils by hitting them at the backplate due to feeding excessive power to them.

I know I have to have more boxes but that does not mean that they should be limited incorrectly.

In my case, I think I am going to go with 1000W treshold for Program limiter and +3db for peaks

98% of the people in the sound industry have never encountered banging the voice coils on the back plate. So their experience under your circumstance is minimum to none. I have experienced such a phenomenon decades ago so, I know exactly what you are talking about.


Your situation is due to either feeding the driver's voice coil frequencies they are incapable of producing or, the loudspeaker is unloading in the box.


All loudspeaker voice coils have a limitation on what frequencies they can produce from a dB Standpoint. It is relative to the TS Parameters as a whole in addition to your experience level on what the numbers mean as a whole.


As we know xmax has no influence on the calculated TS Parameters. It is the reason why you may see woofers moving back and forth as a piston with no sound being heard. That is due to the limitation of the TS Parameters and/or the Enclosure. The cone will still move back and forth regardless, as it is operating solely as a piston at this point.


Horn Response is not sophisticated enough to give you that answer. To find the answer, you would need a dB measurement microphone. Preferably non-C weighted.


Unloading is when the cabinet loses control of the loudspeaker and the driver is operating at the given frequencies as if it is not in a box. Erratic movement of the cone is a prime example of unloading. Again, Horn Response is not sophisticated enough to give that answer. A dB measurement microphone would give you the answer on where you stand from a sound pressure level.


The above is relative to how much power/wattage you are feeding the drivers and, how the loudspeaker is reacting to the box, based on it's TS Parameters at that given moment and time.


So your method is a good starting point. You must take your experience level of your own sound system into play. Use your own gut feeling. If you feel the drivers can handle more, then increase the limitation higher.


Best Regards,


Elliot Thompson
Back to Top
Augusts View Drop Down
Registered User
Registered User
Avatar

Joined: 07 January 2017
Status: Offline
Points: 150
Post Options Post Options   Thanks (0) Thanks(0)   Quote Augusts Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 12:54pm
I shall be doing a toaster test
Back to Top
Augusts View Drop Down
Registered User
Registered User
Avatar

Joined: 07 January 2017
Status: Offline
Points: 150
Post Options Post Options   Thanks (0) Thanks(0)   Quote Augusts Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 1:00pm
Thanks. After I discovered that the amp has a Vpeak limiter, I set it to RMS +3db and the woofers now are not hitting the backplates even if the limiter is never switching off.

I just wanted to clarify what is all this AES, Program and Peak values because my gut feeling is telling me that something is still not right

I guess a toaster test without limiters and then with limiters on and cranked all the way up would give me realistic data 


Edited by Augusts - 27 December 2022 at 1:01pm
Back to Top
Earplug View Drop Down
Old Croc
Old Croc


Joined: 03 January 2012
Location: Europe
Status: Offline
Points: 6917
Post Options Post Options   Thanks (0) Thanks(0)   Quote Earplug Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 1:30pm
Yes, as I mentioned before, just relying on the maths - or models is not going to do it. And trying to decipher some of the data given out by some manufacturers is a waste of time. The only way is to test, use your ears and definitely measure everything you can.

I worked for a while with a carpenter/box builder down here that really made some amazing cabs (I still use several on a regular basis), but I don't think that any came out on the first try. They all needed tweaking (he had SMAART - a great tool if you can get hold of it).


Have you looked at this thread:-


An example of why proper measurement can save a lot of pain & wasted resources!


And:-

"the loudspeaker is unloading in the box"

Definitely something to check. Your filters may need tweaking to avoid that happening again.



Earplugs Are For Wimps!
Back to Top
toastyghost View Drop Down
The 10,000 Points Club
The 10,000 Points Club
Avatar

Joined: 09 January 2007
Location: Manchester
Status: Offline
Points: 10861
Post Options Post Options   Thanks (0) Thanks(0)   Quote toastyghost Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 3:44pm
Originally posted by Augusts Augusts wrote:

Thanks. After I discovered that the amp has a Vpeak limiter, I set it to RMS +3db and the woofers now are not hitting the backplates even if the limiter is never switching off.

I just wanted to clarify what is all this AES, Program and Peak values because my gut feeling is telling me that something is still not right

I guess a toaster test without limiters and then with limiters on and cranked all the way up would give me realistic data 



Your gut is correct. You need to go back and re-read most of the posts & links here.

You’re also misunderstanding the toaster test procedure if you think running it with limiters (or processing) engaged is a good idea.

This peak-to-average ratio is still far too low. You may prevent the coil hitting the back of the gap, but have dramatically increased the risk of thermal failure.

Sitting in the peak limiter robs the signal of dynamic range or crest factor that is in the original content. This will then raise the average power, regardless of where your RMS voltage limiter sits.

The same problem can occur if the HPF is set too high, robbing the cabinet and driver of cooling airflow due to movement at the diaphragm and the exchange of pressure and velocity at the pole piece vent or port(s).

Effective loudspeaker limiting is a balancing act between thermal, excursion and distortion limits, and needs to be coupled with sensible use of high pass filtering for excursion and low pass filtering for thermal buildup, particularly in woofers.

Really, the combination of empirically derived acoustic and electrical high pass responses plus an absolute zero-attack brick wall soft 'clip' limit is the best way to control excursion damage.

The easiest way to achieve this is by choosing a sensibly-sized amplifier in the first place, which can deliver the RMS voltage and unclipped peaks to handle the crest factor of the signal. Which is always going to be greater than 6 dB in practice!

https://www.prosoundtraining.com/2016/07/19/power-amplifier-calculator/

https://www.prosoundtraining.com/2010/03/10/why-all-the-fuss-about-power-the-equivalent-amplifier-size-part-2/

https://www.prosoundtraining.com/2010/03/20/target-spl-and-required-amplifier-size/

The peak limiter is then used to prevent the speaker from sounding bad or becoming significantly non-linear.

https://www.prosoundtraining.com/2010/03/21/using-limiters-to-help-protect-loudspeakers/

No limiter system will ever be completely fool-proof, and speakers will fail if they are driven into their limits - electronic or physical - for long periods of time without correction by an operator. This is why it's a good idea to include an additional automatic gain control system with permanent installations where an engineer won't be present. Even better if that is driven from acoustic measurement in the venue, like the higher end DateQ models.

Despite what Elliot says, a well-qualified model is quite accurate for predicting a subwoofer's excursion and such, so long as the loudspeaker is driven in such a way that it is linear time-invariant.

That generally means small to medium drive voltage signals, which avoid significant nonlinear behaviour.


Most of these things can be calculated to get you started, but when it comes to measurements all of the functionality of Smaart (& software tools which are way more appropriate for loudspeaker design) can be had for free, using REW, VituixCAD, OpenSoundMeter, or ARTA.

Edited by toastyghost - 27 December 2022 at 3:47pm
Back to Top
Augusts View Drop Down
Registered User
Registered User
Avatar

Joined: 07 January 2017
Status: Offline
Points: 150
Post Options Post Options   Thanks (0) Thanks(0)   Quote Augusts Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 4:15pm
Thanks all of you guys for very in depth responses, links, opinions and pointers

If we dont meet any time soon - have a happy New Year!

A
Back to Top
4D View Drop Down
Old Croc
Old Croc


Joined: 13 November 2008
Location: Winchester
Status: Online
Points: 4178
Post Options Post Options   Thanks (0) Thanks(0)   Quote 4D Quote  Post ReplyReply Direct Link To This Post Posted: 27 December 2022 at 4:19pm
Beginning of the month we had a visiting engineer confound both a drawmer sp2120 & a xta dp limiter section  Thoroughly cooking several 18" drivers that hitherto had survived countless hours of relentless EDM unscathed.  How ? Pushing the gain higher but avoiding clipping through the application of heavy compression on the M3 master output.. 



Edited by 4D - 27 December 2022 at 4:19pm
DMZ. "The bass was intense. Girls were literally running up to stand next to the subs"
Back to Top
 Post Reply Post Reply Page  <1234 6>

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.05
Copyright ©2001-2022 Web Wiz Ltd.

This page was generated in 0.063 seconds.