ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: 1 [2] 3  All   Go Down

Author Topic: System limiters in series  (Read 1306 times)

David Allred

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1571
Re: System limiters in series
« Reply #10 on: February 22, 2018, 04:24:13 pm »

http://forums.prosoundweb.com/index.php/topic,160926.msg1479766.html#msg1479766

http://forums.prosoundweb.com/index.php/topic,159553.msg1465845.html#msg1465845

http://forums.prosoundweb.com/index.php/topic,158775.msg1464129.html#msg1464129

http://forums.prosoundweb.com/index.php/topic,159553.msg1465725.html#msg1465725

All good info, but I am still confused by the 1/3rd power.  (1/3rd of what?)

From one of Ivan posts....
"Since the cabinets are 1700W @ 4 ohms, that would be 82V.

I would run the thermal around 1/3rd power  with a attack around 3 seconds.

This would be around 50 volts"

The math (82 * 1/3 = 27) does not equal 50.
82 down 1/3rd = 54.  That's close, but obviously not 1/3rd power of 82.
Help me understand, please.
thanks
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 8677
  • Atlanta GA
Re: System limiters in series
« Reply #11 on: February 22, 2018, 05:17:57 pm »

All good info, but I am still confused by the 1/3rd power.  (1/3rd of what?)

From one of Ivan posts....
"Since the cabinets are 1700W @ 4 ohms, that would be 82V.

I would run the thermal around 1/3rd power  with a attack around 3 seconds.

This would be around 50 volts"

The math (82 * 1/3 = 27) does not equal 50.
82 down 1/3rd = 54.  That's close, but obviously not 1/3rd power of 82.
Help me understand, please.
thanks
Voltage and power are different.

1/2 the VOLTAGE =1/4 the power.

For 1/2 power, you would use the continuous voltage x .707=V

I have since updated my "opinion" on thermal limiters.  I now say 1/4 power (1/2 voltage) for "abusive" situations.  Typically 3 seconds attack time for large powerful woofers, down to 1 second for moderate powered woofers.

This is based on the CONTINUOUS rating of the loudspeaker, which should be 6dB down from the peak rating.

mids and highs are MUCH shorter.

The main reason for the change of opinion is the long term usage, not just signal levels.  1/4 to 1/3 power is fine for extended points in the signal, but you can run into issues if this goes on for hours and hours.

Some events are 72 hours straight with NO break in the music.  This can be HELL on drivers.

So far, the 1/4 continuous power (NOT music power) seems to be a better choice for thermal limiters.

I still like to have program limiters (set at the continuous rating of the driver with an attack time equal to the highpass filter duration) and a peak limiter.  I usually set this at the program voltage, rather than the peak voltage.

In my opinion trying to get a couple of extra dB out of a short peak is simply asking for trouble.

NOTE: some amplifiers have a peak voltage setting, but it is NOT measured by the RMS value of the signal but rather the peak of the signal. 

So if you want a RMS peak to be 100V, you need to set the "peak" limiter to be 140V.

It is important to understand which way YOUR particular amp determines this voltage.

Of course the BEST way, is to simply measure and set accordingly.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Nathan Riddle

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1438
  • Niceville, FL
    • Nailed Productions
Re: System limiters in series
« Reply #12 on: February 22, 2018, 05:20:41 pm »

Ivan beat me to it. His voice is superior. Listen to the master ;)

Ivan...

Sorry,

Thermal limiter = 1/3rd of the continuous power.
RMS limit (rms power doesn't exist) = continuous rating (or program, if not as worried about protecting the speaker(s))
Peak = xMax (@ lowest impedance)

For the sub in question (TH118):
7200w (peak)
3400w (program)
1700w (continuous)
566w (1/3rd of continuous)

V = Sqrt( P * R )

[Thermal Limiter] 47.6 (50v) = Sqrt (1700w / 3  * 4ohm)

[RMS limiter] 82.46v = Sqrt(1700w * 4ohm)

[Peak Limiter] 169v = Sqrt (7200w * 4ohm)

[Real Peak Limiter, lowest impedance] 146.9v = Sqrt (7200w * 3ohms)
Logged
I'm just a guy trying to do the next right thing.

This business is for people with too much energy for desk jobs and too much brain for labor jobs. - Scott Helmke

David Allred

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1571
Re: System limiters in series
« Reply #13 on: February 22, 2018, 05:56:55 pm »

Thanks Ivan and Nathan.
I had a bit of an aha moment on the way home, and thought that I should not be dividing volts by 3, but watts, then converting to volts.  Turns out I was right.
I dove into the limiter and compressors in the MR18 and 360, and hit a stumbling block.
The max attack time in the 360 is 200ms, and the MR18 is 120ms.  How much does that limit (PUN) me.  What are my best options, considering?
2 to 3 hrs is about as long as I go in dance mode. 
As a reminder my sub amp (PLX3602) is 775w into a 600w cont sub (UCS1).
Logged

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 8677
  • Atlanta GA
Re: System limiters in series
« Reply #14 on: February 22, 2018, 07:49:37 pm »

Thanks Ivan and Nathan.
I had a bit of an aha moment on the way home, and thought that I should not be dividing volts by 3, but watts, then converting to volts.  Turns out I was right.
I dove into the limiter and compressors in the MR18 and 360, and hit a stumbling block.
The max attack time in the 360 is 200ms, and the MR18 is 120ms.  How much does that limit (PUN) me.  What are my best options, considering?
2 to 3 hrs is about as long as I go in dance mode. 
As a reminder my sub amp (PLX3602) is 775w into a 600w cont sub (UCS1).
As usual, it is a matter of "it depends", and "specifics matter".

In YOUR particular case, I would turn the clip eliminators on the amp ON.  That should take care of the "peaks".

Since the rating of the amp is just a little bit (1.1dB) over the continuous rating of the driver, I would not worry about a continuous limiter.

So that leaves a thermal limiter.  But you don't have a long enough attack time.

I would set the attack at max (200ms).  The continuous voltage for 600 watts @ 8 ohms = 69V.  The 1/4 power point would be half that voltage, so 34V.

But since you don't have a long attack time, I would suggest a higher voltage/power, so maybe around 45V.  This is a balance between voltage and time.

If you wanted, you could set the limiter in the MR18 as a continuous limiter.  An attack time of 20ms and voltage of 69V would be a good start.

The problem with setting a limiter in the MR18 is that if there are any changes or boosts of freq in the DSP, the limiter voltages can be pretty meaningless.

It is for this reason the limiters should be the LAST thing in line.

If given a choice, I prefer to have the longest attack time (lowest voltage) first, and the peak limiter last.

NOTE: the attack time I suggested for the MR18 is based on a 50Hz highpass (or there abouts.

Some people prefer to set the attack time for twice the time for the xover freq

So that would be 40-50ms.  I would not worry to much about it.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

David Allred

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1571
Re: System limiters in series
« Reply #15 on: February 22, 2018, 09:02:00 pm »

Thanks a ton Ivan.
Logged

Chris Grimshaw

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 822
    • Grimshaw Audio
Re: System limiters in series
« Reply #16 on: February 23, 2018, 03:37:51 am »



I have since updated my "opinion" on thermal limiters.  I now say 1/4 power (1/2 voltage) for "abusive" situations.

Makes sense.
Power ratings are measured with pink noise, clipped to a 6dB crest factor.
That means a square wave with half the peak voltage would have the same power delivery.


When it comes to 72-hour gigs, that's some serious thermal abuse as everthing will end up saturated with heat - driver frames, screws, all that stuff. Things like tapped horns stand a fighting chance since they've got the magnets close to the outside world, with lots of airflow. Sealed mid-high boxes will really suffer.
I guess the solution there is to bring way more rig than would "needed", so you can run well away from the clip lights and keep everything fairly cool.
IIRC, Pi Speakers came up with a good way of getting heat out of the drivers from a front-loaded horn, involving an external aluminium heatsink, and a tube to connect it to the driver's polepiece. Thermal performance was much improved. Could be something to look at for boxes that enclose the drivers completely.

Chris
Logged
Sheffield-based sound engineering.
www.grimshawaudio.com

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 8677
  • Atlanta GA
Re: System limiters in series
« Reply #17 on: February 23, 2018, 07:37:54 am »


IIRC, Pi Speakers came up with a good way of getting heat out of the drivers from a front-loaded horn, involving an external aluminium heatsink, and a tube to connect it to the driver's polepiece. Thermal performance was much improved. Could be something to look at for boxes that enclose the drivers completely.

Chris
The problem with aluminum plates is  that when you stack them (on some cabinets), the plates are next to each other can therefore cannot radiate to the outside world.

I did a test on a FLH cabinet that had a 1/4" aluminum plate that was very close to the driver magnet (at one point anyway), vs a wooden cover.

This was a double driver cabinet, with both drivers getting the same signal. 1 of each type on each driver-that were in different parts of the cabinet.

I put temp probes on the top part of the magnets and ran them outside so I could measure the temp of the magnet housing.

I drove them with two signals at the same time.  pink noise mixed with heavy EDM.  So it was very abusive.

What I found was that the aluminum plate driver ran cooler than the wooden plate driver-about 2 to 3 * C.  This was very consistent from just after the start of the test up to where I stopped the test, which was when the magnets reached around 90*C.

What I found interesting was watching the "cool down".  I had stopped the signal and was just watching the temp on the magnets as they cooled down.

The wooden plate magnet cooled down a good bit faster than the aluminum plate.

There was no air flowing, inside or outside the cabinet.  The cabinet was inside an isolated room, so the ambient temp of the room had risen a fair bit during the test.

The aluminum plate did not provide enough of a temp advantage to make it worth while.  In my opinion anyway.

Something else to consider about aluminum plates.  They can remove heat, just as well as they can transmit heat to the driver in hot outdoor situations.  It is not a one way street.

You need to look at the whole package/usage when considering certain things, not just one part of it.
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!

Dave Garoutte

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1282
  • San Rafael, CA
Re: System limiters in series
« Reply #18 on: February 23, 2018, 12:46:36 pm »

If the ambient temperature is very high, the enclosure and driver are likely to be at the same temperature anyway, so the heat sink wouldn't be introducing any extra heat into the driver.  It is interesting that the 'plated' speaker took longer to cool off.  Perhaps the plate was re-radiating back to the driver.
Logged
Nothing can be made idiot-proof; only idiot resistant.

Events.  Stage, PA, Lighting and Backline rentals.
Chauvet dealer.
Inventor.

Ivan Beaver

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 8677
  • Atlanta GA
Re: System limiters in series
« Reply #19 on: February 23, 2018, 01:06:46 pm »

If the ambient temperature is very high, the enclosure and driver are likely to be at the same temperature anyway, so the heat sink wouldn't be introducing any extra heat into the driver.  It is interesting that the 'plated' speaker took longer to cool off.  Perhaps the plate was re-radiating back to the driver.
At the point of "cool off", there was no signal, so no air movement-inside or outside the enclosure.

I probably should have redone the test, but just lowered the drive level a goo dbit and see what happened.

Oh well, hind sight is 20/20
Logged
A complex question is easily answered by a simple-easy to understand WRONG answer!

Ivan Beaver
Danley Sound Labs

PHYSICS- NOT FADS!
Pages: 1 [2] 3  All   Go Up
 


Page created in 0.073 seconds with 22 queries.