Speakerplans.com Homepage
Forum Home Forum Home > General > General Forum
  New Posts New Posts RSS Feed - Setting limiters help
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Setting limiters help

 Post Reply Post Reply
Author
Message
Aulert View Drop Down
New Member
New Member


Joined: 23 January 2022
Location: Melbourne
Status: Offline
Points: 8
Post Options Post Options   Thanks (0) Thanks(0)   Quote Aulert Quote  Post ReplyReply Direct Link To This Post Topic: Setting limiters help
    Posted: 23 January 2022 at 1:49am
Hi everyone just finished building my first rig and can’t get my head around limiters even with searching the internet for hours and hours. We’re taking it out next week and while we’re off having fun at other stages we want to protect our system from DJs cranking the trim and master up running into clip.

We’re usinga pioneer rx2 and ultradrive.

I’m running 6 wsx rcfs 600 watts rms off 2x nx6000 which at max roughly puts 900 watts rms into each driver.

For tops I have 2 w8cs running off 2 Yamaha p5000s. The Yamaha’s put out 500 watts rms @ 8hm with the mids rated at 300 watts.
For the highs Yamaha’s will be putting out 250 @ 16 ohm and these are rated at 150 rms.

Any help appreciated thanks
Back to Top
Conanski View Drop Down
Old Croc
Old Croc
Avatar

Joined: 26 January 2006
Location: Ottawa, Canada
Status: Offline
Points: 2515
Post Options Post Options   Thanks (0) Thanks(0)   Quote Conanski Quote  Post ReplyReply Direct Link To This Post Posted: 23 January 2022 at 4:34am
How are you connecting 6 subs to two amplifiers?
Assuming you have figured out a setup that will give each driver the same power, start with the subs and set the limiters to just stop the amps from clipping. Then bring up the mids and highs to produce balanced sound with them. To have a safe setup all the amp gains have to be set to max or you need to install covers over the gain controls, Yamaha supplied clear plastic covers just for this purpose if you bought them new. 
I don't know what the input sensitivity is on the NX series but the Yamahas need a pretty hot signal(+6dbu) for full output, but that may work to your advantage because you probably don't need all the power available. 
In the end you want the limiters for all outputs to engage at the same time... you don't want the mids and highs to keep getting louder after the subs hit max output, so figuring out the exact settings to supply full power to the mids and highs is of little value. 
Back to Top
Aulert View Drop Down
New Member
New Member


Joined: 23 January 2022
Location: Melbourne
Status: Offline
Points: 8
Post Options Post Options   Thanks (0) Thanks(0)   Quote Aulert Quote  Post ReplyReply Direct Link To This Post Posted: 23 January 2022 at 9:12am
I’m just running 2 subs per amp channel and one amp just has a spare channel. 

I can set limiters that way but atm I have the mixer trims at 4 o’clock roughly to get a clean sound. When DJs just turn it past that will the volume still go up? Or will it be limited? 


Edited by Aulert - 23 January 2022 at 9:23am
Back to Top
monkeypuzzle View Drop Down
Young Croc
Young Croc


Joined: 25 January 2005
Status: Offline
Points: 923
Post Options Post Options   Thanks (0) Thanks(0)   Quote monkeypuzzle Quote  Post ReplyReply Direct Link To This Post Posted: 23 January 2022 at 12:14pm
What mixer are you using? By trims do you mean the gains on the front panel or small trim pots on the back?
blah blah blah blah blah......
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: 10920
Post Options Post Options   Thanks (2) Thanks(2)   Quote toastyghost Quote  Post ReplyReply Direct Link To This Post Posted: 23 January 2022 at 1:46pm
You only have one limiter on the Ultradrive, which is best treated as somewhere between an RMS and thermal limiter.

Assuming you're using 8-ohm nominal drivers, then that's a 70 V RMS threshold for a 600 W AES rating. In reality, you'll find that a driver on a long front loaded horn like the WSX will have quite a different electrical impedance in pass band, but this is a good place to start.

If you want to do it ‘properly’ then use your mixer, driven as hard as you’d expect into the Ultradrive inputs with a sine wave tone on loop, at any frequency from 50 to 70 Hz. You can generate a WAV of this from the free REW software’s signal generator, if needed.

Use a multimeter to measure the output signal of the mixer at full blast. Just touch the XLR pins on the back with the probes. You want a level of 7.2 Volts AC, or as close as you can get. That's +20 dBu, and is about as loud as signals get.

Keep the output channels muted, remove any EQ, high pass or low pass filters, and set to 0 dB gain. The limiter value for the channel should be at its max to start with. Connect the mixer to the DSP, and unmute the DSP output channel. Use the multimeter again to check the level at the DSP output XLR. It should be close to the value you measured at the mixer, a little less.

Make sure the amp's sensitivity knob or trim lots are fully open. That way, once set, people with idle hands can only make things quieter and safer, rather than more dangerous for your speakers.

Hook the outputs to the amp, but do not connect the speakers.

Use the multimeter and connect the probes to the relevant pins of a spare SpeakOn connector with the back removed. Be careful with your hands!

Put the meter somewhere you can see the display, and set it to read out AC Voltage.

Unmute the channel, and note the amplifier's output voltage reading on the multimeter. The amp might be clipping. Don't worry for now. Turn the DSP channel's limiter value down until the clipping stops. This is the maximum output voltage of the amplifier (without load), note this down, along with the dB reading for the DSP limiter that got this output level.

Now continue to reduce the limiter value until the meter reads 60 V, for the case of your WSX.

Note this limiter value from the DSP, stop the test signal, mute everything, and then reapply your filters for the speaker.

A sine wave has a lower crest factor (is less dynamic) than any music signal, and using a 60 Hz tone allows the use of any multimeter. If you want to verify further, generate a pink noise signal from REW, and drive the mixer fully into the red using it. That's pretty representative of most music played by a bad DJ.

You can just repeat this process for the other DSP outputs and amplifiers, using the same frequency sine wave, but if you want to be super safe it would be good to verify afterward using something within the intended frequency range for that box. However you'll need a better multimeter with a broader range of frequency accuracy, and ideally a TrueRMS reading for that.

For peak limiter thresholds, you need to have at least twice the voltage of the RMS signal, to handle most 12 dB crest factor music without it sounding real bad when limiting. However, if your notes from before show that the amplifier only stopped clipping at a reading that is below 3x to 4x your measured RMS limit, then you have two choices:
1. Set the peak limiter to the measured value below clip, keeping the measured RMS value - but be aware that you'll suffer compression and potential thermal damage from reduced crest factor when running into the RMS limiter for long periods.
2. Set the peak limiter to the measured value, and reduce the RMS limit to a value of 1/3 of the measured maximum voltage before the amp clips. You can measure again to find what dB limiter value achieves this voltage output.

Remember that these limiters have zero ideas about the complex impedance versus frequency, and therefore the current or power being consumed. That's fine, because the drive signal is the voltage, and that's what you're interested in for the most part.

Some simple Ohm's law will get you the relevant voltage from the nominal impedance value, and from there it's just some maths or some easy measurement with cheap tools to get the values by something other than eyeballing or guesswork.

The thermal or RMS limiter attack and release times are best based on some understanding of the cabinet's thermal compression threshold, but for subs values of 2 seconds attack and 3 seconds release are often appropriate.

For suitable attack and release values of separate peak limiting, this is typically based on the frequency range of the passband. The idea is that you want to avoid limiting before one cycle of the lowest frequency is complete, and often this value as the attack.

That's easy enough to calculate, just:
1000/f
Where f is the lowest frequency of the output channel. For the release, a longer release value is common for lower frequencies, typically 16 times the attack value. This reduces as you move up in octaves, and 2x or 4x is more common for HF limiters. You can find plenty of chat about that online, but try to check white papers or manuals from respected DSP manufacturers like XTA rather than forums.

If you get it right - on a better DSP - you should see the peak limits flick well before the RMS limits. That’s a good sign that you’re outta headroom, and to stop turning it up - bring more rig for the venue next time.

Edited by toastyghost - 23 January 2022 at 1:54pm
Back to Top
APC321 View Drop Down
Young Croc
Young Croc


Joined: 24 August 2013
Location: West Midlands
Status: Offline
Points: 680
Post Options Post Options   Thanks (0) Thanks(0)   Quote APC321 Quote  Post ReplyReply Direct Link To This Post Posted: 24 January 2022 at 8:29pm
To quote the author of this topic:

"Hi everyone just finished building my first rig and can’t get my head around limiters even with searching the internet for hours and hours. We’re taking it out next week and while we’re off having fun at other stages we want to protect our system from DJs cranking the trim and master up running into clip."


Whilst I agree that that your limiters should be set correctly, I don't think that this is hitting the root cause of the your worries.

So you have a system that is quite big (6 x wsx), that you are taking out for the first time next week, and you are going to be leaving it unattended whilst you are "off having fun at other stages".

I'm sorry but that is the wrong attitude, and will only come back to bite you.

If the "djs" who are playing on the system are at all likely to do as you say "cranking the trim and master up running into clip" then they are not real djs at all but babies, and need to be treated as such.

As we all know babies should never be left alone, even for short periods, without a babysitter.

Do yourself, your system, and most importantly the crowd listening to your system a favour. Arrange a schedule between you and your team so that that the system is never left unattended.

Work with your "djs". Help them out. Earn their respect. Keep an eye on their levels and physically intervene if necessary.

You might then get re-booked if the event happens again.
Back to Top
Pinyorouk View Drop Down
Young Croc
Young Croc


Joined: 31 January 2011
Status: Offline
Points: 550
Post Options Post Options   Thanks (0) Thanks(0)   Quote Pinyorouk Quote  Post ReplyReply Direct Link To This Post Posted: 29 January 2022 at 8:15pm
[Excellent explanation. Will be certainly be saving this for future use or reference. 

QUOTE=toastyghost]You only have one limiter on the Ultradrive, which is best treated as somewhere between an RMS and thermal limiter.

Assuming you're using 8-ohm nominal drivers, then that's a 70 V RMS threshold for a 600 W AES rating. In reality, you'll find that a driver on a long front loaded horn like the WSX will have quite a different electrical impedance in pass band, but this is a good place to start.

If you want to do it ‘properly’ then use your mixer, driven as hard as you’d expect into the Ultradrive inputs with a sine wave tone on loop, at any frequency from 50 to 70 Hz. You can generate a WAV of this from the free REW software’s signal generator, if needed.

Use a multimeter to measure the output signal of the mixer at full blast. Just touch the XLR pins on the back with the probes. You want a level of 7.2 Volts AC, or as close as you can get. That's +20 dBu, and is about as loud as signals get.

Keep the output channels muted, remove any EQ, high pass or low pass filters, and set to 0 dB gain. The limiter value for the channel should be at its max to start with. Connect the mixer to the DSP, and unmute the DSP output channel. Use the multimeter again to check the level at the DSP output XLR. It should be close to the value you measured at the mixer, a little less.

Make sure the amp's sensitivity knob or trim lots are fully open. That way, once set, people with idle hands can only make things quieter and safer, rather than more dangerous for your speakers.

Hook the outputs to the amp, but do not connect the speakers.

Use the multimeter and connect the probes to the relevant pins of a spare SpeakOn connector with the back removed. Be careful with your hands!

Put the meter somewhere you can see the display, and set it to read out AC Voltage.

Unmute the channel, and note the amplifier's output voltage reading on the multimeter. The amp might be clipping. Don't worry for now. Turn the DSP channel's limiter value down until the clipping stops. This is the maximum output voltage of the amplifier (without load), note this down, along with the dB reading for the DSP limiter that got this output level.

Now continue to reduce the limiter value until the meter reads 60 V, for the case of your WSX.

Note this limiter value from the DSP, stop the test signal, mute everything, and then reapply your filters for the speaker.

A sine wave has a lower crest factor (is less dynamic) than any music signal, and using a 60 Hz tone allows the use of any multimeter. If you want to verify further, generate a pink noise signal from REW, and drive the mixer fully into the red using it. That's pretty representative of most music played by a bad DJ.

You can just repeat this process for the other DSP outputs and amplifiers, using the same frequency sine wave, but if you want to be super safe it would be good to verify afterward using something within the intended frequency range for that box. However you'll need a better multimeter with a broader range of frequency accuracy, and ideally a TrueRMS reading for that.

For peak limiter thresholds, you need to have at least twice the voltage of the RMS signal, to handle most 12 dB crest factor music without it sounding real bad when limiting. However, if your notes from before show that the amplifier only stopped clipping at a reading that is below 3x to 4x your measured RMS limit, then you have two choices:
1. Set the peak limiter to the measured value below clip, keeping the measured RMS value - but be aware that you'll suffer compression and potential thermal damage from reduced crest factor when running into the RMS limiter for long periods.
2. Set the peak limiter to the measured value, and reduce the RMS limit to a value of 1/3 of the measured maximum voltage before the amp clips. You can measure again to find what dB limiter value achieves this voltage output.

Remember that these limiters have zero ideas about the complex impedance versus frequency, and therefore the current or power being consumed. That's fine, because the drive signal is the voltage, and that's what you're interested in for the most part.

Some simple Ohm's law will get you the relevant voltage from the nominal impedance value, and from there it's just some maths or some easy measurement with cheap tools to get the values by something other than eyeballing or guesswork.

The thermal or RMS limiter attack and release times are best based on some understanding of the cabinet's thermal compression threshold, but for subs values of 2 seconds attack and 3 seconds release are often appropriate.

For suitable attack and release values of separate peak limiting, this is typically based on the frequency range of the passband. The idea is that you want to avoid limiting before one cycle of the lowest frequency is complete, and often this value as the attack.

That's easy enough to calculate, just:
1000/f
Where f is the lowest frequency of the output channel. For the release, a longer release value is common for lower frequencies, typically 16 times the attack value. This reduces as you move up in octaves, and 2x or 4x is more common for HF limiters. You can find plenty of chat about that online, but try to check white papers or manuals from respected DSP manufacturers like XTA rather than forums.

If you get it right - on a better DSP - you should see the peak limits flick well before the RMS limits. That’s a good sign that you’re outta headroom, and to stop turning it up - bring more rig for the venue next time.[/QUOTE]


Edited by Pinyorouk - 29 January 2022 at 10:33pm
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: 10920
Post Options Post Options   Thanks (0) Thanks(0)   Quote toastyghost Quote  Post ReplyReply Direct Link To This Post Posted: 29 January 2022 at 9:54pm
Thanks. And apologies - make that 1/2 the Volts, not 1/3. At least for subs or low frequency. M-Noise found a higher crest factor in the HF of most music, up to 20 dB versus the normal 12 dB. So you might want to be more ‘loose’ with your peak limiters above 4 kHz.

Twice the voltage is four times the power, for a given resistance. I’m clearly too used to the Powersoft system, where you type in the relevant units for the type of limiter thanks to the continuous measurement of real-time amplifier output status.

Edited by toastyghost - 29 January 2022 at 9:56pm
Back to Top
jack.bergquist23 View Drop Down
Registered User
Registered User


Joined: 09 March 2021
Status: Offline
Points: 34
Post Options Post Options   Thanks (0) Thanks(0)   Quote jack.bergquist23 Quote  Post ReplyReply Direct Link To This Post Posted: 13 February 2022 at 6:04am
Hi

First, thanks for the great run through there really appreciated. 

Not sure if you can also help with this question but would a class ab amp provide a lower voltage reading without load for some reason?

I've got a pair of lab 2002s (pretty old school) but reading the spec sheet it appears they're rated at 1400w per channel at 4ohm, however both are only reading 46v max input at clip (60hz sine wave driven from REW). Comparing this to my QSC GXD8 which was sitting at 65v at clip. I'm using a multimeter with a TrueRMS reading 

No the end of the world if they're not as powerful as I thought, am looking at some upgrade options, but just want to make sure I'm not reading something wrong before I use them so I don't overload the drivers. For reference they'll be driving some kappa Pro LFCs (600w rms), so will be underdriven if this max output is correct.

Cheers

Jack
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: 10920
Post Options Post Options   Thanks (0) Thanks(0)   Quote toastyghost Quote  Post ReplyReply Direct Link To This Post Posted: 13 February 2022 at 5:07pm
The second page of the Lab Gruppen 2002 spec sheet shows that the rated power is with an EIA-standard shaped stimulus, measured at 1 kHz.

See this page for an explanation of power ratings and the stimulus types used to determine them
https://www.doctorproaudio.com/content.php?2260-loudpeaker-power-handling&langid=1



The EIA standards use a shaped noise with 6 dB crest factor, so you'd need to use that same stimulus to get the same reading. The sine wave you're using has a 3 dB crest factor.

This isn't really a problem or a 'lie' as many people seem to think. Music is very, very rarely pure tone sine waves. The EIA, IEC and AES standards use shaped noise with a crest factor that is representative of most music. M-Noise is a modern take on that, and is free to download as a WAV with a 'max linear SPL' testing guide from https://m-noise.org/ if you want to go a step further.

Your 600 Watt drivers will be rated using one of these noise stimulus, not a sine wave. That’s very important. They also only generally measure the voltage, and calculate the power rating from that. The ‘real power’ that can be dissipated long term is much lower, especially with a sine wave stimulus. This is also why I suggest setting it lower than the peak value you can obtain from the amp - and hence the mention of testing again with a noise stimulus to verify in my previous posts.

You might also want to check that the MLS switches on the amp are set for full output, and try again with a 1 kHz sine. REW can also generate noise signals if needed. Just be aware it won't have the same shape as the EIA curve unless you also configure that in your DSP or somewhere else.

Edited by toastyghost - 13 February 2022 at 5:12pm
Back to Top
jack.bergquist23 View Drop Down
Registered User
Registered User


Joined: 09 March 2021
Status: Offline
Points: 34
Post Options Post Options   Thanks (0) Thanks(0)   Quote jack.bergquist23 Quote  Post ReplyReply Direct Link To This Post Posted: 13 February 2022 at 6:14pm
Thanks again for the extra detail there, that's a really useful link too.

Just to really try and cement my understanding here at this stage could I follow up with the following. 
1) for the lab amps, I'll retest with a 1khz sine wave and see what I get (btw I checked the MLS switches previously and they were set for max output). Would this suggest though that this amp is less capable for sub duty vs the GXD8 that did output a voltage in the expected range using the 50hz input?

2) As my processor (dB mark xca28) only offers a peak limiter and a compressor for each output channel, my thinking was to follow your guide (finding the limiter level required to reduce the input voltage to provide a 50v  output signal), then reduce the gain level on that channel to a point that it maintains the 50v output with the limiter removed. Then lift the peak limiter up to a level that corresponds with the music program limit of the driver. What I'd like to achieve is to have a max level starting point for the system (for a given mixer output voltage) which I can then detune appropriately for the setting at hand. Am I following a sensible path here or making a mess of it?

3) I've caught some other comments that if you're using a system for dance music (the rig will be pretty exlusively used for dnb, techno, deep house), that due to the compressed nature of the music you should aim to set your average levels at 2/3 the average power / voltage rating of your driver. Would you agree with that suggestion?

Sorry for the million questions here but very keen to build my understanding here based around the hardware limitations at hand in terms of processing /limiting. Also appreciate any of these dry setup tasks also need to be balanced with using my ears on the day as an additional step to check I'm not driving anything too hard.

Cheers

Jack
Back to Top
 Post Reply Post Reply

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.06
Copyright ©2001-2023 Web Wiz Ltd.

This page was generated in 0.109 seconds.