RYAN LOUDMUSIC JENKINS Messages: 2547 Registered: August 2005 Location: Surface of the Sun, It's ...
Has No Life
So I jsut changed the amp sensitivity from 1.4 V to 26db gain. A huge difference in the noise at idle. I raised the gain and limiters on the DSP by 11.5 db to get me back to where I was. Sound very clean but it is a little too early to crank this puppy up and hear it at real SPLs.
I have never run any of my amps at 26db gain. I have always run at 1.4v sensitivity. What other advantages other than what I have just experienced with this are there to running 26db gain?Ryan Jenkins
"Two days until the end of when I don't know what to think. Three days until I start the cycle all over again!"
Langston Holland Messages: 885 Registered: August 2004 Location: Pensacola FL
Has No Life
Ryan –
The Issues:
There are two primary failure modes for loudspeaker drivers, one is mechanical through over excursion which results from too much voltage being applied, and the second is over heating from too much average power being absorbed by the voice coil. The first can be dealt with directly with limiters; the second requires something between the ears of the operator. Don’t run your system with the processor or amp limiters engaging steadily (or clipping) for very long or something is going to melt even though the driver may never see too much peak voltage. If it doesn’t sound right, back off.
The Goal:
To correctly limit how much voltage is allowed to drive the four bandpasses in your system and to use your head to keep things from melting when you are running the system near its maximum output. Making things sound as nice as possible in the midst of this is good too.
Limiter Thresholds:
Since your processor and amplifier’s limiters only understand voltage, we will convert the loudspeaker’s power ratings into voltage to find out where to set the limiters. In those cases where the amp runs out of steam before the processor limiter threshold is exceeded, I still set the processor limiter as if the amp could produce the full voltage swing and allow the amp’s limiters to take care of bidniz. This of course is a bad idea if the amp’s limiters sound bad or don’t do the job correctly, but I know from experience that is not the case with the Vortex 6.
We will convert the loudspeaker voltage limits into dBu (dB referenced to .775 volts) to make the math easy. My rule of thumb is to set this limit at the driver’s max. continuous rating +3dB, which many folks, including TCS Audio, call “program” rating. +3dB is a doubling of power. If a driver has the ability to handle a maximum of 40dBu “program”, and your amplifier is providing 30dB of gain, you would limit the output from the processor to the amplifier to 10dBu. It’s that simple, kind of.
Other Limiter Parameters:
Most limiters have adjustments other than the dBu threshold. This is a good thing and allows you to tailor the way the limiter engages and disengages to better suit the bandpass it is protecting. Low frequency drivers sound bad with really fast attack and release times and high frequency drivers aren’t protected well enough with slow attack and release times. So how do you know what to do? I found a Turbosound article about this years ago and noticed that it lined up fairly well with EAW’s MX8750 settings, etc. Experimenting has proven to me that Turbosound’s suggestions (or whoever they got them from) work very well. Don’t worry if your processor doesn’t allow these exact settings, just approximate with a bias toward faster times for sake of protection.
Converting Watts to dBu:
Math: 20∙log [√(W∙Z) ÷ .775] where W = watts, Z = nominal bandpass impedance in ohms.
This got more discussion than I expected, but here’s my take. Signal is good, noise is bad, spinning the pretty little knobs on the amp clockwise without your knowledge is worse. In your case, you’re safe because the Vortex 6 amps have a magic gain switch on the back with 26dB and 32dB settings. You can pretty much rule the world with those two amp gains. When possible use 26dB to reduce noise, which should only be hiss or “white noise”, which is almost exclusively audible from the highs. The only downside remaining is that someone could turn the pretty knobs counterclockwise, which won’t hurt anything other than your feelings. Amps like the I-Tech with controls that can be locked really need to happen throughout the industry. There, somebody said something positive about those crummy amps. :)
Limiter Thresholds Part Deux:
There’s a (very expensive) rule of thumb that says your system ought to have 20dB of headroom over the normal levels that you run to allow the peaks of (classical mainly) music to get to the listener without compromise. Hogwash. 20dB headroom is 100 times the normal (some say RMS) levels! So, 95%+ of your gigs are pop/rock with 10dB or so wiggles (crest factor) over RMS, which represents a headroom of 10 times the power. That I can afford, barely.
What’s the point? uh... Oh yeah, I aim at getting the limiters to begin engaging when the console’s output is around +10dBu because I pretend all audio electronics clip at +20dBu (my world is a happy place), thus 10dB headroom (did that in my head). With this approach, the type of music most of us do just happens to make mechanical VU meters hover around one o’clock on the dial when the limiters start engaging, which I think is pretty.
PS: I’m looking forward to comments and/or corrections.
PPS: Ryan… Dude! Your tops are heavier than your subs – what’s up with that?! :)
RYAN LOUDMUSIC JENKINS Messages: 2547 Registered: August 2005 Location: Surface of the Sun, It's ...
Has No Life
Hey Langston,
Thanks for all the info there. You settings at the bottom for gain and limiters is very similar to how I have mine. I don't have Vortex 6 amps yet for the subs, so I have no choice there for gain setting on the amp. I got that one figured out. The numbers you came up with were 1/2 db off from where I was. I actually set my limiter just a tad lower just in case someone else mixes on my rig. I also didn't think about the +32 db setting for my lows. I will try that. The +26 db setting for the highs and mids worked quite well. I spent more time on the calculator yesterday just to be sure that I didn't throw the driver through the front of the box. Using the Vortex 6 really makes a difference in the sound at the top end.
Oh yeah, The tops are heavy! They suck to lift! They sound great though!
And then there is the 20 db of headroom deal.... I usually do have about 20 db of headroom for most of my gigs. Some of the louder bands will be about 7-10 db of headroom on the really loud parts. I don't mix particularly loud and these boxes can get really, really loud so I don't need a bunch of juice to make it work. Even my subs give me 12-20db of headroom!Ryan Jenkins
"Two days until the end of when I don't know what to think. Three days until I start the cycle all over again!"
Mac Kerr Messages: 9717 Registered: April 2004 Location: Westchester County, NY
Has No Life
Langston Holland wrote on Fri, 10 November 2006 11:40
There are two primary failure modes for loudspeaker drivers, one is mechanical through over excursion which results from too much voltage being applied, and the second is over heating from too much average power being absorbed by the voice coil. The first can be dealt with directly with limiters; the second requires something between the ears of the operator.
Langston, this is the most important part of this discussion. You can break anything if the operator doesn't use his brain.
I have not looked at the processor settings for the TCS speakers, do the processor settings require 6dB less gain for the highs than the other bands? If not, the gain needs to be added to the output of the high band output at the processor, adding back all the noise you may be trying to eliminate by turning down the amp. Most manufacturers require equal gain per band for the processor settings to be right. Available amp output power has nothing to do with this. The amp gain and processor settings are to get the spectral balance right. Amp output power can be factored in via amp selection, or limiting, but not gain. Under normal conditions, even without protecting the high drivers with limiting, it is unlikely they will be overdriven, because the spectral balance of the sound driving them will not include that much energy. If the mixer is deaf, and turns up the highs on all the inputs, or the system goes into oscillation, all bets are off.
RYAN LOUDMUSIC JENKINS Messages: 2547 Registered: August 2005 Location: Surface of the Sun, It's ...
Has No Life
The settings for the bands are pretty vague with regard to gain settings. I have all the EQs set up just like they recommend. I had to adjust the gain settings to match the amps being used for the correct "Spectral Balance."
Mac Kerr Messages: 9717 Registered: April 2004 Location: Westchester County, NY
Has No Life
Ryan, what did you have to change? I would assume from those specs that the amp gains all need to be the same. Did it sound wrong with all the gains the same? It is a pretty vague spec, with ranges for all the crossover frequencies.
Mac Kerr Messages: 9717 Registered: April 2004 Location: Westchester County, NY
Has No Life
Patrick Tracy wrote on Thu, 09 November 2006 23:18
I guess it makes me a little nearvous to run a really big amp wide open to a low-power driver. One bad connector getting some vibration could send career-ending voltage spikes to the poor HF driver. I'm much less concerned about some noise or hum getting in, though that's a consideration as well.
Whether you use a small amp or a big amp, the gain is going to be the same, or else the sound will change. Whatever pop or screech you are afraid of is going to amplified by an amount equal to the total gain of that part of the signal chain. If it is loud enough to drive the smaller high amp into clipping, that will limit the instantaneous power to the high frequency drive, but it will be a clipped signal with a high average power. At that level, the larger amp will not be clipping, and may be less dangerous to the driver.
Patrick Tracy wrote on Thu, 09 November 2006 23:18
I'm assuming that you mean something like "The gain settings in the DSP need to be as specified by the manufacturer for the given amplifier/speaker combination." Now, what is the difference between lowering the voltage at the DSP's output and lowering it in the amp, as long as the output of the DSP is high enough to keep the S/N reasonable while not clipping at the amp's input? Inquiring minds want to know...
There is no practical difference between adjusting the gain at the amp input, or the processor output. The only point of noise pick up between them is the the line level cable in between, an unlikely source of noise. My point is that the manufacturers usually expect that the gain of all amplifiers will be the same. The processor's limiters will protect the drivers from over voltage from a high output capable amplifier. The bigger amp just has a higher point for the onset of clipping. It won't be driven to a higher power than a small amp.
RYAN LOUDMUSIC JENKINS Messages: 2547 Registered: August 2005 Location: Surface of the Sun, It's ...
Has No Life
Mac Kerr wrote on Fri, 10 November 2006 10:38
Ryan, what did you have to change? I would assume from those specs that the amp gains all need to be the same. Did it sound wrong with all the gains the same? It is a pretty vague spec, with ranges for all the crossover frequencies.
Mac
Not all of my amps had the same amound of gain or sensitivity and therefore I had to adust a little to get it right. As far as the Crossover frequencies go, I adjusted those for a while until I found what sound good to my ears. It is almost right in the middle of the ranges that they give. Ryan Jenkins
"Two days until the end of when I don't know what to think. Three days until I start the cycle all over again!"
Mac Kerr Messages: 9717 Registered: April 2004 Location: Westchester County, NY
Has No Life
RYAN LOUDMUSIC JENKINS wrote on Fri, 10 November 2006 13:10
Mac Kerr wrote on Fri, 10 November 2006 10:38
Ryan, what did you have to change? I would assume from those specs that the amp gains all need to be the same. Did it sound wrong with all the gains the same? It is a pretty vague spec, with ranges for all the crossover frequencies.
Mac
Not all of my amps had the same amound of gain or sensitivity and therefore I had to adust a little to get it right. As far as the Crossover frequencies go, I adjusted those for a while until I found what sound good to my ears. It is almost right in the middle of the ranges that they give.
I thought this whole thread started because you had all Camco V6 amps. The important point is that the gain af all the amps needs to be the same for the manufacturer's processor settings to work. If you have an amp with a different gain you can correct for it at the processor output. This has nothing to do with amp power. You protect your drivers from too much amp power with the limiter. Set the limiter for the correct power as Langston described above. The sensitivity of the amp isn't really relevant. It is just the voltage needed to reach maximum output. With a large amp on the highs, and properly set limiters, you will never reach maximum output.
Patrick Tracy Messages: 2187 Registered: February 2006 Location: Boulder, CO
Has No Life
Mac Kerr wrote on Fri, 10 November 2006 10:56
Patrick Tracy wrote on Thu, 09 November 2006 23:18
I guess it makes me a little nearvous to run a really big amp wide open to a low-power driver. One bad connector getting some vibration could send career-ending voltage spikes to the poor HF driver. I'm much less concerned about some noise or hum getting in, though that's a consideration as well.
Whether you use a small amp or a big amp, the gain is going to be the same, or else the sound will change. Whatever pop or screech you are afraid of is going to amplified by an amount equal to the total gain of that part of the signal chain. If it is loud enough to drive the smaller high amp into clipping, that will limit the instantaneous power to the high frequency drive, but it will be a clipped signal with a high average power. At that level, the larger amp will not be clipping, and may be less dangerous to the driver.
Patrick Tracy wrote on Thu, 09 November 2006 23:18
I'm assuming that you mean something like "The gain settings in the DSP need to be as specified by the manufacturer for the given amplifier/speaker combination." Now, what is the difference between lowering the voltage at the DSP's output and lowering it in the amp, as long as the output of the DSP is high enough to keep the S/N reasonable while not clipping at the amp's input? Inquiring minds want to know...
There is no practical difference between adjusting the gain at the amp input, or the processor output. The only point of noise pick up between them is the the line level cable in between, an unlikely source of noise. My point is that the manufacturers usually expect that the gain of all amplifiers will be the same. The processor's limiters will protect the drivers from over voltage from a high output capable amplifier. The bigger amp just has a higher point for the onset of clipping. It won't be driven to a higher power than a small amp.
Mac
So, given that the amps have identical (voltage?) gains, they will all perform identically up to the point where the input voltage results in an output voltage that equals the maximum the smallest amp can deliver, as determined by rail voltage. Above that input voltage the smaller amps will clip while the bigger ones continue to reproduce the full waveform.
Or, given that the amplifiers represent a consistent voltage reference point it makes more sense to accomodate the varying needs of the different frequency bands, in terms of voltage gain, upstream of the amps at the DSP. For the same reason the limiters can be set accurately at the DSP to protect the speakers.
Current flowing through the voice coils then becomes a function of known voltages and (nominal) impedences, and power a function of the now known voltage and current. This ignores the vagueries of frequency dependent impedence, but am I close?
Langston Holland Messages: 885 Registered: August 2004 Location: Pensacola FL
Has No Life
Mac wrote on Fri, 10 November 2006:
I have not looked at the processor settings for the TCS speakers, do the processor settings require 6dB less gain for the highs than the other bands? If not, the gain needs to be added to the output of the high band output at the processor, adding back all the noise you may be trying to eliminate by turning down the amp.
That's true for the most part, though as you indicate elsewhere, in the unlikely event you get some noise in the snake between FOH and the amp racks, it will be reduced by 6dB with the 26dB gain setting vs. 32dB. Secondly, the noise floor of the processor itself is constant regardless of the digitally effected output gain, thus it makes sense to improve its S/N ratio (by 6dB in this example). Two minor issues, to be sure, but OCD wins!! :)God bless you and your precious family - Langston
Mac Kerr Messages: 9717 Registered: April 2004 Location: Westchester County, NY
Has No Life
Langston Holland wrote on Fri, 10 November 2006 15:13
Mac wrote on Fri, 10 November 2006:
I have not looked at the processor settings for the TCS speakers, do the processor settings require 6dB less gain for the highs than the other bands? If not, the gain needs to be added to the output of the high band output at the processor, adding back all the noise you may be trying to eliminate by turning down the amp.
That's true for the most part, though as you indicate elsewhere, in the unlikely event you get some noise in the snake between FOH and the amp racks, it will be reduced by 6dB with the 26dB gain setting vs. 32dB. Secondly, the noise floor of the processor itself is constant regardless of the digitally effected output gain, thus it makes sense to improve its S/N ratio (by 6dB in this example). Two minor issues, to be sure, but OCD wins!!
1) In the OP's case, there is no snake run as the processor is in the amp rack.
2) The noise of the processor may be constant, but it will also be negligible. The noise from the rest of the system upstream of the processor will be boosted by the increase in processor output level. There will be little if any net noise improvement in this situation. You may be able to improve S/N by lowering all the amp gains, and running the inputs of the console hotter, but this will mean raising the point of limiter onset the same amount. Changing gain in the last 2 items in the signal chain isn't going to do much.
My question about the 6dB difference was whether or not the other settings you specified took into account the lower gain of the high frequency amp.
RYAN LOUDMUSIC JENKINS Messages: 2547 Registered: August 2005 Location: Surface of the Sun, It's ...
Has No Life
Mac Kerr wrote on Fri, 10 November 2006 11:34
I thought this whole thread started because you had all Camco V6 amps. The important point is that the gain af all the amps needs to be the same for the manufacturer's processor settings to work. If you have an amp with a different gain you can correct for it at the processor output. This has nothing to do with amp power. You protect your drivers from too much amp power with the limiter. Set the limiter for the correct power as Langston described above. The sensitivity of the amp isn't really relevant. It is just the voltage needed to reach maximum output. With a large amp on the highs, and properly set limiters, you will never reach maximum output.
Mac
I am not quite there yet but I will be really soon. I now have my tops powered with Vortex 6s and the subs are with TCS TA2400s bridged until I can get some more Vortex 6s. I am in the process of removing some old amps, Mackie & E&W, and that is why I had the question regarding the setup with the bigger amps.
This has actually turned out to be a much more enlightening thread than I ever expected. Ryan Jenkins
"Two days until the end of when I don't know what to think. Three days until I start the cycle all over again!"
Mikael Holm Messages: 61 Registered: September 2006 Location: Turku, Finland
Should Get Out More
Mac Kerr wrote on Fri, 10 November 2006 22:23
Changing gain in the last 2 items in the signal chain isn't going to do much.
Actually with digital processors it allows us to run DSP at higher resolution. If possible, i try to match the relative gains at DSP but many times the bandpass equalization makes it impossible.
Ryan Lantzy Messages: 2155 Registered: December 2004 Location: Johnstown, Pennsylvania
Has No Life
Langston Holland wrote on Fri, 10 November 2006 11:40
We will convert the loudspeaker voltage limits into dBu (dB referenced to .775 volts) to make the math easy. My rule of thumb is to set this limit at the driver’s max. continuous rating +3dB, which many folks, including TCS Audio, call “program” rating. +3dB is a doubling of power. If a driver has the ability to handle a maximum of 40dBu “program”, and your amplifier is providing 30dB of gain, you would limit the output from the processor to the amplifier to 10dBu. It’s that simple, kind of.
Langston, these methods prove valuable. I always seem to complicate things and I like your approach to this.
That said. If setting the limiters based on the "program" rating, does this assume that the amp you have attached is capable of producing that power or greater?
For instance, many people recommend sizing amps at 1.5-2x rms of the driver they are powering. In my system, 2 of the band passes have 2x the power, but one has only 1.5x the power.
Should I set my limiter just below that? Or just let the limiter in the amplifier take care of it (note, nothing fancy like DSP in the amps, just a QSC RMX/PLX)?
Ryan Lantzy
"In the beginner's mind the possibilities are many, in the expert's mind they are few."
Forum Home
Login
Forum Search
Register
FAQ
) 2 Vote(s)
