ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB Lounge => Topic started by: chris broadway on February 04, 2014, 09:16:18 AM

Title: AMP clipping
Post by: chris broadway on February 04, 2014, 09:16:18 AM
I am continually in a quest to slowly optimize an existing system I am now supporting.  The full system is the Yamaha LS9, Driverack DBX-260, and crown 600 amps.  In the beginning, there was not a proper gain structure performed, amps were turned all the way up, and the LS9 console master fader set to about -20.  I wanted to do a gain structure to set the clipping level on all three devices.  I performed the gain structure to the best of my ability as the DBX 260 outlined. After I did this, I marked clipping with tape on all three devices and then lowered the master fader level and the amps.   A couple weeks ago I wanted to fix the master fader and individual channel fader level.  The master fader was -20, the individual channels were between -20 and -30, and the individual channel gain was very high.  I thought it would be better to increase the master fader to -5, the individual channel fader to around 0, and lower the channel gain significantly.  Doing this gives me more playing room when making channel fader adjustments.  Here is the problem.  Ever since I did this, I have periodic clipping at the amp and need to lower the level.  The first day I adjusted the console I had to decrease the mid/high amp by half, the next week was fine, then I had to lower it again this week.  The overall volume in the room is still fine and decreasing the amp is not creating an issue.  My question is, does anyone know why this is happening.  To fix it do I need a professional to come in a do a gain structure?  Is a gain structure at this point even needed?  If so, What would I gain?
Title: Re: AMP clipping
Post by: RYAN LOUDMUSIC JENKINS on February 04, 2014, 09:26:32 AM
Who cares where the master fader is if the volume is correct.  Use your ears, not your eyes.  FYI, I really don't like the Driverack's way of setting gain.  I would rather have the gain on all if my amps set for either 26 or 32 dB gain and have all the amp pots turned up all the way.  Make the adjustments in the DR260.  By doing this you can set the level where you want it without adjusting the amps from that point on.
Title: Re: AMP clipping
Post by: Tim McCulloch on February 04, 2014, 09:58:59 AM
Because your mixer clips at +22dBu.  In the analog world, the meter pointer would be slammed against the stop at +4dBu.  The analog mixer probably clips at about the same level, but the mechanical meter can't show that level.

Which "Crown 600" amps?  There are several models with the number 600 in the name.  That asked, I'd submit that your amps were probably shipped with input sensitivity of 1.4v.  That means that 1.4 volts of input level will drive your amp to FULL TILT BOOGIE and probably represents 32dB of voltage gain.  Note that my figures are speculative since we don't know exactly what amps you use... but mostly even that can be irrelevant WRT system noise floor.

Is this a mostly academic exercise or do you have a system noise problem?
Title: Re: AMP clipping
Post by: Paul G. OBrien on February 04, 2014, 10:28:00 AM
Sounds like you completely missed the most important part of setting gain structure, it's the signal level at each gain stage that matters not the control positions. You set gain structure starting at the input gains and working your way through the console to the outboard gear and then the amplifiers with the goal of having the signal level hot but not clipping at all stages. In general you don't want any stage ahead of the amplifier to ever reach clipping and especially not any stage in a digtal processor like the DR260 so be sure to stay well below 0dB full scale in that device. So instead of using clipping as the reference I suggest you use 0dB or +3db on the meters as your target for all stages right through to the amplifiers, and set their gain controls wide open and use the output level controls in the Driverack to dial back the levels enough to prevent clipping. With that done you should have a reasonable amount of travel in the main fader to use as a volume control.
Title: Re: AMP clipping
Post by: TJ (Tom) Cornish on February 04, 2014, 10:31:13 AM
I am continually in a quest to slowly optimize an existing system I am now supporting.  The full system is the Yamaha LS9, Driverack DBX-260, and crown 600 amps.  In the beginning, there was not a proper gain structure performed, amps were turned all the way up, and the LS9 console master fader set to about -20.  I wanted to do a gain structure to set the clipping level on all three devices.  I performed the gain structure to the best of my ability as the DBX 260 outlined. After I did this, I marked clipping with tape on all three devices and then lowered the master fader level and the amps.   A couple weeks ago I wanted to fix the master fader and individual channel fader level.  The master fader was -20, the individual channels were between -20 and -30, and the individual channel gain was very high.  I thought it would be better to increase the master fader to -5, the individual channel fader to around 0, and lower the channel gain significantly.  Doing this gives me more playing room when making channel fader adjustments.  Here is the problem.  Ever since I did this, I have periodic clipping at the amp and need to lower the level.  The first day I adjusted the console I had to decrease the mid/high amp by half, the next week was fine, then I had to lower it again this week.  The overall volume in the room is still fine and decreasing the amp is not creating an issue.  My question is, does anyone know why this is happening.  To fix it do I need a professional to come in a do a gain structure?  Is a gain structure at this point even needed?  If so, What would I gain?
Clipping at the amp (assuming you mean output clipping, not input clipping which can sometimes happen depending on the amp) means that you are asking more of your system than it can give.  Gain structure won't fix this, you need to turn it down.
Title: Re: AMP clipping
Post by: chris broadway on February 04, 2014, 11:05:11 AM
Sounds like you completely missed the most important part of setting gain structure, it's the signal level at each gain stage that matters not the control positions. You set gain structure starting at the input gains and working your way through the console to the outboard gear and then the amplifiers with the goal of having the signal level hot but not clipping at all stages. In general you don't want any stage ahead of the amplifier to ever reach clipping and especially not any stage in a digtal processor like the DR260 so be sure to stay well below 0dB full scale in that device. So instead of using clipping as the reference I suggest you use 0dB or +3db on the meters as your target for all stages right through to the amplifiers, and set their gain controls wide open and use the output level controls in the Driverack to dial back the levels enough to prevent clipping. With that done you should have a reasonable amount of travel in the main fader to use as a volume control.


I performed the gain structure per the instructions in the DBX 260 manual.  I just didn't push knobs to "0" and call it a day.
Title: Re: AMP clipping
Post by: Tim McCulloch on February 04, 2014, 11:07:20 AM

I performed the gain structure per the instructions in the DBX 260 manual.  I just didn't push knobs to "0" and call it a day.

Again, I will ask:  Is this an academic exercise or are you trying to solve an audible problem?
Title: Re: AMP clipping
Post by: Jordan Wolf on February 04, 2014, 11:14:42 AM
Chris,

What are you trying to achieve by resetting the gain structure of this system?

Also, here's a good read (http://www.prosoundweb.com/article/how_do_you_set_system_gain_structure/) on the topic.  There are a few different trains of thought, and each one is a compromise.

Title: Re: AMP clipping
Post by: Paul G. OBrien on February 04, 2014, 01:51:21 PM
I performed the gain structure per the instructions in the DBX 260 manual.  I just didn't push knobs to "0" and call it a day.
Yet your original posts talks almost exclusively about control positions and never once mentioned signal level except when you said the amp was clipping, nothing about that suggests to me that you did anything correctly.
Title: Re: AMP clipping
Post by: chris broadway on February 04, 2014, 04:19:02 PM
Chris,

What are you trying to achieve by resetting the gain structure of this system?

Also, here's a good read (http://www.prosoundweb.com/article/how_do_you_set_system_gain_structure/) on the topic.  There are a few different trains of thought, and each one is a compromise.

when I did the gain structure the original goal was to have clip on the console = clip on the driverack= clip on the amps.  This way everything was level set to some point.  I marked on all three devices where my clip level was.  then I turned the Crown CTS 600 amps down to get a good blend in the room.  Since I turned the amps down after the gain structure to get the correct output of the speakers for my room, I thought I would never have clipping.  After that, the goal I was trying to achieve with the faders was to be able "ride the faders" when needed.  Since the channel faders where so low (-30 to -20) it was hard to make small increment adjustments.  I wanted to lower the channel gain to lift the channel fader to be able to operate between -10 and +5.  Also my in ears avioms get their level from the channel gain and I think there is flood light interference causing a buzz in the avioms.  I wanted to increase the master fader to allow even further lowering of the channel gain.  It sounded good in theory....
Title: Re: AMP clipping
Post by: Nathan Lehouillier on February 04, 2014, 09:53:46 PM
1) PFL/Solo each channel set gain to a level so there is enough head room to not clip.
2) Set all channels to 0 then start bring up the main when you have a good level to start
mixing stop.
3) Start building your mix.
If you are worried about where your main fader is your not mixing with your ears.
If your amps are clipping your just out of amplifier power it has nothing to do with
gain staging. At this point you need more "RIG FOR THE GIG" I think that was Geri O.?
A CTS 600 is a very small amp but I don't know what your really doing so its hard to help.

Nate
KDS&L
Title: Re: AMP clipping
Post by: Glenn James on February 04, 2014, 10:01:54 PM
If your amps are clipping first, then you have all you are going to get out of your system. If you clip anything else prior or at the same time as your amps, you have lost too much headroom in your signal chain.
Set it up so your amps clip before full output of any channel or output on your desk or processors.
Title: Re: AMP clipping
Post by: Jean-Pierre Coetzee on February 05, 2014, 06:49:55 AM
What did you use to create your clipping point while doing gain staging? Noise, if so which noise? Music?(Lord help you), or something else? Also are you saying that your amps clip before the rest of the system and this is what you want to solve or is the entire system clipping at the same time?

Another point is what Tim was asking and that is why did you have to redo the gain staging? Was it for sh*ts and giggles or was there an inherent problem...

For your Aviom problem I am pretty sure Aviom uses a digital A-Net link which is immune to noise(well it will drop out if there was excessive noise not introduce that noise to the output signal), the problem is from your Direct outs(which I believe you would be using, are you using the A-Net card in the LS-9, does it have an A-Net option?) and what have you got after your controllers on stage(A-16ii i will assume). If you have IEMs like it seems you do the interferance is likely being introduced right there since as far as I know the A-Net signal come off digital straight from the LS-9 and therefore the noise cannot be anywhere else...

Is the buzz on the mains as well?
Title: Re: AMP clipping
Post by: Patrick Tracy on February 05, 2014, 01:59:04 PM
1) PFL/Solo each channel set gain to a level so there is enough head room to not clip.
2) Set all channels to 0 then start bring up the main when you have a good level to start
mixing stop.
3) Start building your mix.
If you are worried about where your main fader is your not mixing with your ears.
If your amps are clipping your just out of amplifier power it has nothing to do with
gain staging. At this point you need more "RIG FOR THE GIG" I think that was Geri O.?
A CTS 600 is a very small amp but I don't know what your really doing so its hard to help.

Nate
KDS&L

Here's one way I do it.

1. Amps off
2. Set input gains per manual
3. Set lead vocal fader to 0
4. Set master fader to 0
5. Set DSP input (or other gain stage between mixer and amps, or amp inputs*) to -∞
6. Turn amps on
7. Bring up DSP input (or...) until lead vocal sits on top of stage volume
8. Mix other inputs to that

*Using amp inputs or the output side of the DSP to trim the system may make DSP limiter thresholds incorrect.
Title: Re: AMP clipping
Post by: Brad Weber on February 06, 2014, 04:23:30 PM
Who cares where the master fader is if the volume is correct.  Use your ears, not your eyes.  FYI, I really don't like the Driverack's way of setting gain.  I would rather have the gain on all if my amps set for either 26 or 32 dB gain and have all the amp pots turned up all the way.  Make the adjustments in the DR260.  By doing this you can set the level where you want it without adjusting the amps from that point on.
What are you adjusting when you adjust the DR260 output levels, is it the digital signals pre A/D or the analog signals post A/D?  Are the maximum analog input and output levels of the DR260 the factory default +22dBu or may they have been set for +14dBu or +30dBu via the internal jumpers?
 
If you are actually adjusting the analog output of the DSP then you can adjust the output levels of the DSP to match the input sensitivity of the amps and whether you adjust the levels at the DSP output or at the amp inputs provides basically the same effect.  However, if you are instead adjusting the digital signal level of the DSP, which I think you are, then you may want to use the amp input attenuators for account for how the maximum DSP analog output relates to the amp input sensitivity. 
Title: Re: AMP clipping
Post by: Jean-Pierre Coetzee on February 07, 2014, 05:57:19 AM
Well here is my take on decent gain structure after the console.

Find out the maximum(clipping) input and output voltage/level in dBu/dBv/dBm(you have old gear 0_o) of all your equipment, let's go with what I have in one of our rooms.

Yamaha MG32fx: +24dBu output
DBX 223xs +22dBu input: +21dBu output
QSC RMX2450 input +4dBu

Now to start off with we can see that the input sensitivity of the MG32fx and the DBX223xs is very similar and to adjust the input attenuators  of the DBX unit down 2 db is really a pointless exercise IMHO but can be done. The biggest thing here is the output of the 223 is 21dBu but the input sensitivity of the RMX2450 is +4dBu, this will cause the clip point of the amp to sit way below that of the rest of the system, the easiest way to fix this would be to turn down the output of the 223 by 17 dB or to turn the input attenuators on the RMX2450 down by 17dB... I would honestly do this at the 223 instead of the amp but as far as I know it shouldn't make a hell of a lot of difference...

You would if you had limiters in place set the limiter to not allow the amp to receive enough input voltage to allow the speakers to exceed their peak power rating but that is a whole other story, you should also not be slamming into your limiter that whole time, it is there to protect your speakers from sudden spikes like a dropped mic, not there to control your fingers.

If there is any mistakes in this please call me out and I will edit it.
Title: Re: AMP clipping
Post by: Robert Weston on February 07, 2014, 10:45:44 AM
Chris Broadway -

where are you located?  Perhaps someone on the forum is near you and can stop by to help you with the system.

There's a lot of variables involved with setting levels (input -> output).  Simple setup procedures (as outlined in the DBX manual) are only starting points.  Consider this... when setting gain levels for a band (all things being equal), would you set the same peak-level on a vocal as you would for a kick drum?  Setting the gain levels thoughout the signal chain starts with the input.
 
Title: Re: AMP clipping
Post by: Patrick Tracy on February 07, 2014, 01:45:34 PM
The biggest thing here is the output of the 223 is 21dBu but the input sensitivity of the RMX2450 is +4dBu, this will cause the clip point of the amp to sit way below that of the rest of the system, the easiest way to fix this would be to turn down the output of the 223 by 17 dB or to turn the input attenuators on the RMX2450 down by 17dB...

You're comparing a peak output level with a nominal input level. I'd be willing to bet that the 2450's output will clip before the input does.
Title: Re: AMP clipping
Post by: Ivan Beaver on February 07, 2014, 03:21:58 PM
You're comparing a peak output level with a nominal input level. I'd be willing to bet that the 2450's output will clip before the input does.
Maybe not.  It is entirely possible (and I have measure this on a couple of power amps) that by turning down the input gain enough-that the actual front end (before the level control) can clip or exceed the voltage rails of input stage.

I don't know about the 2450-but if the rails are low (say +-15V) and the sending device has higher rails, then the input can clip before the output does.

That's the reason a physical pad is the best way to attenuate the signal.
Title: Re: AMP clipping
Post by: Patrick Tracy on February 07, 2014, 03:50:41 PM
Maybe not.  It is entirely possible (and I have measure this on a couple of power amps) that by turning down the input gain enough-that the actual front end (before the level control) can clip or exceed the voltage rails of input stage.

I don't know about the 2450-but if the rails are low (say +-15V) and the sending device has higher rails, then the input can clip before the output does.

That's the reason a physical pad is the best way to attenuate the signal.

Yeah, if you turned down the amp level controls enough and they follow the first active stage you could clip that first stage before clipping the output. But wouldn't that be when the drive signal is abnormally high and the target volume is abnormally low?

In my experience (with one PLX, a CE series, a Stewart and some Peaveys, plus various installed systems) with normal average drive levels around +4dBu that doesn't happen. Rather than "making everything clip at the same level" I aim for +4dBu average levels through the system. If I do that I seem to have adequate headroom in spite of different peak capabilities of the gear.
Title: Re: AMP clipping
Post by: Ivan Beaver on February 07, 2014, 06:06:03 PM
Yeah, if you turned down the amp level controls enough and they follow the first active stage you could clip that first stage before clipping the output. But wouldn't that be when the drive signal is abnormally high and the target volume is abnormally low?

In my experience (with one PLX, a CE series, a Stewart and some Peaveys, plus various installed systems) with normal average drive levels around +4dBu that doesn't happen. Rather than "making everything clip at the same level" I aim for +4dBu average levels through the system. If I do that I seem to have adequate headroom in spite of different peak capabilities of the gear.
The problem is that manufacturers of "up stream" devices (mixers-processors etc) are in a race to have the highest output voltage.  +24 to +30 is not extreme.

Yet many "power amp" will deliver full output with something in the range of +4 to +10 or so.

So the upstream devices can easily overdrive any power amp when they are pushed a little.

Yes you may have some extra "headroom" in those devices-but it is not "usable" since the rest of the system would already be at max output.

The only way to get more "real headroom" is to use larger amplifiers and/or more sensitive loudspeakers.
Title: Re: AMP clipping
Post by: Patrick Tracy on February 07, 2014, 09:03:17 PM
The problem is that manufacturers of "up stream" devices (mixers-processors etc) are in a race to have the highest output voltage.  +24 to +30 is not extreme.

Yet many "power amp" will deliver full output with something in the range of +4 to +10 or so.

So the upstream devices can easily overdrive any power amp when they are pushed a little.

Yes you may have some extra "headroom" in those devices-but it is not "usable" since the rest of the system would already be at max output.

The only way to get more "real headroom" is to use larger amplifiers and/or more sensitive loudspeakers.

I'm not suggesting using all that headroom. I'm suggesting keeping levels right around +4dBu, with peaks maybe 6 or 8dB above that, through the system. And I'm betting the inputs on most amps designed for a +4dBu world won't clip before +12dBu. Then you can trim back the level as needed at the amp.

If the OP is clipping amp inputs his gain structure is wrong. If he's clipping the outputs it's a not-enough-rig situation. He could be doing both. Solving the input side of things will reveal if there's a problem on the output side.
Title: Re: AMP clipping
Post by: Jean-Pierre Coetzee on February 08, 2014, 07:55:09 AM
I will be opening up the 2450 to clean it out in the next few days then I will let you know where in the chain the input attenuators are sitting, in one of the other amps I opened recently they were directly after the input jacks so in that case it would make no difference where it is attenuated.
Title: Re: AMP clipping
Post by: Ivan Beaver on February 08, 2014, 08:46:52 AM
I will be opening up the 2450 to clean it out in the next few days then I will let you know where in the chain the input attenuators are sitting, in one of the other amps I opened recently they were directly after the input jacks so in that case it would make no difference where it is attenuated.
I have never seen an amp (at least not an amp with balanced inputs) that had the attenuator right after the input jack.

The signal FIRST has to go through the balanced to unbalanced circuit (where the distortion can occur due to the limited voltage rails) THEN to the actual physical level control.

Would you care to name the model numbers of the amps in which the level/attenuator controls were right after the input jack?

I just find it hard to believe-and am curious.
Title: Re: AMP clipping
Post by: Brad Weber on February 08, 2014, 09:47:50 AM
You're comparing a peak output level with a nominal input level. I'd be willing to bet that the 2450's output will clip before the input does.
The input sensitivity for an amplifier is the input signal level that results in full rated output.  If the input sensitivity of an amplifier is +4dBu then that is the input signal level that would result in the full rated output of the amplifier.
 
That is different than the rated maximum input level for the amplifier and most amps will accept signal levels well above that needed to create full output.  Addresssing those differences in level, the typically higher maximum input signal level to the lower level needed for full amp output, is one reason why you often use the amplifier's input attenuators.  If you have a +22dBu peak input signal and the amp's input sensitivity is +4dBu, then having 18dB of attenuation at the amp input would seem to match the peak input level to the full amplifier output.
 
The alternative is to provide that same attenuation via a fixed pad located between the amp and prvisuos device or to reduce/limit the peak output of the device before the amplifier (provided you are affecting the actual analog output).
Title: Re: AMP clipping
Post by: Jean-Pierre Coetzee on February 08, 2014, 09:54:33 AM
I have never seen an amp (at least not an amp with balanced inputs) that had the attenuator right after the input jack.

The signal FIRST has to go through the balanced to unbalanced circuit (where the distortion can occur due to the limited voltage rails) THEN to the actual physical level control.

Would you care to name the model numbers of the amps in which the level/attenuator controls were right after the input jack?

I just find it hard to believe-and am curious.
I completely agree with you Ivan was just using this particular amp as an example and it has unbalanced inputs... I adopted it when I got this position and have removed it from service but have opened it up for shits and giggles. Its pretty much a no-name amp which is about as far from linear as you can get...

PS: The new 2450a's input clip point happens to be 22dBu which correlates to that of the rest of the system, not that it matters since I use the 223 to attenuate.
Title: Re: AMP clipping
Post by: Patrick Tracy on February 08, 2014, 12:17:56 PM
I have never seen an amp (at least not an amp with balanced inputs) that had the attenuator right after the input jack.

The signal FIRST has to go through the balanced to unbalanced circuit (where the distortion can occur due to the limited voltage rails) THEN to the actual physical level control.

That agrees with the last couple of block diagrams I looked at.