Crossover Outputs or Amp Attenuators

[Dr. J]

I am bringing up this topic because on some other forums where I hang there is somewhat of a debate on this. I wanted to see if any of you have dealt with this in the field and IF it even matters.

I am not partial to this one way or the other BUT I do have some questions. I currently run with my amp attenuators pulled back a bit from clipping. This means my XO outputs are all set a ZERO.

I could make the adjustments the other way around as well BUT does it matter?

Amp attenuators wide open means no one can turn them up any further. May be nice when the system is being used by several different engineers that have different view points. I could also see it being a bad thing too.

On the DR 260 (probably any DSP Device) there is signal loss after it is split into its specific frequencies range -- high, mid & lows. This is very noticeable on the output meters of the 260. They are all at different levels.

In your opinion -- should the outputs on the DSP match the inputs at this stage or does it matter as long as you aren't clipping at the amp?

This may be a completely dumb question so I will take one for the team here...

[Dr. J]

Anyone at all????

Kip -- I know you have something to say.

Ok let me ask this: When you guys are called out to a show -- just how involved do you get? what I mean by this -- is it typical for a Smaart guy to get involved in the Gain Structure of the system OR do you just tap in and do what you have to do to align it?

[Arthur Skudra]

Ok let me ask this: When you guys are called out to a show -- just how involved do you get? what I mean by this -- is it typical for a Smaart guy to get involved in the Gain Structure of the system OR do you just tap in and do what you have to do to align it?

OK, I'll bite. Depends on how much you get paid! Just kidding!

My philosophy is that it IS the responsibility of the system tech to set the gain structure and limiters of the system. The gain structure part is relatively straight forward with digital systems, a bit more involved with analog systems. I still carry around my piezo tweeter for just those occasions. I use a spreadsheet to calculate limiter thresholds (which means you need to know the power rating of the amplifier and speaker in question), or rely on the manufacturer's processors to get those numbers straight. In the absence of either, I will try to find the most dynamic music in my collection to set the limiters, do some further tweaks during the show, albeit a rather imprecise method that is open to risk of failure.

As to where to set those amplifier attenuators, there are lots of different opinions out there, no particular one is right or wrong, just one is more appropriate than another in a given situation. Some prefer to leave the amp attenuators wide open, and trim things back on the output of the DSP, this way if an amplifier's attenuator knob is accidentally "bumped" or sent off for service, the proper level can be restored. Others prefer keeping levels high out of the DSP, and adjust the amplifier's attenuation knob to an appropriate level, some claim that you hear less "noise" out of a system doing it that way (which implies that things may not be right further up the signal chain, or maybe something wrong with your amplifier choice). Note that I stated that the knob on the amplifier is an attenuator, anything less than full output is subtracting dB. An important concept to understand.

[Dr. J]

Thanks Arthur! You always jump out there....

I did a few searches on this topic and man does it differ. It seems that many go with the "protective" side of the issue so no matter what a drunk DJ does (sorry DJ's) their gear is protected.

I however, would like it to be correct as far as signal to noise ratios are concerned. To me that means -- my outputs on my 260 should be in balance with the Inputs. Since there is signal loss going thru the XO -- I will have to adjust those bands.

In the end -- the amp attenuators probably won't be wide open. I am OK with that since I am the only one running the system.

Thanks again Arthur. I thought for sure more would jump out there on this.... Maybe more later...

[Chris Tsanjoures]

I don't like to turn down the attenuaters on the FOH amps. However, I often turn down the HF attenuater on individual MON amp channel outputs. At a recent gig, where many of the companies drivers varied in age and manufaturer... I found the best solution to balance them out was using the amps output, then fine tuning and ringing with the graphic. The company uses digital amps with internal processing, the racks where networked together and it would have been inapropriate for me to change the companies XO settings...but no one will be upset about turning down a knob a few DB here and there.

[Jay Barracato]

I am bringing up this topic because on some other forums where I hang there is somewhat of a debate on this. I wanted to see if any of you have dealt with this in the field and IF it even matters.

I am not partial to this one way or the other BUT I do have some questions. I currently run with my amp attenuators pulled back a bit from clipping. This means my XO outputs are all set a ZERO.

I could make the adjustments the other way around as well BUT does it matter?

Amp attenuators wide open means no one can turn them up any further. May be nice when the system is being used by several different engineers that have different view points. I could also see it being a bad thing too.

On the DR 260 (probably any DSP Device) there is signal loss after it is split into its specific frequencies range -- high, mid & lows. This is very noticeable on the output meters of the 260. They are all at different levels.

In your opinion -- should the outputs on the DSP match the inputs at this stage or does it matter as long as you aren't clipping at the amp?

This may be a completely dumb question so I will take one for the team here...

I think total gain structure is part of the SE's role and is part of optimizing a system.

However, with that said, gain is gain, and if no part of the system is clipping it shouldn't make a difference if you are adjustijng at the dsp output or the amps input. For simplicities sake, it makes sense to have one set in a semipermanent position and adjust the other as needed. I tend to leave the amps wide open and adjust in the DSP.

That way no one came come along and just "turn up the volume" on the amp, and I find it easier to match levels in the digital dsp than on the dial type input of most traditional amps (out the door with modern digitally controlled amps).

[James Woods]

I just found this thread, an find it interesting, since this is one of my favorite topics. Normally, when doing a gig the amount of stuff I do depends enormously as to what needs to be done, and how much time I have to do it. In my Smaart classes I virtually always end up going over things like system gain, and although it doesn´t have much to do with making Smaart work it does have a lot to do with system optimization, and that is what Smaart was designed to help us do.

So, If you want to maximize your signal to noise ratio, then you need to start thinking about how your system is distributing its signals: your gain structure.

Are you using analog multicore returns or digital?
Is your system processor at FOH or in the racks on stage?
What are your nominal and maximum levels of the other components in your system?
Are you expecting noise from lighting and video or other sources?
What kind of a ground do you have, and is everyone on the same one?

So without going into too much detail, if I am using analog returns, I want to maximize the natural dynamic range of the show. If I am using a digital console with +18dBu outputs (analog "nominal" would be -18dBFS), I want those to maintain the dynamic range at the processor (with +24dBu inputs) and adjust it´s input to read -18dBFS too. This means I would be boosting the signal input by 6dB to maintain the same headroom as the console.

Let´s say that the processor´s outputs are also +24dBu and I am feeding those outputs through an analog snake back to stage to the amp racks.

Let´s say the amps (all of them) have a sensitivity of +4dBu (1.23V). This means that at +4dBu input I am going to get the maximum power output of the amplifier. So, If I want 20dB of headroom on my system I would have to either drop the processor outputs by 20 dB, and thus decrease the signal to noise ratio back to the amps running through the analog lines (and close to lighting power lines) making my signals a prime target for hum and buzz.

However, if I ran the processor outputs at their normal level ("0") and dropped the amp attenuators by 20 db, I would only be lowering their senstivity (not the amplifiers power rating), and would be keeping my signals 20dB higher that the noise floor induced on the multicore returns.

All of this said, if the processores are in the amp rack on stage, and I have a computer link to them for adjusting them with short analog cables from the processor outputs to the amp inputs, it would be practically the same dropping the gain at the processor outputs or the amplifier inputs. It would also make sense then to do it in the processor to keep it simple for the guys starting up the system (who may not have a good working knowledge of the system gain structure).

In a digital world, forget all of above. You will be more interested in clocking your entire system to a master clock, and making sure your network is working flawlessly, but this is another story altogether.

I have never boosted the levels on the individual outputs of my processor so that they are all the same level, and then dropped the amplifier gains by the same amount of processor output boost. There is different energy there in each band because the signal being mixed has different instuments and spectrum content. Also, we need to keep in mind that the efficiency of the different drivers in the system are different too.

In general, the simplest solution is usually the most elegant and the most reliable.

[Langston Holland]

Nicely written James. :)

One of the things that can force me to compromise gain structures in portable systems is the fact that the amp gain knobs may get bumped, thus setting them anywhere other than fully clockwise invites driver damage.

It irritates me that most touring quality amps still have gain knobs that can't be locked out. On non-processed amps, a recessed 4 position dip switch (1 to bypass the gain knobs and 3 to set gain) on the rear panel with 3dB increments spanning the range of 17dB to 35dB would be lovely.

Amp mfg's need to wake up and realize that it's not the 70's anymore.

[Ferrit37]

Hey Guys,
my understanding was that calculating the limit points to match the amp gain to the drivers requirements needs the gain of the amp to be at full, if you turn the amp gain down aren't you upsetting this relationship?

[pepe ferrer]

Hey Guys,
my understanding was that calculating the limit points to match the amp gain to the drivers requirements needs the gain of the amp to be at full, if you turn the amp gain down aren't you upsetting this relationship?

Hello,

I agree with Ferrit, if we modify the amp, we are modifying the voltage gain, then we must recalculate the input sensitivity for the calculation of the limitation.
I think so., But I can be wrong.

Greetings.