Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Bob Lee (QSC)]

Andrew,

andrew wrote on Tue, 31 August 2004 00:43

The Y cable would be used BEFORE the amp - ie, line level signal gets split into two. In theory you get a 3db loss at the input of the amp by doing this but in practice, you're not going to notice it.

No, that's not true.

Quote:

The other common misconception with the gain knob on a power amp is that it's actually a gain knob. In reality it's a reduction knob. With the amp on full, the input signal goes to the power amp unadulterated. As you dial down what you're actually doing is reducing the volume of the input signal to the amp - it's not like you're keeping the same input volume and reducing the amount of work that the amp does.. amp always works the same way you're just cutting the input signal. There is a school of thought that says you set the system gain from mic, desk, eq - and then adjust the power amp setting so that it maintains the ideal. What I mean by this is that in theory, if you were only pushing enough volume that's say about 1/2 the mixer's maximum volume and 1/2 the speakers maximum volume, then it would be cleaner to push the mixer to around 90% of it's maximum volume and turn the power amp down to compensate. You get a better noise floor. In practice... well.. I guess if you're doing the sorts of gigs where you'll be concerned and also be able to measure noise floor - then fine. But otherwise, I'd simply set at maximum. Easier to work with. And if you can tell the difference in most live venues between a mixing desk at full power and amp at half, vs mixer at 1/2 power and amp at full - you're doing really well !

It's not a misconception. It's a gain control, and in most amps it is located after the input section (and processing section, if there is one) and before the output section.

When the control is turned all the way down, the amp channel's gain approaches 1/infinity. When it is all the way up, the channel's gain is whatever its designer and manufacturer have made it to be: 20x, 30x, 40x, 60x 80x, et al.

One common misconception is to refer to fader or gain control settings in terms of "power" or "volume," as in "half power" for a fader or gain control set midway up.

[Andy Peters]

andrew wrote on Tue, 31 August 2004 00:43

The Y cable would be used BEFORE the amp - ie, line level signal gets split into two. In theory you get a 3db loss at the input of the amp by doing this but in practice, you're not going to notice it.

WRONG.

--a

[John Roberts {JR}]

Andy Peters wrote on Tue, 31 August 2004 14:30

andrew wrote on Tue, 31 August 2004 00:43 The Y cable would be used BEFORE the amp - ie, line level signal gets split into two. In theory you get a 3db loss at the input of the amp by doing this but in practice, you're not going to notice it. WRONG. --a

Well he did say you wouldn't notice the loss....  

Perhaps he was thinking of a constant impedance (600 ohm) transformer split, but that would be half voltage (-6dB) in each input due to the split.

But then again, when was the last time you saw a unit with 600 ohm source impedance and two other units with 600 ohm input impedances in the same building, at the same time, let alone hooked up to each other with an impedance matching transformer?

Nah, I think you got it right.

JR

[andrew gissing]

Wow.. i've learned a lot.

My understanding of amp gain controls was completely wrong - and i've got 4 QSC amps too !

With the 3db loss thing - where am I going wrong ? I was under the impression that if you took a mic and split it, you get a loss of about 3 (maybe 6?) but you can compensate for that by using gain knob on mixer. And extending that, I thought same would happen for Y splits later on.

Am I wrong here ? Please tell me if I am - and what the correct theory is !

[John Roberts {JR}]

No theory but there are laws or rules that define how signals act under different terminations.

A microphone when hardwire "Y"'d into multiple inputs will be loaded down somewhat by the additional terminations. However if the preamps are properly designed (bridging or 10x the mic's source impedance) the loss will be minimal (1.5 dB or so).

In the case of a line level hardwire "Y" the following inputs should also be relatively high impedance so loading will be insignificant.

There are measurable calculable losses in both cases, but less than a few dB and not significant.

JR

[Kent Stallard]

Thanks to everyone for the responses thus far.

nullset: Thanks for the formula on impedance.

analog Tom: Great post and excellent advice!  I'm getting the pencil and paper ready...

Bob Lee: First of all, Bob, thanks for overlooking the fact that I'm using Crown amps...  

Quote:

Power amps usually have more gain than most situations require. For optimum signal-to-noise, you would turn the amp down and compensate by sending a more nearly normal line-level signal into it.

OK, I understand that concept.  However, what would you recommend for a starting point on the gain controls--center ('detent') position, or all the way 'down'??  

[Lee Patzius]

Kent Stallard wrote on Tue, 31 August 2004 23:31

OK, I understand that concept.  However, what would you recommend for a starting point on the gain controls--center ('detent') position, or all the way 'down'??

You need to set gain structure. I noticed you had a dbx DRPA. dbx describes gain structure as setting gains throughout the whole system to clip simultaneously. But simultaneous clip has been described as a bad idea for many reasons. 1) It would sound most horrible if ever driven to whole system clip. 2) And most likely it wouldn't happen anyway, because the weakest link in the chain will clip first, unless you had a perfectly matched signal chain, with equal intensity clipping points, coupled with a perfectly balanced sub/mid/highs power amp and speaker ratio, dialed in precisely for the music style.

In other words, a lot of music is driven kick and bass heavy...Power hungry subs may clip first...Depends on your system, and it will vary somewhat between conditions. Don't get too caught up in the everything must clip at the same time method. But for starters, I'll WILL start with simultaneous clip within the dbx DRPA first.

I've written a procedure a while back on dbx's site that I use for DRPA users. The following has a few cut and paist inserts...and I noticed a little extra redundant statements thrown in here too. It's just that I editted my article with quick add-ons, only to find I just said that elsewhere.


Here we go:

YOU MUST AGREE to be the type of person who will turn down each of the power amp trim knobs from full blast as needed, otherwise the following method I'm suggesting is useless.

Also, you should have a carefully selected and balanced ratio of power amps and speakers per each low, mid, and high frequency range. If you have a mismatched system, such as under-powered subs, the method I describe below will still work, but you'll need to set your gain structure when the weakest link clips first.

One more thing, in the DRPA, I prefer "custom" amplifier settings, even though my crown CE2000 IS in the menu settings.

To start with, I use the drpa's internal pink noise generator as an initial noise source WITHOUT an rta mic connected, to set my crossover gains inside the drpa. Repeat: Without an rta mic connected. Try pushing the rta mic button without going into the wizard. You will notice the screen immediately goes to auto eq mode.

The objective is to make all the crossover output LED's clip at the same time when you crank the pink noise knob, (with no rta mic connected). This is my personal method of getting maximum gain and dsp resolution out of each of the drpa's crossovers.

Turn down the power amp gains real LOW, for monitoring the pink noise out of the dbx.

Set your dbx crossover gains to zero dB, filter slopes and peq to speaker specs.

Turn ALL the extra stuff like anti-feedback suppression (afs), sub synth, compressors, limiters, delay, to OFF (for now).

Set your EQ to flat. VERY IMPORTANT!!!!!!!!
Repeating: Set your EQ (geq) to flat. This effects the un-mic'ed pink noise generator to start flat, instead of starting with last known geq settings.

Push the rta mic button, do NOT connect a mic, and do NOT push the wizard button. Go through the menu (right arrow) and rotate the pink level up slowly, like up to 0dBu or so. The pink noise generated within the drpa will be flat because the EQ (GEQ) is flat. It is looking for the rta mic which ain't there, so it'll hold a steady flat-lined pink level. NOTICE: The drpa's input leds will NOT be illuminated...No mic that's why.

Look at your crossover output leds. Are they all equal in intensity? Rotate the pink level up higher. Do they all clip at the same time? Take note of weak and strong leds and push the rta mic button again, to bail out and stop the pink noise. Go into the crossover section and adjust weak gains up, and powerful gains down. Repeat all the previous steps from the start over and over again while experimenting with, and adjusting your crossover gains. Make it so all the crossover output gains clip at the same time with flat-lined pink noise, with manufacturers recommended crossover slopes and peq's inserted.

Now it's time to set your power amp input trims using the above method (with dbx's "flat-lined pink noise" generator, NO rta mic connected) AND no speakers connected, to make your power amps clip precisely when its associated crossover's output led clips. Remember, this is ONLY if you have a perfectly balanced and matched system! (Not all power amps may clip at the same time.)

Crank the pink noise to occasional clip on all of the drpa's output leds. One at a time...slowly bring each power amp input trim up until the amp clips. (REMEMBER: NO SPEAKERS CONNECTED)

At this point, you have now set the amp's input trims, and crossovers, using pink noise with a flat line eq, and this is your absolute starting point.

From here on out, when you add in, or color your changes, like sub synth etc., go back and make the crossover output led's dance in equal power again. You will need to constantly go back and re-balance your crossover gains until you've finalized a gain structure that satisfies quick auto-eq sessions. Note: It is recommended to do these gain structure (and auto eq) baseline settings outdoors first.

Next...You must set the limiter to extinguish the power amp clip lights. Turn on your limiter in the menu, but this time crank pink noise in from another flat-lined pink noise source, such as "Bink's" test cd, and run it through a flat EQ'd channel on your mixing console, PFL'd to 0dBu, and cranked to 0dBu on the outputs. Look at your dbx input led's, make sure your test cd level on the dbx INPUT leds are near equal in intensity as the console's OUTPUT leds. You may have to push the +4/-10dB button in back. If possible, bring up the pink noise to +20dBu on the input make the led's occasionally flash the clip. Your power amps should be doing the same. Adjust each of your output limiters down from infinity, until your power amp indicator just stops clipping. (remember no speakers yet)

NOTE: From the drpa to the power amps, your gain structure is set. But you still have to make a decision based on your mixing console's output intensity. Different manufacturer's console outputs vary dramatically from its output verses the drpa's input, there is no input trim on the dbx unit except for the +4/-10dB button. So there may be a discrepancy between console clip, and drpa clip, if the console output doesn't match drpa's 20dBu max input.)

Now it's time to connect your speakers.

Run your auto-eq wizard WITH an rta mic and view the pattern, (even aborted rta's too). Make inversely proportional adjustments to your power amp input trims as needed, to reduce the "heavy handed" portions of your EQ patterns. It may take some experimenting with crossover gains and slopes too, especially if you experience EQ patterns that cannot be corrected with power amp input trims alone.

You'll find that when you make adjustments, it effects the whole signal chain. Gain structure should be like step-by-step and half-step process, with smaller and smaller adjustments as you repeat the process and get closer to unity and max gain structure.

Watch out for drpa's "whole-system compressor" menu, they can throw you off sometimes, especially during a new system setup. Keep them off during gain structure procedures.

Gotta go it's late.

[Kent Stallard]

Lee:

Thanks a lot for that!  I appreciate you taking the time to post it. I've printed up a copy to use at the p.a. cabinet.

I'll likely have some questions after I digest it all.

[Lee Patzius]

lee patzius wrote on Wed, 01 September 2004 03:18

Crank the pink noise to occasional clip on all of the drpa's output leds. One at a time...slowly bring each power amp input trim up until the amp clips. (REMEMBER: NO SPEAKERS CONNECTED) At this point, you have now set the amp's input trims, and crossovers, using pink noise with a flat line eq, and this is your absolute starting point.

Final Notes:

1) The above step is using the "all clip at same time" gain structure mentality as suggested by dbx in their manual.

*Instead* of flashing red clips on dbx's cross over output led's, you may want to bring the pink noise up to a lower reference point, such as in the yellow "warning" led zones (+15dB). You can even set the outputs to a more safer green level too (+5dB or +10dB) if you wanted to, especially if you find that your power amp's input trim is too far counter clockwise. Depends on alot of factors.

Most importantly, it would be ideal that the power amps clip before the dbx unit does.

2) Your limiters are of a concern now because you have an easily rotatable power amp input trim knob. After you set your limiters for baseline gain structure at flat EQ, with a certain input trim clocking position, you will find that different venues will produce different Auto-EQ patterns, which may throw your system "taste" out of balance.

For example purposes *ONLY*: After an auto eq, lets say your subs are weaker...You *CAN* manually raise the power amp input trim up to the point of occasional clip, then back it down slightly to the safe limited zone, *remember* the limiter is still set from baseline gain structure settings. Manually adjusting trim knobs for venue taste is much nicer instead of going into dbx menus.


I'll add in more later gotta go again.

[RobertOziemkowski]

Kent Stallard wrote on Tue, 31 August 2004 06:06

1. I've been told by other musicians (not necessarily sound engineers) that it's always best to set the output levels on the amplifiers to maximum; that doing otherwise limits the effectiveness of the amp.  Is that true? 2. Our band is currently using an old Crown PB-2 amp to power only our two passive monitors, which are both 8 ohm impedance and daisy-chained in series for a mono signal.  I was told that I should bridge the output since the other channel is not being used.  But I also understand that when the amp is bridged, the load becomes more critical.  I plan to eventually add a third 8 ohm monitor.  Is bridging the amp a good idea for use with two 8 ohm monitors?  three?? (I admittedly don't know the formula for impedance when loads are chained in series) The PB-2 is rated at 320W/Ch for 8 ohm stereo, and 965W/Ch for 8 ohm bridged mono.   I'd appreciate any information and advice offered.

Are you getting enough power to the monitors when running both on 1 ch?

If not, run bridged for the 2 mons. But you'll need another amp if you add a third!

If yes, you're set to go.

NO "Y" necessary! On the back of the amp is a switch. Set to Parallel-Mono, use ch1 input!

Now you are sending the same input to both sides. Connect each mon to a side.

When you get a third mon, run 2 on one side @ 4ohms, 1 @ 8ohms. Now use your "gain", or "trim" to balance the volume.