Postby Marty McCann » Wed Sep 02, 2009 12:01 am
When you ask for help regarding presets or general applications, it is very important that you inform us of every piece of equipment that you plan on using in the audio chain.
In order for presets to work, you MUST have power amplifiers that have the same amount of fixed gain. I have to deal with this issue several times a day. I am beginning to get tired and frustrated myself. Not enough however to not write this at 10:30 pm from my home computer.
All Peavey amplifiers have a fixed Voltage gain of Times Forty (X's 40) which is +32 dB of gain on the decibel scale. The Voltage gain in some models of CREST are X's 40, some are X's 20 (+26 dB), and some have switchable gain settings that include common sensitivity of 0.775 Volts.
The sensitivity of a power amplifier is the specified Voltage at the input(with the level control wide open) that will cause the amplifier to reach full rated power. An amplifier that has a rated sensitivity of 1 Volt at 4 Ohms and has a Voltage gain of X's 40 will then deliver 40 Volts into the 4 Ohm load. W = V^2 / R so 40 x 40 / 4 = 1600 / 4 = 400 Watts.
It is also important to understand that even though the amplifier may have a fixed Voltage gain, the input sensitivity will be different depending on the impedance of the load. This is due to the fact that the available output Voltage changes to a lower value in Volts when you lower the load impedance.
I will use a CS-4000 as an example. The CS-4000 is rated at 800 Watts per channel at 8 Ohms and 1350 Watts per channel at 4 Ohms. So to find the input sensitivity (again level control wide open) you must use an inversion of Ohms Law to find the Voltage that will produce the rated power into the stated impedance load. Ohm’s Law says that the power in Watts is equal to the Voltage squared divided by the resistance or impedance of the load (V^2 / R = W). So the inversion would be the Voltage squared is equal to the power in Watts times the load resistance (V^2 = W x R).
So to find the sensitivity of the CS-4000 when it delivers 800 Watt into 8 Ohms; you multiply 800 x 8 = 6400. The Sq Rt of 6400 = 80 Volts. Now since this amp has a Voltage gain of X’s 40 we divide 80 / 40 = 2. So the sensitivity into 8 Ohms is 2 Volts.
To find the sensitivity in Volts to produce 1350 Watts into 4 Ohms: 1350 x 4 = 5400. Sq Rt of 5400 = 73.485 Volts. 73.485 / 40 = 1.837 Volts. Now this would often be rounded off to 1.8 Volts, however 1.8 x 40 = 72 and 72 x 72 / 4 = 5184 / 4 =1296. So you can see that when using math that involves squaring of numbers, it is best to not round off too soon or too much.
So with the above mathematical foundation you should be able to use the published specifications of any brand of power amplifier to figure out for yourself what the Voltage gain might be. I would suggest that each of you urge manufactures to publish the sensitivity into all rated load impedances. Ink is cheap, electrons and bits on the Internet, are cheaper yet.
Now for the bad news (for some of you). There are manufacturers that sell power amplifiers with Voltage gains that are NOT fixed within a model line, “As they should be” (my quote) in our 21st century where we routinely employ digital signal processors. One way to quickly is to examine the specifications. If an model line has say four models with 4 different power levels (example 500W, 760W, 1065W, and 1350W into 4 Ohms), yet if all four models have the identical input sensitivy of say 0.775 Volts or 1.2 Volts, then the gain is NOT fixed within that model line. If the input sensitivity were 0.775 Volts, then in order the Voltage gains would be: X's 57.7, X's 71, X's 84.2, and X's 94.8 respectively. Drive these four amps with a DSP that is expecting a common fixed Voltage gain, and you will have severe frequency response and reliability issues.
Two of our biggest competitors are of course Crown and QSC. Well these are of course very reputable manufacturers, however (I am sorry to report) that many of there products within a given model line do NOT have a fixed Voltage gain. Other manufacturers resemble this remark also.
I will edit this tomorrow and give some examples, but my slow a** dial-up connection will not allow me to download the references that I need. I can only say at this time that I have seen models with a Voltage gain as high as X’s 126 ( and higher). 20 Log V1/V2 = X dB difference in gain. 126 / 40 = 3.125 Log 3.125 = 0.49485 = 9.897 dB. So if you had one of those manufacturers amplifier models that had a X’s 126 Voltage gain, instead of a X’s 40, your factory preset would be off by +10 dB. On more than one occasion I have talked to customers that had just such an amplifier on the high frequency ribbon drivers in the VersArray VR-112 enclosures.
Now, if you must use amplifier that are NOT fixed in Voltage gain within a line; LEARN TO DO THE F’n MATH, or buy a Peavey or a CREST power amp.
Okay, okay don’t blow a gasket, I’ll teach you how to measure and compute any manufacturers power amplifiers gain structure. I would not expect every Saturday night warrior band to have a multimeter. But if you are a regional sound system provider and you do not own or understand the operation of a basic multimeter, then I am sorry my friend, you are simply not yet a Professional. The multimeter is the professional sound mans best friend. Now us old farts may still refer to them as VOM’s or Volt-Ohm-Meters. They are not very expensive, as the cheapest are under $20.00 U.S. Yes they are cheap, but better a cheap one than none at all. I recommend the Fluke model 10.
So you have a VSX-26, VSX-48, or Digitool MX and you want to find the gain of some power amplifier that you would like to use in your system. Use the signal generator function in the above DSP processors and assign a 1 kHz sine wave to an output. Measure the output Voltage of the sinewave and adjust the output level of the generator until you measure 1 Volt RMS between pin #’s 2 & 3 in the balanced cable going to the amp.
(You say your amplifier does NOT have a balanced input? Danger, Danger Will Robinson!!! Many DSP devices will not drive an unbalanced input. Unbalanced is Unprofessional.)
With the amplifier’s level control set to wide open; measure the Voltage at the output of the amplifier, if it reads 20 Volts you have a X’s 20 gain, 40 Volts = X’s 40, and so on. So you measured and read a Voltage gain of X’s 62 (or some other). Simply turn down the sensitivity or level control on the amplifier until you measure 40 Volts on your trusty multimeter. You now have a X’s 40 Voltage gain. Borrow your wife’s or girlfriends fingernail polish and mark a small dot, as a benchmark, so you will know the level setting for X’s 40 gain in the future. You might also want to mark the X's 20 position on the knob also. (Oh I forgot, some of you guys can use your own nail polish).
In case you need to drive the amplifier harder for lighter applications (such as Jazz versus Rock & Roll), the X's 20 position will allow you to drive or hit the DSP inputs with a hotter signal as X's 20 (+26 dB gain) is a -6 dB attenuation or pad below X's 40 (+32 dB gain). This forces the DSP engine to use more bits in the A/D D/A conversions. More Bits = More Resolution.
You do NOT need to have a load on the amplifier to make this measurement. Please be advised also that turning down the input level control on any amplifier does NOT reduce the output power capability of that amplifier. I am still amazed of the number of tech’s with years of operator experience, that are still so naive as to believe that turning down the level control limits the power to the loudspeaker. It does NOT limit the amplifiers full power capability, it only changes the amount of Voltage at the input that will drive it to full power. Example; you have a 500 Watt power amp that has an input sensitivity of 1 Volt RMS, so you turn the input level control down -6 dB; the amplifier is now sensitive to 2 Volts RMS, and WILL reach 500 Watts at 2 Volts of input signal.
Okay, it is now 11:50 pm, and I am going to quit for now, even though I will edit this in the office sometime tomorrow. Guys and girls, I do not mean to hurt anyone’s feelings. If you want to play with the big guys, you simply have to learn everything that you can.
Generally I try and have a lot of patience with customers, but sometimes it can be very trying. One example, I recently had a customer (who has the job title of “Systems Integrator”) call up and complain that his Architectural Acoustics IDL-1000 was muffled and distorted. He said that it even sounded bad when it was in bypass mode. In the IDL-1000, when it has the power is removed, the internal (N.C.) relays connect the inputs directly to the outputs. Now some of you got it real quickly. If you didn’t “get it”, that is equivilent to a hardwired by pass, it’s the same as taking the Male Input XLR in a system and plugging it directly into the Female Output XLR. Although in this case it has Phoenix or Euro Block connectors for ins and outs.
Now I knew that this “systems integrators” problem was somewhere up stream in his system, so I asked him what was feeding or driving the IDL-1000. He told me that it was a TOA Mixer (D 901). I asked him if there was a graphic EQ between the output of the mixer and the input of the IDL-1000. He said no. It was like pulling teeth but with more Q&A, he revealed to me that the TOA mixer was feeding a Shure processor (DFR-22), that was then feeding Bose processor (Panarray). All three of these products were devices with DSP engines. And he was trying to feed an IDL-1000, which is a digital delay DSP. How about A to D + A to D + A to D + A to D? Eight, count em, Eight conversions. When you are dithering the dither (d^2), and dithering again (d^3), and dithering those dithers (d^4), I believe even Mr. Nyquist would have something negative to say about that.
When I was young I thought that a Dither was an Appalachian instrument. I also thought that Group Delay meant that the band was going to be starting late. ;>)
Okay it’s now 12:10 am. I am out of here for now. I will make some edits tomorrow,
(Yawn!)
9-02-09 11:38 am. I edited the above (mainly for typo's and some content), I also plan on editing some more, but I am tied up today with a pretty big church install ($260K).
Oh and if the post time doesn't match, our Website is off by 1 hour.
Kevin C?
Last edited by
Marty McCann on Wed Sep 02, 2009 11:37 am, edited 1 time in total.