Print

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

[John Roberts {JR}]

Your experience explains the rationale behind the 1850. Somewhat less peak power but the ability to deliver what it can for a longer duty cycle.
------------
Allow me to offer some perspective here that may bring some of this into focus. Once upon a time, power amp companies didn't have multiple series of power amps, and X watts meant X watts. The old CS800 would put out 800W all day and all night, which was fine for back when $1 a watt was a good price, but since then the competitive marketplace has changed somewhat. Some enterprising amplifier maker discovered that they could carve a few pounds out of the transformer and heatsink, use a thermal breaker instead of a fuse, open up the current limiting, and voila that 800W power amp can now make 1200-1500W (for at least a few minutes   ).  The real pro's weren't fooled but a surprising number of customers were and apparently still are satisfied with the reduced duty cycle of this new breed of amplifier. I euphemistically call this leaning out the window contest "Value" amplifiers, but they are just the pinto's of the amplifier world. These days all the major amp manufacturer's have their "Value" lines and "Professional" lines. This doesn't mean that the value products are bad per se. They just are what they are. Understand what you need or can afford and buy appropriately.  

Back before Bink's big power amp shootout I mused about the benefit of developing a test to quantify thermal headroom but was a mob of only one and stopped after gaining no traction. I still believe it would be useful, and probably confusing to some, to have a number for that metric.  

JR

[DELETED]

[Bob Lee (QSC)]

RL wrote on Wed, 22 December 2004 14:32

True, but the same result could be achieved by not driving a 2450 quite as hard, and having some headroom in the process. (and more options for uses down the road).  I'm seeing a very small price difference between the two so why not go for the 2450, and run it conservatively?

A similar result, but not quite the same. The RMX 2450 in that situation would still run a little warmer than the RMX 1850HD, but, as you said, with a little more headroom.

In practice, if you get the sound levels you need without clipping the amp, you don't need more power for the situation. But if you are designing your system and are unsure how much power you'll need, it's better to err a little high than a little low.

Quote:

Also, remember that today's power amps are benchmarked to different standards than those of yester-year.  Many amplifiers of years gone by were rated using continuous power output of a sine wave or some other power sucking waveform.  Today, amplifiers are benchmarked using pink noise, and even then, music is so much more dynamic than anything that can be simulated.  They are also tested with a continuous sine wave for safety ratings and maximum current draw ratings.

No, pink noise is not a suitable signal for amp maximum power measurements; sine waves are. Pink noise is used for current draw and thermal loss measurements.

[Bob Lee (QSC)]

RL wrote on Tue, 21 December 2004 17:38

I realize the 1850 is a class H like it's bigger sibling the 2450.  The 1450 is class AB.  According to the articles here on PSW, that doesn't make it much more efficient however.  I do see that QSC lists the same power requirements for the 1850HD as the 1450.  That tells me there is probably a more eff. power supply in it.

It is not a more efficient power supply that accounts for current draw of the RMX 1850HD in relation to that of the RMX 1450. It is, in fact, the class H output circuitry that is more efficient.

What article says that class H is not more efficient than a comparable class AB circuit? It would be in error, because the only reason for class H's existance is to boost efficiency.

[John Roberts {JR}]

RL wrote on Wed, 22 December 2004 16:32

True, but the same result could be achieved by not driving a 2450 quite as hard, and having some headroom in the process. (and more options for uses down the road).  I'm seeing a very small price difference between the two so why not go for the 2450, and run it conservatively?

Not unless by not driving as hard you mean running the 2450 into a higher impedance speaker load. As previously explained, the 1850 is a 2450 with lower rail voltages. For similar speaker loading at any signal level below clipping the 2450 is dissipating noticeably more heat because of the higher rail voltages (15% of the 2450 rail voltage times the output current) so will thermal out sooner. Of course you don't get something for nothing and the 1850 will clip before the 2450 on peaks, making this compromise only worth consideration for less dynamic sub applications.  

Quote:

Also, remember that today's power amps are benchmarked to different standards than those of yester-year.  Many amplifiers of years gone by were rated using continuous power output of a sine wave or some other power sucking waveform.  Today, amplifiers are benchmarked using pink noise, and even then, music is so much more dynamic than anything that can be simulated.  They are also tested with a continuous sine wave for safety ratings and maximum current draw ratings. The transient response of an amplifier (probably a closer approximation to most music) can be very different and is a harder thing to measure.  The narrower the transient the more infinite the bandwidth becomes. With a properly designed PA, that is thought out, limited, filtered, and carefully run, I don't think thermal shutdown is that much of an issue.  If it has become an issue, then your needs are outclassing your gear, generally speaking.

AFAIK power amps are still specified using sine wave power albeit not continuous. Perhaps you're thinking of loudspeakers which use band passed noise sources with known crest factors for power specifications.

What do you have in mind for transient response? Are you talking about "rise time" or "slew rate". The amplifier slew rate wars came and left back in the 70's. I prefer rise time to slew rate as it allows for some real world band passing of the amplifier's transfer function rather than slew rate which is usually measured in an overdriven, essentially fault mode. During slew rate limiting the negative feedback is not even active, while rise time is constant and independent of drive conditions. IMO the bad news is that most manufacturers still spec slew rate instead of rise time and most consumers still think more is better without much sense for how much is enough.

Regarding thermal shutdown not being an issue, that's pretty much what this thread, and many threads on the LAB are about, precisely because modern amplifiers designed with sharp pencils and run down to 2 ohms just don't hang like the good old amps with more iron/copper/aluminum and only running 4 ohm loads.

JR

PS: This was a test right..? you were trying to bait me make wise cracks about your post... It didn't work.  

[DELETED]

[DELETED]

[Chuck Safrit]

All i was looking for was a sugestion on a brand and a series of amps that were inexpensive but still of quality. as far as wattage goes and which one is wired to handle how much head room is best left for the people who are more advanced than i am. All i know is that i cant aford anything right now on my budget that will blow my speakers. I was just looking for a series that would be able to get the most bang for the buck as i think somebody in like the first or second reply answered. Im probably going to go either with the QSC RMS series or maybe the XLS series from crown as i can get real good prices on thoes from where i work (i work for a music dealer i get great prices   well anyway thanks for all the help guys. I think its time to stop the fighting now.

Chuck

[DELETED]

[John Roberts {JR}]

IMO the article is flawed in that it oversimplifies the characterization of efficiency.

Imagine if we were talking about a car's gas mileage. This is like making a claim that a car will deliver precisely 25 miles per gallon and another 30 without consideration for conditions they were driven at.

Such statements at best are useful for crude comparisons but flawed if the test conditions don't mimic the real life application. In the case of music, looking at efficiency at only one power level is not very meaningful.

Since he doesn't state the conditions that he is using to make the comparison I will guess that he may be talking about max clean sine wave power. Although this isn't a perfect fit to all of his examples I don't know what he is using as reference class D or class H designs. He seems to have ignored class G which will be similar to H in efficiency.

While this could spiral into minutia exploring all the ways it could be wrong, lets jump to perhaps more useful generalizations about efficiency (FWIW over the years I have written on the subject of amplifier classes a few times and I think one article from several years ago may still be on the PV website).

First lets eliminate the losers. Class A is a power hog and not even worth considering for much beyond 10W tweaky hifi and transistor radios... I think Class C only works in some resonant modes and only useful for broadcast type applications.

The primary contenders appropriate for audio use are Class AB, Class G (or H), and Class D. Class AB is the simplest (read cheapest) so commonly used up to power levels where the cost savings (due to reduced heatsinks and smaller power transformer) of the higher efficiency exceeds the cost of the increased circuit complexity of Class G/H. Class D promises to be the most efficient and most expensive based on today's technology, but at some point in the future Class D will likely encroach downward on Class G/H and eventually AB as device technology improves (read gets cheaper).

To momentarily switch back to the gas mileage analogy, we can think of the class A/B amplifier as a truck with a one speed transmission. The Class G/H as a truck with a 2 or 3 speed transmission, and the Class D as a truck with a continuously variable transmission. If you pick the right test condition they might all display the same gas mileage but music is like driving over hills and through valleys at different speeds. The truck with a multiple speed transmission will kick ass over the one gear truck, while the digital truck will be even better. But if your truck with the digital transmission breaks you're screwed   . The digital truck really rocks but there's a gotcha in that the complexity of the transmission causes it to waste a little more power while standing still so it needs to be moving to exhibit it's true benefit.

Getting back to reality, well close enough, audio. We will find class A/B effective up to approx 1500W (4ohm bridged) above which class G/H becomes cost effective. Class D for now at least, will remain a higher power product not because of efficiency thresholds but because the cost to carry the technology would not be supported at lower price/power points.  

There is no simple single number that precisely characterizes the efficiency of a power amplifier class, and even the published curves using sine waves at various levels is not music. There is a little bit of "art" or more precisely informed guesswork involved in selecting the intermediate rail voltages in class G/H as this will significantly impact the real world efficiency of such designs when used with different crest factor music. This is a bit like the best transmission for drag racing being different than the best for hauling hogs. While it is actually possible to regulate this intermediate voltage to tune the efficiency for actual conditions AFAIK this isn't done probably because of complexity, cost, and just being "something else to break".

If you review curves for efficiency vs. power level for different class designs this may give you some sense for their performance but again this all must be factored for the complexity of real music.

This is not as clear as I'd like but my dinner is getting cold..

JR