The unofficial power amp thread

[Gruuve]

I'm searching for a power amp at the moment, and thought it might be a good idea to capture data and commentary for future use in one thread.

 

BEHRINGER EP-2500:

Recently, I picked up a Behringer EP-2500 at a dirt cheap price to power a new Acme LowB4. Either this power amp was defective, or it just really stunk. Alledgedly, it'll deliver 2400 watts into 4 ohms in bridged-mono mode, but it clipped so easily that I couldn't get it as loud with bridged 2400 watts as it would go with stereo/parallel 800 watts. (Now, maybe I don't know *exactly* what I'm doing, but I think I'm pretty close.) These are cheap (since they reverse-engineered QSC's RMX design), but the old adage of "you get what you pay for" seems to apply in this case. Anyway, I ended up returning it. I'm curious if anyone has had similar results with another EP-2500 or with the QSC RMX-2450 (which the EP-2500 was copied from). Both the 2U-size QSC RMX-2450 and the Behringer EP-2500 copy are quite heavy at about 35 lbs.

 

YORKVILLE AP3400 (discontinued):

I stumbled across some info on an older Yorkville AP3400. Although it's quite heavy at 42 lbs, it might be worth the weight. According to it's manual it should offer some serious transient response. Here's a sample of the specs for stereo mode with both channels driven:

LOAD ContAVG BurstAVG PeakInstant

8ohms 750w 1200w 3000w

4ohms 1200w 2175w 6000w

 

That looks pretty impressive to me. Bridged-mono only supports an 8-ohm load, but with that kind of transient power available, it doesn't seem like there'd be much need to run it in bridged-mono mode. It's quite heavy at 42 lbs, but might be worth the weight since it could easily power two 4-ohm cabs (or four 8-ohm cabs...ouch!) with plenty of headroom.

 

Here's the URL to the full manual: http://www.yorkville.com/downloads/other/omap3400.pdf

 

Is anyone familiar with these Yorkville models?

 

OTHERS:

Many folks also swear by the Peavey DPC-1400, Stewart World 1.2, 1.6, and 2.1 models, and the Mackie 1400i to my understanding, but I'll let folks who've owned those provide commentary.

 

You folks with lots of power amp knowledge please add your wisdom that will help me and others in the future in power amp hunts!

 

Dave

[rumpelstiltskin.]

i can wax intellectual about amps. i have a little experience with them. tnb has recently asked me a very similar question through PM, so after i answer him, i'll post my answer here.

 

if you're ok with a heavier amp for reliability and tone, go out and buy a crest CA6 and be happy. trust me. and if you're into knowing the designer, go out and buy a crest CC 1800. that'll sound pretty good if you're into more vintage bass tones. it's not as tight as the CA6.

 

robb.

[Gruuve]

Hey Robb:

 

I was hoping you'd pop up with some good info!

Dave

[Gruuve]

Two parameters that I see listed with most power amps are slew rate and damping factor. As I understand it, damping factor has to do with how well the amp controls speaker cone movement, so presumably a higher value is better. No idea what slew rate means.

 

Can anyone explain what those mean, how they are relevant to choosing a power amp for a bass rig, and what are good values to look for? And in cases where these aren't listed, what values should we assume? Presumably if it's a really desirable value, it would be listed. (I have a feeling Rob will give us some valuable info on this one!)

 

And finally, what's the best measurement that would tell us about headroom and transient response? The peak power output rating (vs the continuous rating)? What else? Also, in cases where the peak power isn't listed, what's the typical value in relation to the continuous rating? Or does the peak output vary so much by amp design that it's not possible to make any assumptions about it?

 

Dave

[alexclaber]

Slew rate and damping factor both sound important but they rarely are.

 

Slew rate is the rate of change of voltage, so basically how quickly the transistors can react to a change in input voltage. This seems important to deal with bass transients but actually it rarely is, because any amp that can put out full power at 20kHz is going to have to have a high slew rate because the higher the frequency, the higher the rate of change of voltage.

 

Damping factor is the ratio of output impedance the speaker impedance, and it defines how well the amp can control the speaker's movement in terms of restricting overshoot and counteracting momentum. However, in reality the speaker cable impedance also counts as part of the output impedance, and this sends the damping factor through the floor. Unless there is an order of magnitude difference between damping factors of the amps you're comparing I don't believe it'll make much difference.

 

Just wait for robb to start contributing, he knows far too much about all this!

 

Alex

[Dave Brown]

Robb's the man.

 

As I understand it, slew rate is the speed (measured in volts per milliseconds?) the amp can respond to instantaneous change in voltage. That is to say, an amp with a high slew rate is preferable to bass players. Because of the massive changes in power as we articulate, an amp with a slow slew rate dirtys up that articulation.

 

Anyway, I'm no expert. Maybe someone can shed some light. It was my understanding that the Class D (so called digital amps) have extremely fast slew rates, which is why the Acoustic Image and Euphonic Audio amps are so great.)

[picker]

What would be a fast slew rate as opposed to a slow one? What figure or range of figures should you be looking for?

[_Sweet Willie_]

Other stats that may be relevant in choosing power amps are THD (total harmonic distortion -- I think) and how much input voltage they're looking for from the preamp.

 

THD, I believe, is not always reported in consistent ways from manufacturer to manufacturer, so this may be a problematic stat to get a handle on.

 

The voltage thing is about matching your preamp and power amp -- does your preamp have enough oomph to really get the power amp to give you its full power?

 

I may not being writing all of that using the right terminology correctly, but those who are more knowledgeable can expand on the gist.

 

Peace.

--s-uu

[alexclaber]

Originally posted by Dave Sisk:

And finally, what's the best measurement that would tell us about headroom and transient response? The peak power output rating (vs the continuous rating)? What else? Also, in cases where the peak power isn't listed, what's the typical value in relation to the continuous rating? Or does the peak output vary so much by amp design that it's not possible to make any assumptions about it?

This is the difficult one. I think the best thing to go on is the mass of anecdotal evidence from bass players and PA system (particularly for subwoofer usage) operators. But take it all with a pinch of salt.

 

Your best bet is to check the RMS rating from 20-20000Hz at low distortion - that at least gives you the minimum specs. Peak power has to be at least root2 times RMS power. However, some amps will just meet their specs whilst others will exceed them, and the difference at low frequencies can be large.

 

Of course, if you're using an amp with a large amount of power vs the cab's rating, then do you want it to exceed its specs? Once you start getting way above the cab's power handling then thermal compression and excursion limitations make extra power all but redundant.

 

Alex

[_Sweet Willie_]

There's some nice background info on various things power ampish in this thread:

 

 

Peace.

--s-uu

[Gruuve]

Of course, if you're using an amp with a large amount of power vs the cab's rating, then do you want it to exceed its specs? Once you start getting way above the cab's power handling then thermal compression and excursion limitations make extra power all but redundant.

Good point! I think it's most relevant though in deciding between (say) a power amp that can deliver 1200 watts continuous and 1800 peak versus one that can deliver 800 continuous and 2400 peak. Which one would *sound* better? (I dunno, but based on the suggestion that we typically use about 10% of the amp's power for non-transient stuff, I'd hypothesize that the 800 cont/2400 peak would actually *sound* better and probably clip a lot less. I could be very wrong though.)

 

Dave

[tnb]

Hey Rob, go ahead and start talking here, I'm all ears. Besides, someone will come up with better questions than I would have thought to ask.

[Aldena]

I'll second the CA6. I've used mine for both bass and PA duties. Plenty of power, sounds good, never gets hot, but it is rather weighty.

[lug]

I have the Mackie 1400i and have had exactly zero problems with it in about 10 years. I also have a PEavey DPC 1000 (precurser to the 1400) Sounds great but has shut down on me once. Came back after a couple of minutes and never did it again. Don't have nearly as much time on it as the Mackie, but it sure is a lot more fun to carry.

[alexclaber]

Originally posted by Dave Sisk:

I think it's most relevant though in deciding between (say) a power amp that can deliver 1200 watts continuous and 1800 peak versus one that can deliver 800 continuous and 2400 peak. Which one would *sound* better?

This is where you come up against the specs problem. What frequency is that peak at, and what %THD? And even if two amps have equal RMS and peak power levels at equal %THS across the whole frequency spectrum, if one amp mostly creates 2nd order harmonics when distorting it will sound a lot clearer than one producing lots of 5th order harmonics.

 

The Aguilar DB-750 is renowned for being an insanely loud 750W, which I swear is down to massive peak LF power, and possibly because any distortion that does occur is centred around the even order harmonics which are much harder to hear as distortion and tend to just makes things sound louder in a good way (to a point).

 

There generally seems to be a strong correlation between the price/weight of an amp and the amount of apparent power you get in the real (bass) world for the RMS rating that is quoted. The Crest CA6/CA9 is not expensive but it ain't cheap either, and weighs enough to anchor a substantial vessel. The PLX series amps are similar in price but weigh far far less and anecdotal evidence suggests that they are less fat sounding - which to me suggests that they simply don't have as much peak LF power, which in turn suggests that you might as well get a higher rated PLX and save your back!

 

Alex

[way2fat]

Originally posted by lug:

I have the Mackie 1400i and have had exactly zero problems with it in about 10 years. I also have a PEavey DPC 1000 (precurser to the 1400) Sounds great but has shut down on me once. Came back after a couple of minutes and never did it again. Don't have nearly as much time on it as the Mackie, but it sure is a lot more fun to carry.

I have a DPC-1000, too. I burned it up driving a 4 ohm load in the bridged mode that I thought was an 8 ohm load. It was replaced with a Stewart 1.6 that sounds quite a bit better and weighs 16 lbs.

[rumpelstiltskin.]

Originally posted by Dr. Sweet Willie:

Other stats that may be relevant in choosing power amps ... how much input voltage they're looking for from the preamp.

s-uu wins some sort of prize!

 

 

without a doubt, the most overlooked feature in power amps is input sensitivity. and it's not that sensitivity is a critical feature -- there's no magic number or "right" answer -- but you need to know it to know how an amp will work in your system. input sensitivity is the voltage input required to drive the power amp to full output. some amps require more voltage than others, and on a system level, that can become an issue. for example, many preamps produce less than 1V output. the QSC PLX 2402 has an input sensitivity of 1.3V, with a fixed gain of 40x.

 

what does that mean? say your preamp puts out peaks of .7V. that is just over half of the input senstivity of 1.3V. so instead of sourcing the full 700W into 4O per channel, you'll get only 196W. how can that be? .7V x 40 = 28V. 28V into 4O is 28^2 / 4 = 196W.

 

the good news in this is that it's really easy to know whether your preamp has enough output to fully drive your power amp. if you are able to light up the clip lights on your power amp, you are able to fully drive your power amp. if not, you're not getting full power from your amp.

 

as a side note, this was the case in my system -- a crest pro 7200 power amp and a warwick quadruplet preamp. the input sensitivity of the crest is 1.58V, with a fixed gain of 40x. i actually modified my preamp to double its output, and now i can drive my power amp fully without diming the volume knob.

 

sensitivity is part of the reason why so many people love the crest CA series. across the line, the crest CA amps all have the same low .775V input sensitivity, which is easy for just about anything to drive. the gain changes with each model in the CA series.

 

there are two approaches to input sensitivity, each with benefits and downsides. one design theory utilizes the same fixed gain across the live, with higher powered models requiring more voltage to drive them to full power. the benefit here is that regardless of which amps you have on hand, if one amp fails, you can replace it with another amp of equal or greater power without having to restructure your signal chain and gains. if you want it to go louder, turn up the input signal. the hard part, then, is if you run out of gain on the way into the amp, you'll never get full power out of it.

 

you can see how useful such an approach is in concert sound systems.

 

the other design theory utilizes the same fixed input sensitvity across the line, with higher powered models having higher fixed gain to achieve to full power. the benefit there is that if you want more power, you can buy a bigger amp, and you don't have to change a thing about the signal chain upstream. for truly high powered applications, though, the fixed gain of the amp can get really high -- approaching 100x (with 40x being a very common number in the industry). another downside is that if your amp fails, you have to have a direct replacement or risk restructuring your gains and signal chain upstream.

 

but you can see that for an instrument setup how this approach can be preferrable for most musicians. (of course, having knowledge makes the difference relatively minor.)

 

so does all of that make sense? funny how the mundane but important can get buried under sexier, seemingly more important specs like power, THD, damping factor, and slew rate. but i have plenty to say about all of that, too. eventually.

 

robb.

[lug]

Originally posted by way2fat:

Originally posted by lug:

I have the Mackie 1400i and have had exactly zero problems with it in about 10 years. I also have a PEavey DPC 1000 (precurser to the 1400) Sounds great but has shut down on me once. Came back after a couple of minutes and never did it again. Don't have nearly as much time on it as the Mackie, but it sure is a lot more fun to carry.

I have a DPC-1000, too. I burned it up driving a 4 ohm load in the bridged mode that I thought was an 8 ohm load. It was replaced with a Stewart 1.6 that sounds quite a bit better and weighs 16 lbs.

The difference between the 1400 and the 1000 is that the 1400 is rated for a 4 ohm bridged load. The power specs are the same for other loads. I tend to not run bridged but just run 2 cabs in stereo mode. It's not like I need the extra power unless I'm trying to drive cockroaches out fo the place.

[tnb]

without a doubt, the most overlooked feature in power amps is input sensitivity. and it's not that sensitivity is a critical feature -- there's no magic number or "right" answer -- but you need to know it to know how an amp will work in your system. input sensitivity is the voltage input required to drive the power amp to full output. some amps require more voltage than others, and on a system level, that can become an issue. for example, many preamps produce less than 1V output

Wonderful. So I should pick a preamp then a power amp?

 

say your preamp puts out peaks of .7V. that is just over half of the input senstivity of 1.3V. so instead of sourcing the full 700W into 4O per channel, you'll get only 196W. how can that be? .7V x 40 = 28V. 28V into 4O is 28^2 / 4 = 196W.

mongo want know how to make bass loud. mongo confuse easy, eyes all glazed over.

 

Maybe I should have taken more math and science, but for some reason this is getting easier. Do preamps manufacturers publish output of their preamp? If the do, what spec is it under?

[TimR]

The slew rate is important because it predicts how different amps behave compared with each other regarding the high frequency response. Although the amps will probably be capable of reproducing sine waves faithfully at 20Kz, drum hits etc and more complex waves will approximate square waves and these are what will cause distortion.

[alexclaber]

Originally posted by TimR:

Although the amps will probably be capable of reproducing sine waves faithfully at 20Kz, drum hits etc and more complex waves will approximate square waves and these are what will cause distortion.

My understanding is that any wave, including a completely square wave, can be built up of sine waves of varying frequencies and magnitudes. Therefore, if an amp can produce full power at 20kHz then it will be have a sufficiently high slew rate to produce any musical information at any power up to full power, right up to that frequency. I could be wrong though...

 

robb's right about input sensitivity. If you can't make your amp clip, then you will not be able to fully use the amp's power. Most power amps and preamps get on fine - many of those that don't can be easily modded so they do. But if they don't then your 2000W power amp could be reduced to a 500W amp, and that would be rubbish.

 

If you are intending to bridge your amp into 4 ohms this effectively increases your input sensitivity compared to parallel mono at 8 ohms. I think the difference is about 6dB.

 

Alex

[JPJ]

I'm a Stewart guy myself. I've used a 2.1 as my primary gigging amp for the last 4 years or so. I wish that there was a way that I could turn the fan off, though.

[Gruuve]

On the front of mapping the preamp output level to the amp input level...I gather than 1.4V is a common input spec for power amps right? Some preamps (like the BBE BMax) have the maximum output spec given in dBu's rather than volts. For instance, the BMax maximum output is +22 dBu. How do you convert dBu's to volts?

 

Dave

[rumpelstiltskin.]

Originally posted by tnb:

Wonderful. So I should pick a preamp then a power amp?

not necessarily. it's more that you should be aware of the impact of your choices. most gear won't have a problem working with other gear. but it happens sometimes, and you have to be prepared. even my preamp didn't match well with my power amp, so i chose to modify one. there is a QSC dealer who will modify the fixed gain of the amplifier for you if you want them to. apparently because it's a very popular service that they offer.

 

mongo want know how to make bass loud. mongo confuse easy, eyes all glazed over.

 

Maybe I should have taken more math and science, but for some reason this is getting easier. Do preamps manufacturers publish output of their preamp? If the do, what spec is it under? [/qb]

sorry for the math-y-ness. i tend to try to make my explanations visual. what i was trying to demonstrate was that if you're not driving the power amp fully, you can lose significant amounts of power. if you're not aware of such issues, you'll be stuck wondering why your 700W amplifier is quiter than someone else's 300W amplifier. in that case -- a real-world example -- if you don't have a high enough output from your preamp, a 700W amplifier was turned into a 200W amplifier.

 

typically, an output spec is not particularly in demand. most manufacturers generally bury it in the literature somewhere, though. for example, BBE's Bmax-t has a max output of +22dBu (9.75V) according to the manual specifications. ampeg publishes a gain spec for their preamps (23dB, or about 14x, for the SVP-CL), but you have to know the output from your bass to do the math. even SWR, without have a specs page in the interstellar overdrive manual, makes note of the +24dB output. assuming that's +24dBu, the output can go as hot as 12.3V.

 

How do you convert dBu's to volts?

if you don't like math, you'll just have to take dB on faith.

 

for those unfamiliar with how the decibel (dB) scale works, it can be very confusing. a dB is a way of describing a level relative to another level. if you're speaking in terms of dB, there's always a reference value. so if you want to express x in terms of dB, you have to compare it to a reference value. the math looks like this:

 

x in dB = 20 * log ( x / reference)

 

in words, that's twenty times the log of x divided the reference value. to complicate matters, there are several different reference values that can be used, depending upon the application.

 

but that's what the lowercase "u" is for at the end of dBu. that is one of the several standards, and it means the reference value is .775V. so to find out what +22dBu is in actual volts, you can use the equation with .775V as your reference value:

 

+22dBu = 20 log ( xV / .775V)

 

[algebra]

 

x = .775 * 10 ^ 1.1 = 9.75V

 

so +22dBu = 9.75V.

 

the other very common reference value is 1V. this is notated using dBV. that's a capital "V" instead of a lowercase "u". most of your home stereo stuff is -10dBV sensitive. that means it takes .316V to drive them fully. you run into that more with recording equipment. my sound card has switchable levels between +4dBu (1.23V) -- a studio professional standard -- and -10dbV (.316V).

 

i looked at manuals from ampeg, swr, and bbe. all three gave some sort of evidence to output levels. the BBE listed a maximum output of +22dBu on its specs page. ampeg lists the total gain in dB, which means you have to know your instrument's level to figure how loud it can get, but who knows when it clips! ven SWR, without have a specs page in the interstellar overdrive manual, makes note of the +24dB output. assuming that's +24dBu, the output can go as hot as 12.3V. if it were +24dBV, the maximum output would be 15.8V.

 

does that make sense?

 

robb.

[alexclaber]

0 dBu = 0.775 volt (which equals 1mW into a standard 600 ohm load)

 

0 dBV = 1 volt RMS

 

So 22dBu = 10^(22/20)*0.775 = 9.76 volts

 

Plenty hot enough to drive any power amp into clipping.

 

Alex

[way2fat]

I can make my amps lights turn on by cranking the gain. Is there any connection to input sensitivity?

[rumpelstiltskin.]

yes. in general the knobs on amps are not gain, but attenuators. in other words, they only turn the amp down, so the best way to set them is at full. unless you need to throttle back your power amp for some reason, it's usually best to set its knobs at full.

 

robb.

[tnb]

Very cool.

 

Next question.

 

What's this I've been hearing about don't plug your power amp into a power strip or one of those Furman style power strips?

[alexclaber]

Originally posted by tnb:

What's this I've been hearing about don't plug your power amp into a power strip or one of those Furman style power strips?

Power amps demand a lot of current and many of these 'power conditioners' cannot cope with such requirements, and thus end up restricting the flow of power into the amp.

 

Any power amp worth its salt should be tough enough to be plugged straight into the mains - good protection circuitry is a key feature of most power amps.

 

If you really want to baby your amp with a power conditioner make sure that the power conditioner can handle the power. My QSC PLX 3002 will draw 9.5A with 1/8 power pink noise (equivalent to program power with occasional clipping) on US 120V mains (900W @ 4 ohms).

 

With 1/3 power pink noise (equivalent to severe program with heavy clipping) it draws 17A. At absolute max into 4 ohms it draws 33A. It'll take a massive power conditioner to cope with these demands and although you're unlike to see much more than 10A average draw from similar 2000W amps in normal use, the peak draw could be a lot higher.

 

Alex

[Gruuve]

Ah...here's a good find, although it's a bit dated. Bass Player mag's power amp shootout from 1997 (looks like '97...this is apparently scanned, and I can't quite read the small print date, although I think it's 1997).

 

http://www.stewartaudio.com/pdf/paShootout.pdf

 

Dave