Print

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

[John Roberts {JR}]

Chuck Fry wrote on Thu, 07 February 2008 12:43

I found this "smackdown" thread via a link from a DIYaudio.com thread: http://srforums.prosoundweb.com/index.php/m/229083/19784/ Now what would explain the lower power output at low frequencies of 3 of the 4 amps tested? The PLX 3402 puts out literally twice the power at 20 KHz that it does at 20 Hz. Even the Crest 9001 has a dip in the power curve at 50 and 200 Hz - in the heart of subwoofer country. The burst test should be pulling the same energy from the power supply irrespective of frequency, shouldn't it? Answer that question and maybe we'll have a clue what we're really hearing from amps on sub duty. (BTW this is not intended as bashing any particular amp maker - I happen to own a PLX 3402 and have been happy with it as part of a bass guitar rig.)

While I am not familiar with than particular amp, the general mechanism that explains why an amplifier has lower power output at 20 Hz than 20kHz has been well described here several times.

The PS capacitors in conventional PS are charged at a rep rate of 2x mains frequency (100 or 120Hz). Pure tones much higher than that charging rep rate pull roughly half the time from the + supply and half the time from the - supply. Pure tones that are much lower than the charging rep rate appear to alternately pull from one supply for several charging cycles, then from the other. This results in an effective 2:1 difference in short term PS draw.

Max power is defined by clipping, which in these cases is limited by the bottom of the ripple voltage trough. Depending on how the PS caps are sized it's easy to see a measurable difference between frequency extremes.

For LF applications, amplifiers should be specified based on their rated LF output. If that's still confusing, I can't help you.

Note: real world waveforms are rarely pure tones, but in a narrow band sub application the "one or the other" PS draw phenomenon occurs. It's also worth note, bridging draws from both supplies all the time and will deliver slightly more power output than dual mono...   Now please lets not bring that up again too...... A dB here or there will not make a huge difference, except on paper.

JR

[Chuck Fry]

John Roberts  {JR} wrote on Thu, 07 February 2008 11:08

While I am not familiar with than particular amp, the general mechanism that explains why an amplifier has lower power output at 20 Hz than 20kHz has been well described here several times. The PS capacitors in conventional PS are charged at a rep rate of 2x mains frequency (100 or 120Hz). Pure tones much higher than that charging rep rate pull roughly half the time from the + supply and half the time from the - supply. Pure tones that are much lower than the charging rep rate appear to alternately pull from one supply for several charging cycles, then from the other. This results in an effective 2:1 difference in short term PS draw.

In the referenced test, the PLX is driving a 4 ohm load bridged. This loads both power supply rails equally. So I don't quite understand how your description explains the test results. I could see that the peaks of the 20 Hz tone appear as a continuous load on the power supply where the higher tones are seen as a lesser load.

I should probably ping you and Bob Lee offline to try and puzzle this out.

[Bob Lee (QSC)]

Alexander B Larsson wrote on Tue, 05 February 2008 10:46

And that amps that on paper are indistinguishable still do not sound the same, even within their linear operation (as in no clipping/current limiting).

Among the biggest problems are that when we compare amps …

• they're seldom compared side-by-side.
• they're seldom compared with gains matched--i.e., at the same SPL from one to another.
• they're seldom compared blind--i.e., without knowing which one we're listening to.

Yes, you can take two amps that are indistinguishable on paper, even two of the same model, and they'll sound different if they're set to different gains.

[Pascal Pincosy]

Bob Lee (QSC) wrote on Thu, 07 February 2008 12:35

Among the biggest problems are that when we compare amps … they're seldom compared side-by-side. they're seldom compared with gains matched--i.e., at the same SPL from one to another. they're seldom compared blind--i.e., without knowing which one we're listening to.

I did an amp shootout a few months ago where we did all of the above properly. The results? Entirely inconclusive. Buy amps based on price, feature-set, weight, and brand rep. Don't bother trying to find the best-sounding amp. Your results will change based on the source material used, the speakers used, and the phase of the moon.

http://srforums.prosoundweb.com/index.php/t/27380/9721/

PS I am, however, in full support of Langston's smackdown testing and I'd like to see it reproduced for other amps on the market. If an amp puts out half its rated power at a frequency you're trying to reproduce, you're losing 3dB of output. That's a significant loss.

[SteveKirby]

Pascal Pincosy wrote on Thu, 07 February 2008 16:05

Bob Lee (QSC) wrote on Thu, 07 February 2008 12:35 Among the biggest problems are that when we compare amps … they're seldom compared side-by-side. they're seldom compared with gains matched--i.e., at the same SPL from one to another. they're seldom compared blind--i.e., without knowing which one we're listening to. I did an amp shootout a few months ago where we did all of the above properly. The results? Entirely inconclusive. Buy amps based on price, feature-set, weight, and brand rep. Don't bother trying to find the best-sounding amp. Your results will change based on the source material used, the speakers used, and the phase of the moon. http://srforums.prosoundweb.com/index.php/t/27380/9721/

That shootout did notice a subtle different when used for subs.  That actually helped me make my choice to get a Crest Pro9200, which has worked very well for me.  In spite of it being one of those new fangled switchers.  If we weren't on opposite coasts, it might be fun to compare it to Bob's big iron.

[Pascal Pincosy]

SteveKirby wrote on Thu, 07 February 2008 14:10

That shootout did notice a subtle different when used for subs.  That actually helped me make my choice to get a Crest Pro9200, which has worked very well for me.  In spite of it being one of those new fangled switchers.  If we weren't on opposite coasts, it might be fun to compare it to Bob's big iron.

Yes there were subtle differences, and they seemed to me to be entirely dependent on amplifier/loudspeaker combos and on source material. There was no one amp which sounded "best". One amp would sound good with a specific speaker, and not with another. And I gamed the participants enough to realize that an amp that was considered to sound "bad" at one point in the testing would end up sounding "good" a few minutes later.

Not to say that I didn't have certain personal preferences, but I certainly would take them with a grain of salt.

[Steve Kunkle]

I am in complete agreeance with Pascal. I have found that its all about application.  In my case, I run KF 650's over SB 528's.  I switched from Crown Macrotechs to QSC PL's(6.0 II's on subs, mids and high mids, 230 on highs).  I found that the highs sounded maybe a little crisper... maybe.  But for my application I do miss my 5002's on the subs.  In my case, they did sound better with my speakers, even they have a power rating a bit less than the 6.0.  I don't think thats enough to make me switch back, but if you are running subs of similar nature (something with the p300 driver) it may be something to consider.  In conclusion though, the best wisdome that has been shared in this thread has been to rent, borrow, or otherwise comandeer numerous amps to see which one pleases YOUR ear the most on YOUR RIG.  We are all our own toughest critics when it comes to our own systems, so if you can find what you like and are happy than you have won the battle.

[Richard Rajchel]

I couldn't agree more. Everyone likes to believe they have somewhat golden ears. When we all hear as much with our eyes and preconceived notions is closer to the truth. In blind tests with output matched and no "voltage" shutdowns etc... I doubt anyone could tell you which amp was a Lab and which one was an RMX. The results would be completely random. Aren't volts volts and amps just amps? There are most certainly nuances, and maybe one amp has a sag at a certain frequency, but with just about any two amps behind a curtain driving identical speakers with the identical program material at the identical voltage I can't imagine anyone could tell. Reliability, weight, and how an amp handles overloads or under powering would be far better criteria than how you think an amp might sound better than another.

[SteveKirby]

Ah, the old straight wire with gain, all amps sound the same debate.

Someone mentioned it before, if two pieces of gear measure the same, but sound different, you're measuring the wrong thing.

It's kind of difficult to set up listening tests in SR since you need a source and venue and can't readily a/b things.  When I worked at Dolby, one of the golden ears there (had been asked to review various famous recordings for potential alterations among other things) told me that in his opinion given the compression drivers and horns in cinema or SR, the amps didn't make much difference.  But in a good home system or controlled listening test you could tell.  Which I know I've done.  I was even comparing some amps at home once and an old GF who had no interest in audio or electronics (but was a pianist and singer) could easily tell which amp was playing.  She had no idea which cost more or had the better reputation.  But consistently preferred one out of the three and recognized which of the other two were playing as well.

[Pascal Pincosy]

Richard Rajchel wrote on Thu, 07 February 2008 15:32

I couldn't agree more. Everyone likes to believe they have somewhat golden ears. When we all hear as much with our eyes and preconceived notions is closer to the truth. In blind tests with output matched and no "voltage" shutdowns etc... I doubt anyone could tell you which amp was a Lab and which one was an RMX. The results would be completely random. Aren't volts volts and amps just amps? There are most certainly nuances, and maybe one amp has a sag at a certain frequency, but with just about any two amps behind a curtain driving identical speakers with the identical program material at the identical voltage I can't imagine anyone could tell. Reliability, weight, and how an amp handles overloads or under powering would be far better criteria than how you think an amp might sound better than another.

The testing at my amp shootout was completely blind and I went so far as to change the numbering of the amps they were listening to several times, in order to trick people who developed a preference for a certain numbered amp. It seemed pretty clear to me that the participants were able to pick out the "better" sounding amps most of the time ie the results were not random. Further tested showed that different amps would sound "better" with some loudspeakers than they would with others. Testing with 4 different cabinets (the second set of tests was not blind) showed different "winners" for passive top cab, front-loaded sub, and horn-loaded sub. No one amp was the 'best". And FWIW, the RMX performed very well and actually improved on the Lab Gruppen's sound on top cabinets.

Was my testing statistically valid? No, but it's good enough for horseshoes. If somebody wants to hire me to sit around and test audio gear all day I'll be happy to come up with some definitive results though.