Print

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

[Hilton]

Hi all,

One of my velodyne subs has a problem with the amp. I've replaced all the caps like I did with the other one a couple years ago but it hasn't fixed the problem in this case. It plays fine upto about 70% volume and then makes those horrible sounds you never want to hear from your speakers.  I've swapped the speakers over between cabinets and its not the speaker. Anyway...

So it's time to replace the SUB amps with something like a Behringer iNUKE NU3000DSP 3000W Power Amplifier w/ DSP. 

I'll probably replace the drivers at the same time with something with a little more power handling, matched to the cabinets of course, though my 2 x 120W RMS velodyne setup has been more than ample for 10 years.

Im not ready to go all "Mojave" on my system as it has to be family friendly for a 4 years old and technophobe wife.  However.. I'm tempted to start looking at a full AMP/AVR replacement down the track and go naked AVRless JRiver to AMPs.

I'm about to buy a new house, and one of the requirements is a dedicated theatre room, even if its only a 4M x 6M room.

So here are my thoughts....

While I'm at it, I'll be rebuilding the HTPC in a nice proper HTPC case with a volume knob and possibly some type of display.

I assume from what I've read around the place that the Behringer Amp will do the job for my sub amp replacements. ?
I think i'd like the DSP version to be able to set limiters on the output, though I understand from reading that the DSP isn't the best for quality.

I'll probably keep the Asus Sonar Essence STX II 7.1 for a while but will likely change this in the future if it doesn't over complicate operation of the system for the wife and daughter.
Maybe a Motu..

I don't mind a complicated setup and configuration, as long as it's easy to use once its setup.

I'd still like to be able to hook up a playstation for audio and the occasional dvd or bluray.

My projector has two inputs so I can go direct for video from the playstation and optical into the PC at a pinch.

We do use Youtube, Spotify and Netflix a fair bit but I assume JRiver WDM/loopback will work fine for that.

I have 7 matched speakers, 6 of them full range floor standing that have bi-amp capability and I'd like to take advantage of that at some point too.

My questions....
If im going AVRless will the iNuke 3000 do the job for Subs? (assume that's a no brainer)

Will the Asus STX do an ok job.

Will a few more iNukes for 6 channels with at least fronts bi-amped do the job for home cinema duties.

If I wanted to go all Audiophile and get a dedicated 2 channel AMP what would be the best way to accommodate?

Am I nuts or just going "Mojave"?   

[mwillems]

My questions....
If im going AVRless will the iNuke 3000 do the job for Subs? (assume that's a no brainer)

Seems likely, it depends on what your goals are, but they've got plenty of power and acceptable distortion specs for bass reproduction.

Will the Asus STX do an ok job.

I'll give that a firm maybe.  My experience with the Asus cards is that they're quite nice in the right computer and with the right speakers/gain structure, but not so much otherwise.  I have an ST + H6 (the immediate ancestor of the STX II) and a DX.  Both pick up quite a bit of distortion from the PC power supply.  They do so to different extents in different PCs, but it's always an issue in a way that really isn't a problem with any USB cards I've tried.

Now that issue may be a non-issue for you depending on your gain structure and speaker sensitivities.  With a conventional high gain power amp and speakers with sensitivities in the 90's I could hear the noise pickup from my listening position with amp volume set to listening level, but it wasn't very loud.  With my 113dB/1 watt sensitivity compression drivers I could hear the noise pickup loudly anywhere in the room.  If your speakers have sensitivities in the high 80's low 90's, it probably shouldn't be a serious issue unless you have an unusually high gain power amp.  You may also have a "lower noise" PC than my living room model (or they might've fixed some of these issues with the STX II).

The Asus cards also have unbalanced outputs which means you might get mains hum (it's hit or miss).  People spend truly hilarious sums to try and eliminate ground hum, but the 100% solution is to use balanced interconnections, which would require getting another DAC.

I'd try the Asus and see, it may work perfectly. But you might have to invest in something better isolated/lower noise.  FWIW I'm still using my ST/H6 direct to a power amp to drive my study speakers, and it works just fine with them (but they have an 87dB sensitivity).

Will a few more iNukes for 6 channels with at least fronts bi-amped do the job for home cinema duties.

Opinions will vary, but I would strongly encourage you not to use inukes for regular non-sub speakers (unless those speakers have a power handling capability in the thousands of watts). Most amps have the lowest distortion and best efficiency towards the middle of their power curve.  If your speakers are only rated to handle 200 watts (for example), you'd always be using the inukes at a fairly low percentage of their capacity and not really getting their best from them in terms of distortion performance or power efficiency.  And, in fact, for most of your listening you'd be using them at less than 1% capacity:

Nelson Pass (a famous amp designer) has noted that the vast majority of normal home music listening is done at below 1 watt of output power.  For example, if your speakers have a 90dB sensitivity (not uncommon), unless they're playing louder than 90dB (which is mighty loud indoors) you're only using a watt or less of power.  With those same 90dB sensitivity speakers playing back the loudest peak on a properly calibrated theater system (105dB), they'd still only be using 30 watts or so at peak.  

You do want some headroom, of course, but in my own experience the distortion and noise performance in that "first watt" is much more important than having a huge watt output on the amp.  My "main" system has amps that can put out a few hundred watts, but I use 20 watt amplifiers around the house for satellite systems, and they play really stunningly loud (louder than you'd ever want).  My advice would be to look for amps that have exceptional distortion performance and have a max wattage output around your speaker's RMS power handling.  That will ensure that you'll be able to push your speakers nearly as far as they can go if you want to, but that you'll still get respectable distortion performance/fidelity.  It will also provide some "insurance" in case something goes wrong (a full volume pop won't wreck your speakers, etc.)

Opinions will vary of course and the inukes are nice for what they are (very high wattage amplifiers for sub or PA use), but the published distortion specs aren't stellar. The spec measurement (.3%) is probably taken at a fairly high wattage; in practice you can expect significantly more distortion than that for home listening.  I think they're probably both too much and not enough for your application unless your speakers have a sensitivity in the low 80's and can handle a few thousand watts.

[mattkhan]

fwiw there is some distortion data on the inuke in http://www.avsforum.com/forum/155-diy-speakers-subs/1780914-measuring-amplifiers.html as well as http://forum.speakerplans.com/behringer-inuke-nu6000-vs-kam-kxd7200-bench-tested_topic69202_post688959.html#688959 and http://forum.speakerplans.com/behringer-inuke-nu6000-vs-kam-kxd7200-bench-tested_topic69202_post688986.html#688986

I think the generally accepted interpretation of this data is that an inuke is not a great choice for full range use.

[mwillems]

Thanks for those; I had seen one of those links before, but not the others which have some rich data.  Unfortunately they suffer from the problem I described above in that the first discrete measurement (the "low power" measurement) is at a 50W output which, on an average main speaker would be putting out well more than 100dB (which is pushing into concert levels).  So it may have moderately good distortion performance when quite loud, but the effective THD+N is probably orders of magnitude higher at normal listening levels (i.e. 1 watt or less) especially given the effect of fixed noise and certain other kinds of fixed distortion.  

Your last link includes a THD+N vs. power output graph (which is super important and got my hopes up) but it has a 500 watt linear interval on the X axis which makes it less helpful: the 1 watt (or even the 10 watt) region of the distortion graph is invisible/misleading.  The graph shows uniform decline in that (tiny) region for most frequencies tested, which is never true as fixed noise causes the THD+N as a percentage to unavoidably rise at very low powers.  It's probably a consequence of the fact that the relevant region of the graph is too small to chart the uptick.  So the graph presents useful info for high-power output use cases, but not much useful info for home listening (again, unless you're using very insensitive speakers or listening exclusively at festival levels).

To really get a meaningful depiction you'd need to see that graph with a logarithmic scale on the x-axis so you can see how good (or more likely bad) the distortion/noise performance is at low power.  To see an example of what I mean check the graph on page 4 of this Hypex spec sheet http://www.hypex.nl/docs/UcD180HG_datasheet.pdf, which shows excellent low power distortion/noise performance.  It's surprisingly hard to find a graph like that for most amps (whether done by the manufacturer or a hobbyist), but it's literally the single most important piece of information you can have about an amp in my opinion.  

But I'll confess that I'm just kind of jaded after looking for an amp for my 113dB sensitivity compression drivers and demoing amp after amp after amp and finding that the published distortion specs really didn't apply at all below a watt (much less in the milliwatt range where speakers that sensitive typically run), which made them uselessly noisy for me (lots of hiss and loud crunching with nothing playing).  That's an unusual sensitivity, obviously, so it exaggerates the issue, but even with less sensitive speakers (in the 90's) I've had issues with distortion measurements that look great at 50 watts but look terrible in the first watt.  

FWIW, the end result of my search was just home-building hypex amps because they were the only commercial amp makers I could find in my price range with high-quality measurements available.  That's why I'm not offering any amp recommendations, unless you want to make DIY amps, in which case I can't say enough nice things about Hypex (black silence when nothing's playing, the best and quietest amps I've ever heard)

[Hilton]

Thanks Guys. pretty much what I was expecting but I just wanted to make sure.  They'll make good efficient SUB amps.
Ill consider some other options.

I don't know if you you've heard of holton precision audio?
I've had these pure class -A modules book marked as a fantasy build for a couple months waiting for the right moment.
I discovered his amps a couple years ago when my sony blew an output tranny at a party. I repaired my amp so never ended up doing anything.

All his designs are brilliant and raved about globally.

If your talking low distortion at 1W, these are to die for.

20W of pure class A standard or 40W pure class A with a variation he has.

20 Watts RMS sine wave into 8 Ohms Per Channel with 1.7 amps bias
40 Watts RMS sine wave into 4 Ohms Per Channel in Class AB
Biased into Pure Class A
Power Bandwidth 3.5hz to 350khz -1dB
THD+N 0.003% 1khz, 10khz 0.005% @ 1 watt - 8 ohms
THD+N 0.0035% 1khz, 10khz 0.0047% @ 15 watts - 8 ohms
THD+N 0.0055% 1khz, 10khz 0.008% @ 22 watts - 8 ohms
Dampening Factor At 100hz 8 Ohms = 800
RCA Single ended Input impedance 22k Ohms
Single Ended Input Gain = 27.0dB Position One.
Single Ended Input Gain = 24.0dB Position Two.
Single Ended Input Gain = 21.0dB Position Three
Single Ended Input Gain = 18.0dB Position Four
Signal to Noise Ratio -103dB @ 27dBGain


http://holtonprecisionaudio.com/collections/amplifier-module/products/hpa-nxa20l-pure-class-a-mono-amplifier-module

[Arindelle]

haha with tha 40watt class A amp you can probably heat your living room in the winter, too! 

[stevemac]

I don't know if you you've heard of holton precision audio?

Have been trying to workout how to get minister of finance approval for 5 channel using the NXV200 or NXV300 modules (& failing).  Might retry for 2 channel & expand over time.  The amplifier modules look fantastic, as do the soft-start & power management modules.  I especially like the following testimonials

http://holtonprecisionaudio.com/blogs/testimonials/18260403-ralphs-custom-built-hpa-nxv300-auckland-nz
http://holtonprecisionaudio.com/blogs/testimonials/18260347-leifs-tower-of-power-hpa-nxv800-mono-block-tower-amplifiers

Steve

[jjazdk]

20W of pure class A standard or 40W pure class A with a variation he has.

While I think Holton makes a lot of nice stuff, I find it hard to qualify a 20W RMS amp for anything than a kitchen radio, at least if you want capability to play at somewhat realistic levels.

Most real life speakers has a sensitivity of less than 90dBSPL/W, especially since a lot of manufacturers use 2.83Vrms as input instead of 1W, effectively giving a false high sensitivity.

An example:
1W = 90dBSPL
2W = 93dBSPL
4W = 96dBSPL
8W = 99dBSPL
16W = 102dBSPL
32W = 105dBSPL
64W = 108dBSPL

Well, a 20Wrms amp just does not cut it, not even close...

[mwillems]

While I think Holton makes a lot of nice stuff, I find it hard to qualify a 20W RMS amp for anything than a kitchen radio, at least if you want capability to play at somewhat realistic levels.

Most real life speakers has a sensitivity of less than 90dBSPL/W, especially since a lot of manufacturers use 2.83Vrms as input instead of 1W, effectively giving a false high sensitivity.

An example:
1W = 90dBSPL
2W = 93dBSPL
4W = 96dBSPL
8W = 99dBSPL
16W = 102dBSPL
32W = 105dBSPL
64W = 108dBSPL

Well, a 20Wrms amp just does not cut it, not even close...

Most people tend to seriously overestimate how loud their music listening is.  I think if you were to take a dB meter and measure how loud your normal listening actually is, you might realize that 102dB (which a 20WRMS amp could easily do in your example) is really, really loud.  Probably much louder than most people ever listen at home for extended periods.  Exposure to sound over 100dB for sustained lengths of time will damage your hearing.  The theater reference level calibration points to an average level of 83 or 85dB and allows for peaks up to 103 or 105dB, but playing back sound even at that level in a home is really, very loud.

If you're talking about reproducing an orchestra at concert volume, a 102dB RMS capable amp-speaker combo could easily handle it for most pieces unless you're trying to reproduce the SPL on the stage or in the pit.  See e.g. http://techtalk.parts-express.com/showthread.php?218593-Symphony-orchestra-performance-SPL where measurement of an orchestra showed an average dB in the 80's and peaks in the mid 90's or https://books.google.com/books?id=MZDu1Sti69kC&pg=PA188&lpg=PA188&dq=orchestra+spl+measurements&source=bl&ots=89qqzUiZHy&sig=XgnR9INKaEzVdrqi3qryeruLwp0&hl=en&ei=RlY-TJq_L43-nAfwtfWpBQ&sa=X&oi=book_result&ct=result#v=onepage&q=orchestra%20spl%20measurements&f=false showing typical instantaneous peaks in the high 90's or at 100dB, the most extreme recorded peak by the author being at 104dB.

I've taken a series of SPL measurements over time, and FWIW the vast majority of my listening takes place at or below 80dB or so, with my "loud" listening pushing up into the low to mid 80's for an average level with peaks into the 90's.  I've tried played music at 100+dB and it's a lot like being at a modern rock concert: uncomfortably loud without earplugs.  

I think if an amp can provide an RMS level above 100dB with one's speakers that's plenty of power as one should only ever really need intermittent peak power up there anyway.  If you're a belt and suspenders type shoot for 105dB RMS, but more than that for indoor home use is really unnecessary (and potentially a hazard) unless you have truly enormous rooms or are having house shows.  

And as for going even higher, having spent time around 110dB and 120dB sound, I wouldn't want to be in any room with sound that loud playing without ear protection for more than five minutes.

[jjazdk]

I disagree.

I've taken a series of SPL measurements over time, and FWIW the vast majority of my listening takes place at or below 80dB or so, with my "loud" listening pushing up into the low to mid 80's for an average level with peaks into the 90's.

If you listen to music and measure an average of say 85dBSPL, then your peaks will be in the +100dBSPL range (though most run-of-the-mill SPL meters are not able to measure peaks accurately). Crest factor (peak divided by RMS) is important, both for reproducing speech and music (unless we are talking modern pop-music that is compressed beyond belief), and I would never aim for less than 20dB crest factor in my reproduction chain.

My point is, an amplifier that has enough RMS output for whatever sustained level you won't to achieve, will NOT have a high enough peak power output to sustain a decent crest factor. That, IMO, is the primary reason why plenty of watts is important. Perhaps I whould have stated that in the first post.

Yes, I have measured SPL plenty of times. I have developed loudspeakers for the commercial market for +10 years, worked 5 years in electro-acoustic R&D for two-way radios for Government & Public Safety (where speech intelligibility is paramount) and presently I am employed in electro-acoustic R&D developing hearing aids (besides my own little company).
My credentials does not give my opinion value over others, but is meant to prove that what I say is not speculation but comes from a scientific point of view.

My own personal view is that one can never have too much headroom, and I absolutely hate the sound of stressed components (amplifiers or loudspeakers).

[mwillems]

I disagree.

If you listen to music and measure an average of say 85dBSPL, then your peaks will be in the +100dBSPL range (though most run-of-the-mill SPL meters are not able to measure peaks accurately). Crest factor (peak divided by RMS) is important, both for reproducing speech and music (unless we are talking modern pop-music that is compressed beyond belief), and I would never aim for less than 20dB crest factor in my reproduction chain.

We'll have to agree to disagree about some parts of this.  FWIW, my measurements were taken with a fairly high quality calibrated microphone which is more than capable of measuring very brief peaks (since it's fed into a computer I can recover almost arbitrary time resolution through FFT).  The vast majority of music does not have peaks 20dB above the average level and you can confirm this using JRiver's audio analysis tools or looking at the DR database.  

In fact of the 82000 or so albums in the DR database, about 650 or so albums (well less than 1%) even contain a single track with a crest factor of 20 or above, and fewer than 50 albums have an average DR of 20 or above.  The average is closer to 10 or 12dB (with pop music showing an average closer to 6 or 7dB).

Regardless, where we do agree is that a 20dB headroom allowance is a sensible one, but I think that allowance already builds in an excess of caution so I don't think failing to achieve it by a dB or two is particularly relevant. Most domestic music listening has an average level well below 85dB (which is the theater calibration standard), and an amp that can reproduce 102dB RMS could probably manage brief 105dB peaks (most amps I've seen have a peak power rating close to double their RMS rating).  As I conceded above, a belt and suspenders approach is to have an amp with an RMS rating that will produce 105dB with your speakers.  Such an amp will reproduce audio at theater calibration reference levels (complete with 20dB peaks) quite happily (except for subs, which are a separate story).

For a speaker with a sensitivity above 93dB (of which I have several) a 20 watt amp would be more than enough, and a 20 watt amp would be pretty close to enough for theater calibration of your 90dB sensitivity hypothetical (give or take one or two dB).  More importantly though, it would probably be more than enough for an 85dB average level with the vast majority of recorded music.

My own personal view is that one can never have too much headroom, and I absolutely hate the sound of stressed components (amplifiers or loudspeakers).

You can easily have "too much" headroom such that your getting suboptimal distortion performance, suboptimal power efficiency, and are putting your ears and speakers at risk if you have a power fault or bump up against an interconnect the wrong way. Choose an amp that has as much headroom as you want with 3dB of extra capacity if you're risk averse and call it a day.

As for "stressed" components I think that's only an issue if you don't bother calibrating your system.  With a digital source you can know in advance what your maximum peak level will be. You can then calibrate your system so that the maximum peak will not go beyond your amp's capacity.  It may or may not be as loud as you want, but avoiding component "stress" is trivial if you have the ability to take even basic SPL measurements.

If I had to pick an amp with .01% 1st watt performance that could only play my speakers to 95dB RMS and an amp with .1% distortion in the first watt (which is not uncommon) that could play to 115dB, I know which one I'd pick.

[mojave]

Im not ready to go all "Mojave" on my system as it has to be family friendly for a 4 years old

I have a 1 year old, 3 year old, 5 year old, 11 year old, and almost 15 year old. The 3 year old is learning the system and will be writing a JRiver manual soon for other kids.   I just realized that four of my kids can be considered "lifetime" users of JRiver.

I have a Asus Essence ST and H6 daughter card in my work computer, but have been using a MOTU Ultralite instead the past two months. I prefer the Ultralite and love its multiclient ability and ability to route to other programs. It doesn't have the ESS Sabre32 chips like the other MOTU AVB audio devices so there are pops on startup/shutdown and sample rate changes. In addition to 8 outputs, the Ultralite is also a mic preamp with phantom power, guitar preamp, and headphone amp. One can listen to something on headphones that is different than the analog outputs, too.

The iNuke amps have extremely noise fans so if you do get one, you don't want it anywhere near the listening area.

The vast majority of music does not have peaks 20dB above the average level and you can confirm this using JRiver's audio analysis tools or looking at the DR database.

Both R128 and DR throw out the highest peaks when coming up with their numbers.

[mwillems]

Both R128 and DR throw out the highest peaks when coming up with their numbers.

R128 does ignore the highest peaks, but my understanding was the the DR database tool compared average loudness to actual peak volume (crest factor), which has generally been the case in the wave forms I've examined.  That's why I mentioned Dr instead of R128.

Can you point me to an explanation of the TT-DR methodology that confirms that point?

[mojave]

I was waiting for you to ask. 

http://www.dynamicrange.de/sites/default/files/Measuring%20DR%20ENv3.pdf

[mojave]

The 2 channel track of the song "Dirty Laundry" from Eagles Farewell 1 Tour has an R128 value of 4 and a DR value of 10. This is probably the most dynamic song I've ever heard in my system. The drums are amazing. I was playing it back recently and used Omnimic to measure the max and peak SPL. This is my right main speaker from about 1m.

[mwillems]

I was waiting for you to ask. 

http://www.dynamicrange.de/sites/default/files/Measuring%20DR%20ENv3.pdf

Thanks for that.  I stand corrected.  It definitely still doesn't discard as much peak information as R128 does, but it does discard some for sure.

The 2 channel track of the song "Dirty Laundry" from Eagles Farewell 1 Tour has an R128 value of 4 and a DR value of 10. This is probably the most dynamic song I've ever heard in my system. The drums are amazing. I was playing it back recently and used Omnimic to measure the max and peak SPL. This is my right main speaker from about 1m.

How far are you sitting from the speaker?  I'd assume at least a fewmeters if those measurements are accurate (130dB can cause almost instant hearing damage).

Out of curiosity what was the average level? 

[mojave]

How far are you sitting from the speaker?  I'd assume at least a fewmeters if those measurements are accurate (130dB can cause almost instant hearing damage).

Out of curiosity what was the average level? 

I sit about 3m from the speaker. Gunshots can expose one to ~160 dB and I've shot many rounds without hearing protection when hunting.

The max of 121.7 dB is the average level.

[mwillems]

The max of 121.7 dB is the average level.

Got it, so the actual headroom needed for that track over the average level is about 15dB, so the DR is "understated" by about 5dB in terms of power needed.  Good to know.

3M distant would only be about ~4dB lower than the measurement at your listening position! You clearly like things loud 

My amps have enough power to run my main speakers up to about 120dB RMS give or take a dB.  I just tried running a track up as loud as I could stand it, and I wimped out at about 108dB (it was borderline painful for me).  

But I'm also the weenie wearing earplugs at rock concerts (and while shooting). I used to go to shows without plugs, but very intermittent tinnitus has me taking precautions (lest it become permanent).

[natehansen66]

The op might look at gen 1 Crown xls amps for sub duty. I think they're all on sale now to make room for gen 2. I have an xls2000 on my dual 15" subs. I listen at about 95dB average max at 3 meters and the fans have never come on. Having never looked at its dsp capability I can't speak to that.