|
|
Post by emotifan on Jul 13, 2020 18:42:23 GMT -5
I have a rudimentary understanding of how amplifiers work and have always equated lots of capacitance with high reserves and current capability. Can someone who understands this stuff explain why the current XPA amps need so little?
|
|
|
|
Post by garbulky on Jul 13, 2020 19:00:22 GMT -5
I have a rudimentary understanding of how amplifiers work and have always equated lots of capacitance with high reserves and current capability. Can someone who understands this stuff explain why the current XPA amps need so little? The switching power supply or so they say |
|
|
|
Post by leonski on Jul 13, 2020 19:02:11 GMT -5
If you are referring to the latest, they are using s switching PS. So instead of running at US power freuquency of 60hz, it runs MUCH higher frequency. 100khz or higher would not be unheard of.
This means the PS 'recharges' a LOT faster. A capacitor will only charge to its limit. The PS than recharges when its voltage goes above that of the cap.....With very low levels, a very short period of time
and only at the very peak of the waveform....power line, that is. As demand increases and the cap 'sags', it's time of recharge gets longer.
To my understanding, the downside of Switchers is little or no dynamic power. Does this matter? Very probably not. Keeping that high frequency from the output of the amp is also a consideration.
As for reserves from a conventional PS? That's not always the case. Look at RMS vs Dynamic Headroom and you will only rarely see a figure even approaching 3db. Half that is more common.
Current claims for amps is IMO, a red herring and nearly valueless. Such current would be at a VERY low voltage. A 35 amp claim IF it were at even a modest 30 volts would be over 1000 watts.....
And for how long?
|
|
|
|
Post by JKCashin on Jul 13, 2020 19:05:36 GMT -5
Capacitors act like batteries... storing power the amp is not using so it can be used to handle a sudden demand for power, a drum kick for instance. The idea is that if you make the power supply able to deliver just above the AVERAGE need for power, the excess can be stored so it can punch when it needs to without needing a larger power supply. The unused excess is stored (to a limit) and it can be used instantaneously. After the drum hit or whatever dynamic impulse that wasy needed, the power supply is delivering enough power to maintain the current demand, so the excess charges up the capacitors again... then, anotehr drum hit.
Think of it like this. The power supply is a hose, providing a stream of water. On average, it delivers more water than you need, and there are some buckets (capacitors) that store the overflow. Once the buckets are full, they are full and excess water (power) is thrown away. Now you need a quick pulse of water... you can take it from the main hose, but if you needed more than the hose can provide, the buckets make up the difference. As long as your quick peak demands are spaced out a bit, you're good. But you're often wasting water.
But what if you had a power supply that instead of delivering a steady amount of power that was the average for the music, it was able to dynamically respond to the demand for power. TO use the hose analogy, the hose would be designed to be able to respond directly and quickly to demand, dynamically, whether that demand is high, or low. You only get the water you need. The power supply in the recent Emotiva amps, a switched mode power supply, works like this... they are capable of responding dynamically to need, varying the on-tap power in real-time according to demand.
I'm grossly over simplifying it, but I think it comes pretty close to answering the question. I hope!
|
|
|
|
Post by emotifan on Jul 13, 2020 19:36:41 GMT -5
Wow, I asked and I received! Thanks guys for the technical and simplified answers. I already pretty much understood the linear method with large transformers and caps but JKCashin's description solidified my understanding. It amazes me that a switching PS can respond fast enough to not chop the signal into distortion. I'm considering -upgrading?- to one of the DR options from an XPA-2 Gen 2. But the new and foreign 'to me' architecture causes me to be cautious. I don't claim to have golden ears but I know good sound when I hear it, which my current Gen 2 does give me. I won't do anything untill I get my Magnepan 1.7is mounted on Mye Stands which arrive next week and assess the affects of those for a few weeks. I do invite further insite and experience from others on this subject in the mean time.
|
|
|
|
Post by davidl81 on Jul 13, 2020 19:51:43 GMT -5
Wow, I asked and I received! Thanks guys for the technical and simplified answers. I already pretty much understood the linear method with large transformers and caps but JKCashin's description solidified my understanding. It amazes me that a switching PS can respond fast enough to not chop the signal into distortion. I'm considering -upgrading?- to one of the DR options from an XPA-2 Gen 2. But the new and foreign 'to me' architecture causes me to be cautious. I don't claim to have golden ears but I know good sound when I hear it, which my current Gen 2 does give me. I won't do anything untill I get my Magnepan 1.7is mounted on Mye Stands which arrive next week and assess the affects of those for a few weeks. I do invite further insite and experience from others on this subject in the mean time. So I went from a XPA-2 Gen 2 to the DR-2. Honestly at first there was no difference, this is when I was running a Marantz AV8802A. Since then I have upgraded to a RMC-1 and that processor paired with the DR-2 (everything being fully balanced) is an amazing setup. I sit here as we speak with glass of wine listing to Amazon Music HD and I honestly could not be happier with this setup. That being said the XPA-2 is a damn fine amp, so don’t expect to plug in the DR-2 and it just sound drastically different. |
|
|
|
Post by JKCashin on Jul 13, 2020 21:13:19 GMT -5
Wow, I asked and I received! Thanks guys for the technical and simplified answers. I already pretty much understood the linear method with large transformers and caps but JKCashin's description solidified my understanding. It amazes me that a switching PS can respond fast enough to not chop the signal into distortion. I'm considering -upgrading?- to one of the DR options from an XPA-2 Gen 2. But the new and foreign 'to me' architecture causes me to be cautious. I don't claim to have golden ears but I know good sound when I hear it, which my current Gen 2 does give me. I won't do anything untill I get my Magnepan 1.7is mounted on Mye Stands which arrive next week and assess the affects of those for a few weeks. I do invite further insite and experience from others on this subject in the mean time. Thanks! The laymans explanation is the power supply responds "coarsely" and the amplifier then modulates that input as clean output. Back to the water analogy, think of the new hose being able to deliver water in steps of 1GPH (gallon per hour), and you want to spray from 1GPH to 10 GPH, back to 1GPH, ramping up from 1GPH in a very linear fashion. At first the input is 1GPH and you need 1GPH. No problem. Now you need 1.1GPH, so the input gives you 2GPH. You use 1,1, throwing away 0.9; then you use 1.2, throwing away 0.8, etc., until you get to 2.0GPH and you're using what the input gives you. Now you need 2.1GPH so the inputy ramps up to it's next level. Again, a gross oversimplification... in reality I believe the supply is pulse width modulated, and that is then smoothed, but I am not sure. In water terms, the input is pulsed to come close to what you want as output. |
|
|
|
Post by leonski on Jul 13, 2020 21:33:51 GMT -5
output of an amp to a speaker is simply a moduleated power supply.
One thing which works against simply cranking in more capacitor is the SOA or Safe Operating Area of the output devices.
They can only take so much. And generally, the longer the stress, the shorter the time it will work. Or maybe the shorter the stress, the worse it can be without bad things happening.
All sort of things play into this from the specs of the output device to the amount of heatsink provided.
|
|
|
|
Post by emotifan on Jul 14, 2020 0:13:41 GMT -5
One advantage I suspect might come from the smps and class H is longer life of the components from less heat and that is one of the claims Emotiva makes for these new amps. That's an important consideration seeing as I'll be retiring in a few years and will be on more of a fixed income.
|
|
|
|
Post by thezone on Jul 14, 2020 3:24:46 GMT -5
Interesting read. I have been pondering this type of "upgrade" of late. I am about to install 2 ceiling speakers which means I am 2ch down on amplification. At present I have a 5 channel Rotel RMB-1095 which is approaching 12 years old. I have two options:
1) Buy a bas-x 2ch cheapo amp to get the job done.
2) Sell the Rotel and choose a "make your own" config fro the Emotiva line up of new amps.
1) Will result in a bigger footprint.
2) Will result in the same footprint but I will be replacing a reliable old school amp with one based on new technology which I cant possibly audition before I buy.
|
|
|
|
Post by audiobill on Jul 14, 2020 5:20:58 GMT -5
When you get the DR, please send the XPA-2 to Garbulky for proper disposal.....
|
|
|
|
Post by davidl81 on Jul 14, 2020 9:17:56 GMT -5
Interesting read. I have been pondering this type of "upgrade" of late. I am about to install 2 ceiling speakers which means I am 2ch down on amplification. At present I have a 5 channel Rotel RMB-1095 which is approaching 12 years old. I have two options: 1) Buy a bas-x 2ch cheapo amp to get the job done. 2) Sell the Rotel and choose a "make your own" config fro the Emotiva line up of new amps. 1) Will result in a bigger footprint. 2) Will result in the same footprint but I will be replacing a reliable old school amp with one based on new technology which I cant possibly audition before I buy. You do get a 30 day trail on the Emotiva amps. If you return it all you pay is return shipping, so maybe $50 ish. In my setup I had and XPA-5 G2 and a BasX A-700 for mains and Atmos speakers. I replaced those two amps with a XPA-9 G3 and I have been very happy with the results. I didn’t lose anything at all, and although hard to tell it may all sound better. Plus having all of these channels in one box s really cleaned up my cabinet. |
|
|
|
Post by garbulky on Jul 14, 2020 11:51:59 GMT -5
When you get the DR, please send the XPA-2 to Garbulky for proper disposal..... I will make sure it’s properly uh “disposed of”. |
|
|
|
Post by leonski on Jul 14, 2020 15:06:14 GMT -5
One advantage I suspect might come from the smps and class H is longer life of the components from less heat and that is one of the claims Emotiva makes for these new amps. That's an important consideration seeing as I'll be retiring in a few years and will be on more of a fixed income. Don't forget that the signal / power chain from wall plug and source to speaker is maybe 1% efficient. MIGHT be 2% in extraordinary circumstances. ALL the rest is heat. OK, maybe something is wasting power vibrating....but that's gonna be mere noise compared to the remainder. |
|
|
|
Post by donh50 on Jul 14, 2020 16:39:48 GMT -5
A regular linear supply with a full-wave bridge (normal power supply scheme) takes 50/60 Hz wall frequency and charges the capacitors 100 or 120 times a second. A switch-mode power supply typically operates around 400 kHz and so charges the capacitors 400,000 times a second. Delivering the same current (and output power) the capacitors in the new designs can be about 4000 times smaller. It does not quite work out that easily, but hopefully that helps explain why the caps can be so much smaller.
|
|
|
|
Post by leonski on Jul 14, 2020 18:19:48 GMT -5
As I said above:
One problem with switchers? Keeping that noise OUT of the output.
Even good SMPS has Ripple.
The additional complexity of a switcher means longer development time and more possible failure modes.
For high power amps or Class 'A' amps, the main problem with a linear supply is that pesky Boat Anchor of a transformer.
|
|
|
|
Post by thezone on Jul 14, 2020 18:25:50 GMT -5
I cant possibly audition before I buy. You do get a 30 day trail on the Emotiva amps. If you return it all you pay is return shipping, so maybe $50 ish. In my setup I had and XPA-5 G2 and a BasX A-700 for mains and Atmos speakers. I replaced those two amps with a XPA-9 G3 and I have been very happy with the results. I didn’t lose anything at all, and although hard to tell it may all sound better. Plus having all of these channels in one box s really cleaned up my cabinet.+ Ah except I live in Australia so return shipping will be $250 ish with insurance. Not too mention the $100 shipping to get it in the first place. And because Emotiva pulled their Oz distributorship out there is nowhere to go to audition anymore : ( Sigh, first world problems. |
|
|
|
Post by leonski on Jul 14, 2020 19:32:16 GMT -5
You do get a 30 day trail on the Emotiva amps. If you return it all you pay is return shipping, so maybe $50 ish. In my setup I had and XPA-5 G2 and a BasX A-700 for mains and Atmos speakers. I replaced those two amps with a XPA-9 G3 and I have been very happy with the results. I didn’t lose anything at all, and although hard to tell it may all sound better. Plus having all of these channels in one box s really cleaned up my cabinet.+ Ah except I live in Australia so return shipping will be $250 ish with insurance. Not too mention the $100 shipping to get it in the first place. And because Emotiva pulled their Oz distributorship out there is nowhere to go to audition anymore : ( Sigh, first world problems. Ever consider a NON-Emotiva solution? Plenty out there without the extreme hassles or perhaps better Oz representation? |
|
|
|
Post by thezone on Jul 15, 2020 2:18:50 GMT -5
Ah except I live in Australia so return shipping will be $250 ish with insurance. Not too mention the $100 shipping to get it in the first place. And because Emotiva pulled their Oz distributorship out there is nowhere to go to audition anymore : ( Sigh, first world problems. Ever consider a NON-Emotiva solution? Plenty out there without the extreme hassles or perhaps better Oz representation? Local prices of hifi gear in Aus is painfully high, we have such poor buying power everything is over priced. the Emo stuff is actually good value even with shipping. And i like the idea of having high power modules for the LRC and lower powered stereo modules for the smaller surround and ceiling speakers. or am i being bamboozled by smoke and mirrors here? Otherwise you get a very high powered 7ch amp which is overkill for the small speakers or moderate powered 7ch amp which is underwhelming for the LRC. |
|
|
|
Post by leonski on Jun 2, 2021 14:46:22 GMT -5
Just getting back to this.....
In general? Power needed is related to speaker sensitivity AND desired loudness AND room size in CUBIC FEET.
I'm contemplating 96db sensitive speakers and a 25x2 amp. Plenty.
For my Low-80s panels? Tough to purchase TOO MUCH power.....
My new mantra? ONE AND DONE. I wish I'd known what I know now....but back in the mid-70s when I began in stereo.
Get gear of lasting value and don't subject yourself to annual 'upgrade itis' HT is a deep hole from which few escape intact.
Shipping costs are GOING THRU THE ROOF. So keep That in mind!
|
|
