|
|
Post by emotifan on May 21, 2018 11:18:44 GMT -5
After emailing Emotiva to clear up, in my mind, some differences in their published specs between Gen 2 and Gen 3 X series amps, Keith responded with some information that answered my questions but made me more curious at the same time. I responded back asking him if their new amps had any similarity to Carver's old "Magnetic Field Coil" power amps of the 80s. He hasn't gotten back to me yet but I think I found the answer myself. Following is a link to an old article written by Bob Carver explaining the differences between how his amps and traditional linear power supply amps work and the benefits of his design. It does a really good job of explaining why a large amount of filter capacitance isn't necessary with a switch mode power supply, which was the basis of my question that Keith answered. Specs for their older amps always included how much capacitance each amp had, which is not as crucial with a SMPS. A Wikipedia article explains that Carver's "Magnetic Field Coil" power supply had nothing to do with coils and was in fact a switch mode power supply. His article does shed a lot of light on the advantages of using this type of power supply, so for the audio nerds like myself out there, here you go. carvermk2.com/Docs/Carver%20Magnetic%20Field%20Whitepaper.pdfen.wikipedia.org/wiki/Bob_Carver |
|
|
|
Post by socketman on May 21, 2018 15:43:03 GMT -5
I notice that the Gen2 has a big torodial power supply and lots of capacitance none of which is in t he Gen 3. I will be using mine only for atmos overheads so not too concerned as my main amp is a Gen2. The Gen 3 isnt really getting rave reviews is this due to the topology?
|
|
|
|
Post by emotifan on May 21, 2018 15:53:13 GMT -5
I notice that the Gen2 has a big torodial power supply and lots of capacitance none of which is in t he Gen 3. I will be using mine only for atmos overheads so not too concerned as my main amp is a Gen2. The Gen 3 isnt really getting rave reviews is this due to the topology? I'm not sure why that is. Herb Reichert in Stereophile really loved it driving Magnepan .7 speakers. Something I've noticed over the years is how magazines will love a new product that delivers stellar performance at a value price and then they become hyper critical of the brand going forward. This could be what's happening to Emotiva right now. Since you have both amplifiers I'd like to hear your impressions swapping both amps to listen to your front main stereo setup. What kind of speakers are you using for your mains? www.stereophile.com/content/emotiva-xpa-gen3-two-channel-power-amplifier |
|
|
|
Post by socketman on May 21, 2018 16:34:51 GMT -5
Too be entirely honest my hearing is not what it once was so i doubt i could add much value. I may get ambitious one day and swap things around and see if my boys notice any difference, kind of a blind test if you will. The hard part is that all my XLR's are made to length so a bunch of fiddling will be required, eye sight is sh** too lol. Dont get old its a trap. I am reading over some reviews now that i have read before to gain some more insight but as you say Stereophile and HomeTheater secrets both love the Gen 3. My observations are just that, observations more curiosity than anything. Less shipping weight is always good and i use my amps in the theater for movies only . I will see if i hear any difference but i fear i will be guilty of confirmation bias.
|
|
|
|
Post by Gary Cook on May 21, 2018 19:16:13 GMT -5
Something I've noticed over the years is how magazines will love a new product that delivers stellar performance at a value price and then they become hyper critical of the brand going forward. This could be what's happening to Emotiva right now. Could it be that the "price" isn't so "value" anymore? For example my XPA-5 Gen 1 was $899 and the XPA-5 Gen 3 is $1599, that's close to twice the price, does it sound close to twice as good? Does it even sound as good? Cheers Gary |
|
|
|
Post by socketman on May 21, 2018 19:31:14 GMT -5
Good point G but also realize made in America adds some cost. I was dinged 200 just for customs fees which is absurdly high due to Purolator and or UPS doing their magic. My previous amps had not duty either time they were sent ups. Times the are a changing. No real jobs out there so now people want to make big money delivering parcels,this will be the achilies heel of online buying as we move forward. Trump is putting the clampers on Amazon and god knows who else. I know this is off topic but these big corps need to brought into line,enough of the corpotate welfare provided by government to subsidize big corps so they can screw over the everyday worker.
|
|
|
|
Post by Gary Cook on May 21, 2018 19:42:57 GMT -5
Good point G but also realize made in America adds some cost. I was dinged 200 just for customs fees which is absurdly high due to Purolator and or UPS doing their magic. My previous amps had not duty either time they were sent ups. Times the are a changing. No real jobs out there so now people want to make big money delivering parcels,this will be the achilies heel of online buying as we move forward. Trump is putting the clampers on Amazon and god knows who else. I know this is off topic but these big corps need to brought into line,enough of the corpotate welfare provided by government to subsidize big corps so they can screw over the everyday worker. Strangely, "made in the USA" saves us Aussies 5% as we have a free trade agreement.  Which unfortunately doesn't apply to Emotiva stuff made in China.  Cheers Gary |
|
|
|
Post by leonski on May 21, 2018 19:46:59 GMT -5
After emailing Emotiva to clear up, in my mind, some differences in their published specs between Gen 2 and Gen 3 X series amps, Keith responded with some information that answered my questions but made me more curious at the same time. I responded back asking him if their new amps had any similarity to Carver's old "Magnetic Field Coil" power amps of the 80s. He hasn't gotten back to me yet but I think I found the answer myself. Following is a link to an old article written by Bob Carver explaining the differences between how his amps and traditional linear power supply amps work and the benefits of his design. It does a really good job of explaining why a large amount of filter capacitance isn't necessary with a switch mode power supply, which was the basis of my question that Keith answered. Specs for their older amps always included how much capacitance each amp had, which is not as crucial with a SMPS. A Wikipedia article explains that Carver's "Magnetic Field Coil" power supply had nothing to do with coils and was in fact a switch mode power supply. His article does shed a lot of light on the advantages of using this type of power supply, so for the audio nerds like myself out there, here you go. carvermk2.com/Docs/Carver%20Magnetic%20Field%20Whitepaper.pdfen.wikipedia.org/wiki/Bob_CarverMy Carver Cube, with the tMod had an early version of what he called the 'tracking' power supply. It would turn on/off basically 'in time to the music'. It emitted a low-level 'putting' noise, which had something to do with the CHOKE, which is indeed a 'coil'. As near as I can tell, the main DISadvantage of SMPS is lack of 'dynamic' power. 1.5db to 2.0db above the RMS power is a good idea, generally. BTW, these are the 'Magnetic Field' amps......and a magnetic field IMPLIES a coil of wire. The choke, in this case, a largish inductor, was large but not outrageous in size. The magnetic field, IIRC, was used to 'store' energy for ONE CYCLE.....and than there was the TRIAC (SCR?) used as a high-speed switch. I think the PS also had at least 2 rails, just like the new EMO stuff. Let me read the article......Carver is a pretty smart guy. Read article and I don't think it is technically and fully correct, at least to do with the power supplies. Here is link to a BUNCH of Carver Patents, some of which might still be active or farmed out to other companies. Very interesting read: patents.justia.com/inventor/robert-w-carver |
|
|
|
Post by Ex_Vintage on May 24, 2018 21:57:46 GMT -5
Comparing a linear supply and an SMPS are 2 different animals. The linear supply requires large support caps for 2 reasons. 1st it receives the rectified 120Hz line voltage after it goes through a rectifier. In order to keep the DC voltage variance low, the capacitance needs to be high. 2nd purpose of the large caps is to feed the high dynamic demand from the amplifier producing the music. Music dynamics may demand 2-3 times the "rated" power in order to reproduce high level low frequency content. An SMPS "Switches" a rectified DC voltage at a high frequency (well above audible frequencies) and feeds that switched DC voltage to a set of output capacitors. The output caps for an SMPS do not need to be that large due to the higher switching frequency (maybe 40-60 Khz instead of 120 Hz). The only question is can the SMPS be dynamically responsive enough to respond to the high level low frequency music content. If it cannot, then it still needs larger output capacitors as the energy well. There is no thing as too much output capacitance when it comes to reproducing audio... and it should come from quality audio grade capacitors that have very low ESR and ESL (equivalent series resistance and inductance).
|
|
|
|
Post by leonski on May 25, 2018 2:09:31 GMT -5
Comparing a linear supply and an SMPS are 2 different animals. The linear supply requires large support caps for 2 reasons. 1st it receives the rectified 120Hz line voltage after it goes through a rectifier. In order to keep the DC voltage variance low, the capacitance needs to be high. 2nd purpose of the large caps is to feed the high dynamic demand from the amplifier producing the music. Music dynamics may demand 2-3 times the "rated" power in order to reproduce high level low frequency content. An SMPS "Switches" a rectified DC voltage at a high frequency (well above audible frequencies) and feeds that switched DC voltage to a set of output capacitors. The output caps for an SMPS do not need to be that large due to the higher switching frequency (maybe 40-60 Khz instead of 120 Hz). The only question is can the SMPS be dynamically responsive enough to respond to the high level low frequency music content. If it cannot, then it still needs larger output capacitors as the energy well. There is no thing as too much output capacitance when it comes to reproducing audio... and it should come from quality audio grade capacitors that have very low ESR and ESL (equivalent series resistance and inductance). What you are noting is the 'refresh rate' which for a switcher is orders of magnitude faster then a linear supply. Caps recharge at maybe 100,000 hz or higher in a switcher. don't forget that a capacitor ONLY recharges when the voltage given to it is HIGHER than the voltage IN the capacitor. For partially discharged caps? Maybe only the upper 10% or 20% of each 'cycle' will be available for 'recharge'. Swithers have more and more frequent opportunities for such recharge. And since a switcher operates differently than a linear, I suspect overall 'poor' dynamic power. I remember an early SMPS amp with ZERO dynamic headroom RMS was it. My convention linear amp has an advertised 1.5db dynamic headroom while some earlier NAD designs had as much as 6db headroom. That's a LOT. And there ARE 'practical' limits to PS Capacitance. Too much can produce a starup surge which can pop a breaker or fuse. A Thermister or 'slow start' circuit helps this a LOT. You can also only reduce ripple so much to the point where you reach diminishing returns and additional capacitance has no benefit in that regard. I also suspect that using capacitance, say 120volt caps, and an unregulated PS running at maybe 100volts but of very low CONTINUOUS output might put out a big jolt of power, but at the expense of continuous or RMS behavior. The idea would be an amp with a maybe 100va transformer, at 100 volts and a capacitor bank the size of a steamer trunk. Turn it on maybe 15 minutes prior to use to give it a chance to start up and stabilize. Huge jolts of power available and 25 or 30 watts a side, continuous. At some really drug-down voltage. But I WILL agree that good quality caps can't be beat. However, I'm told that excessively LOW ESR can open another can of worms. Dont' forget that caps in paralled ADD capacitance but begin lowering resistance, just like any resistors in parallel. I'm told this can have stability effects. |
|
|
|
Post by metaldaze on May 25, 2018 11:19:28 GMT -5
Post by DYohn on Mar 6, 2012 at 11:58am
'Dynamic headroom can be calculated from the nominal continuous power rating and the max peak power capacity of an amplifier. The formula is: headroomdb = 10*log(Pmax/Pnom) If the Pmax value is not published, the best case value can be derived from the nominal power supply rail voltage, which can be measured if it's not published. The formula for Pmax in a stereo amp is ((Vrms^2)/2)/load resistance in ohms.'
I found this in another thread but it relates to the XPA-1, probably a gen 1 or 2.
Reading this thread got me interested and your last post Leonski implied you think there isn't much headroom in the gen 3 amps either given their topology. The power supply is regulated which helps but I'm getting weird numbers when trying to run this calculation. Any idea what the headroom may be for the new gen amps?
A few posts later in the referenced thread the educated guess is approx 1.5db
|
|
|
|
Post by leonski on May 25, 2018 12:59:41 GMT -5
You're right. I SUSPECT not much dynamic power from the amps with SMPS. But That's NOT a reason to skip them. Most persons drastically overestimate the power they need. Also, few have a real idea how much power
is used at 'normal' listening levels. In some of the HT systems I see listed, 3 or 4 watts per speaker would be Loud, especially given the 'burst' or 'dyanmic' power needed is maybe 10x to 20x that.
Someone with 95db sensitive speakers telling me they use 50 to 70 watts continuous is either going on DEAF or the eviction notice is in the mail.
Dynamic power can ALSO end up as a numbers game. Under-reporting your amps nomimal continuous rating will TEND to inflate the dynamic 'range' of the amp. You may also get high dynamic power with a smallish
transformer which won't supply a huge amout of continuous current but will run to a high enough voltage to keep a good capacitor bank charged....for dynamic bursts. I don't see Switchers as having the sheer 'bulk' of capacitor I'd expect
for huge bursts of power though operating at the high frequencies they do is GOOD for recharge rate and of course, efficiency.
I own low sensitivity Magnepan. I'm not a nose-bleed level kind of guy. But when watching a movie I DO really like some 'slam' and so tend to value the dynamic power of an amp. Nothing beats the sound of every dish in the house rattling at once
as some very loud event happens on screen.
|
|
KeithL
Administrator 
Posts: 9,938
|
Post by KeithL on May 25, 2018 13:21:30 GMT -5
Keith here..... Here's a copy of the reply I sent to emotifan on his desk case:
The early Carver amps actually included a variety of interesting technologies. Their "magnetic field power supply" was an early sort of switching supply that used a triode to chop AC power from the AC line directly through a transformer. Many of the amps themselves were Class G designs (which is sort of like Class H); they ran on three separate sets of power supply rails. Later Sunfire amps used what they called "a tracking downconverter", which was basically a single switching supply whose voltage varied to track the signal. (Think of a Class D amplifier powering a Class A/B amplifier.) Modern switch mode power supplies work a lot better - but the general principle is the same. Our new XPA models also use our Class H topology - which is the modern version of Class G.
|
|
KeithL
Administrator
Posts: 9,938
|
Post by KeithL on May 25, 2018 13:36:51 GMT -5
Let me clear up the question of "dynamic power" a little bit.
As has already been noted, dynamic power is just another way of saying "the difference between what an amplifier can deliver for a short time but can't deliver continuously". And, yes, it's just another way of looking at the same numbers. At its absolute top speed my Nissan can probably do about 100 mpH, but I normally drive only 60 mpH, so it has "dynamic headroom" of about 2 dB. However, a racing car, which has a top speed of 500 mpH, and normally is driven at 500 mpH, has no dynamic headroom at all. But we all know which is faster.
In the old days, power supplies were expensive, so many amplifiers had a somewhat undersized power supply that could only deliver a lot of power for a very short time. So they ended up with a low power rating, but a high "dynamic power" rating. Many modern amplifiers can actually deliver their full power continuously - which means that their steady state power is much higher. However, because they can already deliver almost maximum power all the time, they can't deliver much more than that short term.
The switch mode power supply we use in the XPA Gen3 can deliver about 3kW...
You're right. I SUSPECT not much dynamic power from the amps with SMPS. But That's NOT a reason to skip them. Most persons drastically overestimate the power they need. Also, few have a real idea how much power is used at 'normal' listening levels. In some of the HT systems I see listed, 3 or 4 watts per speaker would be Loud, especially given the 'burst' or 'dyanmic' power needed is maybe 10x to 20x that. Someone with 95db sensitive speakers telling me they use 50 to 70 watts continuous is either going on DEAF or the eviction notice is in the mail. Dynamic power can ALSO end up as a numbers game. Under-reporting your amps nomimal continuous rating will TEND to inflate the dynamic 'range' of the amp. You may also get high dynamic power with a smallish transformer which won't supply a huge amout of continuous current but will run to a high enough voltage to keep a good capacitor bank charged....for dynamic bursts. I don't see Switchers as having the sheer 'bulk' of capacitor I'd expect for huge bursts of power though operating at the high frequencies they do is GOOD for recharge rate and of course, efficiency. I own low sensitivity Magnepan. I'm not a nose-bleed level kind of guy. But when watching a movie I DO really like some 'slam' and so tend to value the dynamic power of an amp. Nothing beats the sound of every dish in the house rattling at once as some very loud event happens on screen. |
|
KeithL
Administrator
Posts: 9,938
|
Post by KeithL on May 25, 2018 13:52:57 GMT -5
You should note that the rail voltage is the absolute maximum value for the PEAK voltage. So the maximum RMS value of a sine wave driven by that rail will be the rail voltage divided by 1.4 .
Vrms is Vpeak / 1.4
The power formula is (Vrms squared divided by the impedance).
Post by DYohn on Mar 6, 2012 at 11:58am 'Dynamic headroom can be calculated from the nominal continuous power rating and the max peak power capacity of an amplifier. The formula is: headroomdb = 10*log(Pmax/Pnom) If the Pmax value is not published, the best case value can be derived from the nominal power supply rail voltage, which can be measured if it's not published. The formula for Pmax in a stereo amp is ((Vrms^2)/2)/load resistance in ohms.' I found this in another thread but it relates to the XPA-1, probably a gen 1 or 2. Reading this thread got me interested and your last post Leonski implied you think there isn't much headroom in the gen 3 amps either given their topology. The power supply is regulated which helps but I'm getting weird numbers when trying to run this calculation. Any idea what the headroom may be for the new gen amps? A few posts later in the referenced thread the educated guess is approx 1.5db |
|
|
|
Post by leonski on May 25, 2018 16:15:13 GMT -5
Keith here..... Here's a copy of the reply I sent to emotifan on his desk case:
The early Carver amps actually included a variety of interesting technologies. Their "magnetic field power supply" was an early sort of switching supply that used a triode to chop AC power from the AC line directly through a transformer. Many of the amps themselves were Class G designs (which is sort of like Class H); they ran on three separate sets of power supply rails. Later Sunfire amps used what they called "a tracking downconverter", which was basically a single switching supply whose voltage varied to track the signal. (Think of a Class D amplifier powering a Class A/B amplifier.) Modern switch mode power supplies work a lot better - but the general principle is the same. Our new XPA models also use our Class H topology - which is the modern version of Class G.
My understanding was that Carver used a CHOKE as a ' magnetic storage device'. When the voltage dropped, the field collapsed and sent a 'kick' of juice down the line. Providing power. And YES on the Triac, similar to an SCR, but bi-directional. My M400t, 'the cube' made a very low-level 'putt-putt' noise for its entire lifespan. I believe that amp ALSO had at least 2 voltage rails. It ran my Magnepan MG-1 panels to redline while at the same time 'flickering' the house lights in time to the music. |
|
KeithL
Administrator
Posts: 9,938
|
Post by KeithL on May 25, 2018 17:31:48 GMT -5
Basically you are correct..... A choke is simply a transformer with a single winding. The simplest switch mode power supply is what's called "a buck boost regulator". Basically, you close a switch, allowing current to flow through an inductor - "storing energy in the magnetic field". You then open that switch, cutting off the feed of current, and close another switch, allowing the inductor to "dump the stored energy through the load". The choke is indeed used for short term energy storage as the device cycles.... but you still need to use a capacitor as a "reservoir" to filter the output power. Note that the size capacitor you need for a given current output is inversely proportional to the frequency involved... To filter power derived from a 60 Hz line frequency requires relatively large capacitors... By utilizing a much higher frequency, a typical switch mode supply can deliver equivalent performance using much smaller valued capacitors. Here's a link to a rather detailed explanation of the Carver Cube.... including schematics.... www.angelfire.com/sd/paulkemble/sound8d.htmlIt basically works on that principle.... using the triac to switch the power going into the inductor. However, by substituting a transformer with multiple windings, the power supply can deliver three separate sets of rail voltages. (The transformer also provides isolation, so you don't get electrocuted if you touch the speaker terminal and a water pipe... which the safety police consider important.) The "choke" is simply a transformer whose magnetic properties have been optimized for this application. (A transformer in a normal power supply passes a smoothly varying magnetic field; the one in the Carver Cube passes something more like a pulsed magnetic field.) Operationally, a typical power supply recharges the energy resorvoir whenever the output of the rectifiers exceeds the voltage of the storage capacitor. This results in energy being drawn at the peaks of each power cycle. In contrast, the Carver circuit monitors the resorvoir voltage, and "tells the triac to bite off some juice" whenever it detects that the level is getting low. As you can see, the Carver used three sets of voltage rails. Keith here..... Here's a copy of the reply I sent to emotifan on his desk case:
The early Carver amps actually included a variety of interesting technologies. Their "magnetic field power supply" was an early sort of switching supply that used a triode to chop AC power from the AC line directly through a transformer. Many of the amps themselves were Class G designs (which is sort of like Class H); they ran on three separate sets of power supply rails. Later Sunfire amps used what they called "a tracking downconverter", which was basically a single switching supply whose voltage varied to track the signal. (Think of a Class D amplifier powering a Class A/B amplifier.) Modern switch mode power supplies work a lot better - but the general principle is the same. Our new XPA models also use our Class H topology - which is the modern version of Class G. My understanding was that Carver used a CHOKE as a ' magnetic storage device'. When the voltage dropped, the field collapsed and sent a 'kick' of juice down the line. Providing power. And YES on the Triac, similar to an SCR, but bi-directional. My M400t, 'the cube' made a very low-level 'putt-putt' noise for its entire lifespan. I believe that amp ALSO had at least 2 voltage rails. It ran my Magnepan MG-1 panels to redline while at the same time 'flickering' the house lights in time to the music. |
|
|
|
Post by leonski on May 25, 2018 21:28:07 GMT -5
Choke V Transformer is sort of like Transformer V Autoformer.
|
|
