I dont want or need this, its a curiosity question. I was talking to someone on another forum about my amp project. He was argueing that I wasted my money and should have purchased a used QSC or Crest amplifier, becuase I could have had 10,000 watts for 500 dollars, or whatever. I said that, first, I did it because I like to DIY, second, I don't like proaudio amplifiers because of the fans, and third, I doupt highly he was getting 10,000 watts from his amplifiers. I'm curious, because he is now saying he tested his amp to 18,000 watts and is running more than one of them, is that even possible for peaks? I mean, it would require over 150 amps of current at 117 volts. More current at 100 volts, which is probably closer to the rails of such an amp, but I don't know. Last amplifier I used rated at even 5,000 watts RMS used 240 lines because of the current problems. His amp uses a switchmode supply, now those don't ever seem to have much reserve capacitance, so am I right to think that the outlet would need to supply most of the current? Wouldn't even a brief peek at 150-200 amps be enough to not only trip the breaker but potentially melt the insulation on 14 guage Romex?
Now this probably is a very flawed test, but I have used an AC Inductive current measuring meter from Fluke to measure the amount of current that is drawn at startup before of many of my amplifiers. This method may be too inaccurate or too slow to get an accurate reading, but doing this, I've never even seen 100 amps through the A/C lines when it does trip a breaker. It seems like physics alone dictates that 10kw from a normal household outlet is next to impossible. I would even question if some of the amps he is mentioning have enough output devices to support that.
Now this probably is a very flawed test, but I have used an AC Inductive current measuring meter from Fluke to measure the amount of current that is drawn at startup before of many of my amplifiers. This method may be too inaccurate or too slow to get an accurate reading, but doing this, I've never even seen 100 amps through the A/C lines when it does trip a breaker. It seems like physics alone dictates that 10kw from a normal household outlet is next to impossible. I would even question if some of the amps he is mentioning have enough output devices to support that.
10KW
I used to run two Carver 4.0t's bridged on the same 120V AC line, they could easily put out 1200W into my 8 ohm speakers all day and would pop the 25amp breakers long before the peak lights would ever flicker. Lets just take some really simplified math and basic amp designs, said 10KW amp is bridged using 120VDC+/- rail for a total of 240VDC * 2 for a bridged swing of 480VDC. 10,000/480=20.83amps. Now the amp is stereo so you have two channels 41.66 amps, now factor in the power loss of the amp we will even give it a very efficent class H at a thirty percent loss total 54.158amps!! and thats very conservative.
The kind of power exceeds most home mains in the US let alone standard wiring.
I used to run two Carver 4.0t's bridged on the same 120V AC line, they could easily put out 1200W into my 8 ohm speakers all day and would pop the 25amp breakers long before the peak lights would ever flicker. Lets just take some really simplified math and basic amp designs, said 10KW amp is bridged using 120VDC+/- rail for a total of 240VDC * 2 for a bridged swing of 480VDC. 10,000/480=20.83amps. Now the amp is stereo so you have two channels 41.66 amps, now factor in the power loss of the amp we will even give it a very efficent class H at a thirty percent loss total 54.158amps!! and thats very conservative.
The kind of power exceeds most home mains in the US let alone standard wiring.
Using the formula on Wikipedia for peak power, no idea if this is a correct formula, it says, take the peak to peak voltage, then find the peak amplitude it can drive, I assumed it was half the peak to peak, square then, then divide by the impedance, and multiply by the efficiency rating.
120 volt rails X 2, 240 volts P-P, bridged as you put it is then 480, but then divided in half for peak amplitude. Thats 240 squared divided by 8 ohms is 7,200 watts, or 14,400 watts rms. 14,400 watts for 1 second equals 14,400 jouls. Now I understand that a peak impulse could last less time yet, and the only thing I could find was that a single milisecond would multiply that times 250. I couldn't find any goo d charts, just an article that said 2 jouls equals 500 j/ms. If an outlet can produce a maximum of say 3000 jouls, and we reduce this to 1 ms, then I do start to see how an amplifier could potentially produce some insane numbers, if it could utilize that much energy. I mean, thats 750,000 watts per millisecond then. However I have no idea if thats correct or the right way of looking at it. I can't find any good way of calculating an amplifiers instantaneous peak, and because its looked at as a useless number, it seems to be difficult to find.
120 volt rails X 2, 240 volts P-P, bridged as you put it is then 480, but then divided in half for peak amplitude. Thats 240 squared divided by 8 ohms is 7,200 watts, or 14,400 watts rms. 14,400 watts for 1 second equals 14,400 jouls. Now I understand that a peak impulse could last less time yet, and the only thing I could find was that a single milisecond would multiply that times 250. I couldn't find any goo d charts, just an article that said 2 jouls equals 500 j/ms. If an outlet can produce a maximum of say 3000 jouls, and we reduce this to 1 ms, then I do start to see how an amplifier could potentially produce some insane numbers, if it could utilize that much energy. I mean, thats 750,000 watts per millisecond then. However I have no idea if thats correct or the right way of looking at it. I can't find any good way of calculating an amplifiers instantaneous peak, and because its looked at as a useless number, it seems to be difficult to find.
Thanks, they don't seem to have any technical white papers, which is what I thought you were telling me. However, in the specs they do mention the peak output voltage and the peak current, which implies over 11,000 watts per channel in the largest FP13000. However, nothing is given to suggest how this is accomplished. I see that it can run on 120 volts, but can it produce that much output from 120, and if so, how much current draw. I do see however that its power supply is much more robust than is typical of such a supply. It has quite a bit of capacitance it looks like, which would probably help for peaks, though I still don't see enough capacitors, not with such high rails, to store enough joules of energy that much power.
Ok reading the manual did enlighten me a bit, they rate it for use with a 115 volt twist lock nema plug rated at 30 amps. They also say that to reach just 1/4 of its rated power requires 58 amps from the outlet. However you can run it on a standard 220 line rated at 15-20 amps and it will only draw 28 amps to reach 1/4 its rated power. Still doesn't give me much info in regard to understanding or measuring an amps peak output, other than it does look like using even a 30 amp outlet, these amps could not possibly reach there peak output other than for just brief moments.
It's normal to derate a circuit breaker to 80%, so:
@120Volts
15A=12Acontinuous=1440Watts (120*12)
20A=16Acontinuous=1920Watts (120*16)
30A=24Acontinuous=2880Watts (120*12)
30A@240v=24Acontinuous=5760Watts (240*24)
So, you'd need about a 60 amp breaker@240V for 10kw continuous (100% amp efficiency)
@120Volts
15A=12Acontinuous=1440Watts (120*12)
20A=16Acontinuous=1920Watts (120*16)
30A=24Acontinuous=2880Watts (120*12)
30A@240v=24Acontinuous=5760Watts (240*24)
So, you'd need about a 60 amp breaker@240V for 10kw continuous (100% amp efficiency)
Sure, 10,000W is possible. It just requires 208 3 phase at 30 amps. I'm sure many know of this beast:
http://www.crownaudio.com/pdf/legacy/k10684.pdf
http://www.crownaudio.com/pdf/legacy/k10684.pdf
I think you are making this more complicated than it needs to be. This is a matter of simple arithmetic.
Here are the power formulas -
where
P = Power in Watts
I = Current in Amps
E = Voltage in Volts
^2 = a number squared
sqrt() = square root of number in parenthesis
P = IE ...(Current times Voltage)
P = E^2 / R ...(Voltage squared divided by Resistance)
P = I^2 R ...(Current square times Resistance
(link to Ohms Law Pie Chart, scroll down to bottom of link)
http://www.hvacwebtech.com/Ohm's%20Law.htm
[sorry the link won't work because it has a 'space' in it (%20), but I think if you cut and paste the whole thing into your browser, it will work.]
To find the necessary Current applied to an 8 ohm load to produce 10,000 watts -
I = sqrt(P/R) = sqrt(10,000/8) = 35.355 AMPS
To find the necessary Voltage applied to 8 ohms to produce 10,000 watts -
E = sqrt(P x R) = sqrt(10,000 x 8) = sqrt(80,000) = 282.84 VOLTS
That is not from the power grid or electrical outlet, you have to apply that Voltage and that Current to an 8 ohm load to get 10,000 watts.
To find the current necessary from the electrical outlet, assuming 120 volts, we again use I = P/E
I = P/E = 10,000/120 = 83.33 AMPS
83.33 amps from the wall outlet, I don't think so.
Even if we increase it to 240V, that's still 41.667 AMPS, and again, I don't think so.
I'm not even aware of the existence of a 10,000 watt amp. Yes, certainly Rock Concerts use more that 10,000 watts to drive their PA systems, but that is a combination of many amps working together, not a single amp.
The Lab Gruppen amps referred to by someone else are professional concert power amps. One of their larger amps is 2 Channels at 3200 watts per channel for 6400 watts total. That is a rare and expensive amp, I just don't see it going for $500. Their biggest amp is 4 Channels at 2500 watts per channel for a total of 10,000 watts. Both amps into a 2 ohm load.
I like to know how your friend determine his amp was 10,000 watts. No offense, but personally I think he is full of crapola. Especially, when he claims you can buy it for $500. I don't think so.
How about giving us the Brand name of this Magic 10,000 watt amp that sells for a magic 500 dollar bill?
Steve/BlueWizard
Here are the power formulas -
where
P = Power in Watts
I = Current in Amps
E = Voltage in Volts
^2 = a number squared
sqrt() = square root of number in parenthesis
P = IE ...(Current times Voltage)
P = E^2 / R ...(Voltage squared divided by Resistance)
P = I^2 R ...(Current square times Resistance
(link to Ohms Law Pie Chart, scroll down to bottom of link)
http://www.hvacwebtech.com/Ohm's%20Law.htm
[sorry the link won't work because it has a 'space' in it (%20), but I think if you cut and paste the whole thing into your browser, it will work.]
To find the necessary Current applied to an 8 ohm load to produce 10,000 watts -
I = sqrt(P/R) = sqrt(10,000/8) = 35.355 AMPS
To find the necessary Voltage applied to 8 ohms to produce 10,000 watts -
E = sqrt(P x R) = sqrt(10,000 x 8) = sqrt(80,000) = 282.84 VOLTS
That is not from the power grid or electrical outlet, you have to apply that Voltage and that Current to an 8 ohm load to get 10,000 watts.
To find the current necessary from the electrical outlet, assuming 120 volts, we again use I = P/E
I = P/E = 10,000/120 = 83.33 AMPS
83.33 amps from the wall outlet, I don't think so.
Even if we increase it to 240V, that's still 41.667 AMPS, and again, I don't think so.
I'm not even aware of the existence of a 10,000 watt amp. Yes, certainly Rock Concerts use more that 10,000 watts to drive their PA systems, but that is a combination of many amps working together, not a single amp.
The Lab Gruppen amps referred to by someone else are professional concert power amps. One of their larger amps is 2 Channels at 3200 watts per channel for 6400 watts total. That is a rare and expensive amp, I just don't see it going for $500. Their biggest amp is 4 Channels at 2500 watts per channel for a total of 10,000 watts. Both amps into a 2 ohm load.
I like to know how your friend determine his amp was 10,000 watts. No offense, but personally I think he is full of crapola. Especially, when he claims you can buy it for $500. I don't think so.
How about giving us the Brand name of this Magic 10,000 watt amp that sells for a magic 500 dollar bill?
Steve/BlueWizard
Well He was claiming I could get it on the used market, and he actually claimed he got 18,000 watts on his measurement equipment. I asked how exactly he determined this, as I found it highly unlikely. Then he said, well it was only for a second, and I said, I doubt that highly, so he replied, well I meant millisecond. I'm making this sound worse than it is on his part, I mean, I guess he was trying to imply PMPO, which I personally find a worthless way to look at an amplifier. His argument, is, even if the amp can't produce 10,000 watts, if it only produces 2,000 watts given the power you give it, isn't that better than an amp that can only give 150 watts. He has also danced around the subject of how many circuits, and I get the impression its only one 15 amp circuit for his claimed 60,000 watts of amplifier power he is running to his speakers.
My argument is that, in the end, an amplifiers worth isn't dictated by how much power it can produce with less than 1% distortion. S/N ratio is very important to me, especially noise it adds to the room. This means most if not all pro amps with fans will not be ok, as many of them are in excess of 30 db's of fan noise. I hear people argue that they can't hear the fan noise, and even if I can, I can't hear it while music is playing. I don't buy that, I can hear it, it bothers me, and I figure it becomes what my backdrop is "painted" with. My black background is now a whirring noise. They also cover the inside of the amps with dust, and require routine cleaning to keep them from collecting excessive dust. This is far less of an issue with a normal convection cooled amp. He argued, well you can slow the fan or unhook the fan. With a Labgruppen and many similar "radiator" cooled amps, this wouldn't be possible as the amps are cooled in a way that looks like it requires forced cooling. For others, clearly the amps heat sinks are undersized as they expect the forced cooling, and not only could overheat, but you have no voided the warranty. I blew my SAE amplifier by unhooking its fan, and I blew a Behringer amp by slowing its fan, so I don't agree thats an option.
Crest 10001 is the amp he is claiming does this 18,000 watts peak. He is now stating that the amp is wired with two 120 volt outlets each with a 30 amp breaker.
My argument is that, in the end, an amplifiers worth isn't dictated by how much power it can produce with less than 1% distortion. S/N ratio is very important to me, especially noise it adds to the room. This means most if not all pro amps with fans will not be ok, as many of them are in excess of 30 db's of fan noise. I hear people argue that they can't hear the fan noise, and even if I can, I can't hear it while music is playing. I don't buy that, I can hear it, it bothers me, and I figure it becomes what my backdrop is "painted" with. My black background is now a whirring noise. They also cover the inside of the amps with dust, and require routine cleaning to keep them from collecting excessive dust. This is far less of an issue with a normal convection cooled amp. He argued, well you can slow the fan or unhook the fan. With a Labgruppen and many similar "radiator" cooled amps, this wouldn't be possible as the amps are cooled in a way that looks like it requires forced cooling. For others, clearly the amps heat sinks are undersized as they expect the forced cooling, and not only could overheat, but you have no voided the warranty. I blew my SAE amplifier by unhooking its fan, and I blew a Behringer amp by slowing its fan, so I don't agree thats an option.
Crest 10001 is the amp he is claiming does this 18,000 watts peak. He is now stating that the amp is wired with two 120 volt outlets each with a 30 amp breaker.
Well, it appears that you are right. Though, it probably would have help if you have explained the application you were using these amps for, or at least there design intended application. The Crest Pro 10001 is an extremely unusual amp (presumably) intended for large auditoriums and rock concerts. I don't see the average person coming across one of these in their lifetime.
Here are the ratings on the Crest Pro 10001 -
# 5,000 watts @ 2 ohms (stereo)
# 7,500 watts @ 1 ohms (stereo)
# 10,000 watts bridged into 4 ohms
# 15,000 watts bridged into 2 ohms
http://www.directproaudio.com/product.cfm?directid=53802
http://www.directproaudio.com/images/products/crest10001data.pdf
Price -
Suggested Retail = US$6,272
From DirectProAudio = US$4,704
(now discontinued )
So, in theory, if your friend ran the amp in bridged mode into a 1 ohm load, it might well reach 18,000 watts...maybe. But it will certainly reach 15,000 watts...sort of.
The actual rating is 2 ohm Bridged Mode Mono Power = 10,000 watts 20hz to 20khz at 0.1% THD and 15,000 watts Typcial Music Prog. Material.
Only two things affect available power, one is power supply voltage and the other is load or impedance. Impedance goes down, for a given applied voltage, the power goes up. But, it should be noted that the power can never exceed the available power supply voltage; that is fixed. Power supply voltage never goes up, only down. I suspect that the difference in power specs between the 20hz to 20khz test and the 'typical music prog. material' is a reflection of that very fact. Under the sustained sinewave signal of a frequency sweep, the power supply voltage is drawn down. For 'typical music', the power supply has a very short term reserve of available voltage before it drops, so extremely short term peaks could reach high power levels.
Assuming it doesn't exceed the 'no load' available power supply voltage, your friend might ...might... have gotten 18,000 watts assuming he was driving a 1 ohm load.
So, exactly what are you guys doing with these monster amps, certainly not listening to stereo and watching movies in your living room, ...or possibly you are?
Unless it was a heavily used and abused, and very old amp, can you really get a $4,700 amp for $500? And if it was all those things, would you really want it?
I think I probably agree with you comment about ambient noise from the Amp fans. These amps are meant to be used in an environment where the ambient noise is extremely high, like a rock concert, so fan noise isn't a concern because high ambient noise drowns it out. However, in other settings with low ambient noise, like your living room, it could easily be a problem. But, sadly, a very necessary problem, 10,000watts is no pixie amp. 10,000 watts is a lot of heat that has to go somewhere.
Though you could consider water chilling heatsinks. Many 'gamers' use them on their game computers to keep the processor cool. Essentially, chilled water (actually non-corrosive anti-freeze) is pumped through the heat sink to keep things cool, and the chilling mechanism can be some distance from the power being cooled.
One small problem with water chilling though, is when the amp is not under load, or if the Amp is off and the water chiller is still running, it can cause condensation on the heatsinks.
So, sorry I doubted your friend, but once again, what the heck are you guys doing with these monster amps??????
Steve/BlueWizard
Here are the ratings on the Crest Pro 10001 -
# 5,000 watts @ 2 ohms (stereo)
# 7,500 watts @ 1 ohms (stereo)
# 10,000 watts bridged into 4 ohms
# 15,000 watts bridged into 2 ohms
http://www.directproaudio.com/product.cfm?directid=53802
http://www.directproaudio.com/images/products/crest10001data.pdf
Price -
Suggested Retail = US$6,272
From DirectProAudio = US$4,704
(now discontinued )
So, in theory, if your friend ran the amp in bridged mode into a 1 ohm load, it might well reach 18,000 watts...maybe. But it will certainly reach 15,000 watts...sort of.
The actual rating is 2 ohm Bridged Mode Mono Power = 10,000 watts 20hz to 20khz at 0.1% THD and 15,000 watts Typcial Music Prog. Material.
Only two things affect available power, one is power supply voltage and the other is load or impedance. Impedance goes down, for a given applied voltage, the power goes up. But, it should be noted that the power can never exceed the available power supply voltage; that is fixed. Power supply voltage never goes up, only down. I suspect that the difference in power specs between the 20hz to 20khz test and the 'typical music prog. material' is a reflection of that very fact. Under the sustained sinewave signal of a frequency sweep, the power supply voltage is drawn down. For 'typical music', the power supply has a very short term reserve of available voltage before it drops, so extremely short term peaks could reach high power levels.
Assuming it doesn't exceed the 'no load' available power supply voltage, your friend might ...might... have gotten 18,000 watts assuming he was driving a 1 ohm load.
So, exactly what are you guys doing with these monster amps, certainly not listening to stereo and watching movies in your living room, ...or possibly you are?
Unless it was a heavily used and abused, and very old amp, can you really get a $4,700 amp for $500? And if it was all those things, would you really want it?
I think I probably agree with you comment about ambient noise from the Amp fans. These amps are meant to be used in an environment where the ambient noise is extremely high, like a rock concert, so fan noise isn't a concern because high ambient noise drowns it out. However, in other settings with low ambient noise, like your living room, it could easily be a problem. But, sadly, a very necessary problem, 10,000watts is no pixie amp. 10,000 watts is a lot of heat that has to go somewhere.
Though you could consider water chilling heatsinks. Many 'gamers' use them on their game computers to keep the processor cool. Essentially, chilled water (actually non-corrosive anti-freeze) is pumped through the heat sink to keep things cool, and the chilling mechanism can be some distance from the power being cooled.
One small problem with water chilling though, is when the amp is not under load, or if the Amp is off and the water chiller is still running, it can cause condensation on the heatsinks.
So, sorry I doubted your friend, but once again, what the heck are you guys doing with these monster amps??????
Steve/BlueWizard
I'm doing nothing with these amps, I've never even seen one of these amps. I use a 3 channel amplifier I built with Aussiamps NX150 modules and a CLC power supply I came up with. My amp fits my needs perfectly, but when I shared this on another forum, I was told I wasted my money, as I would have been better off buying a used Pro amp, because it would have put out more power. That was how this whole discussion started. Obviously I took some offense to someone stating that I wasted my money, which he contended, if you like to DIY, then it makes sense for you. I said, its not just that I like to DIY, I can control certain things that are important to me, like S/N, power supply design and size, parts quality, etc. I won't even get into that, but it brought on the discussion of all amplifiers sound exactly a like and science has proven it. I have a very strong opinion on that, which I have said in other posts, but thats a separate argument.
Forgetting even the science of all that, my opinion is, if it doesn't hurt anyone, it makes me happy, and I'm not forcing my views on others, let me be. Unfortunately, I think a lot of people think that don't feel differences exist feel that those of us who do are somehow forcing our views on them, and thus fight and argue back. It really becomes like a religious battle. I have tried plenty of things I felt made no difference, however I have never felt that someone made me waste my money, I'm not that mindless, and I hold no ill will. My feeling is that there are too many confounding variables in such a situation, and who knows, maybe the other guy legitimately heard a difference, and whatever problem it fixed doesn't exist in my system, or is outshined by other bigger problems. Maybe all of its snake oil, I don't know, but I don't care either, because I honestly enjoy playing with all these things, and when I do find something that makes a difference, that is the best part of all of it for me.
Forgetting even the science of all that, my opinion is, if it doesn't hurt anyone, it makes me happy, and I'm not forcing my views on others, let me be. Unfortunately, I think a lot of people think that don't feel differences exist feel that those of us who do are somehow forcing our views on them, and thus fight and argue back. It really becomes like a religious battle. I have tried plenty of things I felt made no difference, however I have never felt that someone made me waste my money, I'm not that mindless, and I hold no ill will. My feeling is that there are too many confounding variables in such a situation, and who knows, maybe the other guy legitimately heard a difference, and whatever problem it fixed doesn't exist in my system, or is outshined by other bigger problems. Maybe all of its snake oil, I don't know, but I don't care either, because I honestly enjoy playing with all these things, and when I do find something that makes a difference, that is the best part of all of it for me.
For those who don't know about Aussiamps NX150 , they are impressive.
http://www.aussieamplifiers.com/nx150.htm
In Short-
"NX150 - Next Generation Hexfet / Lateral MOSFET Power Module.
The NX150 Amplifier is a Solid State Complementary Dual Differential Cascoded Topology running in Class AB. It would be one of the most simplest amplifier modules to assemble and test and would require only the most rudimentary knowledge of electronics to complete. Some skill with a soldering Iron would be desirable though.
Specs -
+-56vdc Voltage Rails (Idle, No Signal)
= 146 Watts RMS into 8 Ohms
= 230 Watts RMS into 4 Ohms
+-63vdc Voltage Rails (Idle, No Signal)
= 185 Watts RMS into 8 Ohms
= 275 Watts RMS into 4 Ohms
Frequency Response: 2hz to 160khz (Input Filter Limited)
Power Bandwidth 2Hz - 500khz -3dB Filter Limited
THD Typically 0.005% @ 8 Ohms 1khz
Damping Factor 400
Input Impedance 34k Ohms
Input sensitivity for full power out into 8 Ohm load 900mv
WOW! I'm impressed...seriously.
AU$160 per module in a kit
AU$180 for a +-63V Linear Power Supply.
About US$292.50 ... plus a little work and a few spare parts.
I'm still impressed.
One last comment, regarding the 10,000watt recommendation, raw power for the sake of raw power, in my opinion, is pointless. What does all that power really buy you but a lot of heat, a lot of noise (all kinds), and a lot of risk? Personally, I think 200 good watts are better than 10,000 sucky watts, but that's just me.
Any chance of a photo?
Steve/BlueWizard
http://www.aussieamplifiers.com/nx150.htm
In Short-
"NX150 - Next Generation Hexfet / Lateral MOSFET Power Module.
The NX150 Amplifier is a Solid State Complementary Dual Differential Cascoded Topology running in Class AB. It would be one of the most simplest amplifier modules to assemble and test and would require only the most rudimentary knowledge of electronics to complete. Some skill with a soldering Iron would be desirable though.
Specs -
+-56vdc Voltage Rails (Idle, No Signal)
= 146 Watts RMS into 8 Ohms
= 230 Watts RMS into 4 Ohms
+-63vdc Voltage Rails (Idle, No Signal)
= 185 Watts RMS into 8 Ohms
= 275 Watts RMS into 4 Ohms
Frequency Response: 2hz to 160khz (Input Filter Limited)
Power Bandwidth 2Hz - 500khz -3dB Filter Limited
THD Typically 0.005% @ 8 Ohms 1khz
Damping Factor 400
Input Impedance 34k Ohms
Input sensitivity for full power out into 8 Ohm load 900mv
WOW! I'm impressed...seriously.
AU$160 per module in a kit
AU$180 for a +-63V Linear Power Supply.
About US$292.50 ... plus a little work and a few spare parts.
I'm still impressed.
One last comment, regarding the 10,000watt recommendation, raw power for the sake of raw power, in my opinion, is pointless. What does all that power really buy you but a lot of heat, a lot of noise (all kinds), and a lot of risk? Personally, I think 200 good watts are better than 10,000 sucky watts, but that's just me.
Any chance of a photo?
Steve/BlueWizard
I've posted pics before, but I don't mind posting them again. I don't have any good completed pics, especially of the power supply. I'm in the process of moving, so once I have moved, I will get to finishing them. Essentially, take this as Beta 1 of the amp, as I just wanted to get it working so I could begin testing what a difference each change I make has. For instance, one I expect a huge difference from is the increased capacitance and transformer size, I expect a large change in dynamic performance. Oh and I'm not using Aussie's power supply, I'm just using his amp modules. My power supply uses Hitachi 22mf capacitors rated at 100 volts. They have an ESR of 7 or 8 milliohms. Mind you, the common Panasonic TC and other similar caps are in the 30 milliohm range, rather than 8. If you use to 10mf TC's you still are near twice the ESR of mine. If I then use two in parallel, I have an even lower ESR yet, soemthing difficult to match with the 10mf caps commonly used. The Ripple voltage at my power supply under idle is around 765 uv's, I.E. around 450 milliamps current, and 63 volts out. Thats less that a mv of ripple voltage. Ripple current across the load is around 450 milliamps. With maximum worse case scenereo current draw being say 14 amps at the load at 42 volts, we have around 18 millivolts of ripple voltage. Changing things to the proposed 44mf's at C1 and 66mf's at C2, with an estimated ERS of 4 milli-ohms at C1 and 2-3 milliohms at C2, things change for the better. Now the Idle ripple voltage is around 130 micro volts, and we have a strong 63.5 volts. Changing to around 1.5 amps of current draw voltage drops to 62 volts and ripple changes to 300 micro volts. Now going to a 3 ohm load, which changes current to around 14 amps at output, drops voltage to 42 volts, and ripple changes to 3 millivolts. To make Andrew T happy as I believe he feels that a better aproximation of real world worst case scenereo, here is the results of a .5 ohm load. It will have 55 amps at output and 27 volts, with 9.6 millivolts of ripple. However, at that point I am massivly exceeding the current ability of my inductors, this probably causing massive heat rise in those, along with producing almost 1500 watts of power, even with 60% efficiency thats 890 watts, or almost 300 watts per amp module. I won't say it can't happen, but I don't see it being a realistic situation, and I'm not sure how to best simulate what a short term peak would look like. I also don't have resistors that could dissipate the power necessary to test that in real life.
One area I'm unsure of with these simulations, as my real world measurements of the current setup isn't quite that nice, is the Transformer source resistance. Using the built in calculater it comes up with 1.8 ohms. However using that causes way too much voltage drop, and isn't what I'm seeing. Going with my best guess of 1 ohm for just one transformer and 500 ohms for two, things look like what I'm showing, and improve greatly. The numbers match the simulations better, accept I haven't done any peak current draw tests yet. Actually with the amps at idle, I'm getting around 65 volts, and with no amps connected, ie no load, I get 67 volts. Anyway, my power supply ripple is extremely low compared with a normal capacitor filtered system, which has lowered the noise floor, but should also improve those typical distortion specs, and reduce power supply spurious quite a bit. I don't have a scope, but I do hope to have one once I finish my move, and hope to show how wonderful the supply is working.
Now I have mentioned this before, but this picture was before I finished the power supply. The wires are cut to propper lengths, twisted together into nice neat bundles, secured into place, and everything is mounted via nuts and bolts. Future plans are to hook a second 600VA transformer I have up into this, which I thought I tried, but had a lead come loose, and never fixed it before packing it away. I also will add a second pair of caps for C1, the caps you see, and if I have room, a pair of caps for C2 in that chassis. Then adding a second pair of caps into the amp chassis, if I can find room to do so. This will create the 44mf C1, 11mh L1, and 66mf C2. Actually because of the inductance and resistance of the wire between the power supply and amp chassis, that wire probably should be modeled. I don't know what the series inductance or resistance is, but it probably is actually a CLCRLC, though I suspect the resistance of the wire is far more important than the minimal inductance it has. I also know that whatever effect it has, other than voltage drop, isn't a bad thing, and isn't something to worry about.
One area I'm unsure of with these simulations, as my real world measurements of the current setup isn't quite that nice, is the Transformer source resistance. Using the built in calculater it comes up with 1.8 ohms. However using that causes way too much voltage drop, and isn't what I'm seeing. Going with my best guess of 1 ohm for just one transformer and 500 ohms for two, things look like what I'm showing, and improve greatly. The numbers match the simulations better, accept I haven't done any peak current draw tests yet. Actually with the amps at idle, I'm getting around 65 volts, and with no amps connected, ie no load, I get 67 volts. Anyway, my power supply ripple is extremely low compared with a normal capacitor filtered system, which has lowered the noise floor, but should also improve those typical distortion specs, and reduce power supply spurious quite a bit. I don't have a scope, but I do hope to have one once I finish my move, and hope to show how wonderful the supply is working.
Now I have mentioned this before, but this picture was before I finished the power supply. The wires are cut to propper lengths, twisted together into nice neat bundles, secured into place, and everything is mounted via nuts and bolts. Future plans are to hook a second 600VA transformer I have up into this, which I thought I tried, but had a lead come loose, and never fixed it before packing it away. I also will add a second pair of caps for C1, the caps you see, and if I have room, a pair of caps for C2 in that chassis. Then adding a second pair of caps into the amp chassis, if I can find room to do so. This will create the 44mf C1, 11mh L1, and 66mf C2. Actually because of the inductance and resistance of the wire between the power supply and amp chassis, that wire probably should be modeled. I don't know what the series inductance or resistance is, but it probably is actually a CLCRLC, though I suspect the resistance of the wire is far more important than the minimal inductance it has. I also know that whatever effect it has, other than voltage drop, isn't a bad thing, and isn't something to worry about.
Nobody seems to mention that music signals exhibit high crest factors, usually in the 6..12dB range.
In practice this means that music can easily drive a "10KW" amplifier into clipping while delivering only 600..2500W average to the load.
On the other hand, in high efficiency (switching) amplifiers the amount of power drawn from the outlet is almost the average power being actually delivered to the load, but plain class AB (or B) amplifiers tend to waste as heat 4..8 times the actual output power while playing music.
In practice this means that a 120V 15A outlet is usually enough to power 10KW worth of "hi-tech" high efficiency amplifiers (or even 20KW), but not 10KW of old school class AB stuff.
In practice this means that music can easily drive a "10KW" amplifier into clipping while delivering only 600..2500W average to the load.
On the other hand, in high efficiency (switching) amplifiers the amount of power drawn from the outlet is almost the average power being actually delivered to the load, but plain class AB (or B) amplifiers tend to waste as heat 4..8 times the actual output power while playing music.
In practice this means that a 120V 15A outlet is usually enough to power 10KW worth of "hi-tech" high efficiency amplifiers (or even 20KW), but not 10KW of old school class AB stuff.
Out of curiosity, exactly how does that work. I mean, if a Class AB amplifier has an efficiency rating of say 60%, and a Switching amp is rated at say 95%, is that 35% difference equal to 4-8 times the waste?
Another thing, when you say, a musical signal has high crest factors of 6-12 db's, are you referring to the difference in signal between the say RMS db's and the peak, or for instance, the sudden dynamic swings that take place? In that case, I believe I have read that the large majority of modern CD's often only have between 3-6 db's of dynamic swing. Which of course isn't the point because they can have much more, the manufacturers simply choose not to use it. Movies do have that and more often it seems. Anyway, to the point, how does a 120v15amp outlet drive a 10 or 20kw amp. I mean, I could see the outlet producing enough power for say very short dynamic swings, but I would thing anything over half a second in duration would be enough to trip the breaker. I would think, looking at RMS or actual sustainable power, that 1500 watts is about the limit for a 15 amp outlet.
I mean, it seems like if you talk only in terms of peaks, that its difficult then to look at a 10kw amp as a 10kw amp. It seems that then many of the overbuilt 2kw amps capable of 10kw peaks are just as much a 10kw amp as the "10kw" pro amp. Not that a lot of these amps exist, but simply that, its a difficult and complex way of looking at things. The other thing here is that, if an amp can produce so far in excess of the power a speaker can handle, it seems like risk of blowing the speakers goes way up during peaks. I mean, yes, I know a speaker can handle many times its rating if its a short 10-20ms peak. However, if that peak sustains for over half a second, then you are talking about the actual peak rating of the driver, which is rarely more than double its RMS rating. In the case of my speakers, the Peak power handling is around 250 watts each. I would think that feeding them more than 250 watts for .5 seconds or more would likely damage them over time, if not sooner. I would think that feeding them 2kw or 10kw or even 20kw for even 30ms would be a risky venture. I suppose this is complicated though by the fact that 12+ db's of dynamic range can require in excess of 10,000 watts, and that most speakers much prefer a short term burst of many times the power clean, than the correct amount clipped. By the way, if you can't tell, I'm not really arguing here, I'm thinking out loud about the merit of such massive amps in a real world system. My guess is, no speaker manufacturer would ever suggest you do something like use a 1kw amp with a speaker designed for 150 watts rms, and yet, no matter what the amp is rated at, its only going to produce what it needs to for a given level for that signal. If durring those dynamic peaks a 200w amp begins clipping at 250 watts, and the 1kw amp is producing 500watts unclipped, the speaker not only probably prefers that, but it probably sounds better. Though I'm not sure a speaker rated at 92db's at 1 watt, when fed 500 watts, would exactly sound good, more painful. I mean, unless I did my calculations wrong, at 119 dbs you are using 512 watts. I consider a full range signal at 119db's to be painful. I'd say its cuz I'm old, but its not true, I'm only 25, maybe just old at heart.
Another thing, when you say, a musical signal has high crest factors of 6-12 db's, are you referring to the difference in signal between the say RMS db's and the peak, or for instance, the sudden dynamic swings that take place? In that case, I believe I have read that the large majority of modern CD's often only have between 3-6 db's of dynamic swing. Which of course isn't the point because they can have much more, the manufacturers simply choose not to use it. Movies do have that and more often it seems. Anyway, to the point, how does a 120v15amp outlet drive a 10 or 20kw amp. I mean, I could see the outlet producing enough power for say very short dynamic swings, but I would thing anything over half a second in duration would be enough to trip the breaker. I would think, looking at RMS or actual sustainable power, that 1500 watts is about the limit for a 15 amp outlet.
I mean, it seems like if you talk only in terms of peaks, that its difficult then to look at a 10kw amp as a 10kw amp. It seems that then many of the overbuilt 2kw amps capable of 10kw peaks are just as much a 10kw amp as the "10kw" pro amp. Not that a lot of these amps exist, but simply that, its a difficult and complex way of looking at things. The other thing here is that, if an amp can produce so far in excess of the power a speaker can handle, it seems like risk of blowing the speakers goes way up during peaks. I mean, yes, I know a speaker can handle many times its rating if its a short 10-20ms peak. However, if that peak sustains for over half a second, then you are talking about the actual peak rating of the driver, which is rarely more than double its RMS rating. In the case of my speakers, the Peak power handling is around 250 watts each. I would think that feeding them more than 250 watts for .5 seconds or more would likely damage them over time, if not sooner. I would think that feeding them 2kw or 10kw or even 20kw for even 30ms would be a risky venture. I suppose this is complicated though by the fact that 12+ db's of dynamic range can require in excess of 10,000 watts, and that most speakers much prefer a short term burst of many times the power clean, than the correct amount clipped. By the way, if you can't tell, I'm not really arguing here, I'm thinking out loud about the merit of such massive amps in a real world system. My guess is, no speaker manufacturer would ever suggest you do something like use a 1kw amp with a speaker designed for 150 watts rms, and yet, no matter what the amp is rated at, its only going to produce what it needs to for a given level for that signal. If durring those dynamic peaks a 200w amp begins clipping at 250 watts, and the 1kw amp is producing 500watts unclipped, the speaker not only probably prefers that, but it probably sounds better. Though I'm not sure a speaker rated at 92db's at 1 watt, when fed 500 watts, would exactly sound good, more painful. I mean, unless I did my calculations wrong, at 119 dbs you are using 512 watts. I consider a full range signal at 119db's to be painful. I'd say its cuz I'm old, but its not true, I'm only 25, maybe just old at heart.
pjpoes said:
I noticed my name being mentioned, give credit to google
as I just stumbled on this thread by accident.
Lets clarify a few things.
1. Here is the original thread;
http://www.icixsound.com/vb/showthread.php?t=38708
2. Where did say you can buy a proamp for 10,000 watts
for $500 ? Let me know, I want one.
This is what was said;
I have two Crest 10001 pro audio amplifiers running on
dual 110VAC 30A dedicated circuits. I did a burst test
to reveal about 18kw. Manufacturer claims about 15kw,
so it's inline with specs.
The Crest 10001 requires dual 110VAC 30A dedicated circuits.
[100VAC - 120VAC = ok]
Burst tests are higher because amplifiers have capacitors,
this is where the big numbers come from. The high voltage
rails with power supply capacitance yields a big score.
The Crest 10001 is conventional power supply. There are
amplifiers in car audio, home audio, pro audio that use SMPS.
There is nothing flawed. You didn't understand my post
and you came here for answers and you didn't even
convey the correct message across. How do you expect
people to answer these questions when everything you said
was in error ? You can have just as easily asked these questions
on the ICX forum.
Wrong!
Read the original thread, I told you two times.
snip ..
I said -> I said "A peak test showed about 18,000w looping on a drum solo. I did a burst test on my Crest 10001 with real music on CD.
Not only that, I told you all the other test variables.
The specifics are;
The seven element load; This load is rated for 35kw rms.
http://home.pacbell.net/lordpk/load/
Where did you get the idea that I used a 15A or 20A breaker? That amplifier
has two hardwired 120VAC lines with a 30A circuit, each leg feeds a massive
toroid.
Here's a pic of the amp guts.
http://home.pacbell.net/lordpk/proamps/Crest_10001-1.jpg
http://home.pacbell.net/lordpk/proamps/Crest_10001-2.jpg
That 4RU amplifier weighs 150 pounds. lol
I used a scope and 2 ohm load in bridged mode to register a 18kw transient.
Trust me, that amp is a beast, class H, IIRC, 32 power transistors per channel. I have 64 power transistors drivng that load with an unloaded
voltage of 320 volts. Under load, the voltage drops, that is normal. The
unloaded rails are +160/-160V..
To even consider testing this beast, I had to make a dual 120VAC 30A
extension cord with 8 awg romex, about 10 foot long that plugs into
my dryer outlet {220VAC, 30A}.
Even with this beefy wire, my unloaded AC droped from 120VAC
to 110VAC. I still got 10VAC of line loss even using big wire.
Fun stuff. :beer:
I read parts of the manual for the Crest Pro 10001. I would agree that peaks in the range claimed are possible, using the dedicated 120s 30a x2 supplies. If I read the manual correctly, they say that the peak current draw per power line going into the Pro 10001 is 35.6A, so if you are really planning on running this sucker at full power, you may want to upgrade to 2 120v 40a power feeds. Either way, I hope you're not using these at home.
Have fun,
Dave
Have fun,
Dave
Those pictures are hilarious...
People are so concerned with capacitors with low ESR but the wires that connect the capacitors with the board (and the ones with the power supply) are size 16AWG!!!
Come on, guys, let's be real, do not worry about UFO's with low ESR when you are doing such a bad job at the connections afterwards... use at least number 12AWG and make them shorter!
You are talking of 7mohms ESR, but you are adding more than 10 times that with the wires!
PS: The maximum continuous power from the wall outlet in US is 15Ax120V=1800W in regular households. Is not safe also - from the National Electric Code point of view, there should be never be exceeded 80% of the breaker rating for this kind of loads.
Of course, always you can plug the amplifier in the dryer socket That will go up to 40A at 240V...
People are so concerned with capacitors with low ESR but the wires that connect the capacitors with the board (and the ones with the power supply) are size 16AWG!!!
Come on, guys, let's be real, do not worry about UFO's with low ESR when you are doing such a bad job at the connections afterwards... use at least number 12AWG and make them shorter!
You are talking of 7mohms ESR, but you are adding more than 10 times that with the wires!
PS: The maximum continuous power from the wall outlet in US is 15Ax120V=1800W in regular households. Is not safe also - from the National Electric Code point of view, there should be never be exceeded 80% of the breaker rating for this kind of loads.
Of course, always you can plug the amplifier in the dryer socket
That will go up to 40A at 240V...- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.