ATI (Amplifier Technologies) Owners Thread

[highfigh]

Yes, that's why I use words like "most", "almost", it is all relative. The simple fact is, heat = I^2*R so it the load is a resistor, most of the heat involved will be in the load, not the amplifier. If the load is highly reactive, then most of the heat will be dissipated in the output devices.

In the latter case, it is too much to try and explain it in an easily understood way as it involves more than Ohm's law lol.., so if you really want to understand the details, please do your own research as I have done in the past.

Below is a couple of places for a good start, without spending too much time on it:
Phase Angle Vs. Transistor Dissipation

And, below is interesting, but lots of people don't realize, when all they care are "power/watts":



As usual, this website covers lots more than the above on power dissipation, but the above article is just a good starting point.

Now the popular EPDR, that also seem to get misunderstood a lot, as it mislead (inadvertently) some to think high phase angle results in higher current drawn, where in fact, Stereophile use this concept to illustrate in terms of heat, the effects would be as though the load draw more current. That's a case of creating new confusion, while trying to make complex matters easier for people to understand. And specifically, in this EPDR thing, Stereophile uses the "equivalent" concept, using lowered impedance, there would result in higher current (Ohm's law) and then people can understand right away, higher current means tougher on amps, when in fact, what makes it tougher on the amp is the higher power dissipation portion in the output devices, current does not really change because if the load is say 4 ohm, if the output voltage is 40 V, then the current is still 40/4=10 A, regardless of the phase angle. It is only that with a larger phase angle, the effects of the 10 A that goes through the output device, would result in more heat in the output devices (usually power transistors) than if the phase angle is smaller. I need to stop now, as I am just writing casually/randomly, and I know I am expressing myself well, so please try the links provided and you will have better time than reading my post.

Equivalent peak dissipation resistance, EPDR | AV NIRVANA

Heavy Load: How Loudspeakers Torture Amplifiers Page 2 | Stereophile.com

I know the Power Cube test was discussed here years ago- I would like to see the results of modern equipment using that- it shows power output WRT resistive loads as well as capacitive reactance and inductive reactance.

 

[TLS Guy]

I know the Power Cube test was discussed here years ago- I would like to see the results of modern equipment using that- it shows power output WRT resistive loads as well as capacitive reactance and inductive reactance.

The problem is, that there are a lot of speakers incompetently designed.

The whole issue revolves around true power and apparent power. When a speaker presents a really dangerous load to the speaker it is because the current for the apparent power has to be supplied by the output, even though it is returned during the cycle. This is what can really heat up output devices and blow them. Not only that there are incompetent speaker designs around where one can surmise the crossover is actually in resonance and presents an extraordinary risk.

The bottom line is that when an amp blows up, the amp manufacturer is likely to get the blame when the fault lies entirely at the door of the speaker manufacturer.

The really dangerous situation arises when the impedance drops below the DC resistance of the driver, or drivers. That is the ultimate malpractice of speaker design and its around, live and well.

 

[Sly]

Ideally, if the load is like a resistor, the heat will be mostly dissipated in the load resistor. If the load is highly reactive, such as a combo of an inductor and a capcitor (speaker loads are obviously somewhere in between), then almost no heat will be dissipated in the load but would end up getting dissipated in the amp’s output devices, resulting in the amp running much hotter and may also result in stability issue.

How effective is current feedback to voltage feedback? The ati 6000 series is the only amps that i can see using this.
From what i looked up current feedback messes up the dc performance and isn't as accurate but lengthens bandwith.
There must be a reason why most amp designers moved away from cfb designs, how come ati uses it on they're top models?

 

[highfigh]

The problem is, that there are a lot of speakers incompetently designed.

The whole issue revolves around true power and apparent power. When a speaker presents a really dangerous load to the speaker it is because the current for the apparent power has to be supplied by the output, even though it is returned during the cycle. This is what can really heat up output devices and blow them. Not only that there are incompetent speaker designs around where one can surmise the crossover is actually in resonance and presents an extraordinary risk.

The bottom line is that when an amp blows up, the amp manufacturer is likely to get the blame when the fault lies entirely at the door of the speaker manufacturer.

The really dangerous situation arises when the impedance drops below the DC resistance of the driver, or drivers. That is the ultimate malpractice of speaker design and its around, live and well.

Right- lots of goofy loads cusing problems, but the Power Cube shows whether an amp is only good with resistive loads, or if it will handle any kind of difficult load. The chart shows a sharp dropoff when the tested amplifier doesn't have the cajones to handle it. This was shown by a couple of car amplifier manufacturers and obviously, they want it know when their amps can handle it, but terms for power output became common among people working with car audio. I'm sure you have been aurally assaulted while driving by someone who just had to show off their big, potent stereo, complete with dimming lights and rattling license plate trim and body panels. Some of those systems not only had massive amplifiers and woofers, they may have had additional batteries, some with 'stiffening caps'. Those amplifiers tend to have power supplies that are largely unregulated, so they need additional help when the demand is high and believe me, they can draw a lot of current. The ones on the chart that fell on their face have no headroom or ability to handle the worst loads.

Some of the power ratings, the number is just for example-

100 Sony Watts.
100W JBF (stands for Just Before Fire)
100 WLS (stands for 'When Lightning Strikes')

MAYbe Bob Carver had the right idea when he performed a 'service'that made solid state amplifiers "sound more like a tube amp" by installing a 1 Ohm resistor in series with the output- it lowered the damping slightly, but it also added that One Ohm to the load and may have saved some amplifiers.

I know someone at KEF, so I'll make sure to ask about this impedance chart.

 

[TLS Guy]

Right- lots of goofy loads cusing problems, but the Power Cube shows whether an amp is only good with resistive loads, or if it will handle any kind of difficult load. The chart shows a sharp dropoff when the tested amplifier doesn't have the cajones to handle it. This was shown by a couple of car amplifier manufacturers and obviously, they want it know when their amps can handle it, but terms for power output became common among people working with car audio. I'm sure you have been aurally assaulted while driving by someone who just had to show off their big, potent stereo, complete with dimming lights and rattling license plate trim and body panels. Some of those systems not only had massive amplifiers and woofers, they may have had additional batteries, some with 'stiffening caps'. Those amplifiers tend to have power supplies that are largely unregulated, so they need additional help when the demand is high and believe me, they can draw a lot of current. The ones on the chart that fell on their face have no headroom or ability to handle the worst loads.

Some of the power ratings, the number is just for example-

100 Sony Watts.
100W JBF (stands for Just Before Fire)
100 WLS (stands for 'When Lightning Strikes')

MAYbe Bob Carver had the right idea when he performed a 'service'that made solid state amplifiers "sound more like a tube amp" by installing a 1 Ohm resistor in series with the output- it lowered the damping slightly, but it also added that One Ohm to the load and may have saved some amplifiers.

I know someone at KEF, so I'll make sure to ask about this impedance chart.

The trouble with Bob's idea, which also applies to tube amps, is that high source resistance makes the amp's FR tend to follow the impedance curve of the speaker. That is why in the tube days, speakers were 15 or 16 ohm. That made the resistance of the secondary winding of the output transformer a much lower percentage of the speaker impedance. That is why tube aficionados make such a deal about matching speakers to amps. In that case it makes a difference. But it is still akin to placing the "tail on the donkey" while blindfolded!

 

[PENG]

Right- lots of goofy loads cusing problems, but the Power Cube shows whether an amp is only good with resistive loads, or if it will handle any kind of difficult load. The chart shows a sharp dropoff when the tested amplifier doesn't have the cajones to handle it. This was shown by a couple of car amplifier manufacturers and obviously, they want it know when their amps can handle it, but terms for power output became common among people working with car audio. I'm sure you have been aurally assaulted while driving by someone who just had to show off their big, potent stereo, complete with dimming lights and rattling license plate trim and body panels. Some of those systems not only had massive amplifiers and woofers, they may have had additional batteries, some with 'stiffening caps'. Those amplifiers tend to have power supplies that are largely unregulated, so they need additional help when the demand is high and believe me, they can draw a lot of current. The ones on the chart that fell on their face have no headroom or ability to handle the worst loads.

Some of the power ratings, the number is just for example-

100 Sony Watts.
100W JBF (stands for Just Before Fire)
100 WLS (stands for 'When Lightning Strikes')

MAYbe Bob Carver had the right idea when he performed a 'service'that made solid state amplifiers "sound more like a tube amp" by installing a 1 Ohm resistor in series with the output- it lowered the damping slightly, but it also added that One Ohm to the load and may have saved some amplifiers.

I know someone at KEF, so I'll make sure to ask about this impedance chart.

I prefer the way Amir does his, the power cube thing maybe used/useful/convenient way back when it was not common to use popular software such as Excel that has been getting more and more powerful, to the point it is no longer used mainly for spreadsheet oriented accounting kind of applications.

Take a look:

Expect to get from a popular class AB AVR amp:
Marantz Cinema 40 AVR Review | Audio Science Review (ASR) Forum

From a cheap made in China class D chip amp (the chip is by TI so presumably made in USA, though that no longer matter much nowadays imo..):

ATI (Amplifier Technologies) Owners Thread | Page 13 | Audioholics Home Theater Forums

One thing I don't like any of those tests is the narrative, that they all use the term "power", that as I have expressed my opinion (as an EE) that it was an unfortunate thing for whoever started over using that term. Even the power cube guys knew, that's why they emphasized just voltage, ideally, voltage and current should be used, without bothering the term "power". As TLSGuy mentioned, there is apparent power, and "real power". For electromagnetic loudspeakers, "real power" really not too relevant. Think about this, at 75 degree (an extreme example just to make the point, though I have seen 75 degrees for some real speaker, so it happened) phase angle, the "real power" isn't going to be much, and the output devices would have to dissipate lots of heat.

Loudspeaker are not base board electric heaters lol, voice coils to produce heat is not even a good idea, it makes sound when the speaker driver cone vibrates to move air, that's it. So, the use of the term power in watts isn't a good idea, but I understand that's just thinking as an engineer.

 

[PENG]

How effective is current feedback to voltage feedback? The ati 6000 series is the only amps that i can see using this.
From what i looked up current feedback messes up the dc performance and isn't as accurate but lengthens bandwith.
There must be a reason why most amp designers moved away from cfb designs, how come ati uses it on they're top models?

I think it depends on the designer's goals and the applications so even in a Marantz amplifier, you may find both are used, in different parts of the circuitry. It is not hard to convert from one to the other with. For in depth comparison and analysis of the pros and cons, again, depending on the specific applications and stated goals, we should perhaps ask @gene to do a couple of video on this topic, with well know designs such as John Siau, Nelson Pass, Bruno Putzeys etc.

 

[highfigh]

I prefer the way Amir does his, the power cube thing maybe used/useful/convenient way back when it was not common to use popular software such as Excel that has been getting more and more powerful, to the point it is no longer used mainly for spreadsheet oriented accounting kind of applications.

Take a look:

Expect to get from a popular class AB AVR amp:
Marantz Cinema 40 AVR Review | Audio Science Review (ASR) Forum

View attachment 74349

From a cheap made in China class D chip amp (the chip is by TI so presumably made in USA, though that no longer matter much nowadays imo..):

ATI (Amplifier Technologies) Owners Thread | Page 13 | Audioholics Home Theater Forums

View attachment 74350


View attachment 74351

One thing I don't like any of those tests is the narrative, that they all use the term "power", that as I have expressed my opinion (as an EE) that it was an unfortunate thing for whoever started over using that term. Even the power cube guys knew, that's why they emphasized just voltage, ideally, voltage and current should be used, without bothering the term "power". As TLSGuy mentioned, there is apparent power, and "real power". For electromagnetic loudspeakers, "real power" really not too relevant. Think about this, at 75 degree (an extreme example just to make the point, though I have seen 75 degrees for some real speaker, so it happened) phase angle, the "real power" isn't going to be much, and the output devices would have to dissipate lots of heat.

Loudspeaker are not base board electric heaters lol, voice coils to produce heat is not even a good idea, it makes sound when the speaker driver cone vibrates to move air, that's it. So, the use of the term power in watts isn't a good idea, but I understand that's just thinking as an engineer.

The word 'power' is what the masses know, Volt-Amps isn't. Who wants to try to re-educate people who don't know or care about the technicalities?

I was thinking about this last evening- the discussion about impedance dropping too low didn't happen often 40-50 years ago because the nominal impedance of most speakers was 8 Ohms and the mose frequent reference to 4 Ohms was seen on the back of receivers and other amplifiers, near the speaker terminals- it showed "Spkr A 4-16 Ohms, A+B 8 Ohm minimum".