Class D


Been thinking of trying a D amp to reduce clutter. Most that I see are not rated at 2 ohms.  My PSB Stratus gold's will drop to 3 ohms or lower at some frequencies. So my question is will these types of amps handle this impedance ?
Thanks in advance. Chris
zappas

Showing 7 responses by jaytor

The primary advantages of GaNFETs (compared to mosfets) are low on resistance and low input capacitance.  The low on resistance allows for higher efficiency, particularly into lower impedance loads. In car audio, available power is pretty limited so it stands to reason that these devices would be making inroads here. 
As @audio2design points out, EPDR does not apply to class D amplifiers. However, the efficiency at which a class D amplifier operates is reduced as load impedance goes down since the output transistors will be conducting more current during their switching and on-phases for the same power into the load. 

Class D amps are theoretically 100% efficient if the output transistors had zero on-resistance, infinite off resistance, and switched infinitely fast. But, of course, this isn't the case. Typical class D amps are 90% or more efficient into 8 ohms (at max output power), but efficiency drops by approximately 40% into 4 ohms, and 40% again at 2 ohms.

Under normal circumstances, this isn't a huge problem because an audio amplifier in a normal home music environment is typically operating at a small fraction of it's peak power. But under heavy demand (or test conditions), the lower operating efficiency into lower impedance loads will eventually cause the amplifier to get too hot and shut down. The good news is that modern class D amps all have circuitry to protect the amplifier under these conditions. 

As @audio2design also points out, class D amps will also incorporate maximum current limiters which will also limit maximum power into low impedance loads. Since this circuitry operates almost instantaneously (checking current on every switching cycle), this is more likely to limit maximum power into low impedance loads since music peaks can often be many times higher than average power requirements. But these current limits are generally quite high. For example, the Purifi module, which is rated at just over 200w into 8ohms, has a 25A current limiter, but this is what ultimately limits the peak power into loads below about 2 ohms.
@audio2design Yes, that is what I meant. Sorry if my language wasn't clear. 
@georgehifi - How is that any different from digital audio, where DACs get by (very well I might add) with switching frequencies well below a Mhz? And in DACs, there isn’t a practical way to do global feedback since the input is digital, so it’s much more difficult to eliminate filter effects.
@georgehifi - I usually try hard not to get personal on these forums, but you’re continued spewing of misinformation is getting on my nerves. A lot of people on this forum count on knowledgeable people to explain how things work. But you just spout ridiculous drivel. And you keep doing it despite a number of experts like atmosphere and audio2design explaining in straightforward terms why your wrong. But instead of trying to understand what they are saying, you keep regurgitating the same old arguments which don’t even apply to what is being talked about.

I think you better quit while you’re behind. Who knows how much more ridiculous your counter-arguments will become.
Class D is a pulse-width modulation technology, meaning that the audio signal amplitude information is all contained in the width of the pulses (which is analog, by the way).

The switching frequency is therefore exactly the same issue as it is with PCM digital audio. The frequency only has to be high enough such that it can be filtered out with minimal amplitude and phase errors in the audio band.

But since the information itself is actually analog (e.g. the width of the pulses) as is the input signal, global feedback can easily be applied to correct for amplitude and phase errors, so achieving a clean response is even easier than it is with digital audio.

So the notion that you need Mhz sample rates to achieve high-quality reproduction is pretty ridiculous, let alone Ghz rates.
It's often the case that extra filters are added to allow the device to pass FCC requirements and/or to improve stability into arbitrary loads.

If you are making modifications for your own use, you can get away with it, but if you are trying to make a commercial product, you will need to make sure the product continues to meet FCC requirements for emissions before offering the product for sale.