Let's stop calling Class D amplifiers "efficient" when they are not.

[wwenze]

I don't make calcs based on AC current ratings especially with audio equipment because the rectifier into capacitor arrangement has poor power factors especially at low loads. Example of an ATX PSU with no PFC:

The maximum theoretical PF is around 0.6-something and it cannot go higher because that would require downstream voltage to fluctuate too much. Since a 100% resistive load has no capacitance. Conversely, the more capacitance you have for the load current, the lower the load and more stable the supply voltage, the lower the PF.

Another win for SMPS in class D modules since they often (tho not always) come with APFC especially in the western markets, with the added side benefit of allowing 100-240V operation.

 

[DonH56]

I don't make calcs based on AC current ratings especially with audio equipment because the rectifier into capacitor arrangement has poor power factors especially at low loads. Example of an ATX PSU with no PFC:

View attachment 379464

The maximum theoretical PF is around 0.6-something and it cannot go higher because that would require downstream voltage to fluctuate too much. Since a 100% resistive load has no capacitance. Conversely, the more capacitance you have for the load current, the lower the load and more stable the supply voltage, the lower the PF.

Another win for SMPS in class D modules since they often (tho not always) come with APFC especially in the western markets, with the added side benefit of allowing 100-240V operation.

I used what I could find. It was QSC's data for input current (rms) for various output conditions, so measured not rated. It would be great if you could provide more accurate data; I'll delete my post.

 

[wwenze]

I used what I could find. It was QSC's data for input current (rms) for various output conditions, so measured not rated. It would be great if you could provide more accurate data; I'll delete my post.

We used the same piece of data on QSC's website when making our respective statements. VA is not Watt, that's all.

 

[DonH56]

We used the same piece of data on QSC's website when making our respective statements. VA is not Watt, that's all.

Unless PF = 1 but I doubt many folk here care to dig that deep. Since QSC did not provide either power factor or instantaneous current and voltage for me to integrate I'll defer to whatever you think the efficiency should be.

 

[SIY]

Unless PF = 1 but I doubt many folk here care to dig that deep. Since QSC did not provide either power factor or instantaneous current and voltage for me to integrate I'll defer to whatever you think the efficiency should be.

IIRC, QSC uses PFC.

(A record for acronym ratio)

 

[wwenze]

Well we can backcalculate from that:

I assumed it did not have PFC due to the big 50Hz transformer, but it has above 0.7 at some points and varies until 0.85 at full load so this has the behavior of passive PFC.

Also strangely enough I'm not seeing the class G-ness of this amp because efficiency at 1/8 and 1/3 is pretty ahem compared to full, where's the region of high efficiency that people keep pointing on the graphs.
Edit: Oh

Edit2: Ok I did the same for GX5 because I just had to see the G-ness

 

[restorer-john]

For comparison, Buckeye Amps stereo 350 W/ch (8 ohms) amp (Hypex NC502MP based) uses 0.5 W at idle and is specified as 92% efficient at full power.

0.5W is standby (soft off), not idle.

 

[DonH56]

Well we can backcalculate from that:

View attachment 379509

I assumed it did not have PFC due to the big 50Hz transformer, but it has above 0.7 at some points and varies until 0.85 at full load so this has the behavior of passive PFC.

Also strangely enough I'm not seeing the class G-ness of this amp because efficiency at 1/8 and 1/3 is pretty ahem compared to full, where's the region of high efficiency that people keep pointing on the graphs.
Edit: Oh
View attachment 379514

Edit2: Ok I did the same for GX5 because I just had to see the G-ness
View attachment 379517

I missed QSC had PFC, not something I looked for. My original idea was a quick calculation, should have known better than try that here.

I earlier mentioned that QSC does not make it clear class of operation; you have to dig into various documents for each amplifier series, and not all amps in the series have the same class. Lower-power amps are often AB, higher-power H or D.

0.5W is standby (soft off), not idle.

Figures.

I think I've screwed up enough for one thread, out.

 

[pma]

Hypex SMPS1200A400 eats 9W at idle. @restorer-john is right that idle and stand-by is not the same.

 

[mhardy6647]

IIRC, QSC uses PFC.

(A record for acronym ratio)

 

[Emlin]

View attachment 379618

Not even acronyms. Initialisms.

 

[mhardy6647]

Not even acronyms. Initialisms.

touche.

 

[DonH56]

PF = power factor
PFC = power factor correction

QSC uses PFC to correct a low PF.

 

[rdenney]

I don’t much care about efficiency as a percentage at low power outputs. If I’m putting a watt into the speakers, I don’t care idf the amp is consuming 10 watts or 1.1. And efficiency will always be zero at idle.

I do note idle power consumption. B&K claimed my amp idles at 60 watts. That’s a significant amount of heat for no output—enough to need significant heatsinks. But not heatsinks larger than at rated power.

The efficiency of Class D matters for heatsink and cooling design, it seems to me, which isn’t much of a thing at idle. Cooling sufficient for rated power will be abundant for idling, it seems to me. My Buckeye, if it is indeed producing 300 wpc, is creating about 50 watts of heat from the devices (plus power supply). It’s creating less than that at idle.

Percentages are statistics that can be really misleading.

Rick “just tell me how many watts are being radiated as heat” Denney

 

[Doodski]

I don’t much care about efficiency as a percentage at low power outputs. If I’m putting a watt into the speakers, I don’t care idf the amp is consuming 10 watts or 1.1. And efficiency will always be zero at idle.

I do note idle power consumption. B&K claimed my amp idles at 60 watts. That’s a significant amount of heat for no output—enough to need significant heatsinks. But not heatsinks larger than at rated power.

The efficiency of Class D matters for heatsink and cooling design, it seems to me, which isn’t much of a thing at idle. Cooling sufficient for rated power will be abundant for idling, it seems to me. My Buckeye, if it is indeed producing 300 wpc, is creating about 50 watts of heat from the devices (plus power supply). It’s creating less than that at idle.

Percentages are statistics that can be really misleading.

Rick “just tell me how many watts are being radiated as heat” Denney

I keep requesting mains wattage draw c/w current metering too so we can see what is occurring with this gear. I had it at one electronics lab I owned and it was so nice to operate when it is all prewired with meters and one just plugs in and also have a variac to it if required but for @amirm's purposes a variac is not required.

 

[Petrushka]

I keep requesting mains wattage draw c/w current metering too so we can see what is occurring with this gear. I had it at one electronics lab I owned and it was so nice to operate when it is all prewired with meters and one just plugs in and also have a variac to it if required but for @amirm's purposes a variac is not required.

Wattmeters are $10 on Amazon.

 

[Doodski]

Wattmeters are $10 on Amazon.

There is a little more to it than a Watt meter. It needs voltage metering and current or watts. I prefer current and then we can make our own Watts calc.

 

[Petrushka]

There is a little more to it than a Watt meter. It needs voltage metering and current or watts. I prefer current and then we can make our own Watts calc.

The ready made ones are recording. It would be useful to know the energy consumption over a week or month of actual use.

 

[Doodski]

The ready made ones are recording. It would be useful to know the energy consumption over a week or month of actual use.

The week or month is a difficult metering because the unit needs to be operated and used for that period otherwise it sits there quiescent. To get Amirm to meter the voltage and current will be a hurdle... LoL.

 

[Petrushka]

The week or month is a difficult metering because the unit needs to be operated and used for that period otherwise it sits there quiescent. To get Amirm to meter the voltage and current will be a hurdle... LoL.

But that’s the point. Annual energy consumption.