Measuring Class D output

[Gaz]

Are there any special considerations measuring a Class D amp’s output power? I’ve got a 700W (into 4 Ohms) Class D bass amp and I wanna check its output. I have a scope and 1000W 4 Ohm dummy load, to do so. When I measure clean power at clipping with a 1khz signal Like Inwoild with a tube amp I get 26VAC RMS. That’s just 169W. Am I missing something or do I need any special equipment to measure output power accurately with Class D?

[Enzo]

The load has no idea what kind of amplifier made the power.

Make and model of amp? Does it say 700wRMS into 4 ohms? or does it say something like peak power or music power?

700 watts into 4 ohms would need about 52v if I figure correctly.

[J M Fahey]

Agree and add: HOW are you measuring those 26V RMS?

* With your scope?
Where do you clip hot and ground scope probe?

* With a hand held multimeter?
If so:
where do you put meter probes?

a) from speaker out Hot to ground
b) across load resistor, meaning black to one end, red to other end

I think whatīs hapening but donīt want to lead you, so please answer these questions first.

Of course now Enzo knows what I suspect, certainly others too, but please letīs wait for your answers

[xtian]

This is cruel. I only learned the answer recently, as I've spent the last decade learning tube amps, not solid state.

Cruel, I say.

[eschertron]

So this is the classic: "when is ground not ground?" Zen
or maybe the classic: "what is the sound of one ground clapping?" Tao

[TimmyP1955]

Float the scope ground from mains ground. The amp may be a pair of smaller amp channels in bridge mode.

[g1]

And Enzo's 52V is precisely double the 26V that was measured.

[J M Fahey]

I still would love to hear gazīs answer from his own lips

[Gaz]

You guys make my dumb questions more fun Ok I got 52V now measuring from across the dummy load instead of measuring from ground. Can someone explain?

[drewl]

The output isn't grounded.

[pdf64]

You guys make my dumb questions more fun Ok I got 52V now measuring from across the dummy load instead of measuring from ground. Can someone explain?

The amp may be 2 power amps operating in bridge mode, ie both conductors to the load are live, in opposing polarity, eg http://www.circuit-finder.com/catego...wer-amplifiers

[J M Fahey]

Yes, I wanted you to find it

Typical modern Class D amp architecture is bridged amps, donīt know why because itīs not really needed, I suspect snowflake "3.3V and 5V world" designers are *terrified* at dangerous high voltages ... such as 40V rails .... deadly 80V end to end and higher so they try to minimize that.

Going also to the extreme of using 2 ohm and even 1 ohm speakers (just check JBL EON and similar) so as to get high power with puny voltage.

So your 700W poweramp is *definitely* a bridged output one, so each speaker terminal, both "+" and "-" (canīt call it "ground" , can we?) swings 26V *relative to ground* and what you measured.
Since they are out of phase, you get full 52 V end to end, as predicted by Enzo.

Many go to an extreme trying to put scope across that floating output signal, from chopping Security Ground to try and get a differential front end (think old Tektronik and such) scope.

Me? ... I just conventionally scope one end to ground, and use my second scope channel to monitor the other.
Or even scope just *one* ... and then switch the clip and monitor the other, with scope ground clip staying where it belongs: ground.

[Dave H]

Many go to an extreme trying to put scope across that floating output signal, from chopping Security Ground to try and get a differential front end (think old Tektronik and such) scope.

Me? ... I just conventionally scope one end to ground, and use my second scope channel to monitor the other.
Or even scope just *one* ... and then switch the clip and monitor the other, with scope ground clip staying where it belongs: ground.

We used to have Tektronix scopes with differential plug-ins at work. To measure bridge outputs at home I use a two channel scope in A+B mode with one channel inverted, Channel A and B inputs to the amp outputs and both probe ground clips to amp ground.

[xtian]

Modern DSO scopes have differential mode.

[glebert]

Yes, I wanted you to find it

Typical modern Class D amp architecture is bridged amps, donīt know why because itīs not really needed, I suspect snowflake "3.3V and 5V world" designers are *terrified* at dangerous high voltages ... such as 40V rails .... deadly 80V end to end and higher so they try to minimize that.

I have wondered why they would do the bridged outputs, adding the cost of the extra FETs and the transformer/choke/whatever to couple the two sides, but I think it is either for cost (which drives almost every decision in design) or the thermal solution necessary. Even with 90% efficient class D operation you would still have 70W to dissipate from this amp. Using more devices spreads the thermal energy over a larger area, which will keep your Tj down and may allow you to use a simpler heatsink and fan approach. I have an 800 watt Ashdown amp that is smaller than the heatsink alone from my old 150 watt Peavey head. Then again the Peavey still works after 25 years and the Ashdown is on the fritz