10% THD?

And the only 3% distortion you'll see from a modern solid state amplifier is CLIPPING distortion.

Graham

For most modern amp designs, the curve of distortion versus power output at 1 kHz looks like a lopsided valley: At very low outputs, there is moderate crossover distortion, which essentially is a fixed-size discontinuity near zero so it becomes a smaller percentage of the total at output power rises. So distortion decreases linearly until some mid-power region (1-10 watts, say). Then as you continue to increase power, you soon hit a point where the distortion starts to rise rapidly with power, as you get near to (and into) clipping. (The distortion can rise before the signal actually hits the hard clip point, beacuse near clipping the amp's negative feedback has less elbow room to work its magic.)

At any rate, the manufacturer has to make a judgement call about what power level to claim. He can pick a high level if he thinks the customer will be impressed by watts and not know or care about distortion. Or he can pick a lower power and correspondingly lower distortion to appeal more to purists. (The ability of the amp to withstand the power without damage is not typically an issue here. Amps dissipate the most power at intermediate levels.)

As others have pointed out, music often has transient peaks that can be clipped a fair amount before they become objectionable. When the signal level is set so that the tallest peaks are just hitting the 100W point, the average power is probably more like

Best regards,

Bob Masta DAQARTA v3.50 Data AcQuisition And Real-Time Analysis

"Bob Masta"

The characteristic rise in published THD curves at power levels below about

NOT any mysterious x-over region effects !!!!!!!!!!

The general definition of THD *** includes *** hum and noise - which tends to be a fixed level at all power outputs. So the PERCENTAGE increases the signal level reduces !!!

Obvious - right ?

The typical power output difference between the 0.1% THD level and 0.01% THD level is trivial for the majority of hi-fi quality SS amps. Maybe 102 watts as compared to 100 watts.

Only by doing THD measurements on * real amplifiers * can one have any insight into the facts of the matter .

Which I do all the time.

While Bob does not.

...... Phil

spec.

its

these

sheet

nobody

THD...

Well, maybe I should change it to nobody in their right mind would BUY such a lousy amplifier. You certainly can do better than 10%THD but at what price, I don't know. This thing might fit your needs if sound quality is not a big issue.

Four ohms is going to make things worse because if the amp struggles with 8 ohms, 4 will require more current loading the amps even more. However, a sub-woofer may not care, the speaker will not respond to the higher harmonics anyway.

In my mind, a bigger issue is intermodulation distortion. If it produces high harmonic distortion, you can bet it has high IM as well. This is also a nonlinear distortion and is curiously not even spec'd on the data sheet. What does that tell you?

I think you're missing the point.

I have little doubt that this amplifier is capable of 0.1% THD at say 60W.

The 10% THD figure has simply been used in order to produce an inflated power rating. This is considered normal practice with car audio.

Graham

such

not

power

Yep, apparently that must be the case. However, this isn't car audio. Here's the device:

Also, can this POS deliver 100 Watts of its distorted power on all five of it's channels simultaneously? I'll bet not. The rating must be one channel at a time run for just the few seconds it takes to get a measurement and no longer.

IIRC, the mannual says the power supply consumes 155W, 200VA.

I'm guessing the power supply won't appreciate being driven on all 5 channels with subwoofers.

Don't you love it when a device is spec'ed at 500W output, with less than 200W input. I explained it to my wife this way: how can a business collect only $200 a day, but spend $500 a day, and stay afloat?

But this isn't unique to JVC. Sherwood's unit does something similar. (Power cable says 200W, and it's rated at 500W or more.)

Which is why, I guess, if I want 500W per channel, I'd better build the darn thing myself. ;-)

Michael

I've learned recently that speaker manufacturers are similar. The good ones, when specifiying frequency response, will tell you that it's, e.g.,

This is actually VERY good compared to most car stereo boom things.

Right. Thanks for reminding me about the hum and noise. It's the same basic concept I was taliking about: A constant that is a bigger percentage of a smaller total.

(I'm not sure why you are calling crossover distortion "mysterious"... I suspect it is the main source of the residual distortion at the bottom of the valley around 1 watt or so.)

Modern methods using synchronous averaging and FFTs (with a sound card and the right software... hint, hint!) can measure the distortion alone, by removing noise and cancelling hum. But I admit it is purely academic at these low levels, and surely not worth the effort for most purposes.

I sure won't argue. It depends on how steep the curve is for the amp. (My suspicion is that this amp may not be in the "hi-fi quality" department.) At any rate, they must have had *some* reason to pick the 10% THD point to state their power rating. "100 watts" certainly would appeal to Marketing better than, say, "97 watts".

Of course, it *could* be a typo. I can just picture Moe telling Larry "Tenth percent THD" and Larry telling Curly "Ten percent THD".

Agreed!

No fair, you peeked! Yes, sad to say that software development has really been keeping me away from the test bench for far too long.

But I still have my monster dummy load bank (4 x 8 ohms, 200 watts each) sitting there gathering dust. Some day, Real Soon Now....

Best regards,

Bob Masta DAQARTA v3.50 Data AcQuisition And Real-Time Analysis

Yo Phil.... theoretical question.... I have two hi quality sine generators and a THD meter. If I set generator 1 at 1V 1KHz and generator 2 at .03V 3khz, thats the only 'distortion component', 3rd harmonic thats 3% of the fundamental. I guess I could square it and sqrt it, but since thats the only harmonic, would it show 3% THD on the meter? Thanks.

I just realized that this statement probably marks me as an out-of-touch dinosaur, since it is based on experiences with bipolar output devices... now driven to near-extinction by MOSFETs as the preferred output stage.

"Back in the day", the bias of the output stage was a critical setting. Too little and the amp ran as class B, with heavy crossover distortion as the positive and negative devices handed off their share of the load near 0 volts. So the bias was adjusted for class AB, where each half carried over just a bit across zero, overlapping with the other half in a linear region around zero. That way, there was always somebody there to carry the load. (Still true today, but read on...)

The problem with bipolars, however, is/was thermal runaway: As they get hotter, they conduct harder and that causes them to heat up still more. So you had to be very careful, and have good temperature compensation for the bias circuit, or the output stage would go up in smoke. (That's not *absolutely* true: With enough negative feedback you can get away with pretty meager bias, but there are other problems with that approach.) I tended to be cautious, and not too concerned about what I couldn't hear anyway.

But MOSFETs (which I've never used in an output stage) have the opposite behavior from bipolars: As they heat up, their resistance goes up and they conduct less, causing them to cool down. So I would not be at all surprised to find that bias is nowadays set much more agressively, so that crossover distortion may indeed be a thing of the past. Mea culpa!

Best regards,

Bob Masta DAQARTA v3.50 Data AcQuisition And Real-Time Analysis

"Bob Masta" "Phil Allison"

Allow the measured s/n ratio to be 90dB unweighted, then the s/n at 10 mW is only 50 dB and -50 dB equates to 0.3 %.

The distortion residual at 10 mW output ( as seen on a scope after removing the fundamental) is usually much lower than 0.3 %.

QED.

....... Phil

"BobG"

Once the 1kHz wave is removed, the 3 kHz is left alone.

So it will give a 3% reading, like you say.

But so would some 60 Hz hum or a supersonic signal.

...... Phil

Got it! Thanks, Phil. For decades I have apparently been laboring under a misconception that there was always some residual crossover distortion, but now that you point it out it seems totally unjustified... and I'm more than glad to be rid of it!

Best regards,

Bob Masta DAQARTA v3.50 Data AcQuisition And Real-Time Analysis

you used to see this kind of spec all the time for junk-grade audio. but jvc? sad.

there are 6 "mosquito" ringtones from 15 khz up to 20khz. i've tested them all (vzw tones, motorola krazr) on a 5 year old and an 11 year old that didn't know they were being tested, and neither even looked up.

uch

not

wer

exactly. this may well be a fine amp at 60 watts or 70 or whatever. some middle management moron undoubtedly decided "this amp needs to be

ones,

off topic: just ran into this:

"A favorite demonstration of mine was to plug the [McIntosh] ML-4 woofer section into the 120-volt wall outlet. This really captured the attention of visitors who questioned the power handling of the system, and it was also very loud. "

This guy's whole site captured more for an afternoon.