Measuring audio amp output power

On 26/07/2015 13:55, Cursitor Doom wrote:


I use a bank of 10 vitreous resistors spaced apart on tag board with a  
60V sub 1W bulb across (for any DC problems etc) and a well droppered  
small low-wattage monitor speaker (for sound quality monitoring)

If you stick to 400Hz sine ,and a constant resistance, for general power  
monitoring purposes , a good quality DVM on the AC scale gives a good  
RMS reading. I only use a scope if there is quality of sound issues

says...


I guess a 1kW unit for 110V would need ca 10A and hence be ca 10 ohm;  
not far off. But a 230V element would have higher resistance and several  
in parallel might be needed for each channel for a 8 ohm load...

Mike (retiring exhausted after all that mental arithmetic!).

The headlight will not have enough wattage for that amp.  A 100 watt light  
bulb may seem like a good subistute.  Only problem is the resistance changes  
with 'brightness' of the bulb.  When dark the resistance is very low and  
gets higher as the bulb lights up.
Every time the power changes, the resistance of the bulb will change due to  
the heating of the filiment.

On Sun, 26 Jul 2015 12:55:17 +0000 (UTC), Cursitor Doom



Have you searched the web for an 8 ohm dummy load?
<https://www.google.com/search?q=audio+8+ohm+dummy+load
<http://www.ebay.com/itm/like/201057901616
Note that the resistors should be bolted to the biggest aluminum heat
sink you can find.  Mine lives at a local auditorium and weighs about
20 Kg.  It doesn't need to be that heavy to handle the heat, but heavy
does tend to discourage those that would want to walk away with my
dummy load.  My collection of phosphorescent test cables lasted about
a week.
<http://www.homedepot.com/p/Rust-Oleum-Specialty-10-oz-Glow-in-the-Dark-Spray-267026/204209388

Also, this load consists of 4ea 8 ohm resistors in series parallel.
That was suppose to be useful to produce a 2 to 32 ohm load, for
stereo or mono, none of which I've ever needed.  It also had a pair of
meters across the loads, but one of the stage gorillas stepped on one
and I've never bothered to replace it for lack of a matched pair.

Hint:  Leave room for one or two thermometers, which will help you
determine when it's safe to handle.

 Cursitor Doom wrote:




** 10 watt rated resistors will handle 100 watts if submerged in a container of water.  


....   Phil  

On Sun, 26 Jul 2015 20:18:28 -0700, Jeff Liebermann wrote:



Is that inductiveness seriously likely to be a problem at audio  
frequencies?

I just did a quick check with a 10 ohm 60 watt wire wound resistor and a  
couple of other low value resistors.
Used a 100 MHz scope, function generator and Fluke 87 meter.

From low audio up to 20,000 Hz and a sine wave with the 10 ohm WW there did  
not appear to be any noticable difference.  When I switched to a square wave  
I noticed a large spike on the leading and trailing edges, especially as I  
went higher up in frequency.  I did not see the spike with the resistor  
removed from the test leads.

From this rough test, I would say that if using sine waves you could get a  
close to true test, but if music or other odd ball waveforms then you would  
want the noninductive types.

Ralph Mowery wrote:




 ** And this is a sufficient test for the purpose.




** Which has harmonics into the medium and HF bands.




** All due to the harmonics way above audio range.




** Music signals stop at 20kHz.





** An audio amplifier reproducing a square wave will not pass harmonics much above 50kHz top the load -  so you are wrong again.

  

....  Phil  

Do try and keep up.  This is about the output of the amplifier.  It does not  
mater what the amp is going to pass.

From a quick test with a couple of wire wound resistors an audio frequency  
square wave was producing a spike on the leading and trailing edges of the  
ww resistor.  It did not do that with a carbon resistor.

Ralph Mowery wrote:



** Huh ????

Since amplifiers cannot pass the harmonic frequencies, they will NOT appear at the output.




 ** An "audio frequency square wave"  must have its harmonics filtered out above the audio band.

However, what you have is a *WIDE BAND* square wave with its fundamental at an audio frequency.

I understand your test and a similar one to compare low value WW resistors for such inductance. Normally you see overshoot and ringing at frequencies in the low MHz range on the scope - using a 100kHz square wave.



...  Phil  





...  Phil  


 It did not do that with a carbon resistor.

On Mon, 27 Jul 2015 12:50:31 +0000 (UTC), Cursitor Doom




Disclaimer:  I am not an audiophile or expert on audio equipment
beyond a few occasional repairs and some long past work at a recording
studio and several radio stations.

Well, that depends on what you're trying to accomplish with the dummy
load.  If it's performance tests for a data sheet with calibrated
equipment, then yes, every divergence from exactly 8+j0 is important.
However, if you're using the load to simulate a real loudspeaker to
test for ringing, oscillations, crossover distortion, resonant peaks,
and such, then there's no way a purely resistive load is even close to
a real world loudspeaker impedance:
<https://www.google.com/search?tbm=isch&q=loudspeaker+impedance

Let's do some measuring and math.  Digging through my junk box, I find
a dummy load that I think was used to test power supplies:
<http://www.11junk.com/jeffl/pics/drivel/slides/dummy-load-inductance.html
That's 2ea 2.5 ohm and 2ea 5.9 ohm inductive RH type resistors in
series for a total of 16.8 ohms.  Measured inductance of 16 uHy at
1KHz yields:
  Xl = 2*Pi*f*L = 2 * 3.14 * 1000 * 16*10^-6 = 0.1 ohms
So, this load looks like:
  16.8+j0.1
Good enough.
However, if you were making measurements up to 100 KHz, where the load
would look like:
  16.8+j10
methinks a non-resistive load might be useful.

Considering that your original question was about testing a used
amplifier to see "what it was capable of", y'er right.  You can do
that nicely without using overpriced non-inductive terminators.  When
you actually build your load, you might want to do the above
measurement and calculations.

Thank you, gentlemen.

So we're all agreed:  

1. there's no problem using wire-wounds for a load provided the power  
output measurement is carried out with sine waves from a suitable  
external generator.

2. Even low rated power resistors can be used for the load provided their  
case temperatures are kept low by some effective means of external  
cooling (something more than mere heat sinks and fans, IOW).

Many thanks.

That is the way I see it.

 You can probably go at least 5 and maybe 10 times the power on the load  
resistor if you put it in some water or oil.  Many times that if you have  
liquid nitrogen.

On Monday, July 27, 2015 at 6:09:33 PM UTC-4, Ralph Mowery wrote:


NO!!!!

Most metallic conductors are Zero-Resistance at cryogenic temperatures. So, a wire-wound resistor may drop off to near-zero in liquid nitrogen.  

The liquid nitrogen was mainly a joke to the extreme of cooling.  However if  
enough power is being used, the resistor its self  would be heated to a more  
normal temperture  and would be way above the temperature of the nitrogen.  
This woudl bring the resistance back up.

snipped-for-privacy@aol.com says...

temperatures.
nitrogen.  

If only superconductors for liquid N2 temperatures (77K, -196C) were so  
common! Even at liquid He (4.22K, -270C) only a minority of metals make  
the transition.

My Kaye and Laby says: "At low temperatures the effects of impurities  
etc become increasingly important and these largely determine the value  
of the residual resistance to which many metals decrease at low  
temperature."

Mike.

Phil: > Scopes have a lot of errors.

 ** That's a a bit harsh.  

Analog scope days and having error sources and effects drilled into my head.
I just really wanted to say, know what your error sources are.  At one point in my life numbers like 2E18 and 7E18 were considered "essentially the same".

On 07/31/2015 11:54 PM, Ron D. wrote:


Former astronomer?

Cheers

Phil "also a former astronomer" Hobbs