Measuring audio amp output power

[Cursitor Doom]

Hi all,

I have an Audiolab 8000 amplifier I picked up at a boot sale and would
like to establish what it's capable of, since subjectively the power
output seems a bit on the low side when driving recommended speaker loads.
Anyway, ideally I'd like to use an 8 ohm 100W power resistor as a dummy
load for each channel and then measure the p-p voltage output across it
on a scope with the amp turned right up. Trouble is, I don't have such a
resistor and was wondering if there's any suitable substitute? I recall
someone somewhere using a car headlamp bulb but I doubt they come as 8
ohm units so some sort of elaborate series/parallel combo of lamps would
be necessary to get that value. Has anyone a better idea?
cheers, CD.

[Gareth Magennis]

"Cursitor Doom" wrote in message news:mp2lbl$9no$2...@dont-email.me...

An electric fire/heater element works well.



Gareth.

[N_Cook]

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)

[MJC]

In article <ZV4tx.58349$Ch1....@fx40.am4>, sound....@btconnect.com
says...

>
> An electric fire/heater element works well.

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!).

[Ralph Mowery]

"Cursitor Doom" <cu...@notformail.com> wrote in message
news:mp2lbl$9no$2...@dont-email.me...

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.

[Gareth Magennis]

"MJC" wrote in message news:MPG.301f0a766...@news.plus.net...

These are really cheap:
http://www.amazon.co.uk/2Kw-Electric-Heater-Overheat-Protection/dp/B004KDRCUO/ref=sr_1_2?ie=UTF8&qid=1437924981&sr=8-2&keywords=electric+fan+heater


The heater element in the one I bought was actually several windings, and
easy to hack and series/parallel/tap etc to get the right ohmage.

I had hoped to use the built in fan to cool the element, but it turned out
the fan motor was actually 110v, and took this from half the 240v heater
element. Doh!
So I bought another one which sits on top blowing cold air over the elements
when required.

(It should be noted that the resistance of the element is not constant, it
increases somewhat with temperature, but I don't need an accurate fixed
resistance load for my purposes)





Gareth.

[N_Cook]

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

[Jeff Liebermann]

On Sun, 26 Jul 2015 12:55:17 +0000 (UTC), Cursitor Doom
<cu...@notformail.com> wrote:

>Trouble is, I don't have such a
>resistor and was wondering if there's any suitable substitute? I recall
>someone somewhere using a car headlamp bulb but I doubt they come as 8
>ohm units so some sort of elaborate series/parallel combo of lamps would
>be necessary to get that value. Has anyone a better idea?

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.

--
Jeff Liebermann je...@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

[Gareth Magennis]

"Cursitor Doom" wrote in message news:mp2lbl$9no$2...@dont-email.me...

Here's another idea:
Get a reel of this -
http://uk.farnell.com/pro-power/ecw0-80/wire-0-8mm-copper-enamelled-125m/dp/1230984

By my calculations it will have a resistance of 4.25 ohms. (117m of the
125m length gets you 4 ohms)
Although it is only rated at 0.8A, I reckon you could extend that massively
by unravelling it all and laying it in a large container of water.
You could even have a flow of water through the container to keep the
temperature down.

It would be interesting to find out what power rating this could actually
tolerate.

Or is this a really stupid idea?! (For some reason this amuses me)



Gareth.

[Gareth Magennis]

"Gareth Magennis" wrote in message news:TP7tx.48715$m04....@fx43.am4...

Just to extend that idea a little further:

You could probably put a wood former into the chuck of an electric drill,
and quickly and easily make a long coil using this enamelled copper wire.
You could then partition the coil into several smaller ones without cutting
the wire, and suspend them vertically in a container of water.

Quick back of a fag packet calculations:
Suppose you made a long coil 10cm diameter. You would then suspend 0.5m
lengths vertically in the container, so the windings are spaced 5mm apart on
average.
This gives 100 turns at 31.4cm per turn giving 31.4 meters of wire per coil.

So you would need 4 of these series coils from the 117m of wire required to
get a 4 ohm load.


Now, how many Watts could this contraption handle?



Gareth.

[Trevor Wilson]

**Lamps are a really bad idea. Don't do it. Incandescent lamps exhibit a
HIGHLY variable resistance, depending on how much current is flowing.

Just buy four of these of these:

http://www.newark.com/vishay-dale/rh0504r000fe02/wirewound-resistor-4-ohm-50w-1/dp/41K9167

Dump them in a bucket of water, or bolt them to a chunk of aluminium and
place a small fan on the aluminium.

--
Trevor Wilson
www.rageaudio.com.au

---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus

[Phil Allison]

Cursitor Doom wrote:

>
> I have an Audiolab 8000 amplifier I picked up at a boot sale and would
> like to establish what it's capable of, since subjectively the power
> output seems a bit on the low side when driving recommended speaker loads.
> Anyway, ideally I'd like to use an 8 ohm 100W power resistor as a dummy
> load for each channel and then measure the p-p voltage output across it
> on a scope with the amp turned right up. Trouble is, I don't have such a
> resistor and was wondering if there's any suitable substitute?

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


.... Phil

[jurb...@gmail.com]

>"Just buy four of these of these:
>
>http://www.newark.com/vishay-dale/rh0504r000fe02/wirewound-resistor-4-ohm-50w-1>/dp/41K9167 "

Why do those Vishays look so much like Dales ?

[Trevor Wilson]

**Success breeds imitation. That's my take. You probably should ask
Vishay (who owns Dale). It would seem that Dale originated the aluminium
clad resistor (I may be wrong though) and others have copied the design.
Engineers don't have to worry about supply, since there are several
suppliers of essentially identical parts.

[Jeff Liebermann]

On Mon, 27 Jul 2015 06:56:53 +1000, Trevor Wilson
<tre...@SPAMBLOCKrageaudio.com.au> wrote:

>Just buy four of these of these:
>http://www.newark.com/vishay-dale/rh0504r000fe02/wirewound-resistor-4-ohm-50w-1/dp/41K9167

The RH variety is inductive. For non-inductive, you want the NH
series, which Newark and Digikey do not seem to carry. Mouser has it
at:
<http://www.mouser.com/Search/Refine.aspx?Keyword=NH0504R000FE02>
for twice the price.

>Dump them in a bucket of water, or bolt them to a chunk of aluminium and
>place a small fan on the aluminium.

An aluminum heat sink will keep my coffee warm. All the water does is
raise the humidity in the lab and create a huge mess when I tip over
the bucket. I'll take the aluminum heat sink, but without the fan,
which makes too much noise.

[Cursitor Doom]

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

> The RH variety is inductive. For non-inductive, you want the NH series,
> which Newark and Digikey do not seem to carry.

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

[Ralph Mowery]

"Cursitor Doom" <cu...@notformail.com> wrote in message

news:mp59en$khn$3...@dont-email.me...

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.

[Jeff Liebermann]

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.

[Trevor Wilson]

**No.

[Cursitor Doom]

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.

[Ralph Mowery]

"Cursitor Doom" <cu...@notformail.com> wrote in message

news:mp681u$8ul$2...@dont-email.me...

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.

[MJC]

In article <mp681u$8ul$2...@dont-email.me>, cu...@notformail.com says...

> 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).

To an extent. Seems to me that the winding temperature will rise because
the insulation between case and wire will not have infinite heat
conduction. At some power level the lifetime must start to shorten...

Mike.

[Phil Allison]

** Not in the slightest.

Liebermann is blowing it out his arse, as usual.


.... Phil

[Phil Allison]

Ralph Mowery 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.

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

> When I switched to a square wave

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

> I noticed a large spike on the leading and trailing edges, especially as I
> went higher up in frequency.

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

> From this rough test, I would say that if using sine waves you could get a
> close to true test, but if music

** Music signals stop at 20kHz.

> or other odd ball waveforms

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



.... Phil

[N_Cook]

2) only if you know the current carrying capacity of their winding wire

[Phil Allison]

N_Cook wrote:

> Cursitor Doom wrote:
> >
> > 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).
> >

> >
>

> 2) only if you know the current carrying capacity of their winding wire

** Huh ??

Wot a classic Kookism.



.... Phil

[Jeff Liebermann]

Nope. A fuse works by heating the wire high enough to where the wire
oxidizes and turns into an metallic oxide which then crumbles. If you
can extract the heat from the wire to maintain a temperature below
this point, the fuse will last indefinitely. Dumping the load
resistor into a bucket of water does this quite nicely. To do this,
the wire is in intimate contact with a refractory ceramic which then
transfers the heat to a metal case and finally to the water. No air
gaps allowed.

You can demonstrate the principle with a common coffee cup heating
coil:
<http://www.ebay.com/itm/171376872247>
If you bypass the internal thermostat and run it in open air, it will
blow up rather quickly. However, if you immerse it in water, the
water will conduct most of the heat away, thus preventing the heater
wire from fusing.

[Ralph Mowery]

"Phil Allison" <palli...@gmail.com> wrote in message
news:0c799d3c-9477-4076...@googlegroups.com...

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.

[Phil Allison]

Ralph Mowery wrote:

> "Phil Allison"

> >
> >> >
> >> > 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.
> >
> > ** And this is a sufficient test for the purpose.
> >
> >
> >> When I switched to a square wave
> >
> > ** Which has harmonics into the medium and HF bands.
> >
> >
> >> I noticed a large spike on the leading and trailing edges, especially as
> >> I
> >> went higher up in frequency.
> >
> > ** All due to the harmonics way above audio range.
> >
> >
> >> From this rough test, I would say that if using sine waves you could get
> >> a
> >> close to true test, but if music
> >
> > ** Music signals stop at 20kHz.
> >
> >
> >> or other odd ball waveforms
> >
> >
> > ** An audio amplifier reproducing a square wave will not pass harmonics
> > much above 50kHz top the load - so you are wrong again.
> >
> >
>
> Do try and keep up. This is about the output of the amplifier. It does not
> mater what the amp is going to pass.

** Huh ????

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

> 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.

** 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

[Ron D.]

Audio load: http://www.mcmelectronics.com/product/28-048 (No mention of tolerance)

Done.

Scopes have a lot of errors. A DVM may work, but pay attention to frequency response. If you stick to a sine wave you won't need a TRMS meter.

Next problem, you need to know the resistor value accurately.

I don't think you want or need want 0.001% accuracy.

[Phil Allison]

Ron D. wrote:

> Audio load: http://www.mcmelectronics.com/product/28-048
> (No mention of tolerance)

** I use the same resistors for all amplifier bench testing.

Submerged in water, they can handle 1000W each.

Using a pair and a simple switching scheme gives 4, 8 and 16 ohms.

> Scopes have a lot of errors.

** That's a a bit harsh.

Many analogue scopes manage 2% accuracy for time and amplitude & modern DSOs
have on-screen displays of the same parameters with better than 1% accuracy.

> A DVM may work, but pay attention to frequency response.

** Most can be trusted to about 1kHz on sine waves.


.... Phil

[Trevor Wilson]

On 31/07/2015 2:32 PM, Ron D. wrote:
> Audio load: http://www.mcmelectronics.com/product/28-048 (No mention
> of tolerance)
>
> Done.
>
> Scopes have a lot of errors. A DVM may work, but pay attention to
> frequency response. If you stick to a sine wave you won't need a
> TRMS meter.

**Digital 'scopes often have true RMS display capability, with excellent
accuracy.

>
> Next problem, you need to know the resistor value accurately.
>
> I don't think you want or need want 0.001% accuracy.
>

**Far and away the biggest source of inaccuracy lies with line Voltage
regulation. Or lack of it. I would certainly not panic about resistor
accuracy nor 'scope accuracy, when faced with typical line Voltage
swings. Naturally, most professionals use a variable auto transformer
(aka: Variac™) and a line Voltage monitor.

[pf...@aol.com]

NO!!!!

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

[Ralph Mowery]

<pf...@aol.com> wrote in message
news:ab86ef38-7eac-4270...@googlegroups.com...

>> 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.
>
> 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.

[MJC]

In article <ab86ef38-7eac-4270...@googlegroups.com>,
pf...@aol.com says...

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

[Ron D.]

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".

[Cursitor Doom]

Just one final point. Is a single sine wave sufficient, or will it have
to be two, harmonically-unrelated sine waves? ISTR with RF power
measurements you need to perform the slightly more complicated 'two-tone'
test; just wondering if the same applies at audio frequencies?

[Phil Hobbs]

Former astronomer?

Cheers

Phil "also a former astronomer" Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

[Phil Allison]

Cursitor Doom wrote:

> Just one final point. Is a single sine wave sufficient, or will it have
> to be two, harmonically-unrelated sine waves?

** A single sine wave is needed for audio power testing.

Amplifier power ratings are based on the *rms* value of a sine wave - hence the misleading term "rms power".

> ISTR with RF power
> measurements you need to perform the slightly more complicated 'two-tone'
> test; just wondering if the same applies at audio frequencies?

** Using two tones tests for intermodualtion, always present where there is non-linearity.


.... Phil

[Cursitor Doom]

On Sat, 01 Aug 2015 18:29:58 -0700, Phil Allison wrote:

> ** A single sine wave is needed for audio power testing.
>
> Amplifier power ratings are based on the *rms* value of a sine wave -
> hence the misleading term "rms power".

Well that would make sense with RMS being, IIRC, equivalent to the DC
heating effect in a resistive load.

[Phil Hobbs]

RMS is an amplitude, e.g. 'volts RMS'. Phil's right that the term is in
error, but there's so many lies and so much snake oil in the audio
business that this minor absurdity seems a fairly small price to pay to
clear away all the "peak intermittent music power" (PIMP) nonsense. ;)

Cheers

Phil Hobbs

[Cursitor Doom]

On Sun, 02 Aug 2015 09:04:54 -0400, Phil Hobbs wrote:

> RMS is an amplitude, e.g. 'volts RMS'.

Yes, I recall 0.707 of the peak signal level. So for example if I set an
amp to some arbitrary output and feed into an 8 ohm load and measure say
10V peak-to-peak on a scope, then I'm getting about 6.25W RMS out.

Phil's right that the term is in
> error, but there's so many lies and so much snake oil in the audio
> business that this minor absurdity seems a fairly small price to pay to
> clear away all the "peak intermittent music power" (PIMP) nonsense. ;)

Fully agree with you on that!

[Cursitor Doom]

Should have put the 6.25W RMS in inverted commas, since there's no such
thing as RMS power, of course. It's really all about *average* power.

[Phil Allison]

Cursitor Doom wrote:

>
> > RMS is an amplitude, e.g. 'volts RMS'.
>
> Yes, I recall 0.707 of the peak signal level. So for example if I set an
> amp to some arbitrary output and feed into an 8 ohm load and measure say
> 10V peak-to-peak on a scope, then I'm getting about 6.25W RMS out.

** The "peak" value is referenced to the zero line, so there are negative and positive ones. The "peak to peak" value has no polarity and is easier to read off a scope screen but to get the rms value of a sine wave one divides by 2.83 ( 2 x sq.rt 2)

So in your example, that value is 3.54Vrms and the power is 1.56 watts.


.... Phil

[whit3rd]

On Friday, July 31, 2015 at 8:42:40 AM UTC-7, MJC wrote:
> In article <ab86ef38-7eac-4270...@googlegroups.com>,
> pf...@aol.com says...
> >
> > Most metallic conductors are Zero-Resistance at cryogenic
> temperatures.
> > So, a wire-wound resistor may drop off to near-zero in liquid
> 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.

Superconducting transition is not the only misbehavior; carbon resistors have a
semimetal/semiconductor transition, and turn high-resistance when cold,
and metal-film resistors move VERY far from their nominal values when
cold. Platinum film RTDs work because they have resistance accurately
proportional to absolute temperature: resistance at 77k is about a quarter of
that at room temperature.