A Test. How much Voltage (power) do your speakers need?

[AndrewT]

1.2Vac with the -12dBfs test frequency?
That means you will never overload a 1½W amplifier from your digital sources.

 

[mark100]

The 107dB is for a 3-way active, designed for 100hz up.

If I add the subs into the spec, overall SPL probably goes up because I usually run subs about 6dB hotter than tops....well, depends on what weighting/SPL you use i guess...(I like flat)...

Methodology for coming up with 107dB on the 3-way is in post #606 in this thread...
https://soundforums.net/forum/low-earth-orbit/diy-audio/12390-60-degree-diy-mid-hi/page41

 

[Pano]

Thanks Mark.

1.2Vac with the -12dBfs test frequency?
That means you will never overload a 1½W amplifier from your digital sources.

Or that he's running about 0.6 volts on a typical dynamic CD with peaks hitting about 5 volts. That's not a lot, for sure.

 

[brig001]

Good question. I think mostly small rooms and decently efficient speakers. Like 92dB/W or more.

My old Altec A5 system generally needed voltage attenuation from line level, not amplification. Current gain was all it needed. That's with sensitivity somewhere in the mid 90 dB/W

My speakers are also mid 90dB/W, so I got rid of my power amps. I re-designed my active crossover to give 5A output current and drive the speakers from that. I bet lots could benefit from a similar approach.

Brian

 

[mark100]

Thanks Mark.
That 107dB @ 2.83 is for the system overall, of some specific part of it, like the woofers?

Hi again Pano, I realize I made a mistake. I have been mapping all my measurements against your test tone measurement, and forgot your tone is -12dB rms.

On my 3-way, the appropriate driver for using your test is the mid section.
I rms limit it at 60v, peak 120v.

So I should have checked 10-20v on your survey.

Now, if we move to subs, I'm back at over 20v...😉

 

[Hearinspace]

OK, I measured again. Measurement this time is 1.6Vrms.

I could go for a little higher than 2 o'clock (Zero's at 6) on the dial but the only time I've ever taken it higher is on low level YouTube videos so not sure why I'd do it here.

The impedance is , well, I don't know what to call it. What would you say? (Red stands for port speaker, green is starboard. ; )

 

[Pano]

I re-designed my active crossover to give 5A output current and drive the speakers from that.

That may be all that most people need. Do you have a description of it somewhere?

 

[Pano]

...forgot your tone is -12dB rms.

Just to be clear, the test tones are RMS 12dB below FULL SCALE. They are 9dB below the loudest permitted sine wave.

You will sometimes find test CDs with sine wave referenced 3dB up from where mine are. That's because they take the 0dB as being a sine wave with its peaks at full scale. A sine like that has an RMS value of -3dB Full Scale. I would call that a -3dBFS tone.

This has caused a good bit of confusion in the thread. The reference levels are 3dB apart as they take 0dB from different places. Both are valid.

 

[AndrewT]

Thanks Mark.


Or that he's running about 0.6 volts on a typical dynamic CD with peaks hitting about 5 volts. That's not a lot, for sure.

That's not the way you explained how your test works.

 

[mark100]

Just to be clear, the test tones are RMS 12dB below FULL SCALE. They are 9dB below the loudest permitted sine wave.

You will sometimes find test CDs with sine wave referenced 3dB up from where mine are. That's because they take the 0dB as being a sine wave with its peaks at full scale. A sine like that has an RMS value of -3dB Full Scale. I would call that a -3dBFS tone.

This has caused a good bit of confusion in the thread. The reference levels are 3dB apart as they take 0dB from different places. Both are valid.

Thanks Pano, I'm used to sine peak being 0dBFS...with sine RMS power being -3dBFS.
The AES method has never made sense to me, but then neither does our continuing use of decibels, or our english measurements 😀
IOW, MY problems !

 

[Pano]

That's not the way you explained how your test works.

Did I get it wrong? Where do you see the difference?

 

[brig001]

That may be all that most people need. Do you have a description of it somewhere?

Hello Pano, here it is http://www.diyaudio.com/forums/solid-state/237027-low-power-amp-adding-current-buffer-preamp.html should finish it off with some photos etc., but don't want to start all the arguments again...

Brian

 

[mcandmar]

I measured a surprisingly low 362mv RMS at both frequencys with my KEF 104ab Speakers. I guess my F4 has plenty of headroom left 🙂

 

[AndrewT]

I measured a surprisingly low 362mv RMS at both frequencys with my KEF 104ab Speakers. I guess my F4 has plenty of headroom left 🙂

That 0.362mVac represents a level of ~16mW into 8ohms for that -12dBfs signal.
16mW is roughly -18dBW, add on the 9dB to get back to the absolute maximum sinewave output required you arrive at -9dBW, or ~0.1W Power Amplifier.
or use the formula given by Pano:
Maximum power for the amplifier = 0.362² = 0.13W into 8ohms.

If the KEF 104ab is rated @ 4ohms over a substantial part of the frequency range then you will need a ¼W amplifier.

Something doesn't seem right to me.

 

[Pano]

That does seem surprisingly low. It's somewhere around the levels I was using on my big Altec horns.
Are you sure of your voltmeter? We can check it on your CD or DAC output, for example.

 

[kgrlee]

Are you sure of your voltmeter?

Loadsa voltmeters ONLY give 'accurate' results up to about 400Hz.

In da old days, the first passive AC voltmeters which gave sensible results from 20 - 20k Hz were the cheapo Japanese ones using Ge diodes.

Towards the end of the 20th century, I was surprised that loadsa 'modern' DVMs had the same restrictions.

Look carefully at your voltmeter spec. If it doesn't specify a Frequency Range, it is almost certainly only good up to 400 Hz ... or maybe even only 100Hz.

 

[Pano]

Yes, well. If you read the beginning of this thread, that is taken into account.

 

[Pano]

We use a sine wave to test with (after setting the level) because it's easy for a voltmeter to measure accurately, unlike music. All the meters I've tested have been accurate at 120Hz and within a small fraction at 220Hz.

 

[mcandmar]

Just double checked for sanity, at the preamp output i am seeing 384mv @120hz and 240hz. 0db test files of 50, 100, and 200 hz measure a hair over 1vac. If i set my player output to -11.9db those files give me a reading of 275mv. So yeah, that is all it needs, still really surprises me.

I also used three meters and got the same results +/- a few mv. Fluke 175, 17B+, and a Uni-T UT61E.

 

[Pano]

Good test, thanks. That all lines up for an RMS output of 1V full tilt from your preamp.
I'm as surprised as you at how little voltage you need at the speakers.