Both are transparent as far as I can tell (and measure). Having the extra power of the '686 is nice - especially with my rather inefficient (87 dB @ 1W,1m) test speakers. That was a surprise, actually. I hadn't expected the additional headroom to matter.
I'm not up on the latest audiophile alphabet soup, so I can't tell you if one has more PRAT, PLOP, or KA-PLUNK than the other.
Very cool. Thank you for the synopsis. I wasn't aware of the FTC (2000) requirement, though I did wonder how the many home theatre receivers managed to meet the "FTC requirement". I figured they did something similar to the old 'specmanship' of using DIN45500 vs DIN45462 (if I recall correctly, the DIN about Car-Fi).
I doubt the various eBay vendors care about the FTC. As far as I can tell they just copy the numbers from the spec sheet of the chips they use and call it good. Never mind the fact that you can easily degrade the performance of a chipamp 10-100x with a poor PCB layout.
Tom
Last edited:
It's always like this, and I dont understand why.
In reality we use no more than 10 watts, even with inefficient speakers (mines are 86db, and my listening peaks are 90db @ 2,7m).
A mystery that science isn't able to explain yet
It's always like this, and I dont understand why.
In reality we use no more than 10 watts, even with inefficient speakers (mines are 86db, and my listening peaks are 90db @ 2,7m).
A mystery that science isn't able to explain yet
I think the explanation is that we use a few watt (average) when we crank it loud. But with a crest factor of 14 dB (average music) or 20+ dB for good recordings of dynamic music the peaks will reach a few hundred watt. So with a flea powered amp, you'll get some clipping. That's usually OK as long as the amp clips nicely and the peaks are short lived. But I do suspect the clipping is what leads to listening fatigue. Up the power of the amp and the amp will clip much, much less ... and you'll get less listening fatigue.
That would be the more scientific approach for sure. Unfortunately I don't have a digital scope, though I think the TEK 2465B I'm using is a digital-analog hybrid. I do occasionally think about dialling up my HP/Agilent/Keysight/Nom-d'Jour and getting a quote for a digital scope. So far I haven't been able to justify the expense, though the exercise of doing "polaroids" off the TEK 2465B is getting a bit long in the tooth.
Having a scope with persistence would be nice for stuff like this. Meanwhile I'll see what I can catch with the scope I have. Maybe the scope display on the AP has a persistence function. That'd be worth checking out.
Tom
I am waiting for this for such purpose
Oscilloscope, Multimeter & Logger that Fits in Your Pocket! by Pokit Innovations —Kickstarter
Oscilloscope, Multimeter & Logger that Fits in Your Pocket! by Pokit Innovations —Kickstarter
Tom, try this:
Connect your scope from one output of the '686 to ground. Set the trigger voltage to the point where the smaller amp would clip (accounting for the 2X between differential output and single-ended). Then crank up the brightness and crank up the music. Even though you may not catch waveforms you can document, the scope should trigger and flash a trace at the points where the smaller amp would clip. If you get lots of triggering, you know what's going on. If you don't get triggering then you can hypothesize why the bigger amp sounds 'better' even if you're not listening at levels beyond the capability of the smaller amp.
Connect your scope from one output of the '686 to ground. Set the trigger voltage to the point where the smaller amp would clip (accounting for the 2X between differential output and single-ended). Then crank up the brightness and crank up the music. Even though you may not catch waveforms you can document, the scope should trigger and flash a trace at the points where the smaller amp would clip. If you get lots of triggering, you know what's going on. If you don't get triggering then you can hypothesize why the bigger amp sounds 'better' even if you're not listening at levels beyond the capability of the smaller amp.
It should be like this but... Did you do the math?
The highest reading I have seen recently is 85db@3m, which is equivalent to about 90dB@1m If I'm correct (please allow rounding of numbers). My speakers are 86dB efficient, hence the power used is about 30W on peaks. Typically I use less than one watt.
I'm sure you didn't even max the modulus-86. Yet we feel this phenomenon.
Maybe there are quicker transients at higher SPL that dont get registered in a poor SPL meter (such that of my iphone8, which is pretty good tbh). Hence the test with ad digital scope is a very good one.
Do you read my website much?
LM3886 chip power amplifier thermal design.
My math says 81 W on the peaks (@ 90 dB SPL, 1.8 m from two speakers, 87 dB/W*m efficient). That's for 14 dB CF. It'd be higher for 20+ dB CF.
I use the Studio Six Digital app for my iPhone. It's pretty good, but microphone in the iPhone is not calibrated. Neither is the app... The apps vary all over the place and the microphones vary ±7-8 dB. You can see a test of many apps by the US CDC (Centre for Disease Control and Prevention) here: So How Accurate Are These Smartphone Sound Measurement Apps? | | Blogs | CDC
Tom
If we know the gain structure between the DAC and speakers there is no need for a scope on the speakers output to know exactly the levels of the peaks: we know what power will result from a given dBFS level so we just have to look at the digital signal.
So for example if we plug a 2V RMS DAC on mod686 amp with the default x20 gain we get max 200W into (min) 8 ohms at 0dBFS (leaving inter sample clipping aside) and we should not clip the amp whatsoever.
So for example if we plug a 2V RMS DAC on mod686 amp with the default x20 gain we get max 200W into (min) 8 ohms at 0dBFS (leaving inter sample clipping aside) and we should not clip the amp whatsoever.
Last edited:
Not much sorry
Anecdotally I have compared a 50W and a 180W (both power on 4 ohm as my speakers are 4 ohm) and didn't experience this phenomenon. If I were to raise the volume a bit I would probably clip the little one, but i really cant, living in a condo.
I used decibelX and I think the paid version can be calibrated, and supports various weighting.
Anyway smartphones today got incredibly good at such things. I tend to trust it more than a cheap SPL meter.
I'll read this too, thanks for the link.
"This follow-up study suggests that using external, calibrated, microphones greatly improves the overall accuracy and precision of smartphone sound measurements, and removes much of the variability and limitations associated with the built-in smartphone microphones."
Wonder how to connect my Earthworks M30 to a lighting port and a mic pre is needed too I guess.
Last edited:
Studio Six Digital has a calibrated version as well. They also offer calibrated microphones that plug into the phone. There's certainly plenty of opportunity to get fancy in that space.
It seems some of the apps use an average response curve for the specific phone model as calibration. That's probably decent.
I'd trust my phone about as much as I'd trust the old Radio Shack SPL Meter that it's supposed to emulate. It's good enough for what I use it for, i.e. to get a handle on how loud the music is.
Tom
That's certainly a valid approach. It'll work well on 0 dBFS recordings but those recording that have one sample at 0 dBFS and everything else at -20 dBFS, such as the original Dire Straits, "Brothers in Arms" CD, you might find yourself wanting higher gain.
Of course, the best place to put this gain is in a preamp, but if you want to go directly from the DAC to the power amp, you'll have to increase the gain of the power amp.
Tom
Tom, this remark was about the mod286 vs mod686 power "debate" a few posts above, and the fact that it was difficult to get the exact peak power needs for a given listening level.
The fact is it can be monitored easily by looking at the source levels (ie analyzing the source file) and doing the maths based on the gain and impedance.
The fact is it can be monitored easily by looking at the source levels (ie analyzing the source file) and doing the maths based on the gain and impedance.
Fair enough.
My board assembly house is done with the first batch of boards. You should see the stock levels on my website increase momentarily. I expect to get the first orders out tomorrow and to have the documentation done over the weekend.
I decided to tweak the mounting bracket a bit so those will arrive around the end of this month.
Tom
My board assembly house is done with the first batch of boards. You should see the stock levels on my website increase momentarily. I expect to get the first orders out tomorrow and to have the documentation done over the weekend.
I decided to tweak the mounting bracket a bit so those will arrive around the end of this month.
Tom
