Home Theater Forum and Systems banner
1 - 17 of 17 Posts

·
Registered
Joined
·
3,069 Posts
Discussion Starter · #1 ·
Hi guys,

I have just ran some THD distortion test on my dual FI Q18's and this is on my middle tune which is 17.5hz.

I did it to these instructions:

Use the RTA Function, set Mode to Spectrum, FFT Length to 131072, Averages to 4, Window to Hann.

Use the onboard tone generator in REW, fire up a sinewave. Set the THD calculator on (It'll open up a THD tab). Make sure the record button is on.


Could someone explain this to me on the results I have got as I think it looks extremely high.

cheers

Graham
 

Attachments

·
Elite Shackster
Joined
·
7,142 Posts
Ive never used the THD function, so I'll be watching this one with very keen interest. Also, you know the inevitable question from me :D. You fancy some sealed comparisons :R

The ported in theory should actually show less distortion around the port tune, and especially so at higher spl levels.
 

·
Premium Member
Joined
·
6,711 Posts
One thing to be wary of is using frequencies where the harmonics coincide with your local mains frequency and its harmonics, as any hum pickup (electrical or acoustic) would skew the results (though those would show up when the tone wasn't playing). That aside, see how the distortion levels change if you reduce the output level by 10 and by 20dB. It would also be worth measuring each of the sub's playing alone.
 

·
Registered
Joined
·
3,069 Posts
Discussion Starter · #4 ·
John,

Here's the -10db and -20db which I have reduced on the MV on the onkyo.

I presumed the distortion levels at this volume would be pretty low to be honest and not what I was reading.

Which distortion reading should I be looking at as there are 2-9.

cheers

Graham
 

Attachments

·
Elite Shackster
Joined
·
7,142 Posts
9 % difference between -10 and -20 is a fair swing. It shows how much high spls demand. I do wonder though how much this would change with a different amp.
 

·
Registered
Joined
·
3,069 Posts
Discussion Starter · #7 ·
Just to throw this in, this is +10db from what the normal setting I had was.

High setting is -16db on the MV
Normal is -26db on the MV
-10db is -36db on the MV
-20db is -46db on the MV

cheers

Graham
 

Attachments

·
Elite Shackster
Joined
·
7,142 Posts
Its probably be down to the amp, so its entirely possible, but having never having done this myself before I couldnt honestly say for sure. The sealed sub will tax the amp more driving distortion up a little more as your results show. You would have to try another amp to see the effects though. As long as your drivers are within xmax, they should contribute little to any distortion produced.
 

·
Registered
Joined
·
3,069 Posts
Discussion Starter · #9 ·
The wierd thing is that it doesn't sound the slightest bit distorted even at high levels so why is the RTA show up crazy distortion levels. How do I found out what the cause is?

cheers
graham
 

·
Elite Shackster
Joined
·
7,142 Posts
How would you know what the specific distortion sounds like without hearing the same setup with less distortion for comparison. If it was that easy to simply hear, we wouldnt need REW. This is the whole point of this kind of program, it shows you these things without the need for masses of experimentation and comparison.

Ive seen it said, that with the lowest frequencies, its almost impossible to detect distortion till it reaches at least 10%.
 

·
Registered
Joined
·
3,069 Posts
Discussion Starter · #11 ·
REW is showing my distortion at 46% at only -26db on the MV and only 93db. Surely I would hear that amount off distortion.
 

·
Elite Shackster
Joined
·
7,142 Posts
Maybe, maybe not :huh:, I cant really speak for you. To detect something like this though, you need a comparison benchmark unless its ridiculous and entirely obvious. You may well be hearing it, you may just not know it. You could be doing something wrong when testing for all I know and getting duff result. We need an RTA expert to jump in here I think to be honest.
 

·
Premium Member
Joined
·
6,711 Posts
The largest distortion component is the second harmonic, which does not sound "distorted" (think tube amp :)). High even harmonic distortion orders are typical of overdriven single-ended amplifiers, look carefully at the levels through the chain to the sub. The other thing to be careful of is what you are using to measure with. SPL Meter? Mic? If mic, be careful of premap gains and levels. If SPL Meter, make sure the range is appropriate.
 

·
Registered
Joined
·
3,069 Posts
Discussion Starter · #14 ·
John,

I am using the Radioshack analogue meter. I had it set to 70db for setting up my levels. So if I am running at a level off 94db then I need to change the radioshack to 90db range setting. Is that correct.

cheers

Graham
 

·
Registered
Joined
·
12 Posts
Maybe not relevant to such "huge" distortions, but I usually tick "Lock frequency to FFT" in "Generator" - it helps to measure tweeter (or amp) distortions, where we deal with tenth of percent's.
 

·
Premium Member
Joined
·
6,711 Posts
I am using the Radioshack analogue meter. I had it set to 70db for setting up my levels. So if I am running at a level off 94db then I need to change the radioshack to 90db range setting. Is that correct.
Yes, the meter is easily overloaded so needs to be in the right range. You should also run through the SPL level calibration after you change the meter range so that REW understands the overall gain change that has happened (otherwise your plots will be drawn lower than they should be). You can just use your usual test signal (e.g. a receiver test tone at 75dB), enter in REW the reading that would give on your meter when your meter is on the 70dB range, but leave the meter set to the 90dB (if that makes sense) - you are basically telling REW that even though the electrical signal it is seeing is lower, it still corresponds to 75dB acoustic level.
 
1 - 17 of 17 Posts
Top