Evalulating gear for THD usually consists of running several tests @ various f's, and/or restricted to a few f's mulitone tests.
But real world use comprises of tons of different f's & harmonics. So how to establish Exactly how much THD is being created ?
But real world use comprises of tons of different f's & harmonics. So how to establish Exactly how much THD is being created ?
There is no one number that can be given to an amplifier to describe it's THD under all conditions. The best you can do is say the distortion will be less than X% between Y and Z frequencies at power leverls between A and B.
A multitude (multi-tone) of freqs all at same time will exercise the amp in a little more realistic way. Its a step in the right diretion. but still other tests are needed.
-RM
-RM
This topic gets raised fairly often, perhaps because the amplifier design authors usually mention it in passing. The historical Belcher test comes to mind and jcx has commented here that such multi-tone techniques don't really add significantly to what can be learned by traditional single, dual and triple tone (IMD) testing. I have no idea whether all that is true as I understand it but I think that is the gist of it.
Here's an overview paper of multi-tone technology published at Gedlee's site: http://www.gedlee.com/downloads/THD_.pdf
Here's an overview paper of multi-tone technology published at Gedlee's site: http://www.gedlee.com/downloads/THD_.pdf
99% of the needed distortion information is gathered by a traditional single-tone THD/THD+N measurement, where the source is swept over an appopriate range of frequencies (say 10-100k Hz) and levels (say from full scale down -40 dB of that). There are rare cases where this does not catch a significant nonlinearity mechanism (e.g. near the bandwidth limit of the DUT, where the harmonics may be filtered out), and this can be fixed with a two-tone IMD measurement. Multi-tone measurements are good for high-speed production testing, but I'm not aware that these should be of any interest for R & D.
Samuel
Samuel
What evidence to support at inverse?
The thd of SIN, can give us an idea but it is far from the real work that plays an amp, with the music.
I can say then, that the SIN is perfect for hiding non-linear trajectories hehe!
or, How does a audio power stage, when modulates traits "long" in the unbalanced (from the VCC rail)? have linearity?
- Status
- Not open for further replies.