Of course you have to measure (calibrate) the generator level first to determine 0 dBFS. I can not do all with the tool. It's not an AP ($ 30,000), it's a free progam for $ 0. Hint text of the edit: --> 'Gen H1 level that is shown in the analyzer without DUT/NOTCH/AMP.
I haven't looked at your latest version yet, but the user should input both the fundamental level as well as the HD levels, to have a meaningful result.
Once you have the two values you could even offer the user a selection of ratio designators, dBV, dBFS, dBr, that should be a no-brainer.
That you can do for very much less than an AP ;-)
Jan
Once you have the two values you could even offer the user a selection of ratio designators, dBV, dBFS, dBr, that should be a no-brainer.
That you can do for very much less than an AP ;-)
Jan
If we can measure the spectrum, then we can get good information about the tested audio device. If we calculate THD from the measured spectrum, we will lose a lot of information about DUT. If we need to do something to improve the circuit, then THD cannot help. Two devices with the same THD can sound very different.
High-end audio manufacturers sometimes describe distortions more detailed, for example, individual levels of lowest order harmonics or the spectrum pictures, and it is nice practice in my opinion.
What is the main goal?
High-end audio manufacturers sometimes describe distortions more detailed, for example, individual levels of lowest order harmonics or the spectrum pictures, and it is nice practice in my opinion.
What is the main goal?
That is correct. A spectrum picture is worth a thousand words. But you don't really know what quality level you are on. A number x THD is helpful to estimate the whole thing. Especially if you have little experience with spectra.
hmmm arent we talking about what the referene level is -- the fundemental?
FS would refer to the input... or zero level. If fund is 1v then that is the FS ref to. dBr is OK also.
Carrier doesnt work for me unless I am talking about modulated signals.
-RM
Last edited:
Whenever we state the THD of a device under test, it is imperative that we state, at minimum, the level and frequency of the fundamental coming out of the device. We can then say something like THD is -120 dB (0.0001%) and everyone knows what we mean.
It is important that we not get entangled in semantics that may be mis-interpreted.
Cheers,
Bob
@RNMARSH
I completely agree with your comments.
I also acknowledge that dBc derives from modulation lexicon and dBr would serve equally well. But I still like dBc for its unambiguous flavor. I'm used to looking at dBc as a measure of local oscillator quality where a complete lack of modulation is the ideal.
Best.
I completely agree with your comments.
I also acknowledge that dBc derives from modulation lexicon and dBr would serve equally well. But I still like dBc for its unambiguous flavor. I'm used to looking at dBc as a measure of local oscillator quality where a complete lack of modulation is the ideal.
Best.
The full spectrum analysis also gives us other useful information: yesterday I was testing a boost amplifier that is part of a phono amp I am building. All was going well with the individual op-amp circuits, but one (of the LM4562s) was giving a rise is noise at the bottom end rising to about 20dB above midband at 30 Hz or so. All the others were dead flat. I replaced it with a Burr Brown 4562 and that peculiarity dissappeared. Turned out to be an out of spec. chip.
Regards
Mike
There really have been two versions of the LM4562 (LME49720). The first from National and the second from TI after buying National and shifting it to a newer process. Burr Brown is also part of TI and doesn't have such a chip. They do have other dual opamps with different strengths.
JRC makes a similar 4562 part https://www.mouser.com/datasheet/2/294/njr_njm4562_e-547842.pdf which might be very good.
JRC makes a similar 4562 part https://www.mouser.com/datasheet/2/294/njr_njm4562_e-547842.pdf which might be very good.
@RNMARSH
I completely agree with your comments.
I also acknowledge that dBc derives from modulation lexicon and dBr would serve equally well. But I still like dBc for its unambiguous flavor. I'm used to looking at dBc as a measure of local oscillator quality where a complete lack of modulation is the ideal.
Best.
Well, lets just not mix them. Keep dBc to modulation measurements and acoustic weighting measurements.
But we all still need to know the ref level (dBr) used on the DUT and so that needs to be always stated by everyone submitting their T&M data.
THx-RNMarsh
Last edited:
Maybe they took the number because it has cachet in audioland?? JRC normally makes very good parts. it may be worth trying a few to see how they actually perform.
How you know a semiconductor salesperson is lying? His lips are moving. (Never got pushback from one when I pointed this out. Never saw a new chip finished on time. . .)
How you know a semiconductor salesperson is lying? His lips are moving. (Never got pushback from one when I pointed this out. Never saw a new chip finished on time. . .)
The NJM4562 is a completeley different part than the one from National. Actually, it appears to be much older and the number scheme is a continuation from many previous NJM45xx products (like NJM4560). We can only wonder how come National use the '4562' number...
Never used one of the NJM's personally. But what I know is, 10years ago or so, the LM4562 had huge batch-to-batch variations wrt unbearable popcorn type of noise appearing sometimes on more than 20% of the parts. This, and the poor RFI performance made me stop using the part....
Yup, my mistake,
I was going from memory and the chip next to the LM4562 was a BB DRV134.
The newer LM4562 looks like a National part, judging by the logo.
However, the new part did droped the LF noise down considerably and point being that it showed up pretty quick by using a spectrum analyser function
Regards
Mike
Last edited: