high quality sine wave ?

[Audioguru]

Check out the MAX7400 series.

https://datasheets.maximintegrated.com/en/ds/MAX7400-MAX7407.pdf

Rob

I made a sinewave that had 10 steps and about 6% distortion with a CD4018 and a few resistors. Then I filtered it with National Semiconductor switched capacitor filter ICs that had 8 Butterworth orders and the output distortion was so low that it could barely be measured.

Guys have taken a buzzer sound and filtered it with switched capacitor filters and ended up with pretty good sinewaves.

[nfmax]

Almost as good as the K-H 4402 is the -hp- 339A THD analyzer, which includes a low-distortion generator.  The residual distortion in mine measures about 0.002% (-94 dB) which should be good enough for any normal applicatioin.  Units such as the 339A have the advantage that the analyzer tunes directly to the internal oscillator frequency, and will lock to an external oscillator when the frequency is approximately correct.  Also, the indicated fraction of distortion scales automatically to the input signal.  You can learn a lot about the DUT by analyzing the distortion output signal, which has the fundamental removed.

The oscillator part of the 339A was made as a standalone oscillator model 239A - quite a small dinky little thing, and no fan so they tend to be in immaculate condition internally. I have one and it works nicely, but they are very rare on the secondhand market. As well as low distortion, it has good level setability right down to sub 1mV, and it the output is isolated from ground - two essential attributes for a low distortion audio source.

[Electro Fan]

previous thread might have some related discussion
https://www.eevblog.com/forum/testgear/waveform-generating-displaying-question/

[TimFox]

Will a good oscillator + the HP 334A be just as good as the HP 339A?  I picked up two of the former, which set me on a path for a decent 1kHz sinewave reference.  Looking at the distortion output with fundamental removed was definitely part of my plan.

How does the K-H 4400 compare to the K-H 4402?

BK

The residual distortion in the 334A is not as good as in the 339A.  Both the 333A and 334A will auto-tune the frequency and phase when the manual tuning and phase knobs are close to the correct values for the external oscillator.  Another advantage of the 339A is that the distortion value indicated scales automatically to the input level, while the 334A/333A require a manual adjustment to 100%.