There are plenty of ok-ish arbitrary waveform generators you can get for cheap on aliexpress. This particular one has a 2x 14bit 250MSa/s DAC in it and feature an additional counter input. The original ±20ppm is fine for most tasks, but not as a reference to measure other crystals.
I replaced it with the AST3TQ-T-10.00MHz-28 a TCXO with identical footprint. After aging its accuracy is ±1ppm/ +-3ppm typ. (after 1/ 20 years), which is still far ahead of the original one. Replacement is simple, desolder the old oscillator with hot air and put on the new part. I can not verify the accuracy with my equipment, but there's not much that can go wrong so I trust the datasheet. The frequency counter on my DS1074 reports a spot-on 10MHz output though.
I'm writing this 1 year after modding my device, now the chosen part is "NC" on mouser, but as of writing there are still 157 units left and you should be able to use any other TCXO with similar specs.
With the TXCO I can be sure the output frequency is accurate (enough). This matters, for example, if you want to measure the frequency of an crystal connected to a MCU. As a bonus all generated frequencies will be accurate as well, so you could use this DDS to test a RTC or similiar.
Sorry for the late reply, hackaday.io's notification system is broken and I didn't get an email.
Hi,
What is the real advantage to change the TXCO ?
Thank