EEVblog Electronics Community Forum
Products => Test Equipment => Topic started by: Winston03 on August 30, 2019, 08:17:21 pm
-
Just purchased one of these DDS waveform generators, the 24MHz version. It works, but there is about a 5% frequency error between what the display indicates and what is actually generated. For example, when set to generate a 10.0000KHz sine wave, the actual output is 9.52380KHz. The built in frequency counter is also in error about the same amount when measuring an external signal, but oddly will agree with the (incorrect) reading on the display when measuring itself. Is there some calibration procedure that would correct this situation? I did not see any reference to such in the manual. Or maybe it's an internal reference that's out of spec? With a DDS generator I kind of expect the output to be what is indicated, so it kind of sucks as it is. The output amplitude is spot on though.
-
Assuming your counter you measure with is calibrated, it shows your internal time base is off. I don't know the spec for the device, but I presume this has to be way-off spec. Since it is new, I would suggest you contact the seller for an exchange rather than trying to calibrate yourself. To do that properly, you will need a Rubidium or GPSDO, along with a high precision counter (you can drive that off the GPSDO as well), and that's assuming there is documentation on how to do that properly.
Don't know about Feeltech, but for a quality AFG, I routinely get frequency error under 0.1 ppm if I take effort to calibrate its internal timebase.
-
As previous post you need to exchange it. The frequency accuracy spec quotes +/- 5 X 10 -6. A quick check on my FY3200S set to 10.00000 Khz shows 10.0000 Khz on my DS1054Z which is OK and within specification. Checking against an actual frequency standard is not really worth it for a cheap but useful function generator.
-
The time base is shared with both the generator and counter, that’s why the counter would show 10k with the AFG signal. If you feed a true 10kHz signal into it, likely it will likely show a 10.5kHz count.
-
I just measured the reference clock oscillator frequency as being 23.9997MHz. That was using the counter in my Siglent 1202X-E scope. The oscillator is marked as 24.000 along with a part number and logo I could not identify. Likely Chinese made. So that would be only about -12.5ppm assuming the Siglent is accurate. The FY 3200 was purchased from a Chinese seller on eBay for a really low price. Not sure it would even be worth the bother trying to send it back and get a refund.
-
Problem solved! Seems this is a known issue related to a software bug. I only now just stumbled upon an ealier thread in eevblog discussing this exact problem. The FY3200 can be recallibrated by pressing the PARAM and left arrow buttons together while powering up. Hold for two seconds, release and a calibration display appears. Press the blue function knob, power off and on, and calibration is restored. I did this and my generator now reads perfectly.
-
Great to hear! If it is a software/firmware bug, you may want to keep an eye on future updates. If you buy it from a Chinese vendor, I can appreciate the potential difficulty in exchanging it.
Good luck with your projects.