I have received my FY3206S generator and have had some time to test it so I thought I would share some oscilloscope screen shots and some data.
First thing I noticed is that this generator uses very little power. I powered it through Kilowatt power meter.
10 kHz Sines 10 Vp-p no load: 1 W
10 kHz Sines 10 Vp-p 50 Ohm load: 2 W
6 MHz Sines 20 Vp-p no load: 3 W
6 MHz Sines 20 Vp-p 50 Ohm loads: 3 W
*USB connection does not change power.
Nice low power
I then checked out the performance with a Tektronix MDO3104 1 GHz BW, 5 GS/s oscilloscope. I connected with BNC to BNC cables with the scope set to terminate in 50 Ohm.
Per request earlier in the thread first up is rise time for a square wave. The square is nice and clean with rounded edges, linear slope, and essentially no overshoot. Measured 10-90% rise time is 19 nsec. Perfectly fine for a 6 MHz square. Feeltech spec is "<100 nSec" but 100 nSec makes no sense for a 6 MHz square so I am assuming that spec is a typo in the manual. See attached scope shot "FY3206S MDO3104 Sqr Rise 5Vpp.png".
Second, we have investigation of square wave edge jitter. With the set frequency a simple integer sub multiple of the 250 MHz sample rate I could not see any jitter. I assume if I really looked I would at least see sub nSec jitter. If I picked a non-integer sub-multiple I could see two distinct edges (via scope persistence turned on) 5 nSec apart. For non-integer sub-ratios you would expect one cycle of the 250 MHz clock =4 nSec spaced edges once in a while to get the average frequency to the set point on average. I also looked 9 cycles beyond the scope trigger point to give it more chance of edge jitter. See attached scope shot "FY3206S MDO3104 Sqr Jitter 5Vpp.png".
I then looked at triangle wave linearity and sharpness of the corners at the peaks. At 1 Vp-p set point yielding 500 mVpp at the scope at 1 MHz the slopes are very linear and the corners look pretty sharp. See attached scope shot "FY3206S MDO3104 1MHz Trgl 500mVpp.png".
Next up is 500 mVpp triangle at the full 6 MHz. There is some rounding of the peaks and the p-p measured amplitude drops 15.6% but overall pretty nice triangle waveform for top frequency. See attached scope shot "FY3206S MDO3104 6MHz Trgl 500mVpp.png".
Now moving on to sine wave purity/distortion. Attached is an FFT spectrum of a 10 Vpp set point amplitude at 100 kHz. Measured fundamental amplitude is +4.8 dB or 1.75 Vrms or 4.94 Vpp (low by 1.2%). The largest harmonic was the fifth at -51 dB relative to the fundamental. Based on this spectrum I would say THD is < 0.5% although I don't have a direct THD measurement. Easily below the published spec of 0.8% at 1 kHz. See attached scope shot "FY3206S MDO3104 100kHz Sine Spectrum 5Vpp.png".
Lastly, I checked the noise and distortion while outputting the smallest amplitude set point of 10 mVpp. The noise on the waveform with the scope set to 1 mV/div was a fuzzy width of the sine waveform of about 0.4 mVp-p or about 0.15 mVrms which is close to the scope's measurement noise. Looking at the spectrum the largest harmonic is about 40 dB down. Note that at this small level there are harmonic lines showing up at 140 kHz which is not related to the 100 kHz sine and is either a generator noise and/a very high frequency signal aliasing on the scope's 100 MS/s setting for this measurement. At any event that spectrum peak is only 30 uV rms. All in all, pretty good low amplitude results. See attached scope shot "FY3206S MDO3104 100kHz Sine Spectrum 5mVpp.png".
The one bad thing with this generator is that the output amplifier can not drive a full 10 Vp-p across a 50 Ohm termination load with the set point set to 20 Vp-p. The specification trickily lists no load output as up to 20 Vp-p, which it does, and lists the output impedance = 50 Ohm, which it has, but the specs lists nothing about output amplitude with 50 Ohm load. By my measurements at all frequencies the amplitude maxes out around 7.5 Vp-p across 50 Ohm. It seems fully linear up to about 10 Vp-p set point. At 11 Vp-p set point the third harmonic for a 20 KHz waveform grows by 10 dB instead of the ideal expected 0.8 dB if the wave shape stayed constant and only amplitude grew by 11/10 gain. So, fully usable up to 10 Vp-p set point with 50 Ohm load while larger amplitudes are only useful with less load.
*This limited output drive capability explains why the 6 MHz 20 Vp-p set point power draw with and without 50 Ohm load can both be 3 W given the power doesn't report fractional Watts.
Just for fun I made an 2048 point waveform that is 10 sine cycles. This way when commanding 6 MHz you actually try to output 60 MHz. I would say using this technique you get usable output up to the full FY3224S product of 24 Mhz. Above 30 MHz the amplitude drops and the phase lag increases rapidly. Below are my test results:
Ch1 at 2.00 Vp-p (1.0 Vp-p into 50 Ohm) at 100 kHz for reference output amplitude:
(Phase measured relative to TTL sync pulse output)
Freq Amp Phase
MHz dB Deg
6.0 -0.5 -60
10.0 -1.5 -91
15.5 -3.1 -137
20.0 -4.9 -180
24.0 -6.9 -220
I even tried multiple cycles of a square wave and could get a usable but not "pretty" square wave up to 15 MHz.
I am guessing that if you pay the extra money for the FY3224S version you probably get the exact same HW with the FW letting the max frequency go up to 24 instead of the 6 for the FY3206S model I tested. Maybe there is an capacitor value difference in the DAC/output buffer amplifier low pass filter that allows higher bandwidth but I doubt it.
If there are any questions I will try to answer them. I will also update if further testing shows up anything interesting.
gby
Warning: On my Win 7 64 bit PC I did not install the old USB driver that was emailed to me. I just plugged the unit in and let Windows find the latest driver for me. When using the PC software it seemed to work. Until I tried to send an arbitrary waveform down to the unit. It erased the current waveform and then got stuck when trying to download the new data eventually reporting a Parity Error2. After trying many things, including learning that the USB connection is a Prolithic Technologies USB to serial PL-2303HX bridge chip, I finally beat Windows into submission to forget the new driver it had found and installed the old one emailed to me. In that configuration everything works fine. However, the USB to serial bridge chip is now obsolete and Prolithic states on their web site they will NOT update the driver for that chip for Win 8, 8.1. They recommend using their new pin for pin compatible chip instead.
Home
Help
Search
Login
Register
