| Products > Test Equipment |
| RD JDS6600 25MHz 2-Channel DDS AW Function Signal Generator |
| << < (35/44) > >> |
| Adrian_Arg.:
This model of generator finings, which comes with external power supply, can be changed to a source of better quality, with a 15mhz I can use for hobby or should be higher Mhz. https://www.ebay.com/itm/JDS-6600-60MHz-Dual-channel-DDS-Function-Waveform-Signal-Generator-Counter-CE/372059704084?epid=2210616056&hash=item56a07c8b14:g:z2IAAOSwXGtZpMkois for use in small electronics projects |
| Mr. Scram:
--- Quote from: tjones99 on March 03, 2018, 06:22:36 pm ---Reading through this thread as it is the best I have found so far. Thanks for all the info. One thing I have not seen covered (hope I didn't miss it somewhere) is amplitude calibration. I set the amplitude for 16V square wave (talking 25Hz here so no bandwidth issues) and I find I get 16.7v measured on the scope. First question everyone is going to ask is how do you know the scope is right? The scope was actually calibrated and I fed a second signal into the scope from a power supply that had a voltage of 8.02 volts (measured with an expensive calibrated HP DVM) and the JDS6600 output was higher than +8.02V. To get a true 8V peak output I had to set the JDS6600 to 15.3V amplitude. I was hoping there was a mechanism to adjust (calibrate) the amplitude of the output. There are 4 trim pots on the board... 1A 1D 2A 2D... anyone know what those accomplish before I mess something up really bad :-) I looked through the communication protocol and saw nothing that I though helped. Hoping RD Tech has some input on getting the amplitude closer to true!! --- End quote --- Is your lead properly terminated? At 25 Hz there shouldn't be any trouble anyway, but just to be sure. Youtube is full of people who claim defects or malfunction, while it's obviously operator error. |
| Cliff Matthews:
RD Tech invents words.. As the box says, in their "lastest" video, some "unstruction" on software installation. |
| bugi:
I got mine earlier this week, and now have had time to test the JDS6600 a little bit. I have no idea how old/new version it is. The box was "blank" (unlike some video/photos that I've seen which show text etc. on the box) and it felt like it had been stored in a swimming hall (and smelled like some chemical I have smelt perhaps once in couple years or so, just can't remember what chemical) :P It was overall packaged pretty much watertight, so it must have been affected during storage - or maybe it is just different kind of chinese cardboard. I collected any issues I've seen in this thread/topic or elsewhere, and went through them. I couldn't really test noise/quality stuff at this time, due to, hmm.., "mismanaging" some of my measurement cables (so the setup isn't the best), but most things are obvious even through less than optimal connections. Note: I'm a newb, so some results (whether for better or worse) could be my mistakes. However, I did try my best to note earlier comments about how to test each case (and how not to test). When a case seemed to be ok, I tried higher frequencies than the typical 10k, just in case the issue is too short time to be noticed when looking at slower waveforms. Overall, "good enough" for me, I just have to write a note about the channel 2 phase randomization quirk and tape it on top of the device, so that I don't forget it and waste hours some day later trying to figure out why my circuit is behaving badly, when it is actually the generator quirk doing it. Results (with simple grade of OK/BAD/MEH): (EDIT: "5V" etc. is amplitude, not peak-to-peak.) 1. Change frequency, waveform should remain continuous (no change in phase). Sine 5V 10kHz to 20kHz and 5MHz to 10MHz, on single channel. -> On channel 1, the phase continues smoothly. BUT on channel 2, the phase makes a seemingly random change in the phase at the same time. (See also 2.) Happens every time, not just occasionally. OK / BAD. 2. Change frequency of one channel, unsynced, the other channel should not be affected in any way. Sine 5V 10kHz to 20kHz and 5MHz to 10MHz (and relative phase 0 deg and 90 deg). -> Changing the frequency of a channel results in the expected change in the output freq of that channel, BUT channel 2 makes a seemingly random change in phase at the same time. (Note, whether the adjusted channel is 1 or 2, channel 2 is the one that gets the weird shift in phase.) Happens every time, not just occasionally. BAD. 3. Change frequencies of both channels at once (synced); both should change simultaneously and phases should continue smoothly. Sine 5V 10kHz to 20kHz and 5MHz to 10MHz. -> Both channels change frequency and phases continue smoothly. (Note how sync enabled prevents the phase shift issue on channel 2.) OK. 4. Both channels, running long-term, should keep their relative phase and have no (unexpected) jitter/phase glitches. Square and sine wave, 10kHz synced, 10MHz synced and unsynced, 8MHz and 4MHz unsynced. Using infinite persistence. With channels having the same frequency or at integer ratios in order to be easier to spot phase drifting. -> No sign of glitches or phase drifting over long periods. OK. 5. Sample rate and step size/resolution 5V and 2.5V 50%-50% triangle and sine, 10kHz, 1kHz, 100Hz and 10Hz. Zooming down to 10mV/div and timing to see the "steps" at positive zero-crossing edge. -> Saw about 5mV per step for triangle, about 7.2mV/step for sine (both at 5V amplitude), and about 2.5mV/step for 2.5V amplitude triangle). So, seems to be about 2000 steps per direction (half-waveform going up), or about 11-bits. Weird. Doesn't match directly with either the advertised 12-bit resolution, or the 2048 sample waveform length. It looked like the negative side has slightly larger steps (triangle 5V, ~5.5mV on negative side vs. 5mV on positive side), at least near zero-crossing - perhaps "DAC" (the resistor-thingy) distortion? Not quite what I expected (I somehow thought that for simple waveforms the on-the-fly phase calculation and RAM-LUT would be almost de facto at low frequencies (like the nice DDS-chips do, though with specialized circuitry), which would result in the full resolution being usable). (More than good enough for my purposes; I can add extra filter on the output for lower frequencies if the steps cause trouble, though the possible distortions might not be as easy to handle.) MEH. 6. Channels should be in phase accurately (simultaneously clocked output changes). With sine and square, 10kHz 5V, same output for both channels, looking at rising edge. -> Channel 2 seems to lead channel 1 about 2ns. MEH. 7. Sync frequency sweep, the phase difference should remain constant. Could not figure out how to check it accurately, so relied on "by the eye" and "waveforms overlap within pixel". Checked both at low frequencies (<=10kHz) and higher (3-5MHz). -> Seems to be ok, at least could not see similar relative phase drift as shown ealier in this topic. OK. I also took a look at noises and waveform shapes, but as mentioned, due to measurement setup issues, I can not be sure of the results. Of those, perhaps the most interesting results were 1. The difference in the shape of step-change ringing; channel 1 looked "normal", channel 2 was... weird.. Have to take another look at this later. (Edit: tried with the leads that came with the generator; channel 1 ringing remained about the same, but the channel 2 weirdness turned to just "very very low ringing", so, seems likely it was/is just cable differences combined with generator output and/or scope AFE differences having fun with me.) 2. One channel ringing hard seems to leak to the other channel a bit. If this is by the generator (not by my bad setup), maybe tweaking it by adding a bit more capacitance and/or ferrite somewhere could isolate the channels better. Unless the coupling is in a shared amplifier or such. |
| Kimbo:
Thanks for the tips Takfuji. 30 dB is very useful! Any chance of some photos showing how you implemented these suggestions? |
| Navigation |
| Message Index |
| Next page |
| Previous page |