Products > Test Equipment

Rigol DG800 (long) list of problems and the sad death of a FPGA

<< < (5/6) > >>

thinkfat:

--- Quote from: thinkfat on February 16, 2020, 10:25:08 am ---
--- Quote from: ebastler on February 16, 2020, 08:58:07 am ---
--- Quote from: thinkfat on January 20, 2020, 04:28:00 pm ---Talking about strange behavior, my SDG2042X doesn't hold phase between channels if one of them is at a very low frequency. When setting a 10MHz sine wave on one and a 1Hz PWM on the second channel, phase shifts by 90° every second.

--- End quote ---


--- Quote from: Siglent on January 20, 2020, 09:00:46 pm ---Hello Thinkfat. You can contact our EU support team by writing support-eu@siglent.com and they will be able to assist you. - Jason, SIGLENT North America

--- End quote ---

Hi Thinkfat -- just curious: Did you reach out to Siglent, and did they have a solution? I have the same SDG, and am generally pleased with it. If Siglent were actually able to provide solutions for that kind of issue, I'd be very pleasantly surprised!

--- End quote ---

Yes, I did contact the EU support, but they were not really able to help. They said they could not reproduce the issue and if I wasn't able to solve it I should send the device back for replacement because clearly, then the hardware would be faulty.  :palm:

Of course it wasn't. I eventually found an obscure setting in the Output configuration which synced the two channels and eliminated the phase shift.

Only now, possibly after updating to the latest firmware I have a different issue: Even if I clock the device from an external source (GPSDO) I get a significant frequency offset. I tried feeding a 10MHz signal into my TIC and clocked it against a 1PPS from a GPS and what came out had a very significant drift. I know I did the same test with the factory firmware and it didn't exhibit that problem.

--- End quote ---

I got it worked out, I believe:

- if the two channels are phase-locked, even if the second channel is turned off, phase locking takes priority over frequency accuracy. What puzzles me, though, is that even if the first channel is an integer multiple of the second (e.g. 10MHz, 1Hz), the 10MHz output will be slightly off (couple of mHz). Might be an attribute of the PLL they use for synthesis. What's even more puzzling is, if I tune the second channel to 2Hz, the frequency offset is gone, at 3Hz it's back, etc.
- If the second channel is switched off, the first channels' frequency offset stays whatever it was before you turned the second channel off. Even if you change the second channel frequency while it's off, no change on the first channel. You need to turn the second channel on again to see the effect.
- If you set the two channels to "independent", no frequency offset.

Here's a little table I worked out (Ch.1 set to 10MHz):

--- Code: ---Ch.2[Hz]     1      2      3     4     5     6     7     8     9    10   11   12   13   14   15   16
Ch.1[mHz] -170   +830      0  -500  -240     0     0  +160     0  -240    0    0    0    0    0 +160

--- End code ---

Above 16Hz you get mostly '0', but occasionally there will be a difference of a couple 10 mHz.

Strange, but once you know about it you know how to take care of it.

MarkoAnte:
Update:
I tested and played around with Siglent SDG 1032X.
My thoughts about on comparison of problems from original post.

1) It does not go negative on start up, BUT sometimes it starts ok, sometime it does not, form what I can tell its random. See pictures below. The setup is ch1: square wave, 0 to 3.3V, 1MHz ch2: sin wave, -250mV to 250mV, 1MHz. I triggered on square wave, had sin wave on, square wave off and then turned square wave on.
Ok:

Not ok:

Zoomed pictures also attached.

2)Phase alignment is better, as it at least does not jump around, but it still leave a lot to be desired. (Phase set to 0)

f[MHz]   0.8   0.9   1.0   1.1   1.2   1.3
Angle[°]   14   16   18   20   21   23

3) Did not test communication

4) Phase align is automatic when you change something, presets seem to work, no touch screen so, in my mind, its 10X better to use just because of that.

5) Standard design, better feeling power button, stack-able, has handle and rubber corners/feet, has usb in the front for loading arb. waveforms.

From what I tested the siglent seems to be a better unite (it has less stuff that rub me the wrong way) but still does not star ok (sometimes). If I would be buying a new FG I would get the siglent but I don't know, if its good enough for me to buy it and have it with the rigol. 


@2N3055 I don't think so.
1) I always enter Low/high level pair because I'm lazy and don't want to the hard maths that is 3.3/2.  :-DD
2) I was doing 0 - 3.3V at 1MHz on a FG that does 25MHz. Also I can increase the frequency and it still does 0-3.3V no problem.
3) It does not show the setting change on screen.
4) If that the output voltage would be the problem, then it would presumably not be a transient response on start up but always present.

thm_w:
Response from Rigol:

--- Quote ---I have found  out that new firmware is expected by the end of June that will address this  issue.
--- End quote ---

Not sure why the release time is so long, but nice if it can be fixed.

thm_w:
This offset on power on issue is now fixed in the firmware released on Feb 18 (0.2.4.0.0):
https://www.eevblog.com/forum/testgear/new-rigol-16-bit-function-generators-dg800900-series/msg2950062/#msg2950062

core:
On the paper DG800 it's a good AWG for the price, but with all these issues it's not easy for me to justify the money.

The MSO5k internal AWG have the same problem regarding the offset when the generator is started.
In the picture : square wave, 1kHz, 500mV, offset 250mV.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod