EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope

[matlipinski]

Hi.
I have bought MSO5074 (hw. version 1.00) some time ago, and from the beginning I have overshoot undershoot issues, as well as square signal from arbitral generator has also 20% over and undershoot. After applying newest firmware  01.01.04.08 and lfcal.hex from mable and running self calibration they are still there. After applying mable's lfcal.hex they vanished from 1st channel, almost vanished from second channel, but they are still on 3rd 4th channel. Do You think i should send it to Rigol?

Please look at the screenshots.

[TK]

Hi.
I have bought MSO5074 (hw. version 1.00) some time ago, and from the beginning I have overshoot undershoot issues, as well as square signal from arbitral generator has also 20% over and undershoot. After applying newest firmware  01.01.04.08 and lfcal.hex from mable and running self calibration they are still there. After applying mable's lfcal.hex they vanished from 1st channel, almost vanished from second channel, but they are still on 3rd 4th channel. Do You think i should send it to Rigol?

Please look at the screenshots.

Have you compensated each probe in 10X mode?

[matlipinski]

Yes probes are compensated. It is not probe issue (i have used each of four probes on all channels). I have done everything. I's not my first oscilloscope. Square signal generated on external arbitral generator looks much better, no over or undershoots.

Spikes from generator are also visible on 1054Z (2nd channel arbitral generator, 1st channel 1khz test signal from MSO5000).

[matlipinski]

Could someone dump rights and ownership of /rigol/data directory and files (/rigol/data folder itself too). Thank you.

[thm_w]

Could someone dump rights and ownership of /rigol/data directory and files (/rigol/data folder itself too). Thank you.

This enough?

Code: [Select]

<root@rigol>ls -a -al
total 28221
drwxrwxr-x   12 1000     1000           960 Aug 19 02:51 .
drwxr-xr-x   22 1000     1000          1024 Sep 25  2017 ..
-rwxrwxr-x    1 1000     1000       6692678 Aug  2 05:33 K160M_TOP.bit
drwxrwxr-x    4 1000     1000           288 Jul  4 07:34 Qt5.5
-rwxr-xr-x    1 root     root      22200576 Aug 19 02:46 appEntry
drwxrwxr-x    9 1000     1000           608 Jul  4 07:34 cups
drwxr-xr-x    2 root     root          1432 May 30 04:33 data
drwxrwxr-x    2 1000     1000          1152 Jul  4 07:34 default
drwxrwxr-x    2 1000     1000           872 Jul  4 07:34 drivers
drwxrwxr-x    4 1000     1000           288 Jul  4 07:34 mail
drwxrwxr-x   10 1000     1000          1040 Aug  2 05:33 resource
drwxrwxr-x    2 1000     1000           736 Jul  4 07:34 shell
drwxrwxr-x    2 1000     1000          1336 Jul  4 07:34 tools
drwxrwxr-x   11 1000     1000           744 Jul  4 07:34 webcontrol


Code: [Select]

<root@rigol>ls -a -al
total 776
drwxr-xr-x    2 root     root          1432 May 30 04:33 .
drwxrwxr-x   12 1000     1000           960 Aug 19 02:51 ..
-rw-r--r--    1 root     root           148 Jan  1  1970 Key.data
-rw-r--r--    1 root     root           668 May  6  2019 adccal.hex
-rw-r--r--    1 root     root           156 May  6  2019 adcsmpdelay0.hex
-rw-r--r--    1 root     root           156 May  6  2019 adcsmpdelay1.hex
-rw-r--r--    1 root     root        722716 May  6  2019 cal.hex
-rw-r--r--    1 root     root            92 Jan  1  1970 chDelay.hex
-rw-r--r--    1 root     root           112 Mar 20  2019 dgCal.hex
-rw-r--r--    1 root     root            36 Jan  1  1970 extcal.hex
-rw-r--r--    1 root     root          1572 May  6  2019 hcal.hex
-rw-r--r--    1 root     root            36 Mar 21  2019 lah.hex
-rw-r--r--    1 root     root            36 Mar 21  2019 lal.hex
-rwxr--r--    1 root     root          1164 Mar 21  2019 lfcal.hex
-rw-r--r--    1 root     root            80 Nov  5 02:58 lxi.txt
-rw-r--r--    1 root     root            18 Jan  1  1970 mac.txt
-rw-r--r--    1 root     root            36 Jan  1  1970 precision.hex
-rw-r--r--    1 root     root          1028 Nov  5 03:03 stat.dat
-rw-r--r--    1 root     root           280 Mar 21  2019 sysvendor.bin
-rwxr-xr-x    1 root     root            12 Jan  1  1970 user.conf


[matlipinski]

It is hacked version or stock? Thank you!

[thm_w]

It is hacked version or stock? Thank you!

Thats hacked

[Martin72]

Hi,

In march I did a test with the 40ps pulsegen and with 50Ohm termination or not:

https://www.eevblog.com/forum/blog/new-rigol-scope/msg2270838/#msg2270838

Yesterday my new terminator arrived, from Huber&Suhner.
And there´s a difference between this and the former one.
While the former one (from Telegaertner) produces "high" overshoots, the Huber&Suhner produces significant less.
A little surprising to me.

[TK]

Does anyone else find the output range of +-2.5V (HighZ mode) of the AWG very limiting?  It cannot even generate a 3.3V digital signal.

[thm_w]

When I had AC coupling on today, the signals seemed quite slow to fall into place, I don't have much to compare to but I assume this just means the AC coupling caps are larger than typical values?

Does anyone else find the output range of +-2.5V (HighZ mode) of the AWG very limiting?  It cannot even generate a 3.3V digital signal.

Yep, I am working on a board to amplify the output. Originally it was just going to plug into the USB and the two output BNC's, although now its morphed into something overly complicated. Maybe at some point I will retry that basic design idea (+/-5V with a simple 7660 for low output current for example). Here are some examples of commercial versions:

https://www.aliexpress.com/item/4000338925225.html
https://www.aliexpress.com/item/32894305598.html

I did buy a DG800 which can do +/-10V, but it is nice to use the same instrument to generate and measure.

[thm_w]

Does anyone else find the output range of +-2.5V (HighZ mode) of the AWG very limiting?  It cannot even generate a 3.3V digital signal.

Here is the idea so far:
- +/-8V rails generated from USB should permit easily available opamps to hit +/-5V output into 100 ohm.
- Gain fixed at 2x
- USB pass-through (hopefully it can still detect with additional capacitance on the line)
- USB connection either short pigtail or a 3D printed standoff + male plug

Not sure yet:
- pots to trim gain or input offset, may affect HF performance

Components: THS3091, TPS65130.

[rolfdegen]

Hallo

I am the new owner of a Rigol MSO5104.

I test the web interface of the MSO5104 and its very slow. The refresh rate is about 1 image per second.

I have GBit Lan. Can speed up the web interface.

[mabl]

I have GBit Lan. Can speed up the web interface.

It is my understanding, that the performance of the web interface is limited by the image compression on the scope's CPU before transmission over ethernet.

[nctnico]

I have GBit Lan. Can speed up the web interface.

It is my understanding, that the performance of the web interface is limited by the image compression on the scope's CPU before transmission over ethernet.

Indeed. None of the oscilloscopes will show a difference whether the network speed is 100Mbit or 1Gbit. The bottleneck is always the CPU.

[Martin72]

https://www.eevblog.com/forum/testgear/review-rigol-mso5000-tests-bugs-questions/msg2852332/#msg2852332

Hmpf…
Measuring ripple noise on power supplies is a basic thing scopes should do it.
As mentioned in this link I´ll try it again with a filter, if this doesn´t work….
I was always confident with this one and now such a simple thing won´t work.
Makes me think twice…

[Martin72]

I have to put that into perspective….
Having a look in the specs of the supply, it claims that ripple will be under 1mVrms.
And the rigol just can´t display it, it´s own noise will cover this, regardless of any filter attempts. ( Ripple of a simple supply is no problem, it´s "big enough" to display it)
Will test it again when on work next week, with our lecroy scopes.

[Martin72]

Measured again, with following settings :
Trigger on AC line, acquisition mode average (64times).
Source:

Chinese linear supply WEP603D.
Maximum output voltage 60Vdc, current 2.5Adc...

With these settings, you can probably see the ripple.
Where the offset comes from, I don´t know really...

[2N3055]

With these settings, you can probably see the ripple.
Where the offset comes from, I don´t know really...

Martin,

offset comes from waveform not being symetrical so it has DC component when it gets averaged (integrated) on coupling capacitor.

[Martin72]

Ah....

Thank you Sinisa !

[ebastler]

offset comes from waveform not being symetrical so it has DC component when it gets averaged (integrated) on coupling capacitor.

Umm.. but the offset visible in Martin's screenshot is significantly higher than the peak-to-peak amplitude of the AC signal. I don't see how the signal would drive such a large offset.

Also, didn't Martin state earlier that the offset appears as soon as the probe is connected to the scope, even with no signal connected?

[2N3055]

offset comes from waveform not being symetrical so it has DC component when it gets averaged (integrated) on coupling capacitor.

Umm.. but the offset visible in Martin's screenshot is significantly higher than the peak-to-peak amplitude of the AC signal. I don't see how the signal would drive such a large offset.

Also, didn't Martin state earlier that the offset appears as soon as the probe is connected to the scope, even with no signal connected?

Hm, you are correct, it is larger than amplitude... It cannot be so large just because of asymmetry of what is on screen. I stand corrected, I didn't pay attention..

But this is x64 averaging, so it shows averaged data of many acquisitions. If there are quick high amplitude pulses, you won't see them but it will skew average...

Before using average, I use peak mode and persistence  to get feel of extrema of the signal, and also a quick spectra of the signal so I know what will average and what will be messed up.

For instance, if you have random spaced large amplitude bursts (like WiFi or 4G) you can get all kinds of fun stuff on a scope when it gets rectified in scope input and then averaged with signal you're looking at... Looking at millivolts with 500 MHz bandwidth is all kinds of fun..

Original signal without averaging has no DC offset.

Interesting.

[Fungus]

I have to put that into perspective….
Having a look in the specs of the supply, it claims that ripple will be under 1mVrms.
And the rigol just can´t display it, it´s own noise will cover this, regardless of any filter attempts. ( Ripple of a simple supply is no problem, it´s "big enough" to display it)
Will test it again when on work next week, with our lecroy scopes.

Are you using a 1x probe? It's pointless trying to measure ripple in 10x mode.

(and be sure to turn the 20Mhz bandwidth limiter on, too)

[Martin72]

Hi,

Yes, measuring ripple I´ve "learned" on work, doing only with 1:1 (and Filter-Cap) - Exception: When You measure ripple of a high voltage converter.
Like this for example:

http://www.pantechnology.com.hk/eng/shortcut/ripple.htm

(Linear supply, for switching supplies the cap´s value must be much less, depending on the switching frequency)

[David Hess]

Having a look in the specs of the supply, it claims that ripple will be under 1mVrms.

I would be really cagey about a modern DSO's automatic RMS measurement.  Some measure AC RMS and some measure AC+DC RMS and the terminology is not consistent.  Dave even did a video on the subject.

And the rigol just can´t display it, it´s own noise will cover this, regardless of any filter attempts. ( Ripple of a simple supply is no problem, it´s "big enough" to display it)

Oscilloscopes tend to be pretty noisy.  Sometimes an external low noise amplifier designed for the measurement bandwidth is required to make power supply noise and ripple measurements.

Older oscilloscopes going back to the 1970s often had bandwidth limits lower than 20 MHz to limit noise sufficiently for their most sensitive range.  So the more sensitive or noisy the oscilloscope was, the lower the bandwidth limit was to produce a reasonable display.

[Martin72]

Hi again,

offset comes from waveform not being symetrical so it has DC component when it gets averaged (integrated) on coupling capacitor.

Test it out what happen to the signal displaying when it´s not symmetrical and scope is ac-coupled, see pics below. (No averaging mode, doesn´t take a effect when using it)
Further I´ve noticed a little offset (last pic), where I don´t know whre it´s coming, from the build-in Generator or from the scope itself.

Martin