Rigol MSO 5000 hardware/software revisions

[skander36]

Try this :
Set Default -> Activate G1(default settings) connect a cable from G1out to CH1 input->Press AUTO

Let us know what's happened.
Maybe some scopes are defective .

That displays the sine wave on CH1, the other three channels stay off.

My scope is doing as can be seen here : https://youtu.be/2Uvl-dC13Pc
If it is defective, why is not doing this on Fw. 04.08 and below ?
Calibration does not change anything .

[pmaggi]

Is it possible to back to 00.01.04.08?

[skander36]

Is it possible to back to 00.01.04.08?

Yes . With "SINGLE" key at boot time .

[sb42]

Try this :
Set Default -> Activate G1(default settings) connect a cable from G1out to CH1 input->Press AUTO

Let us know what's happened.
Maybe some scopes are defective .

That displays the sine wave on CH1, the other three channels stay off.

My scope is doing as can be seen here : https://youtu.be/2Uvl-dC13Pc
If it is defective, why is not doing this on Fw. 04.08 and below ?
Calibration does not change anything .

It certainy seems less "defective" than mine going by the noise levels

Have you tried firmware 00.01.02.00.03?

[skander36]

Yes , I said before.
Both 02 & 03 have the same behavior.

[NoisyBoy]

Sorry to hear that, 02.03 fixed the ghost channel problem in my scope, as it has not occurred again after the upgrade.  But I have not heard any other feedback and you are the first to reply.  So I guess the result is inconclusive at this point.

I agree with your observation as my scope never had the ghost issue prior to the 02.02 upgrade, but 02.03 seem to have solved that problem.

[skander36]

Hi  NoisyBoy,
The 03 version has been out very close in time to 02, so I have put it  immediately on scope to see that they have solved .
But nothing changed.
I have tested from then by putting 04.08 then advanced to 02.03 , back to o4.08 ->02.02 ->02.03 and again .
There was one time when I was on 02.02 the things seem to be better in that channel 3 was not activating automatically but only when I hook a cable from G1 .
But when I have moving to 02.03 the things went back and channel 3 is activate by itself again even I have downgraded to 02. So I decided to remain on 04.08 to waiting for a new fw. update .

[NoisyBoy]

Hi skander36, sorry to hear what's going on with your scope, and thanks for sharing that datapoint with us so others can be aware of potential issues with the upgrade.

In your case, I agree that staying at 04.08 would be the best option, as the ghost channel bug is a real pain. 

[pmaggi]

In my scope 02.00.03 also produces the ghost in channel 3.

[skander36]

In my scope 02.00.03 also produces the ghost in channel 3.

Yeap ... I wasn't to call it a bug as vast majority report that it is not manifesting in their scopes, but as long it is present only on specific version of firmaware , but not in the previous and it can be reproduced , I don't know how to name it.
I think that Rigol lower the threshold levels and in some conditions appear more noise that open channel 3.
Maybe someone has better suggestions ...

[skander36]

With fw. 03 I have tested level on which ch.3 is activated and I have found that over 23,5 mV on ch1 , ch3 is activated . Under 23,5 mV ch3 is not activated. So ch1 activate ch3 (and some time ch4) .

[typoknig]

I just got my MSO5074 this week. It is HW rev 01.01.000 and came with FW 00.01.01.04.08. I ran the following tests, updated the FW to 00.01.02.00.03 without performing a self calibration, ran the tests again, then performed the self calibration, and ran the tests one last time.  I got the same results in all three scenarios and never saw a single ghost channel. Only when a grabber was on the probe was that probe's channel activated:

All probes disconnected, rubber caps over BNCs, pressed AUTO 10 times

All probes disconnected, rubber caps off BNCs, pressed AUTO 10 times

Each probe connected by itself with grabber and ground lead, pressed AUTO 10 times

Each probe connected by itself without grabber, but with ground lead, pressed AUTO 10 times

Each probe connected by itself with grabber, but without ground lead, pressed AUTO 10 times

Each probe connected by itself without grabber or ground lead, pressed AUTO 10 times

All probes connected with grabbers and ground leads, pressed AUTO 10 times

All probes connected without grabbers, but with ground leads, pressed AUTO 10 times

Is it possible those of you seeing the ghost channel have had a less than perfect self calibration? Maybe some interference like LED lights or a leaky wireless device caused the self calibration to be inaccurate?

[NoisyBoy]

I'll be damned, typokbnig is on to something I never considered - WiFi interference!

I have a high power mesh network running in my home, I have been seeing a consistent 1+ mV noise on my scope, oddly, only on channel 2 but none of the others.  I was about to send the scope back to Rigol for repair.

When I powered the scope up tonight, channel 2 became the ghost channel that would always come on when I press Auto (which did not surprise me as I know about the Channel 2 noise issue).  Thanks to typoknig's suggestion, I powered down the mesh satellite closest to the scope, and guess what, the noise is gone, and no more ghost channel issue!

I am not surprised about the interference, I should have known better, but I am surprised that it only manifested in a single channel, perhaps it is less shielded than the others.  Anyway, armed with this new finding, I will leave the WiFi satellite off and run a full calibration this evening.

Back to skander36's comment, I have supplied signal of varying voltages to channel one up to 5V p-p, and I have not been able to get any ghost channel to come on.  So the suggestion of trying it outside of WiFi range and LED lights may well worth a try and see if it makes any difference.  However, like you, I have not seen this issue in firmware 04.08, so I would agree Rigol might have lowered the turn-on threshold just enough for this problem to manisfest in our scopes.

Just finished a self cal, noise level is consistent across all 4 channels, right at around 303uV (at 2mV/div) with +/- 3 uV across channels. 

[Gandalf_Sr]

Wow, graet wrok typoknig!

[skander36]

Hi , @typoknig and @NoysyBoy ,
Thank you for your suggestions. I also give that tips to others but I was not able to follow for myself  ...
I moved the scope away from bench (into other room) and after calling default settings (Restore Defaults)  from boot menu (single key), I was doing a self calibration .
Now the ghost channel activation does not appear anymore .

[Gandalf_Sr]

So are there some ideas as to why this is happening?  Like the MSO5000 is picking up some kind of RF transmission and that is messing up the self-calibration?

[NoisyBoy]

skander36, it is great to hear that the WiFi tip works for you as well and was able to fix your problem.

Gandalf_Sr, the Rigol is definitely not great at shielding the frontend from RF interference.  I would not call this a failure or classify it as inadequate yet, but I have not seen similar issue at any other test gear in the same test location.  In my case, the noise it picked up is between 1-2 mV on channel 2, which is almost 6-7X the normal noise floor, so it definitely would mess with self cal.

So if you own or plan to buy a MSO5000, putting it in an area with strong WiFi signal can potentially really mess with your noise floor and cause ghost channel issue.  If you have them, try turn off your WiFi while doing the self cal, or while doing low level measurement.  Again, this interference was only across one channel in my case, so check all your channels and don't base your finding on channel 1.

[skander36]

So are there some ideas as to why this is happening?  Like the MSO5000 is picking up some kind of RF transmission and that is messing up the self-calibration?

And why ch3 is activated by signal from ch1 ?
They should not be separated (isolated) ?
At what level this influence is done ?

[Sighound36]

That possibly explains why I have never had this issue with the 8000 or 5000 we have here.

The wifi is not meshed but far enough away from the router.


When I dismantled the 8000 to rebuild the very average power supply there were RF fingers everywhere, even on the Lan/USB/Aux BNC inputs on the back really nice design as well.

All of the ASIC cans use decent finger arrays to the case, this is maybe because of the higher bandwidth potential of the 8000 but they are in all of the places you would expect them to be placed for a half decent scope.

[NoisyBoy]

Hey Sighound36, thanks for the great input.  Do you happen to know whether it is the same with the MSO5000?  In my case, I did all the tests with nothing attached, so it was not probe related and the interference was picked up internally.

[TK]

One difference Dave noticed during the MSO5000 teardown was the difference in the input stage shielding.  The 5000 has the shields open with aluminum heatsinks sticking out.  I am not an expert but I suspect it cannot be as good shield as one completely closed.

[Martin72]

I am not an expert but I suspect it cannot be as good shield as one completely closed.

I´m not an expert in this too, but I would agree to this.

[typoknig]

One difference Dave noticed during the MSO5000 teardown was the difference in the input stage shielding.  The 5000 has the shields open with aluminum heatsinks sticking out.  I am not an expert but I suspect it cannot be as good shield as one completely closed.

Here is a link to that point in the teardown. Note that the 7000 does have some holes in the cans, just smaller.

https://youtu.be/P5faiEUXbGg?t=772

[skander36]

I think the problem is rather in software .
Don't forget that the problem is not present in fw. 04.08 and still in 02.02 .
Only in 02.03 can be corrected.

[NoisyBoy]

I would say the ghost channel is caused by the firmware upgrade, but the 1+ mV noise I picked up is definitely RF interference.

The larger opening above the ASIC heat sink may very well be the cause of the RF interference.  It is something that we should be aware of while placing the scope.  It seems like Rigol might have focused more on cooling and less on interference while coming out with the design. 

Comparing this to the Siglent 2000X+, https://www.eevblog.com/forum/testgear/siglent-sds2000x-plus-coming/msg2976282/#msg2976282, the Siglent seem to do a better job in providing more comprehensive shielding for the front end from the look of it.