Rigol DS 2072 noise problem

[playfsx]

Hi, before i start i must warning you, my english may hurt your eyes !
Today i get in my hands the brand new rigol ds2072 , i love it ! ... forrrr ... aboutttt 40 minutes and then i discovery ...  bugs and problems!! i mean whaaatt ?? i play with it only 40 minutes ! 

So let's begin, the first bug it is about the usb host , well if you open the scope with the usb stich already inside, there is not mount it , not even unplug it and plug again.  the solution is to open the scope without the usb stick and then plug in , not a big deal but steal annoying to plug an unplug the usb device all the time. I really want it all the time in so i can take print screen and save waves .  RIGOL FIXX IT !

The second problem is what i see spikes even if i don't have connect the probes and that really piss me off , so i take some pics , maybe you can help me with this one.

===========================================================================================================
with probe:

 
a closer look:


===========================================================================================================
touching the tip:


a closer look:


===========================================================================================================
with out probe:


and the screen:


===========================================================================================================
close put not touching:


and the screen:



===========================================================================================================
Hz:


===========================================================================================================
zoom in:


===========================================================================================================

I tried to close any other device in the room (including light bulbs ) i move it to another room but noting ...  seem the problem to be the inside psu ....

so i decide to make a "self calibration", i wait to finish and when it start up, not only there is no have fix it the problem but the expire all the trial version module who i have !! the scope come fully unlock (56Mpts i2c spi ....)  for limited time (about 30 hours) ... all gone without even taste it ..

next problem is ... well .. i try to measure the noise of the onboard regulator (5v) of my arduino and the give me AC sine with 38volts peak to peak ?!??!??!whhhhhhhhhhaaaatt ? i check my probe i check the configuration in the scope and i get my uni-t 61e to check it ! the uni-t give all most 5v with 10mv peak to peak,  i try to get a pic right now and it's reading correctly , soo what going on ?

next ... this is not exactly a problem... this is a epic fail from the rigol with love :
so i tried to install the drivers for this scope , i put in the cd and the only thing who i see is the manual and a readme.txt
with this inside:

Setup Explanation:
1??Download the USB driver from NI website-"NIVISAruntime.msi".
2??Execute "NIVISAruntime.msi" to install the driver program.
3??Execute and install the application software.

well very convenient , a cd with a txt in it who tells you to download the software ... i mean yeah this what the cd meant for ..


so if you can help me with the spikes
update:
I open the scope in the morning and it's working fine !! and the next night the doing the same !! so clearly something happening at the night !!

 

[zaoka]

Take it to frends home and try to make sure there is nothing in the home that may cause that.

[KTP]

It looks like a nice scope...comparable in specs if not slightly better than the DSOX2002 from Agilent for cheaper $$$.

The noise does seem very consistent...I am with you that it is coming from internal and not external.  I will be interested in hearing the solution.

[Rosendorfer]

Hi playfsx

No idea about spikes.... sorry..
But seems You are the first one hire with hands on Rigol DS2000 series...
I would love to hear anything good or bad  regarding this scope.

Rosendorfer

[playfsx]

It's nice scope yes , but well i don't like the software , it can make it a lot better, for example  the fft is pretty poor and really want to use that future ... im going to make a video review for this scope when i found the problem, so mabe that help you decide if worth the money or not
so to the problem now,
im gone try to another home tomorrow, but i don't this it going to change anything.. The only thin make me believe that is internal the problem is the freq. of the signal, is 100hz double the ac freq. in my home so probably is some bridge that it's double the freq. but why in day working correctly  ?  and why when im touch it the amplitude is going crazy, i mean 5v p-p is too much!

[madworm]

If you touch the probe with your finger you get the line frequency sine wave. That is normal.

[playfsx]

If you touch the probe with your finger you get the line frequency sine wave. That is normal.

yes and ... no , i take the 50 ac main with 100hz spikes, the ac of the main it's ok the spikes is the problem 

[EEVblog]

I was initially going to say it was LCD backlight driver (common issue with scopes), but then I looked at the frequency and it's 100Hz.
Clearly it's the full wave bridge rectifier in the mains switchmode, somehow... Maybe the Y class suppression cap is involved?
In any case the values are quite small. But when you advertise 500uV/DIV as a standout feature you want to be sure that stuff isn't visible.

Dave.

[playfsx]

well i don't know for the power supply, i stay awake all the night to see what time it going to stop doing that , so i didn't open or close any device, i don't touch the scope and the best off all is, i make a video so you guys can have a clear picture

  you can see it in HD   and is about 2 minutes

so im thinking maybe the night it's more noise intro the ac lines and the psu can't reject it ? is any easy way to test if that is the problem , i try to measure the peak to peak before and after with the uni-t 61e but i don't see any difference in the ac .

dave thanks for your reply , if you see the video im getting 160mv p-p with the probes, this is not small , i want to trigger low voltage signals and the spikes getting in the way, not mansion how bad is for the fft :/   

[armandas]

Just a sanity check. Is your scope earthed?

[flolic]

Plug a BNC terminator and try again.

[poorchava]

Maybe you've got a neighbor using some heavy duty ghetto construction high power device (like home made welder or something).

Such devices are known to generate strong disturbances in AC lines if there is no proper mains filtering. In some countries connecting such equipment to a public grid line is officially not allowed, you need to have permission and a proper high power connection. Of course what peoplke do in reality is whole different thing.

[robrenz]

Improperly filtered variable frequency drives will also pollute the surrounding area with nasty noise

[playfsx]

Maybe you've got a neighbor using some heavy duty ghetto construction high power device (like home made welder or something).

Such devices are known to generate strong disturbances in AC lines if there is no proper mains filtering. In some countries connecting such equipment to a public grid line is officially not allowed, you need to have permission and a proper high power connection. Of course what peoplke do in reality is whole different thing.

i don't know about that , but im pretty sure what is some pulse in my ac line doing that , is any way to filter my ac power ? i mean i thinking to try a ups but im not sure if this is good idea, is any problem if i power the scope with out earth reference  ? (im gone try disconnected from the ac line )  expect the ups what other device i can use to filter my mains ?

[sotos]

Maybe you have a Dimmer for your lights or electronic transformer for them.

They produce a lot of noise.

I’m thinking, because you said only in the night you have the problem.

[playfsx]

Maybe you have a Dimmer for your lights or electronic transformer for them.

They produce a lot of noise.

I’m thinking, because you said only in the night you have the problem.

im not have dimmer , i don't thing to be any from my device that causing that , i have unplug every single device from all my house include light bulbs and try it, and accusal the spices become some mV more intensive

[KTP]

I think a decent battery backup UPS to power your scope just for testing is a great idea.  Those kind that are meant to work with digital equipment and computers (probably still a modified sine wave though).  You could make a cable that leaves earth connected and disconnects the AC.

[A Hellene]

Nice scope, playfsx! And welcome to the forum!

The oscilloscope seems to be pretty much quiet, with a noise floor level of 520µV! *salivating*

Of course, I certainly do not like these 2.5µs wide, 45/90/165mVpp periodical spikes, seemingly coming out of nowhere. They seem to be associated to the mains frequency.

This is a long shot but, the scope must have an isolated mains zero-crossing or frequency detector circuitry in its PSU PCB, in order to be able to take readings synchronised to the mains frequency: You can see that signal at the older Rigol DS1000E/D PSU schematic sheet at the end of the first post, and how this signal can be selected as a trigger source (being fed into the fourth input of the LMH6574 multiplexer) at the Trigger/Comparator schematic sheet. So, since that signal is there, it could accidentally interfere (through dirt, flux remains, a malfunctioning component, etc.) with its surrounding circuitry, and find its way to the display buffer under certain circumstances (temperature, humidity, etc.).

Since the spikes seem to have a close relation to the mains frequency, try to power the device by a UPS, whose output frequency will eventually deviate from the mains frequency: Then you will know the source of the spikes, whether they are being induced internally or they are being picked up by the probes and/or the input amplifiers.

As suggested previously, try shorting the input with a 50 ohm terminator and see if the spikes disappear. Do not be afraid also of powering the scope by a floating (ungrounded) UPS. Even though you would not be in any danger at all, just try not to touch any of its metallic parts, like the BNCs. Of course, you could always ground the device by grounding the UPS ground line.

Try taking readings, making combinations of all the options above, while the scope would normally introduce those ugly but periodical and symmetrical spikes.


-George



EDIT: More info on the AC line trigger circuitry.

[hans]

I've seen issues with oscilloscopes from other mains powered devices before too, like an airconditioning unit. I was debugging issues with some power supplies, AC coupling, 50 - 100mV/div and triggering on an 4-channel Isotek (I consider Isotek a 'B-brand')  scope. Everytime to airconditioning unit starts or stops I saw spikes (because I was triggering in normal mode)
I don't know how well the things where grounded there with earth, but I imagine quite bad because soldering stations was causing interference with an Aoyue hot air station as well.. If I turned off the Weller solder station, the Aoyue hot air station would briefly pulse on.

So I'm pretty sure it's better to power it with earth reference.

[playfsx]

thanks for your answers 

I don't have a 50ohm terminator and i don't thing this is going to give any result at all, never the less i try to short the bnc connector with a 47ohm 1/4w resistor, and yeah i don't take any spikes , i don't expect anything else i mean is only 4mv with out the probes and with a so small resistor it is going to disappear.

So i try more thinks ! I try to power the scope in another house and sure enough no spikes at all , not in the day not in the night !
I power the scope with a ups in the same house and accusal when unplugged from the mains i see spikes and there is not difference if i ground the ups or not.


This in not surprising i guess because the ups it's not sine wave capable.

this measurement i take with the probe close to ups.

So clearly my ac line have the problem 
I have a idea , what if im gone to accusal measure it, now i can't measure the mains directly but what if i try with a transformer, i mean if i have some spikes going to the primary coil i must able to see it on the secondary right ?
The biggest transformer i have is 220 to 24v and sure enough i found some thing !!

So let's make some math, the p-p of the secondary is 78.2 and the mains is 625v p-p , so the transformer ratio is 625/78.2 ~= 8


The spike in the secondary is 9,28V p-p so in mains is 9,28 * 8 = 74,24V p-p

We now have the reason now how i can fix it ??
My devices not cause that so i must some how filter the mains .

I don't have try the ups of my friend in my home, so i don't know if the are going to fix anything if i buy one , but im thinking the only type ups it would be accusal work is the on-line ups and this is not cheap , is any other way to filter the mains ? preferably cheaper way ? 

[KTP]

Wait.  You tried a 50ohm termination (or you sort of made a 50 ohm termination using a 1/4 watt 50 ohm resistor with short leads at the bnc to ground) and the problem went away?

This sort of points to the problem not being generated internal to the scope, but also means it is not coming into the scope over the mains.

Maybe check with the downstairs neighbor and see if he is building a gocart in his bedroom with a TIG welder...

[A Hellene]

The 24VAC/50Hz transformer is very-very slow in response, and it should filter out those 2.2µs wide (~450KHz) spikes. This means that the spikes are probably created internally by some parasitic element, out of the AC line zero-crossing detector circuitry. Please read my additional material at my previous message.

Try to read the 24VAC secondary of the mains powered transformer again, while the scope is powered by the UPS. Let's see if the spikes follow the peaks of the under test transformer waveform or the peaks of the UPS power supply. That is because the mains and the UPS waveforms will most probably be out of phase; and as all your captures show, the spikes follow the AC power supply peaks with a delay of ~1ms.


-George

[vk6zgo]

To be quite sure of what you are seeing,you need to eliminate the Rigol from the test setup.
Try to borrow another Oscilloscope,have a look at the secondary of the transformer again,without the Rigol plugged in.
My guess is that you will see no spikes,Now plug the Rigol in & look for spikes,still with the other'scope.
If the only time you see spikes is with the Rigol plugged in,it will point to the Rigol as the source.

[playfsx]

Wait.  You tried a 50ohm termination (or you sort of made a 50 ohm termination using a 1/4 watt 50 ohm resistor with short leads at the bnc to ground) and the problem went away?

This sort of points to the problem not being generated internal to the scope, but also means it is not coming into the scope over the mains.

Maybe check with the downstairs neighbor and see if he is building a gocart in his bedroom with a TIG welder...

I make one with 47ohm(i forget to mansion the 47ohm so i edit the last post) resistor and yes it went away, so if understand correctly that you said is that the probes are pick up this noise ?
In my building the are living mostly students and now in the summer it's not here, it is however a small bar close to my house, maybe the have duty speakers doing something, i mean i have login what time the start and what time they stop and so far look like this:
day 1:  21:40 - 5:20
day 2:  21:48 - 4:20
day 3:  ?? - 4:10

Please read my additional material at my previous message.

George i relay try to understand but i don't have so much knowledge, i will however try again tomorow, maybe the clear head helps,

The 24VAC/50Hz transformer is very-very slow in response, and it should filter out those 2.2µs wide (~450KHz) spikes. This means that the spikes are probably created internally by some parasitic element, out of the AC line zero-crossing detector circuitry

maybe you have right, but im pretty curies accusal check it , now i don't have a function generation but i can use a 555 i can get to 500khz easily
so this is my setup:
i power the 555 form a 12v power supply , the output is connected with a 470 ohm resistor to the primary (220v) and i probe secondary (24v) with the ch2


in the scope i use a average times 16 because it's not that stable

so 590khz and not filterer, im pretty sure that the problem it's is not my scope,  if it is im going to sent it back im not going to open it, it take me all most one year to gather the money im not going to lose the warranty ...

Try to read the 24VAC secondary of the mains powered transformer again, while the scope is powered by the UPS. Let's see if the spikes follow the peaks of the under test transformer waveform or the peaks of the UPS power supply. That is because the mains and the UPS waveforms will most probably be out of phase; and as all your captures show, the spikes follow the AC power supply peaks with a delay of ~1ms.

it's hard to take ups of my friend to my house , but i will try .

To be quite sure of what you are seeing,you need to eliminate the Rigol from the test setup.
Try to borrow another Oscilloscope,have a look at the secondary of the transformer again,without the Rigol plugged in.
My guess is that you will see no spikes,Now plug the Rigol in & look for spikes,still with the other'scope.
If the only time you see spikes is with the Rigol plugged in,it will point to the Rigol as the source.

nice point here, but i know only one who have a oscilloscope, it's my university and there is not even one at the million to borrow me one ! 
still i don't thin that the rigol is the source, in the other house it working perfectly and in my home at the day .
maybe i can try other sensitive device HMMmm ..  im thinking my sound card but in not that sensitive and probably i will create a small explosion 

 

[notsob]

This is probably a dumb question - "do you have fluorescent lighting ?", if answer is yes, have you tried turning it off. (I'm looking for a source of power spikes within your house)