EEVblog #1146 - New Low Cost Rigol MSO5000 Oscilloscope

[bd139]

Very true on both counts.

[Martin72]

Maybe my expetations are too high, maybe I shouldn´t compare affordable scopes automatically with the ones I work everyday with in our testfield.
But I do, perhaps in the hope that someday  a "B-Brand" scope launch into the market, which makes older good quality stuff forgettable.

My impressions by playing around with the 5074 are that this one could do it.
If rigol optimize the firmware proper, the hardware can have the capability for it - I´ll see it in the next days at work.
By the way, the 1mV/div range is bandwith limited - a "B" pops up by the choose.
At 2mV/div it disappears.

[0xdeadbeef]

You seem to forget Lecroy is selling Siglent gear. So on one hand you have a company which builds really high-end oscilloscopes and on the other hand sells low end gear

Again, this lacks any dedication to the details discussed here...

Here, I fixed the selective quoting for you so now all the details are within context as it should be.

It's not my lack of quoting skills that was in question.

[glenenglish]

I agree that 2 GSps is an overkill for a 100 meg scope but if I like it I will upgrade it to 350M, and apples for apples, dollar for dollar, it is hard to beat.

It's not brickwall filtering of course. That's the -3dB.

My experience is that the minimum oversample rate of 4x BW gives me the pictures I want if there is a bus fight.. This is quite obvious on my 5GS tek scope's response.

it will be interesting to measure the rolloff rate of the 100meg at 2GSps. IE what Rigol do with the filtering.

beyond 350 MHz, probing accurately becomes more difficult also.

[Martin72]

Some functions are not avaible and "greyed out", perhaps functional with an option/firmware upgrade, seems normal to me.
BUT:
"Options Install", "Online Upgrade" and "Local Upgrade" are also deactivated...

Got a answer from rigol today:


The fields are self-activating.

A detection of a USB-stick with options-license file will activate the "Options Install" Button.
A detection of a USB-stick with a firmware-file will activate the "Local Upgrade" Button and if the scope is online, you got access to the online upgrade button.

[Fungus]

I agree that 2 GSps is an overkill for a 100 meg scope

When you turn on all four channels it's less, proportionally.

When you unlock 350MHz it's about right.

[mrpackethead]

A unconfirmed claim of of the MSO5000 has been made by a chinese student.   

"Well, I have patched the firmware, let it jump out license verify produce. But I can't make it public until next year March. Because Rigol sold out about less than 300 units now.

In fact I'm working on my friend's scope and I havent ordered yet (lack of money...Im just a ungraduated). I m wonder if I make it public prematurely, maybe they will fix it and it can't be cracked anymore.

Btw, there's no keygen for 5000 series oscilloscope because it cant be realize. The only way to crack it is to patch firmware.

The detail of crack this scope I will
publish it to my github when my scope is successfully cracked."

Sadly he does not want to provide the info, I think he is worried that Rigol will patch the issue before he has collected enough money to buy his own.      If he was able to crack it, i'm sure that others will be able to do it as well, pretty quickly.   if he wants the 'claim to fame' of being the guy who cracked it, he will need to publish it before anyone else does i guess.   though it seems he just wants the 350Mhz scope for the 70Mhz price.

[2N3055]

I agree that 2 GSps is an overkill for a 100 meg scope

When you turn on all four channels it's less, proportionally.

When you unlock 350MHz it's about right.

Just to make sure it has 8/4/2 GS/sec for  4ch and 8/2 GS/sec for 2 ch.

More than Nyquist is useful. Most scopes go way more than sticker spec. Keysight wrote WP in which they basically said that you need 5x oversampling to make it work real well with real world filtering and such. If speed and memory is there no harm done. At least some members wont complain that Rigol shows points that don't exist. 

[Martin72]

At least some members wont complain that Rigol shows points that don't exist.

For example ?

[2N3055]

Some time ago was discussion where some members didn't believe that you can reconstruct signal if you don't have many samples per period....
I was joking about that...  Not important..

[Martin72]

Ah, okay...
And it makes sense.
Our older LeCroy Waverunner models ( 1999-2002)showing us everything in the right manner, although they have "only" 500Mpts/Sa by 200 Mhz BW.
A few months ago, a LeCory sales man told me the "right ratio" between samples and bandwith - unfortunenatly I can´t remember the value at the moment, but I´m sure it was less then expected.
Today I´ll take my 5074 to work, test it next week under "real conditions".
As I mentioned before, Rigol itself compares it with the 2000er models from Tek and Keysight.
The 7000s series they compare it with the 3000s series of Tek and Keysight.
Understatement or true sight...

[0xdeadbeef]

The MSO5000 hasn't got "smart" probe interfaces -> entry level as Keysight DOSX2000, R&S RTB2000 etc.
The DS7000 has them, so it's targeted at the "one above entry level" like Keysight DSOX3000, R&S RTM3000 etc.

[nctnico]

At least some members wont complain that Rigol shows points that don't exist.

For example ?

AFAIK on the DS1000Z the sin x/x is implemented incorrectly which hasn't been fixed (yet).

[Fungus]

AFAIK on the DS1000Z the sin x/x is implemented incorrectly which hasn't been fixed (yet).

Nope. There's been entire threads discussing this but you're still spreading anti-Rigol FUD anyway.

Short version: Turning sin(x)/x "off" when you're in dots mode with more than channels turned on at full sample rate (and only in dots mode with more than two channels turn on at full sample rate) doesn't seem to disable it completely, it seems to use an alternate reconstruction filter instead. But we're not 100% sure because you have to go really close to the scope's bandwidth limit to even see it.

That's a completely different thing than "sin x/x is implemented incorrectly".

[Martin72]

First Time measuring at work, spi-decoding….
BIG disappointement…



It decodes from the screen, what you see is what you get, if not all data are displayed, then you couldn´t see all the data in the event table…..

Turn one channel off, all data are lost - Lecroy WS3024 and EVEN the smaller, much cheaper siglent 1104 can do it, they decode it from, the memory….and this one do it like the 1054Z 
Hope they correct it, so much memory inside and then decoding from the screen....
And:
The event table...You have to turn it on in the menu, good.
But:
If you press the "Back" button or if you turn on another channel, it disappears....
You have to go back in the menu and turn it on again....

Rigol, please fix it - Event table on/off only when you press the specific button.
And pretty please rigol....decode the data from the memory......PLEASE.

[dr.diesel]

And pretty please rigol....decode the data from the memory......PLEASE.

They never bothered to fix it in the 4000s.

[rsjsouza]

And pretty please rigol....decode the data from the memory......PLEASE.

They never bothered to fix it in the 4000s.

Nope. The DS4000 series decodes from the acquisition memory and not from the video buffer.
https://www.eevblog.com/forum/testgear/rigol-scopes-serial-decoding/

[nctnico]

First Time measuring at work, spi-decoding….
BIG disappointement…



It decodes from the screen, what you see is what you get, if not all data are displayed, then you couldn´t see all the data in the event table…..

Are you friggin' kidding? How can that happen in a scope released in 2018? If you think you will ever need decoding then return it and buy it again when Rigol has implemented decoding the full memory.

[bd139]

$10 POS aliexpress logic analyser does better than that.

[Fungus]

$10 POS aliexpress logic analyser does better than that.

My advice: Use one of those, then and leave the 4-channel, 350Mhz, 200Mpoint memory, $900 oscilloscopes for the rest of us. 

[mrpackethead]

$10 POS aliexpress logic analyser does better than that.

My advice: Use one of those, then and leave the 4-channel, 350Mhz, 200Mpoint memory, $900 oscilloscopes for the rest of us. 

Allegeledy, 1Ghz

[bd139]

$10 POS aliexpress logic analyser does better than that.

My advice: Use one of those, then and leave the 4-channel, 350Mhz, 200Mpoint memory, $900 oscilloscopes for the rest of us. 

I do exactly that already 

[Fungus]

$10 POS aliexpress logic analyser does better than that.

My advice: Use one of those, then and leave the 4-channel, 350Mhz, 200Mpoint memory, $900 oscilloscopes for the rest of us. 

Allegeledy, 1Ghz

Even if it did the whole memory they'd just complain it couldn't decode all 200Mb without any lag.

Haters gonna hate.


PS: What happens if you zoom out on a long sequence of data? The manual seems to suggest the event table should still work for all the signal on screen, ie. it's not working with just 1200 bytes of downsampled data like a DS1054Z.

Also: Does it display rubbish if you scroll the start of a byte off the screen?

[Fungus]

$10 POS aliexpress logic analyser does better than that.

My advice: Use one of those, then and leave the 4-channel, 350Mhz, 200Mpoint memory, $900 oscilloscopes for the rest of us. 

Allegeledy, 1Ghz

I forgot to mention the huge sampling rate - good for avoiding the Gibbs phenomenon. 

[0xdeadbeef]

I wonder if there is a misunderstanding what sampling from screen or memory means exactly.
When you use 2GSa/s and set the memory to 400kPoints or 200kPoints as shown in the screenshots, the sample memory is all that fits on the screen with 10divs and 20µs/div (400kPoints) or 10µs/div (200kPpoints). IMHO you could only complain about decoding memory vs. screen if the selected memory depth exceeds what is visible on the screen.

E.g. the Lecroy WS3K (and any other Lecroy)  can use the full memory independently of the sample rate, i.e. the memory can exceed what fits on the screen by far. However, even letting the decoding aside, sampling the whole memory lowers the update rate (and increases the dead time) of course. I.e. the time needed to sample 200MPoint at 2GSa/s is 100ms which means the update rate would be decreased to 10Hz even if the decoding and display wouldn't need any time at all. This is exactly why "fast" scopes as the Keysight DSOX only decode or measure what's on the screen and/or don't even let you select memory depth independently from sample rate.

Then again, even with HW based decoding it's somewhat inevitable that decoding 20MPoints will take longer than 100kPoints.
Anyway, I guess people who expect the whole memory to be sampled and decoded without any speed impact didn't really get the whole DSO thing.