R&S RTM3004 The Swiss-Knife-Oscilloscope

[egonotto]

Hallo,

I am interested in inherent noise in the most sensitive mV/div range.

Can someone please post a file with about 1 MSamples in the most sensitive range with a 50Ohm termination without input.

Nice if I get the data from different scopes

This gives me the possibility to calculate some values.


Among other things:
Distribution of the noise (amplitude and frequenz)
Mean oft the noise

Perhaps in "BIN - LSB".

And a short example in "BIN - LSB" and "TXT" (: Text file format) that I can prove whether I understand the "BIN - LSB".

Best regards
egonotto

[blacksheeplogic]

I should have one some time next week - let me know if there's anything you particularly want to see tested.

Not so much tested but your thoughts compared to the Keysight. I'm considering the upgrade from the 3000X to the 4000X.

[Rich@RohdeScopesUSA]

Hallo,

I am interested in inherent noise in the most sensitive mV/div range.

Can someone please post a file with about 1 MSamples in the most sensitive range with a 50Ohm termination without input.

Nice if I get the data from different scopes

This gives me the possibility to calculate some values.


Among other things:
Distribution of the noise (amplitude and frequenz)
Mean oft the noise

Perhaps in "BIN - LSB".

And a short example in "BIN - LSB" and "TXT" (: Text file format) that I can prove whether I understand the "BIN - LSB".

Best regards
egonotto

We don't spec the noise the for the 3000, but we do for the 4000 (it is excellent on both scopes, but the 4000-class typically has more data requested by customers so we put it in the brochure for the 4000).  As you probably know, noise is a function of bandwidth.  Our noise is excellent, especially at the most sensitive settings as you requested - we are typically 1/2 to 1/5 the noise of 4000-class scopes.  The RTA4000 (and RTM3000) have full bandwidth, hardware steps in the front end.  No bandwidth limiting as Tek/LeCroy do and no software magnification like Keysight.  To give you an idea of the noise performance, comparing 1GHz, 1mV/div, 50Ohms we are ~0.8% of full scale, while the 4000X is over 4% full scale.  And while peak-peak is not typically the best way to measure noise, it is the most noticeable.  If you have a RTM3000 or a RTA4000, it is immediately noticeable when you zoom in to 1mV/div and compare the same setting versus a scope like a 3000X or a 4000X. 

Please let me know if you have other questions.

-Rich

[egonotto]

Hello,

thanks Rich.

I read the data in the RTA4000_datasheet.pdf and I am impressed.

Nevertheless I'd like to have the noise sample.

I am interested in noise sample of other scopes  from R&S as well as from other scopes from other manufacturer.

(remembers perhaps on Heinrich Bölls „Doktor Murkes gesammeltes Schweigen“)


Best regards
egonotto

[mikeselectricstuff]

Attached are captures for 2004 and 3004
Highest sample rate & sensitivity, sample mode, single acquisition 100ksamples
2004 : 2.5Gs/sec, 2ns/div, 1mv/div
3004 : 5Gs/sec, 500ps/div, 500uV/div

TL;DR on 3004, peak-peak acquisition mode, peak-peak noise on a full-memory acquisition is slightly under 1mV, 2004 slightly over

on the 3004, even open-circuit at 1M it's well below 1.5mV

[Rich@RohdeScopesUSA]

First bug found - in spectrum/spectrogram mode, it often (but not always) clears the spectrogram display when you stop acquisition using the run/stop button

Hi Mike - we're having trouble reproducing this - any additional info you can send on when you saw it?

-Rich

[mikeselectricstuff]

First bug found - in spectrum/spectrogram mode, it often (but not always) clears the spectrogram display when you stop acquisition using the run/stop button

Hi Mike - we're having trouble reproducing this - any additional info you can send on when you saw it?

-Rich

For example :

https://youtu.be/czeM1ZbVEwM

It does it consistently the first time but not on subsequent run/stops, but I have seen it do it very consistently - in these cases it had been running for a while in spectrogram mode, and it was then clearing every time the trace was stopped, with no other controls being changed inbetween.

[egonotto]

Hallo,

thanks Mike.

Enclosed the histograms.

Best regards
egonotto

[Rich@RohdeScopesUSA]

Attached are captures for 2004 and 3004
Highest sample rate & sensitivity, sample mode, single acquisition 100ksamples
2004 : 2.5Gs/sec, 2ns/div, 1mv/div
3004 : 5Gs/sec, 500ps/div, 500uV/div

TL;DR on 3004, peak-peak acquisition mode, peak-peak noise on a full-memory acquisition is slightly under 1mV, 2004 slightly over

on the 3004, even open-circuit at 1M it's well below 1.5mV

What's the BW of your scopes again?  I think the RTB2000 is 300MHz and the RTM3000 is 1GHz? 

-Rich

[mikeselectricstuff]

Yes, max BW on both

[nazcalines]

Question for anyone that has one of these: Has the glossiness of the display been an issue? Is it glass or plastic?

[Hydron]

I believe that some (e.g. Mike) find it an issue, but it depends on where the scope is and what lighting is used.

The display is glass, and while not of retina-resolution is very nice to use (especially compared to most of the rest of the market).

Matte finish tablet screen protectors work very well to solve any glare issues, though I decided that I liked the glossy look better and took off the protector film I tried (lighting glare isn't an issue in my case).

[mikeselectricstuff]

Shiny screen is an issue if position is such that you get reflections, and shows fingermarks. A matt screen protector works well.

[Hydrawerk]

Is Keysight going to replace these 7 years old DSOX2000 and DSOX3000?

[JPortici]

given the amount of options they have i wouldn't want them to. (3000T, of course)

keysight is the only one with an almost complete set of decodes and triggers in the sub-5k area. otherwise for slightly more exotic protocols you have to go waaay up in price

but i's good to see the longer-than-usual list of options they have in these two new models is there a plan to add more in the future? (as keysight does)

[ws2812b]

@mikeselectricstuff: can you have a look what are the differences with respect to the RTB2004? I would be interested what I can hope (press Rich for  ) for with one  of the next fw releases
@Dave: did you get one as well?

[mikeselectricstuff]

I summarised the differences from the datasheets earlier in this thread, and have subsequently edited for the things that were wrong in the datasheets.
First vid should be done sometime over the weekend.

[Neganur]

Hallo,

thanks Mike.

Enclosed the histograms.

Best regards
egonotto

Hey, what software opens those bin files by the way? Doesn't seem to be the same format as the Keysight's waveform 'save as *.bin'

[mikeselectricstuff]

Quick look & teardown :

[egonotto]

Hallo,

the file 3004.BXM:
"
<?xml version="1.0" standalone="no"?>

<Database FWVersion="01.100" SaveItemType="Data">
   <Group Name="ReferenceCurveAttributes" NofSubGroups="0" NofProperties="10">
      <Prop Name="Source" Value="C1"/>
      <Prop Name="XStart" Value="-1.00000e-05" Unit="s"/>
      <Prop Name="XStop" Value="9.99980e-06" Unit="s"/>
      <Prop Name="BaseYStart" Value="-2.838e-03" Unit="V"/>
      <Prop Name="BaseYStop" Value="2.162e-03" Unit="V"/>
      <Prop Name="RecordLength" Value="100000"/>
      <Prop Name="InterleavedTraceCount" Value="1"/>
      <Prop Name="MultiChannelExport" Value="1"/>
      <Prop Name="DataFormat" Value="UINT16"/>
      <Prop Name="MSBFirst" Value="0"/>
   </Group>
</Database>
"

this tells the info one need.


The file 3004.BIN has 100000 values each 2 byte in format uint16 with MSB

As example the first two bytes in this file was:
00 8E
This mean 0x8e00
Now you need the upper 10 bit. Division by 64 gives the data:
0x8e00/64 = 0x238  = 568

0 correlate to BaseYStart and 1023 correlate to BaseYStop
If
a := (BaseYStop-BaseYStart)/(1023);
m := BaseYStart;
than
x correlate to (a*x + m).

I use octave. The little code gives you the histogram for RTM3004
"
lange = 100000;

fp = fopen("D:\\3004.BIN","rb");
kanal3004 = fread(fp,lange,"uint16");
fclose(fp);

for i = 1:lange
  kanal3004(i) = kanal3004(i)/64;
endfor

hist(kanal3004, 1024, 1);
"



Best regards
egonotto

[maginnovision]

Quick look & teardown :

Despite not needing I'm slightly jealous. Thanks for more R&S coverage, haha, I certainly can't afford to do it myself. Looks like a decent upgrade from the 2004 but if I were to buy a second scope it'd probably end up being the 3000T. Fair amount of bugs in new scopes that I can do without.

[Rich@RohdeScopesUSA]

Quick look & teardown :

Despite not needing I'm slightly jealous. Thanks for more R&S coverage, haha, I certainly can't afford to do it myself. Looks like a decent upgrade from the 2004 but if I were to buy a second scope it'd probably end up being the 3000T. Fair amount of bugs in new scopes that I can do without.

The 3000T is very nice and has a super fast update rate and some decodes we don't (yet) offer, but if/when you are ready to buy a 3000-class scope I'd definitely encourage you to get both in and compare them side-by-side.  I think you'll be impressed at how low the input noise is on the RTM3000, even at full bandwidth on small volt/div settings, the spectrum analysis capability, the 10-bit ADC, the bigger, higher resolution display, how quickly it boots and the significantly deeper memory.  A faster update rate can be nice, but you can make up for it with a little extra time.  There isn't a way to make up for lack of memory when you run out of it.

Bottom line, they are both nice scopes.  I'd encourage you (or anyone else) to try both out before buying.  In the US/Canada, our partners have lots of demos - don't be afraid to ask.

-Rich

[maginnovision]

The 3000T is very nice and has a super fast update rate and some decodes we don't (yet) offer, but if/when you are ready to buy a 3000-class scope I'd definitely encourage you to get both in and compare them side-by-side.  I think you'll be impressed at how low the input noise is on the RTM3000, even at full bandwidth on small volt/div settings, the spectrum analysis capability, the 10-bit ADC, the bigger, higher resolution display, how quickly it boots and the significantly deeper memory.  A faster update rate can be nice, but you can make up for it with a little extra time.  There isn't a way to make up for lack of memory when you run out of it.

Bottom line, they are both nice scopes.  I'd encourage you (or anyone else) to try both out before buying.  In the US/Canada, our partners have lots of demos - don't be afraid to ask.

-Rich

I bought the RTB2004 when it came out for the intro price. Even then the 3000T was on my shortlist but for the price and the features I couldn't go wrong. There are small issues with two decodes being one full duplex decode, I've had some unreproducible lockups. It is a very nice scope though and I would not change that decision. When time comes for a 3000-class scope I'll have to reevaluate but I bet it'll be a tough decision where remote interface and big higher res screen become big factors.

[nctnico]

Despite not needing I'm slightly jealous. Thanks for more R&S coverage, haha, I certainly can't afford to do it myself. Looks like a decent upgrade from the 2004 but if I were to buy a second scope it'd probably end up being the 3000T. Fair amount of bugs in new scopes that I can do without.

The 3000T is very nice and has a super fast update rate and some decodes we don't (yet) offer, but if/when you are ready to buy a 3000-class scope I'd definitely encourage you to get both in and compare them side-by-side.  I think you'll be impressed at how low the input noise is on the RTM3000, even at full bandwidth on small volt/div settings, the spectrum analysis capability, the 10-bit ADC, the bigger, higher resolution display, how quickly it boots and the significantly deeper memory.  A faster update rate can be nice, but you can make up for it with a little extra time.  There isn't a way to make up for lack of memory when you run out of it.

IMHO Keysight needs to come up with something significantly better than the Megazoom based scopes they have currently to keep up. Lack of memory is one thing (with all analog and digital channels enabled you end up with 1/8th of what it says on the badge). R&S OTOH really needs to iron out the issues in the firmware. Things which occur every now and then hint towards parallel thread concurrency issues in the firmware which are difficult to reproduce and therefore very difficult to solve.

[JPortici]

Problem is, despite the shortcomings (memory, noise, can't apply decodes after an acquisition) keysight has a scope with all the options others have only in higher class scopes.
Today.
Not maybe in the future ("we don't (yet) offer"). Today.