Author Topic: new killer scope in town - a true game changer from R&S - RTB2002 & RTB2004  (Read 415979 times)

0 Members and 1 Guest are viewing this topic.

Offline BrianSchmalz

  • Contributor
  • Posts: 19
  • Country: us
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

I'll hook the scope directly to my PC, which is Gb Ethernet, tomorrow, and repeat the experiment.
 

Offline BrianSchmalz

  • Contributor
  • Posts: 19
  • Country: us
Also, I'm *totally* digging the Remote Front Panel. It feels faster than the actual UI in some ways, and I find using the mouse on the web browser to be very quick, smooth, and intuitive. It just feels really snappy this way. I've never used a scope's web interface that I liked before. This is totally different.
 
The following users thanked this post: Someone

Offline gslick

  • Frequent Contributor
  • **
  • Posts: 286
  • Country: us
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

Does the web browser interface also let you save the waveform data in REAL and UINT 8, 16, 32 binary data formats? The formats are described in section 9.8.1 Transfer of Waveform Data in the User Manual. For the same waveform data are the binary formats smaller and faster to transfer than the ASCII format?
 

Offline corax

  • Contributor
  • Posts: 30
  • Country: us
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

I'll hook the scope directly to my PC, which is Gb Ethernet, tomorrow, and repeat the experiment.

I played around a bit with the SCPI commands documented in the manual; over ethernet, the scope accepts SCPI commands on port 5025.
I used netcat on a Linux machine to pull the whole segment acquisition memory for a channel in REAL format:

echo -e "FORM REAL\nCHAN1:DATA:POIN MAX\nCHAN1:DATA?" | nc -q -1  192.168.200.60 5025 > out.bin
(scope is on address 192.168.200.60)

With the acquisition record length set to 10MSa, the resulting out.bin file is 40MB and transferred in about 25 sec.  I'm connected with 1Gbps ethernet.

The SCPI setup commands let you select REAL, UINT, or ASCII output formats as well as selecting either all segment data or only what's on the screen.

« Last Edit: April 06, 2017, 05:37:04 am by corax »
 
The following users thanked this post: NA5WH, vokars

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 17419
  • Country: nl
    • NCT Developments
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

I'll hook the scope directly to my PC, which is Gb Ethernet, tomorrow, and repeat the experiment.

I played around a bit with the SCPI commands documented in the manual; over ethernet, the scope accepts SCPI commands on port 5025.
I used netcat on a Linux machine to pull the whole segment acquisition memory for a channel in REAL format:

echo -e "FORM REAL\nCHAN1:DATA:POIN MAX\nCHAN1:DATA?" | nc -q -1  192.168.200.60 5025 > out.bin
(scope is on address 192.168.200.60)

With the acquisition record length set to 10MSa, the resulting out.bin file is 40MB and transferred in about 25 sec.  I'm connected with 1Gbps ethernet.
So that is 12.8Mbits/s.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline corax

  • Contributor
  • Posts: 30
  • Country: us
So that is 12.8Mbits/s.

I don't know how efficient netcat is, though. Might have to write a small C program to find out where the bottleneck lies.
 

Offline JoHr

  • Regular Contributor
  • *
  • Posts: 81
  • Country: de
So that is 12.8Mbits/s.

I don't know how efficient netcat is, though. Might have to write a small C program to find out where the bottleneck lies.

I can get the same ~12MBit/sec with my RTB.
The law of conservation of bugs states that the total amount of  bugs of an isolated system remains constant. Bugs can neither be created nor destroyed; rather, they can be transformed from one form to another.
 

Offline vokars

  • Contributor
  • Posts: 29
  • Country: de
So that is 12.8Mbits/s.

I don't know how efficient netcat is, though. Might have to write a small C program to find out where the bottleneck lies.

Thank you very much for your tests. After ASCII and real: Could the real conversion be the bottleneck here? I remember that these embedded devices are better with integer than with real. But depends on the mcu.
 

Offline corax

  • Contributor
  • Posts: 30
  • Country: us
Thank you very much for your tests. After ASCII and real: Could the real conversion be the bottleneck here? I remember that these embedded devices are better with integer than with real. But depends on the mcu.

Doesn't look like it.  Trying the same test with UINT 8,16,32 results in about the same bitrate for the transfer by the looks of things.
Could still be limited by netcat though.
 
The following users thanked this post: vokars

Offline JoHr

  • Regular Contributor
  • *
  • Posts: 81
  • Country: de
I testet with various format and been reading 20MSa channel data.

FormatData amountData rate
UINT820MByte12.3MBits/s
UINT1640MByte12.8MBits/s
REAL80MByte12.5MBits/s

For remote postprocessing REAL data a pretty comfortable, but with UINT you can reduce your processing time a lot cause of the lower transmission time.
You  can use CHAN:DATA:YINC?    CHAN:DATA:YOR?   to convert the data on your local PC.
« Last Edit: April 06, 2017, 08:20:58 am by JoHr »
The law of conservation of bugs states that the total amount of  bugs of an isolated system remains constant. Bugs can neither be created nor destroyed; rather, they can be transformed from one form to another.
 
The following users thanked this post: vokars

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 11967
  • Country: gb
    • Mike's Electric Stuff
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

Does the web browser interface also let you save the waveform data in REAL and UINT 8, 16, 32 binary data formats? The formats are described in section 9.8.1 Transfer of Waveform Data in the User Manual. For the same waveform data are the binary formats smaller and faster to transfer than the ASCII format?
I think it may be faster and/or more convenient to transfer data using the USB MTP mode - you can just drag files out of the folder in various formats
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline JoHr

  • Regular Contributor
  • *
  • Posts: 81
  • Country: de
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

Does the web browser interface also let you save the waveform data in REAL and UINT 8, 16, 32 binary data formats? The formats are described in section 9.8.1 Transfer of Waveform Data in the User Manual. For the same waveform data are the binary formats smaller and faster to transfer than the ASCII format?
I think it may be faster and/or more convenient to transfer data using the USB MTP mode - you can just drag files out of the folder in various formats

I tried the USB-TMC Connection ... and get about 13-15MBit/sec on my PC.
The law of conservation of bugs states that the total amount of  bugs of an isolated system remains constant. Bugs can neither be created nor destroyed; rather, they can be transformed from one form to another.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 11967
  • Country: gb
    • Mike's Electric Stuff
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

Does the web browser interface also let you save the waveform data in REAL and UINT 8, 16, 32 binary data formats? The formats are described in section 9.8.1 Transfer of Waveform Data in the User Manual. For the same waveform data are the binary formats smaller and faster to transfer than the ASCII format?
I think it may be faster and/or more convenient to transfer data using the USB MTP mode - you can just drag files out of the folder in various formats

I tried the USB-TMC Connection ... and get about 13-15MBit/sec on my PC.
Quick test shows about 8MBytes/sec exporting binary trace data via USB MTP
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 
The following users thanked this post: vokars

Offline GeorgeOfTheJungle

  • Super Contributor
  • ***
  • Posts: 1748
  • Country: pl
Oh dear, oh dear, the more videos I see of the RTB2004 the more I like it. Please stop because I'm about to pull the trigger. Thank you very much.
« Last Edit: April 06, 2017, 03:27:26 pm by GeorgeOfTheJungle »
int main (void) { while (1) fork(); }
 

Offline nctnico

  • Super Contributor
  • ***
  • Posts: 17419
  • Country: nl
    • NCT Developments
OK, got it. So I set CH1 to take up all 20 Msamples of memory, did a trigger, then went to the save/recall page, and downloaded the entire 206MB text file. I'm only on 100 Mb Ethernet here, and it took about 200 seconds. (about 1MB/s) This was using Chrome.

I'll hook the scope directly to my PC, which is Gb Ethernet, tomorrow, and repeat the experiment.

I played around a bit with the SCPI commands documented in the manual; over ethernet, the scope accepts SCPI commands on port 5025.
I used netcat on a Linux machine to pull the whole segment acquisition memory for a channel in REAL format:

echo -e "FORM REAL\nCHAN1:DATA:POIN MAX\nCHAN1:DATA?" | nc -q -1  192.168.200.60 5025 > out.bin
(scope is on address 192.168.200.60)
I tried the same on my GW Instek. It needs 5 seconds to transfer 10Mpts in 16 bits format (approx 20MB of data) so that is around 30Mbit/s.
« Last Edit: April 06, 2017, 01:18:33 pm by nctnico »
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 
The following users thanked this post: vokars

Offline BrianSchmalz

  • Contributor
  • Posts: 19
  • Country: us
For anyone who has one of these scopes - have you tried running the probe comp app? When I run mine on CH1, I can't get it to work - it says either the signal was not detected, or the amplitude was not high enough.

When I run it on CH2, 3 or 4, I don't get any errors, but no matter how I turn the three adjustments on the probe, the signal never looks close to square - it's very rounded. I feel like I'm missing some important step here. (Yes, I did connect the probe to the ground pin and the P0 pin of the pattern gen)

=====

I did a factory reset, and now CH1 is not giving me errors. However, I'm still not able to get a square shape with any of the probes that came with the unit.
« Last Edit: April 06, 2017, 01:28:13 pm by BrianSchmalz »
 

Offline Octane

  • Regular Contributor
  • *
  • Posts: 95
  • Country: us
My probe comp worked fine on all four probes on all four channels. I only was a little disappointed that the otherwise nice step by step instruction does not tell you what the two screws on the BNC connectors side are exactly, it only shows turning the one that is further away from the scope.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 11967
  • Country: gb
    • Mike's Electric Stuff
For anyone who has one of these scopes - have you tried running the probe comp app? When I run mine on CH1, I can't get it to work - it says either the signal was not detected, or the amplitude was not high enough.

When I run it on CH2, 3 or 4, I don't get any errors, but no matter how I turn the three adjustments on the probe, the signal never looks close to square - it's very rounded. I feel like I'm missing some important step here. (Yes, I did connect the probe to the ground pin and the P0 pin of the pattern gen)

=====

I did a factory reset, and now CH1 is not giving me errors. However, I'm still not able to get a square shape with any of the probes that came with the unit.
Do you have probes switched to x1 ?
Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline BrianSchmalz

  • Contributor
  • Posts: 19
  • Country: us
For anyone who has one of these scopes - have you tried running the probe comp app? When I run mine on CH1, I can't get it to work - it says either the signal was not detected, or the amplitude was not high enough.

When I run it on CH2, 3 or 4, I don't get any errors, but no matter how I turn the three adjustments on the probe, the signal never looks close to square - it's very rounded. I feel like I'm missing some important step here. (Yes, I did connect the probe to the ground pin and the P0 pin of the pattern gen)

=====

I did a factory reset, and now CH1 is not giving me errors. However, I'm still not able to get a square shape with any of the probes that came with the unit.
Do you have probes switched to x1 ?

Uh, yup. Thanks Mike. That was my problem. In 10x the prob comp process worked just fine - all good now!

You would think they could have had that as a text 'hint' in the steps that walk you through the process.

Also, if the colors of each channel are fixed, why not put color printing next to each channel's BNC so that it's clear which probe to plug in where without having to turn on the channel to see it's little button light up with its color. Hmm.
 

Offline mikeselectricstuff

  • Super Contributor
  • ***
  • Posts: 11967
  • Country: gb
    • Mike's Electric Stuff
For anyone who has one of these scopes - have you tried running the probe comp app? When I run mine on CH1, I can't get it to work - it says either the signal was not detected, or the amplitude was not high enough.

When I run it on CH2, 3 or 4, I don't get any errors, but no matter how I turn the three adjustments on the probe, the signal never looks close to square - it's very rounded. I feel like I'm missing some important step here. (Yes, I did connect the probe to the ground pin and the P0 pin of the pattern gen)

=====

I did a factory reset, and now CH1 is not giving me errors. However, I'm still not able to get a square shape with any of the probes that came with the unit.
Do you have probes switched to x1 ?

Uh, yup. Thanks Mike. That was my problem. In 10x the prob comp process worked just fine - all good now!

You would think they could have had that as a text 'hint' in the steps that walk you through the process.
Yes, also look at what the probe settings were
Quote
Also, if the colors of each channel are fixed, why not put color printing next to each channel's BNC so that it's clear which probe to plug in where without having to turn on the channel to see it's little button light up with its color. Hmm.
Cost. And you should have put the colour rings on the probe lead. (would be nice if these were actually the right colours)


Youtube channel:Taking wierd stuff apart. Very apart.
Mike's Electric Stuff: High voltage, vintage electronics etc.
Day Job: Mostly LEDs
 

Offline BrianSchmalz

  • Contributor
  • Posts: 19
  • Country: us
Cost. And you should have put the colour rings on the probe lead. (would be nice if these were actually the right colours)

Absolutely I did! Almost the first thing after I pulled them out of the box. But, which colors go to which BNC? That's what I'm getting at - without turning on the scope, and turning on each of the 4 channels, there are no graphic hints on the front face as to which color probe goes into which BNC.

I guess I can just draw it on with marker or something.  :D
 

Offline MarkL

  • Supporter
  • ****
  • Posts: 1528
  • Country: us
Can you tell how to get the circular segmented recording option without changing option-id-resistors or being sued/shot by Keysight?
It is/was available on the X2000 and X3000 series via the license hack as option SGMC.  You can read about it in the hack thread, starting around here:

  https://www.eevblog.com/forum/testgear/dsox2000-and-3000-series-licence-have-anyone-tried-to-hack-that-scope/msg338637/#msg338637

As far as I know they haven't shot anyone yet for hacking.
 

Offline salviador

  • Regular Contributor
  • *
  • Posts: 61
  • Country: it
    • https://www.youtube.com/user/mancio92M
what is it the meter widget from the menu?
someone has a screen?
 

Offline BrianSchmalz

  • Contributor
  • Posts: 19
  • Country: us
what is it the meter widget from the menu?
someone has a screen?

You can add a little volt meter window (which you can move around the screen) to each channel. It can display DC, AC RMS, or DC + AC RMS volts. See this screenshot.
 

Offline JoHr

  • Regular Contributor
  • *
  • Posts: 81
  • Country: de
what is it the meter widget from the menu?
someone has a screen?

You can add a little volt meter window (which you can move around the screen) to each channel. It can display DC, AC RMS, or DC + AC RMS volts. See this screenshot.

... with up to four measurements ( one each channel).
The law of conservation of bugs states that the total amount of  bugs of an isolated system remains constant. Bugs can neither be created nor destroyed; rather, they can be transformed from one form to another.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf