Author Topic: The Siglent SDG2042X Thread  (Read 264274 times)

Qw3rtzuiop and 2 Guests are viewing this topic.

Offline billfernandez

  • Regular Contributor
  • *
  • Posts: 141
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #400 on: June 12, 2016, 12:47:19 pm »
High-performance sampling system with 1.2 GSa/s sampling rate and 16-bit vertical resolution.

Innovative TrueArb technology, based on a point-by-point architecture, supports any 8 pts ~ 8 Mpts Arb waveform with a sampling rate in range of 1 uSa/s ~ 75 MSa/s.

Where does the limitation of 75 MSa/s come from, while it uses a sampling system with 1.2 GSa/s?

For which signals does it use 1.2 GSa/s? And for which does it use less?

Why you would want to use a variable sampling rate that goes down to 1 uSa/s?
Isn't it better to just always use the highest available sampling rate for best results?

Or is it because of memory limitations, when you want to store longer periods?
But then again, isn't 1 uSa/s a bit on the low side? :)

Siglent tends to combine random good-sounding specs into a single passage, making it hard to figure out where they all come from.  Here are partial answers from my personal notes:

16-BIT RESOLUTION AND 1.2GS/s:

These apply to the DDS (Direct Digital Synthesis) mode of generating arbitrary waveforms.

IN DDS mode the unit pulls 14-bit samples from memory at the rate of 300 million samples per second, then between every two adjacent samples it inserts three interpolated samples that it creates on the fly, resulting in four 16-bit values OUT for every one 14-bit sample IN.  The resulting sample stream is sent to the  digital-to-analog converter at the rate of 1.2 gigasamples per second (four times the rate at which samples are pulled from memory).  This allows the generator to output much cleaner waveforms.  Note that no interpolation is done between the first and last sample if the generator is set to loop the sample set.

75Ms/s, 1uS/s:

These apply to the TrueArb mode of generating arbitrary waveforms.

In TrueArb mode the unit will draw samples from memory at a variable rate from 1 sample per 1 million seconds to 75 million samples per second and deliver them (with interpolation?) to the DAC.

Note that the unit will only let you set a "frequency" of up to 4.577,637kHz.  But frequency of what?  If you divide 75M by 4.6k you get 16,384.  So apparently when you set the frequency you are telling the unit how fast to cycle through 16,384 samples.

So what happens if you have more than 16K samples in your sample set?  I don't know.
 
The following users thanked this post: electrolust

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #401 on: June 12, 2016, 12:51:11 pm »
Good explanation of how "TrueArb" waveform generation works (as well as other methods) can be found in this technical brief from Tektronix.

http://www.tek.com/document/fact-sheet/understanding-signal-generation-methodologies

As far as why you wouldn't always want to use the highest sampling rate. TrueArb generators work on a variable clock, and displays every point in the waveform. To control the frequency, you change the sample rate. While the SDG2000X generators will allow you to switch to frequency adjustment while in TrueArb mode, its really just changing the sample rate to change the output frequency.

This next part is how I understand it, but could certainly be wrong. The max sample rate in TrueArb mode is 75MSa/s. The 1.2GSa/s is when in DDS mode, where the actual sample rate is 300MSa/s, but 4x interpolated to get the effective sample rate of 1.2GSa/s. So the sample rate is much higher in DDS mode, but with TrueArb mode you are guaranteed to have every point displayed accurately. Each method has its pros and cons. Which mode you use will depend on what you are trying to do.

Edit: Good explanation of using interpolation and filtering for signal generation from National Instruments: http://sine.ni.com/np/app/main/p/ap/mi/lang/en/pg/1/sn/n17:mi,n21:42/fmid/3014/
« Last Edit: June 12, 2016, 01:16:48 pm by CustomEngineerer »
 
The following users thanked this post: electrolust, smarteebit

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #402 on: June 12, 2016, 01:05:21 pm »
Note that the unit will only let you set a "frequency" of up to 4.577,637kHz.  But frequency of what?  If you divide 75M by 4.6k you get 16,384.  So apparently when you set the frequency you are telling the unit how fast to cycle through 16,384 samples.

So what happens if you have more than 16K samples in your sample set?  I don't know.

The max "frequency" it will let you set in TrueArb mode is dependent on the number of samples in the waveform. If the waveform has 375points, the max "frequency" you can set it to is 200KHz. 75M / 375 = 200,000. So it would be the same if you had more than 16K samples, the max "frequency" you could set would be lower than 4.6KHz. You are correct, that frequency is essentially telling it how fast to cycle through the points in memory. Your actual output frequency will depend on how many cycles are in your waveform. So if you have an 375 pts waveform, thats composed of 2 cycles of a sine wave, running at 75MSa/s, your actual output frequency would be 400KHz.

Pretty sure those numbers are correct, but definitely open to correction if I'm being stupid.

Edit: I know 375pts was an odd number to pick, but the SDG2000X have several different sized builtin demo arb waveforms, and 375pt was the smallest one. If I hadn't been working on the lab computer this weekend I would have generated an 8pt signal with the EasyWave software and probably used that for my example.
« Last Edit: June 12, 2016, 01:19:41 pm by CustomEngineerer »
 

Offline MasterTech

  • Frequent Contributor
  • **
  • Posts: 862
  • Country: 00
Re: The Siglent SDG2042X Thread
« Reply #403 on: June 12, 2016, 03:16:57 pm »
Also note that the minimum pulse rise time is said to be 8.4ns in the specs (thats in arb mode), but its not, it can go down to 4.5ns in dds mode :-+
« Last Edit: June 12, 2016, 04:53:29 pm by MasterTech »
 

Offline bson

  • Supporter
  • ****
  • Posts: 1539
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #404 on: June 12, 2016, 03:25:09 pm »
Oh, so it doesn't interpolate between the last and first sample in a recurring DDS waveform?  This explains a few things.  That's also braindead IMO.  This means a repeating waveform has to be oversampled into N*4+1 samples and decimated to N+1 before uploading to the generator, when you really want an N sample long repeating waveform.   :palm:
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1404
  • Country: no
Re: The Siglent SDG2042X Thread
« Reply #405 on: June 12, 2016, 11:30:32 pm »
How big are the sample rates with the Siglent SDG1025?

Is it there more consistent and 125 MSa/s both for DDS mode and ARB mode? In that case the SDG1025 performs better than the SDG2042X in ARB mode? Really?

Does the SDG1025 also do interpolation?

Regarding no interpolation on SDG2042X between the last and first sample in a recurring DDS waveform.
Could this be fixed in a firmware update? Maybe this can be reported to Siglent?
 

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #406 on: June 13, 2016, 06:18:08 am »
The SDG1025 doesn't have TrueArb, only DDS. It uses DDS for generating both the builtin waveforms as well as the arbitrary waveforms.

Edit: Good explanation of SDG1000 DDS vs SDG2000X TrueArb sample rates from rf-loop can be found here
https://www.eevblog.com/forum/testgear/best-arbitrary-waveform-generator-for-around-$500/msg756726/#msg756726
« Last Edit: June 13, 2016, 06:26:20 am by CustomEngineerer »
 

Offline Timpert

  • Contributor
  • Posts: 43
  • Country: nl
Re: The Siglent SDG2042X Thread
« Reply #407 on: June 13, 2016, 06:26:51 am »
Could it be possible that the FPGA passes its data to the DAC chip at 300 MSPS instead or 1.2 GSPS? The DAC is equipped with configurable on-chip interpolation filters up to 8x, so getting the sample rate up to 1.2 GSPS requires configuriguration of the on-chip interpolation to 4x and feeding data in at 300 MSPS per channel. Perhaps the folks at Siglent used another round of upsampling in the FPGA to brush up the arb waveform signal fidelity, so that it is usable right up to the Nyquist frequency? Having to carry "only" 300 MSPS (or even 150 MSPS in true ARB mode with the DAC set to 8x upsampling) to the DAC instead of 1.2 GSPS simplifies hardware design and FPGA workload, so I'd more or less expect Siglent to do something like this in a cost-concious product.
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1404
  • Country: no
Re: The Siglent SDG2042X Thread
« Reply #408 on: June 13, 2016, 06:46:51 am »
Thanks for the links to the documents from Tektronix and National Instruments.
I will go through these documents, and come back if I still have questions :)
« Last Edit: June 13, 2016, 07:07:04 am by pascal_sweden »
 

Offline Siglent America

  • Regular Contributor
  • *
  • Posts: 220
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #409 on: June 17, 2016, 10:25:07 pm »
Hello all.

Another nice new firmware update on the Siglent SDG2000X generators. It adds some new features - such as the ability to Combine CH1 & CH2 - as well as fixes some bugs.. Here is the FW link and what it does:

    You can download the firmware from the below URL:
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/firmware/SDG2000X-P21R2.rar

    The new firmware details:
-------------------------------------------
1. Added Waveform Combination function, to support combining CH1 and CH2 from internal to generate complex signals such as sine + noise and two-tone. The set path: Utility | Output Setup | Wave Combine
2. Added mode selection: “Phase-Locked”and “Independent”. In “Phase-Locked” mode the generator resets both channels every time the frequency is changed to maintain synchronization automatically.
In “Independent”mode the generator will not reset any channel when the frequency is changed; the phase difference between channels is random. The set path: Utility | mode
3. Fixed several bugs
        a) Can not remove files with illegal characters
        b) In burst mode, press the soft key for cycles, it changes to infinite right off the bat
        c) External modulation input inverted compared to convention
        d) Can not remember the clock source setting
        e) Some unexpected glitches when enabling the output or change the parameters
 
The following users thanked this post: CustomEngineerer, nugglix, MasterTech

Offline uncle_bob

  • Supporter
  • ****
  • Posts: 2425
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #410 on: June 17, 2016, 10:33:58 pm »
Hello all.

Another nice new firmware update on the Siglent SDG2000X generators. It adds some new features - such as the ability to Combine CH1 & CH2 - as well as fixes some bugs.. Here is the FW link and what it does:

    You can download the firmware from the below URL:
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/firmware/SDG2000X-P21R2.rar

    The new firmware details:
-------------------------------------------
1. Added Waveform Combination function, to support combining CH1 and CH2 from internal to generate complex signals such as sine + noise and two-tone. The set path: Utility | Output Setup | Wave Combine
2. Added mode selection: “Phase-Locked”and “Independent”. In “Phase-Locked” mode the generator resets both channels every time the frequency is changed to maintain synchronization automatically.
In “Independent”mode the generator will not reset any channel when the frequency is changed; the phase difference between channels is random. The set path: Utility | mode
3. Fixed several bugs
        a) Can not remove files with illegal characters
        b) In burst mode, press the soft key for cycles, it changes to infinite right off the bat
        c) External modulation input inverted compared to convention
        d) Can not remember the clock source setting
        e) Some unexpected glitches when enabling the output or change the parameters

Hi

As an owner of one of these generators, let me thank you for posting this. It is definitely a plus to get both a heads up on updates *and* an explanation of what to expect when I do the update.

Thanks!

Bob
 

Offline nugglix

  • Regular Contributor
  • *
  • Posts: 201
  • Country: de
Re: The Siglent SDG2042X Thread
« Reply #411 on: June 17, 2016, 11:38:35 pm »
As an owner of one of these generators, let me thank you for posting this. It is definitely a plus to get both a heads up on updates *and* an explanation of what to expect when I do the update.

Thanks!

+1

Really nice to know in advance what has been changed.
 

Offline bson

  • Supporter
  • ****
  • Posts: 1539
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #412 on: June 18, 2016, 02:30:40 am »
Lots of good fixes there - can't wait to get it installed!
 

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #413 on: June 18, 2016, 02:41:24 am »
Hello all.

Another nice new firmware update on the Siglent SDG2000X generators. It adds some new features - such as the ability to Combine CH1 & CH2 - as well as fixes some bugs.. Here is the FW link and what it does:

    You can download the firmware from the below URL:
http://www.siglentamerica.com/USA_website_2014/Firmware&Software/firmware/SDG2000X-P21R2.rar

    The new firmware details:
-------------------------------------------
1. Added Waveform Combination function, to support combining CH1 and CH2 from internal to generate complex signals such as sine + noise and two-tone. The set path: Utility | Output Setup | Wave Combine
2. Added mode selection: “Phase-Locked”and “Independent”. In “Phase-Locked” mode the generator resets both channels every time the frequency is changed to maintain synchronization automatically.
In “Independent”mode the generator will not reset any channel when the frequency is changed; the phase difference between channels is random. The set path: Utility | mode
3. Fixed several bugs
        a) Can not remove files with illegal characters
        b) In burst mode, press the soft key for cycles, it changes to infinite right off the bat
        c) External modulation input inverted compared to convention
        d) Can not remember the clock source setting
        e) Some unexpected glitches when enabling the output or change the parameters

Thank you Siglent! This sounds like a great update. Adding a very nice new feature that I hadn't specifically heard anyone ask for and making several changes that I have seen people asking for. Very good indication that you are paying attention to us. Clearly you are listening to your users.
« Last Edit: June 18, 2016, 02:43:01 am by CustomEngineerer »
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 15340
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: The Siglent SDG2042X Thread
« Reply #414 on: June 18, 2016, 05:51:01 am »
Lots of good fixes there - can't wait to get it installed!
Some of those fixes ^^^are no doubt from your efforts.  :-+
Thanks.
Avid Rabid Hobbyist
 

Offline pascal_sweden

  • Super Contributor
  • ***
  • Posts: 1404
  • Country: no
Re: The Siglent SDG2042X Thread
« Reply #415 on: June 18, 2016, 06:00:48 am »
Dear Siglent! Thanks for the good work!

What about the fix for no interpolation on SDG2042X between the last and first sample in a recurring DDS waveform.

Could this be fixed in the next firmware update? :)
 

Offline JohnG

  • Regular Contributor
  • *
  • Posts: 182
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #416 on: June 18, 2016, 07:45:15 am »
Installed new firmware, no problems so far.

Thanks very much for keeping this up to date. The output sync and the bug fixes are very helpful to me, much appreciated. Based on my experience and this kind of support, my employer now has 4 of these generators, and is has purchased some additional equipment as well. Please pass this along to the powers that be at Siglent, because I would like this kind of support to continue.

John
 

Online alank2

  • Super Contributor
  • ***
  • Posts: 1984
Re: The Siglent SDG2042X Thread
« Reply #417 on: June 18, 2016, 08:02:07 am »
Updated mine without issue.
 

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #418 on: June 18, 2016, 08:19:07 am »
Dear Siglent! Thanks for the good work!

What about the fix for no interpolation on SDG2042X between the last and first sample in a recurring DDS waveform.

Could this be fixed in the next firmware update? :)

I'm still not sure that this is a bug, it could be working as designed for whatever reason. Has anyone actually reported this to Siglent as a bug? If so, what was their response?
 
The following users thanked this post: mtdoc

Offline bson

  • Supporter
  • ****
  • Posts: 1539
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #419 on: June 18, 2016, 12:25:05 pm »
        d) Can not remember the clock source setting
Still not seeing this fixed... I plug in the external standard, set the clock to external, power cycle, and it comes back up with the source set to internal.  So still need to reach in behind the instrument to unplug it each time I power off the instrument. :(

 
The following users thanked this post: tautech

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #420 on: June 18, 2016, 12:33:24 pm »
Dear Siglent! Thanks for the good work!

What about the fix for no interpolation on SDG2042X between the last and first sample in a recurring DDS waveform.

Could this be fixed in the next firmware update? :)

Alank2 gives a good explanation of how believes the the SDG2000X to behave in DDS mode earlier in this thread. See

https://www.eevblog.com/forum/testgear/the-siglent-sdg2042x-thread/msg794951/#msg794951

If he is correct, and what he says seems to make sense with what I have seen, then the non interpolation between the first and last sample would seem to make sense. Though without asking Siglent I don't think we can know for sure if its a bug or doing what it supposed to.
 

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #421 on: June 18, 2016, 04:35:06 pm »
I updated the firmware with no issues. Also wanted to add, the new helpful displays when enabling/setting the new combine waveform output and phase locked mode are very helpful and very nicely done. Good job Siglent!
 

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #422 on: June 18, 2016, 04:42:31 pm »
Heads up for those of you that have hacked your SDG2000X, Siglent have added a username/password required for logging in through telnet potentially closing the hole. If you have already hacked it before the update you are probably fine, but I couldn't say for sure as mine wasn't hacked.
 
The following users thanked this post: tautech

Online tautech

  • Super Contributor
  • ***
  • Posts: 15340
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: The Siglent SDG2042X Thread
« Reply #423 on: June 18, 2016, 04:59:12 pm »
Heads up for those of you that have hacked your SDG2000X, Siglent have added a username/password required for logging in through telnet potentially closing the hole. If you have already hacked it before the update you are probably fine, but I couldn't say for sure as mine wasn't hacked.
Thanks for the "heads up".  :scared:

Just checked to see if member analogNewbie who found the hole is still an active member and yes he was online yesterday.

Let's see if he spots this FW improvement
Avid Rabid Hobbyist
 

Offline CustomEngineerer

  • Frequent Contributor
  • **
  • Posts: 459
  • Country: us
Re: The Siglent SDG2042X Thread
« Reply #424 on: June 18, 2016, 05:04:17 pm »
Good news for those of you who have hacked your generators. You can easily downgrade to the previous firmware version and the telnet username/password restriction is gone. So its definitely not completely closed for good. I have no idea if you hack it before upgrading to the newest firmware if it remains hacked, sorry.

Edit: new text below

For me the new features added in the most recent firmware are definitely worth it.

Rob, have you heard anything about Siglent every actually selling the bandwidth upgrades? I currently have no use for >40MHz, but if I ever get to that point that was something I was hoping would finally be available. And it seems like with each new firmware Siglent attempts to close the hacking hole, at least partially, so hacking may not be an option at that time. Though I would be perfectly happy paying for the extra bandwidth, just not a completely new generator to get it.
« Last Edit: June 18, 2016, 05:07:57 pm by CustomEngineerer »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf