The Siglent SDG2042X Thread

[xygor]

Arrow's tagline is "Five years out."  Seems like an odd thing for them to say, but I guess they're being honest.

[uncle_bob]

Arrow's tagline is "Five years out."  Seems like an odd thing for them to say, but I guess they're being honest.

Hi

You sort of *hope* that's not going to be the information on the "ship date" of the parts you just ordered through them

Bob

[tequipment]

Interesting how our competitors are speaking for us on here.  I usually never talk bad about our completion because
we are very focused on our customer experience and growing our business. 

We know test equipment better than almost anyone out there.  I personally know Siglent units very well.
We made a choice not to carry Siglent units at this time.

We started this company in 2002 with 3 people.  Here is a little background on me:
http://www.tequipment.net/vice-president-biography/

We often outsell our competitors by 10 to 1 because of our customer experience, good pricing,
great product data, engineers on staff, etc.

Thanks for the business.  It is very appreciated.
Best Regards,
Evan Cirelli

Vice President and CO-Founder
TEquipment.NET

[alank2]

I will add that I've always had fair dealings with TEquipment over the years - and I am a picky customer who expects a company goes that extra mile.  I love the SDG2082X I ordered from you last year, I use it much more often than I thought I would.  I am actually pleased than neither company is digging dirt on the other.

[os40la]

Interesting how our competitors are speaking for us on here.  I usually never talk bad about our completion because
we are very focused on our customer experience and growing our business. 

Who is speaking for TE?  We are just wondering why TE no longer carries Siglent. There was a post that someone said they THINK (which means guess in my book) TE is focusing on other lines. I hardly consider that talking bad about TE.  As far as I can find TE has said nothing about the Siglent thing which is why rumors start flowing. I don't think people are looking for a detailed reason, just something would have been nice so people who just purchased Siglent's from TE don't feel confused about their dealer (at the time) no longer selling Siglent. I use TE for my supplier and have purchased Siglent  from TE a month ago. I don't care the reason why TE is not selling. I just am trying to find out what my options are if I need support.

[smarteebit]

Are we still talking about SDG2042X? 

[uncle_bob]

Are we still talking about SDG2042X? 

Hi

Hopefully now that Siglent / Tequipment has a thread all for its self, the discussion will move over there.

Based on the various bits and pieces, it's sort of a "nothing to see here, move on" kind of thing. The gear will get supported and nobody is getting dramatic about any of it.

Now, if my SDG2042X would just get here !!!! Still scheduled for later this month, just like it's always been.

Bob

[os40la]

After reading thru this tread I noticed that it seems to be ok to install (SDG2000_V200R001B01D01P17R5.ADS) on a 120mhz SDG2042X Is that correct? How many have applied this update and what are your thoughts. I see it is a one way and there is no going back.

I prefer to get my original serial number back. not the 1234567890'ish one. Any thoughts on the process to restore my serial and still keep 120mhz. Should I do this before the P17R5 update or after?.  I still have my original file.

should I restore this file and remove the <license>... line? then do the P17R5 update?

[alank2]

I'll add that when doing the latest firmware update - be patient.  My unit got to 30% and then spontaneously restarted and looked like it was stuck on the startup screen for a long time.  I wait starting to get nervous that it has bricked itself or something, but it did eventually come out of it and it had applied the update.  So, give it time, don't even think of touching that power button, and let it do its thing...

[uncle_bob]

After reading thru this tread I noticed that it seems to be ok to install (SDG2000_V200R001B01D01P17R5.ADS) on a 120mhz SDG2042X Is that correct? How many have applied this update and what are your thoughts. I see it is a one way and there is no going back.

I prefer to get my original serial number back. not the 1234567890'ish one. Any thoughts on the process to restore my serial and still keep 120mhz. Should I do this before the P17R5 update or after?.  I still have my original file.

should I restore this file and remove the <license>... line? then do the P17R5 update?

Hi

Unless you bought the 120 MHz version, it seems to be an either / or sort of thing. You can have your original serial number and original bandwidth *or* you have 12345... and the full bandwidth. Several people have suggested putting your original file back before the upgrade and then reversing the process (save the file, delete the file) after the upgrade. If anybody has spotted a change in the file from the upgrade, I missed them reporting that.

Bob

[RoadRunner]

i have ordered my SDG2042x from aliexpress in mid January , i have received software version 1.17R5 which is listed on silent website  as latest firmware, i verify that it telnet is still working  . but the old hack just remove the license fill will not working any more , silent has fixed it ( or broken it) ,currently you can not downgrade to lower version.

so if you have old version (<1.17R5) of firmware and you upgrade to 1.17R5 or later, it may switch to your original current license or not boot at all.

i have looked through telnet and NSP_system_info.xml file is not important any more , content is same as if get automatically by the app with serial number 0123456789 , application does not seem to care about it at all, i could not able to locate where they are storing the real license and serial now.

i deleted the NSP_system_info.xml when the siglent program was running, and it gets automatically created again.
but when i terminated siglent program , removed the file and restarted. it never get created and it does not matter if you have it or not program just run normal. 

but good news is,I was able to hack it with one other method ,

Hi

Unless you bought the 120 MHz version, it seems to be an either / or sort of thing. You can have your original serial number and original bandwidth *or* you have 12345... and the full bandwidth. Several people have suggested putting your original file back before the upgrade and then reversing the process (save the file, delete the file) after the upgrade. If anybody has spotted a change in the file from the upgrade, I missed them reporting that.

Bob

and you can still retain original serial number,i really want to share new hack  but don't know if i should share into public domain  ,because if silent get to know about this,  they are going to fix this in next update.

[mojoe]

My experience with the firmware upgrade.

On the very old firmware that came with my unit, I renamed the XML file and rebooted. It booted up with full bandwidth and the generic SN. I then applied firmware 15 and everything was still good.

After reading here that firmware 17 changed a few things, but that the mod was still doable, I flashed 17 without undoing the previous mod. Upon bootup, I still had full bandwidth and the generic SN.

I then logged in and copied the original XML file with my SN to the new location. I removed the one line with the license key. Rebooted and I still have full bandwidth. No need to make the folder ro manually after the mod, as it is marked ro on bootup.

I checked the siggen on a SA and the output looks very clean, even at 120 MHz. Checking the frequency with a GPSDO-locked HP counter, it was about 18 Hz high (worst case with temperature drift) at 120 MHz. If I did the math correctly, that is 0.15 ppm. That is pretty damn good for the internal reference.

(Edited to add picture. SN partially obscured.)

[CustomEngineerer]

@Gaurav, the way to hack it on the 17R5 is posted 10 or so posts up. It looks like Siglent tried to close the hole but didn't do a great job of it.

Edit: Actually, probably a little more than 10 posts back, but it is there.

[markone]

Hi to all,

have someone tried to set 1dBm amplitude for sine wave signal output on 50ohm load (50ohm output mode of course) ?

Mine 2042X, regardless the freq., sources 245mVrms (716mVpp), i I would have expected 223mVrms (1dBm).

The fanny thing is that the generator is spot on amplitude wise, if i set 223mVrms i obtain exactly that value, so the dBm scale seems way off.

[rf-loop]

Hi to all,

have someone tried to set 1dBm amplitude for sine wave signal output on 50ohm load (50ohm output mode of course) ?

Mine 2042X, regardless the freq., sources 245mVrms (716mVpp), i I would have expected 223mVrms (1dBm).

The fanny thing is that the generator is spot on amplitude wise, if i set 223mVrms i obtain exactly that value, so the dBm scale seems way off.

Error in your math, not so much in Siglent math.


---------------
Note also that many times nominal 50 ohm is not 50ohm in real world. There is tolerances but also if go out from DC there is reactances and not only pure resistance and finally what we get is sum of errors. I'm also sceptic if Siglent output is 50ohm (+/-1%) impedance over whole its frequency band. Perhaps there is this accuracy resistor but after all parasitic reactances I doubt it is true impedance with this 1% accuracy. (yes I know this is nitpicking but after we sum every single error together result may be more out what we expect. With higher frequencies it is even very fun that many 50ohm cables are not 50ohm impedance. Example Belden " semi crab" RG-58/U  is 52ohm (nominal +/- 2 ohm) and example M17/84 is nominally 50ohm +/-2ohm. Some chinese (Taiwan island) manufacturer "total junk" RG-58/U is specified 53ohm +/- 3ohm    and this is not yet super junk what can find in chinese "car garace" sellers and on the street and eBay.

After then come also oscilloscopes inputs impedances if scope have "50ohm" inputs or if used some external 50ohm terminators. This is other world of errors. 


But for "ideal world"
-1dBm in 50 ohm system is ~199mVrms
0dBm in 50 ohm system is ~224mVrms
1dBm in 50 ohm system is ~251mVrms

0dBm = 1mW (independent of impedance of course)


Here are  nice calculatos
http://www.daycounter.com/Calculators/Decibels-Calculator.phtml

Here is Mini Circuits table
http://194.75.38.69/pages/pdfs/dg03-110.pdf


Then:

...sources 245mVrms (716mVpp)...

Sinewave?

If there  716mVpp pure sine wave it is not 245mVrms  it is 253.1mVrms, not 245mVrms

[Performa01]

Oh yes, rf-Loop has already covered this topic very nicely.

I was about to say that I'm not very good in remembering numbers, nevertheless that 224mV figure for 0dBm into 50 ohms has burned into my brain (it actually is 223.6mV).

So without a calculator, +1dBm has to be some 10% higher, hence my first guess would have been about 247mV. Of course, with the help of a calculator, we get the more exact value of 250.88mV...

[Performa01]

You're confusing 1dBm with 0dBm.

0dBm is 223.6mVrms into 50 ohms, +1dBm is about 10% higher, as already stated.

EDIT: Ooops - the post I was replying to has gone - so we're all in agreement now obviously 

[markone]

Sorry, Got IT, it's my bad 

I do have to not always trust my self at late night 

Anyway, apart the DC offset that seems a little off with some low values settings (at least trusting my BM869), mine 2042X seems very precise, well within specs.

[Performa01]

Anyway, apart the DC offset that seems a little off with some low values settings (at least trusting my BM869), mine 2042X seems very precise, well within specs.

I don’t have a SDG2000X, but I think most waveform generators will be the same in this regard, i.e. very nice low noise signal sources even for DC – much better than any ‘precision’ DC power supply – but an offset error of a couple millivolts is to be expected.
I guess this is just the offset error of the high speed operational amplifiers / buffers in the circuit.

[uncle_bob]

EDIT: Ooops - the post I was replying to has gone - so we're all in agreement now obviously 

Hi

Why is it that every time *I* do those sort of errors (an alarmingly common occurrence), they get quoted and preserved forever and ever in all their un-editible glory ??

One of the features of a catalog I once wrote was a table translating things like dbm, volts rms and volts peak to peak. Conversion errors were pretty common and we would get a lot of calls from people checking out our parts.

Bob

[markone]

Anyway, apart the DC offset that seems a little off with some low values settings (at least trusting my BM869), mine 2042X seems very precise, well within specs.

I don’t have a SDG2000X, but I think most waveform generators will be the same in this regard, i.e. very nice low noise signal sources even for DC – much better than any ‘precision’ DC power supply – but an offset error of a couple millivolts is to be expected.
I guess this is just the offset error of the high speed operational amplifiers / buffers in the circuit.

The error is quite small but not constant, you cannot compensate it with a fixed offset and/or proportional factor.
As soon as a will be able to read my BM869 under Labview (the output data format seems quite ugly and i feel lazy to write the driver) i will do a measure loop to plot the error curve.
 

[markone]

Why is it that every time *I* do those sort of errors (an alarmingly common occurrence), they get quoted and preserved forever and ever in all their un-editible glory ??
Bob

I was fast enough to hidden my second mistake, anyway the first one is still available for any LOL & LULZ

The worrying thing is that i was convinced to be right  i mixed the will for one mW with dB unit annotation as the last nooob of the block. 

[uncle_bob]

Hi

The DAC used on the generator (like all DAC's) has an integral non-linearity spec. Depending on the waveform and where the max error hits, this is likely to give you about four LSB of error (as mentioned earlier in this thread). If you run through the math, that alone with no other errors will get you a millivolt. If there is an opamp or three in the chain, they likely have temperature coefficients to their offset voltage. I very much doubt they put (noisy) chopper stabilized amps in there.

A milivolt or three of variation as you switch between this and that plus a couple of milivolts after cal would be doing very well. Mine doing worse than yours also would not be un-expected. This sort of stuff will vary unit to unit.

Bob

[markone]

I was forgetting, the same behaviour apply for DC output mode.

[Performa01]

I was forgetting, the same behaviour apply for DC output mode.

Yes, of course. DC is just the programmed offset without any variation (AC signal). It is produced by a DAC and runs through all amplifiers/buffers just like any other waveform.