Siglent SDG1025 Arbitrary/function generator under some tests

[aghp]

Updated image. FW .23

I know now three versions.  .15,  .20 and .23. 

FW .23 is compatible least SDG1010 and 1025 ( I have used for  both)

.15:   TFT driving is with one word:  Terrible!
(it maybe that there is (was) totally something wrong with TFT drive in FW level becouse also contrast/wiev angle was "wrong")

.20:   TFT colors are still littlebit like candy store but not this heavy contrast issue with .15
(this image looks something like this or between .15)

.23:   TFT mainly due to reduced candy colors looks good. Also .20 version display cursor cosmetic bug is repaired.
(this have less candystore colors and visuallu more clear to look  and perhaps better for western markets (but of course one like apple and one like jackfruit)

Later I will take picture what show TFT after FW upgrade.[/i][/color]




Here:





I have made small tests with SDG1025

Now first some tests with sinewave and some modulations.

Disclaimer: Yes, I'm also biased and I know its dangers and  I try regognize it and  avoid its affect too much. Im not seller who only sell as many "nice boxes" as can. I'm too honest for this. Also maybe important point is that many of peoples are also my friends (here locally) and I need look these peoples face to face also after sell some equipment. Equipments prices are so low that my faces price is just incredible value over these prices... so..
I have my faces with  tens of years experience with electronic hobby and also tens of years experience with industrial electric and electronics design and also repair and building/developed some new things (today mainly in history).


Outside building quality is better what I expect.  Very good if think price class.
Keyboard works perfect and same for rotary encoder.
Display is ok, but I want adjust it more dim... but can not. (If I design these colors ...  after then it looks more harmonic and more expensive)
Also buttons do not need so much force that it moves whole unit... nice to use.
Also cursors below rotary knob can handle easy with same time as turn know with single hand.
Two separate channels, 14bit vertical resolution for BOTH channels, 16k for arb, 125MSa.
Pity they are not (expensive) isolated floating differential outputs...  they have common GND.

There is fan and it is not silent but, noise level is not  bad and also tone of voice is nice low.


But yes, these are made now in Siglent best factory product lines and they are really NOT Atten factories.. 
(and please do not mix it to Atten!  Atten have example ATF 20B what is nearly as piece of bullshit - one look to details in specifications and it can wonder why they make this? For who?)


Full sets of test images you can find here .
(tests are also continuing but first now this sine..)




Here some random examples:



Pictures are self-explanatory (look carefully also SA settings).
(
Note: noise level is mainly SA noise level with these settings
Note: SA reference level 0 is set for +17dBm
Note: Noise level is more low if reduce RBW but sweep time is so long that waiting time is too long for this purpose.)

2MHz 

7MHz 



Some test with modulations.
Carrier 20MHz


Turn AM on and modulation depth 30% with 800Hz mod frequency (sine, internal (adjustable)


Same as before but DSB-AM   (USB + LSB))  Double sideband (DSB) available, not separate USB or LSB sidebands and not ISB (ISB = DSB so that sidebands are independently  modulated.).
But this is better than nothing.
Sidebands well clearly over +60dBc
All available modulations: AM, DSB-AM, FM, PM, FSK, ASK, PWM, (+sweep and burst)



FM  deviation 25kHz.


Agen some sinewave, now audio area.
(signal +17dBm)

1MHz  to 100Hz (only some samples) +17dBm (@50ohm)
With more low levels there are some better results.
Also, it can see with Keithley 2015THD (20Hz - 20kHz)
Depending about signal level THD is between <0.03 - <0.1%  from 20Hz to 20kHz.

Freq affect only littlebit. Whole level range is divided to several level "bands"
It looks like THD is most low in every band upper borders, specially around 0.7Vrms.
(also when I look this with spectrum analyzer and slowly stepping down level with
small increments, it can clearly see how 2nd and 3th harmonic
level change several dB's just cross over these level bands borders.


20kHz


5kHz



1kHz



100Hz  (this is limit for SA, it can go more low but result is not very accurate)
(also if look carefully here can see small frequency offset, between SDG1000 and spectrum
are both only with they own internal references.
 

My opinion is that overall not bad result with sinewave quality for this kind of function generator.
But it need tightly classify to its own price class.
If this is 3k price, I give lot of negative arguments. 
But for normal entry-level and middle and even higher level hobby use it is good. Also it is usable in many professional use BUT it is NOT even nearly High-End "state of art" equipments. There is no any kind of competition with these.
If you need some special good for some special purpose, many old professional units give maybe best result together with price. (but you need really know these equipments, there are goods and bads.. and then "do not even touch" equipments even if they have been expensive in history. (I do not want name these! and do not ask... I do not tell.. exept in some very rare case)

SDG1000 Better than I have expect.

It have some FW bugs, minor bugs, cosmetic bugs and some more severe but overall it works well and after 5 minute learning.. UI can use very simple in most needs. I have not yet find any total freezes or crashes.

Maybe one of most bad (orange or nearly red class error, some one may classify it as fatal error)

This really BAD  FW "bug" error: if try set signal level value using dBm. (it do not work ok between -1dBm  to  +1dBm  and most bad is that if adjust with knob to this area... it really do some fatal (this is Signal generator)  it may turn output level to full! 

Attention: Before new FW update what (I hope soon!) repair this.. with sensitive electronics tests  do NOT use dBm, use Vp-p or Vrms.

To be continued... mainly there (bottom is link)

[Bored@Work]

But yes, these are made now in Siglent best factory product lines and they are really NOT Atten factories.. 

Oh come on 

Siglent is Atten. Siglent was previously called "DinYoung Technology Company, Ltd." a subsidiary of the Atten group. Sometime in 2008 DinYoung disapeared, and Siglent appeared, and history was rewritten. Some DinYoung information is still on the Siglent site, complete with an Atten logo.

Here is a comparison how DinYoung was giving their history, and how Siglent is giving the company history. DinYoung statements in italics. Even the spelling errors are the same.

< 2002, entrepreneurs of DinYoung Technology Company.,Ltd started to studies  preliminary digital oscilloscope.

> 2002, entrepreneurs of SIGLENT Technologies Co.,Ltd started to studies  preliminary digital oscilloscope.

< 2005.10, DinYoung Technology Company.,Ltd producted the sampling rate of up to 1 G, have EasyHunting and a number of innovative technologies such as the ADS7000 family, EasyHunting makes the digital oscilloscope's precision of DinYoung Technology Company.,Ltd increased 2-4 times than the same level of precision digital oscilloscope .

> 2005.10, SIGLENT Technologies Co.,Ltd producted the sampling rate of up to 1 G, have EasyHunting and a number of innovative technologies such as the ADS7000 family, EasyHunting makes the digital oscilloscope's precision of SIGLENT Technologies Co.,Ltd increased 2-4 times than the same level of precision digital oscilloscope .

< 2007,the series of ADS7000  sales in the United States and Western European , our digital oscilloscope  product sales 10,000 in 2007.

> 2007,the series of ADS7000  sales in the United States and Western European , our digital oscilloscope  product sales 10,000 in 2007.

< 2007.1, DinYoung Technology Company.,Ltd introduced ADS2000 series,it will be the first time that introduce the U disk storage technology and the introduction of printing technology to 1 G sampling of the digital oscilloscope.

> 2007.1, SIGLENT Technologies Co.,Ltd introduced ADS2000 series,it will be the first time that introduce the U disk storage technology and the introduction of printing technology to 1 G sampling of the digital oscilloscope.

< 2008.1, DinYoung Technology Company.,Ltd  begined to the researchs of  bandwidth of 500 M-1G-oscilloscope and plans to empolder products in 2010 of the bandwidth of up to 1 G of the mid-range digital oscilloscope products.

> 2008.1, SIGLENT Technologies Co.,Ltd  begined to the researchs of  bandwidth of 500 M-1G-oscilloscope and plans to empolder products in 2010 of the bandwidth of up to 1 G of the mid-range digital oscilloscope products.

< 2008.3, DinYoung Technology Company.,Ltd producted ADS1000C series, fashion aesthetic appearance, compact size and powerful features.they maked ADS1000C to sharp rise in sales.

> 2008.3, SIGLENT Technologies Co.,Ltd producted ADS1000C series, fashion aesthetic appearance, compact size and powerful features.they maked ADS1000C to sharp rise in sales.

< 2008.8, DinYoung Technology Company.,Ltd introduced real-time sampling rate of up to 2 G's ADS1000CE series.

> 2008.8, SIGLENT Technologies Co.,Ltd introduced real-time sampling rate of up to 2 G's ADS1000CE series.

At that point DinYoung disappeared. It was all suddenly Siglent, an amateurish attempt to rewrite history, and claims Siglent had never anything to do with Atten, although DinYoung had, and for all practical purposes DinYoung is Siglent, an Atten subsidiary. Now, when did Rigol win that lawsuit against Atten? Could that have been in 2008?

[eSaloooo]

compare to rigol's dg1022 or dg4062 wich is better ?

[aghp]

But yes, these are made now in Siglent best factory product lines and they are really NOT Atten factories.. 

Oh come on 

Siglent is Atten.

Yes they are same bacround but. Atten factory is NOT same as Siglent factory!

You need take bus and (or) taxi to travel cross around Shenzhen...
But here is freedom for different religions and I will respect it.
----------

There are thousends of Rigol owners, they can make tests and publish these.
There is lot of opinions in many kinds of kidplay videos and so on but nearly total lack of real data with methods what can also repeat in what ever laboratory. With some reasons opinions "this is nice and this is terrible and so on..." 

But Rigol 4000 series is different, sure better and ín different price class.

Rigol 1022A is nearly same price class or littlebit cheaper, and with reasons..
This Siglent have 2 channels and both 14 bit vertical resolution. Both channels have all modulations!  16k  memory for arb. 125MSa/s sampling.
Afaik Rigol have 14bit CH1 and 10bit CH2 and only CH1 have full set of modulations and 4k for CH1, 1k for CH1. 100MSa/s. This can not call twi channel  generator, it is somehow 1 and half.

But honestly, in many use, they are not se far away each others in features.

Rigol 4000 is different, it is clearly in upper class.

But if someone have real data about Rigol 1022 signal quality it is nice information.
I have set all tests for find problems, not for find nice looking data!

[aghp]

Some tests about jitter. 

Here SDG  use "pulse" mode.

Pulses period 100us and pulse 200ns wide.

Note: Oscilloscope I have forget to 10x but signal come with 50ohm coaxial from SDG output 1.
(so read this vertical as 500mV div)
 

First picture it can see trigger is set for one rising edge before horizontal center line.


Next picture is zoomed to this pulse edge what is image centerline (it is next rising edge after trigger edge. (Agilent have used same method))

Last picture is zoomed to this edge for 2ns/div scale.  Can not find clear jitter.  But can not sure detect and proof  anything under 300-500ps(1).  Note: Oscilloscope is set for infinite perstence for collect enough time all variations to screen.





(1) later perhaps I  check this with other kind of oscilloscope and equipments but this time this now is enough for classify this SDG if it is bird or fish.

Something more can find here.


-aghp

[eSaloooo]

would u pls show some jitter about pulse at 15megaHZ and 25ns pulse width for example~
or the jitter @ frequency is about 1.1133KHZ that is not a integ frequency~~
thanks~

[aghp]

would u pls show some jitter about pulse at 15megaHZ and 25ns pulse width for example~
or the jitter @ frequency is about 1.1133KHZ that is not a integ frequency~~
thanks~

No, I can not show 15MHz.
Pulse limit is 5MHz in all models from 5MHz to 50MHz.

(yes some sellers seems tell it have 10MHz... these sellers do not know what they sell, maybe they have never ever use this equipment in live... many of these sell just "nice boxes" as easy as they can. What value add these sellers  give to customer for this money they get between source and end user. Nothing but in worst case negative value!

With around 1kHz in pulse mode can not produce 25ns pulse width.
(setting increment is related to period (or frequency)

But I can later look with some random frequency some pulses.

Square wave can go to same frew than sine.

Its jitter is highly related to frequency.  Of course due to its basic fixed frequency DDS principle. (it is not Agilent PxP or more fine system) explained littlebit here  http://www.newark.com/pdfs/techarticles/agilent/ComparingFGPerformance.pdf

I have made square jitter tests but due to unclear manufactures specs sheet I will wait manufacturer answer  how they have defined these specifications or how these need "translate" so that they have some real meaning)

Note: Tests are continuing and I will publish the results when they are gradually completed in addition to other tasks with higher priority.

[aghp]

Here is some:

With all frequancies or periods  it use one clock cycle raster what is 8ns in this model. (125MSa/s)
It adjust pulse place but not touch pulse width. If trig from rising edge and look falling edge same pulse, no jitter. (last picture)  In advanced equpment (example some R&S or Agilent High-End models but this is under 0.5k equipment. With littlebit or more complex solution this adjusting  by jumping is eliminated more or less. 

It is well stated  by Agilent:

Conclusion
Direct digital synthesis (DDS) technology has served the function/
arbitrary waveform generator industry very well for many years. This
technology has been used by many instrument vendors to reproduce
a wide range of waveforms, including sine, square, pulse, triangle,
ramp, and user-defined arbitrary waveforms. While DDS is good in
applications requiring dynamically changing frequencies, it has some
serious drawbacks with respect to jitter and distortion.

Littlebit over 1kHz, 1us pulses.

Pulse positions jumping 8ns. (I have not find short time searching any two clock cycle jumps)
(with this principle this is not fail, it do it just as it need in this kind of system = it is natural.
(But specifications tell: Jitter 6ns + cycle 100ppm , I have seen only 8ns jumps or no jump.
Also measured over 1000 pulses also over this period max jump is 8ns (it do not cumulate)


One 1us pulse trigger rising and looked falling edge in same pulse.



Exactly same happend here what also 1kHz area.


8ns jump for adjust pulse position  or no jump.


-aghp

[nctnico]

But yes, these are made now in Siglent best factory product lines and they are really NOT Atten factories.. 

Oh come on 

Siglent is Atten. Siglent was previously called "DinYoung Technology Company, Ltd." a subsidiary of the Atten group. Sometime in 2008 DinYoung disapeared, and Siglent appeared, and history was rewritten. Some DinYoung information is still on the Siglent site, complete with an Atten logo.

Maybe but this generator is also sold by Lecroy with a Lecroy badge. It can't be that bad!

[aghp]

Maybe but this generator is also sold by Lecroy with a Lecroy badge. It can't be that bad!

Looks same, but who know if it is exactly same. I can tell only that LeCroy sell same kind of unit what is propably manufactured by Siglent ODM production for LeCroy. Is it 100% same? I do not know.

Who will try and update LeCroy FW update to Siglent model? I do not want even try.
Siglent is manufacturer. It is not "shop". 

LeCroy do not sell only pure equipment alone, they sell "total" including LeCroy customer care before sale, after sale, they have repair/service and calibration, they have peoples who will answer and help you, they can even assist people to find right solution for some needs. You go to Siglent and ask what all equipment I need for this and please show exactly total solution how we measure this in our product line and please demonstrate how you really can do it in practice. Just this point all people in Siglent stop understanding any word of english.

--
You have read all bored's comments on this forum?
On this forum it is like salt in soup. Just harmless fun. (and sime times some bits of truths like sharp needle) 

[alm]

Maybe but this generator is also sold by Lecroy with a Lecroy badge. It can't be that bad!

Did you read madworm's complaints about the Lecroy-branded Atten WaveAce 224?

[aghp]

Pulse positions jumping 8ns. (I have not find short time searching any two clock cycle jumps)
(with this principle this is not fail, it do it just as it need in this kind of system = it is natural.
(But specifications tell: Jitter 6ns + cycle 100ppm , I have seen only 8ns jumps or no jump.
Also measured over 1000 pulses also over this period max jump is 8ns (it do not cumulate)

This need correct.

Specifications (latest) tell truth.
(it is pity they copypaste other manufacturers specs  including also lies or mistakes)
SDG100 series datasheet  V1.3:
 Pulse: Jitter(pk-pk) 8ns
It is in datasheet and it is also exactly my real tests result.

[Achilles]

apart from all the Atten/Siglent discussion:

I downloaded a version of easywave to try it out and there the max Vpp. is 6V (?). Can you confirm that one?
Then there is the question if the Vpp is limited to 6V in Arb. mode or if there is a workaround for it (like editing the CSV).
At least it is nice to see that the wave is stored as csv then, what makes editing of the signal quite easy and you may not really need the software itself.
Can you just load the csv via USB stick to the Generator? That may be interesting to know as well.

Thanks for your tests and review.....

---------------------------------------------------------------------------------------------------------------

add-it: Just tried messing around with the CSV. It is possible to edit it and go over 6Vpp in that way. The program doesn't give any errors then and shows the Signal quite fine. Anyway...if I set the amplitude to 40Vpp and give some points above the 20Vpp the generator has as Max in it's specs there is still no complaint. So the software is bugged anyway...... would be interesting to know in which way

[aghp]

apart from all the Atten/Siglent discussion:

I downloaded a version of easywave to try it out and there the max Vpp. is 6V (?). Can you confirm that one?
Then there is the question if the Vpp is limited to 6V in Arb. mode or if there is a workaround for it (like editing the CSV).
At least it is nice to see that the wave is stored as csv then, what makes editing of the signal quite easy and you may not really need the software itself.
Can you just load the csv via USB stick to the Generator? That may be interesting to know as well.

Thanks for your tests and review.....

---------------------------------------------------------------------------------------------------------------

add-it: Just tried messing around with the CSV. It is possible to edit it and go over 6Vpp in that way. The program doesn't give any errors then and shows the Signal quite fine. Anyway...if I set the amplitude to 40Vpp and give some points above the 20Vpp the generator has as Max in it's specs there is still no complaint. So the software is bugged anyway...... would be interesting to know in which way

I have not yet made these tests, I have first concentrated to signal quality and accuracy, drifts etc and spectrum purity and this kind of things, also load effect and signal output impedance mismatch effects etc. And for find many these adjustments limits what are nor explained in datasheets/specifications.

When you open new editing it can not change over 6V.

After you open work you can go to "Property" menu.
There you can set vertical example to 20V.
But what it means... after waveform is in generator you adjust voltage. (but now careful... vertical is 14 bit. Do waveform so that it also have its highest or lowest point adjusted to whole vertical range, other way you do not take "all out" from 14 bit system. I think, but I do not know, that  Vpp setting in SDG  is related to waveform Vpp values are full 14bit scale. (good example how difficult simple thing is with foreign language)

I do not exactly know why there is just 6V... there may be something natural behind this but now I can not (or do not want) even guess.


But in "property" menu you can set this  V  (if set 10V it show of course -5 yo +5) what is good for your need so that it is more easy think these levels what are designing in wanted signal. (after ready, output level adjust work  whole specified area independent this "designing". (if you bhave used all 14bit)
Disclaimer: I have not tested this in practice.

But main thing is... what ever voltage you want use for your thinking process.. there is only one truth in machine. this waveform is signed 14bit vertical 16k horizontally (time) long sequential table it handle 125MSa/s. Scaling it to output voltage when user set output levels is SDG inside head ache and same for frequency.   (how to do 30MHz out from 25MHz unit...    )

I have preconception that .csv can move also with USB stick. But not yet tested do it really work.
I believe only what I have really tested. I do not directly and blindly believe any dtasheet or User manual. Exept prhaps in some very very triavial things.

[Achilles]

Yeah sure you can scale the voltage and it's fine as you just increase the mV steps, but somehow I would like taking a measured Signal and rebuild it 1:1 directly instead of measure it, scale it down to fit the 6Vpp and after building it scale it up again to fit it's original Range......at least for me these would be two unnecessary steps.
I asked Siglent via email and the replied it's just a bug in the software and it is actually capable of reading 20Vpp in the Arb. Waveforms.

So if the download of the csv really works there wouldn't be any need of the software anyway which would be a big bonus in my opinion. Just use excel, Matlab, R, or whatsoever and build your waveforms as CSV and you're ready to go...... much nicer and accurate than repainting in some fancy window

[aghp]

Yeah sure you can scale the voltage and it's fine as you just increase the mV steps, but somehow I would like taking a measured Signal and rebuild it 1:1 directly instead of measure it, scale it down to fit the 6Vpp and after building it scale it up again to fit it's original Range......at least for me these would be two unnecessary steps.
I asked Siglent via email and the replied it's just a bug in the software and it is actually capable of reading 20Vpp in the Arb. Waveforms.

So if the download of the csv really works there wouldn't be any need of the software anyway which would be a big bonus in my opinion. Just use excel, Matlab, R, or whatsoever and build your waveforms as CSV and you're ready to go...... much nicer and accurate than repainting in some fancy window

ADD for underlined:
maybe you do not need scale.. (only use SDG level setting for correct level)


Ican set 20V  in property setting (after first open  new work) then its full scale is same as full scale out from SDG (note datasheet exeptions).
 
(it seems be bug that when first open new work sheet there this 6V.)

I have tested.
I can set example 30V. I just made 5 cycle sine arb with 1kHz and 30V.
Using USB to SDG. (this time 10MHz SDG)

What happend.
It is just full 20V if I adjust SDG out 20V.
(so what ever do and use full scale (independent of scale voltages when do waveform) this full scale is also SDG full scale. naturally becouse there is 14bit DAC.)

I test ecxatly this (and some others, example this 30V)
start easywave
file  new
property window as default - ok
open property others as is but 20Vpp and - ok
draw - equation draw - new equation - 10*(sin(5*x)x)   and ok
Now you see 5 sinewave cycle and Vpp 20V (+10 and -10V)
remember this is (now) SDG1010
ok, lets test. USB to SDG and -->
arb - level 20Vpp - freq?  oh I want 1MHz so I set 200kHz and there it is..

Then I do same with easywave but now 30vpp
ok.. then 15*(sin(5*x)x)
and to SDG and full scale is just 20V.

But then, with 20V scale I make only 2Vpp sinewave.
This csv to SDG
agen output is adjustable ~0 - 20Vpp
(so, it seems have littlebit own brain... before it think it do not this...  )

[Rosendorfer]

Hi aghp

I have SGD1025 for few weeks now and as it is my first "real" generator I can't tell as an expert but from my point of view it is very nice  "little thing", and as for my testing and measurements it just works and seems to stick with datashets pretty well.
Having real 2 Channel 25MHz generator with all that modulations and possibilites is funn...

But I have question about feature I'm missing or just can't find.
Is it possible to "link" botch channels the way that we have output ie. square wave at both channels with the linked frequency but with different phases, amplitudes or duty cycles...
There is option to copy setting of one channel to second but not to link channel frequencies..
Any help ..??

And if You could send some request regarding new firmware to Siglent, about new schemes for screen colours  ... In bright day difference between green and yellow is not that big ...

Regards
Rosendorfer

[aghp]

Hi aghp

But I have question about feature I'm missing or just can't find.
Is it possible to "link" botch channels the way that we have output ie. square wave at both channels with the linked frequency but with different phases, amplitudes or duty cycles...
There is option to copy setting of one channel to second but not to link channel frequencies..
Any help ..??

Regards
Rosendorfer

Afaik, there is not this feature.

I have now used it for tests (short time), not so much UI tests and what kind of tricks there is possibe and how exactly all go becouse first I have concentrated to signal quality testings and also try find if there is something what breaks specifications (it need look with specifications sheet V1.3 other versions are obsolete).
 Ch can copy from other to other including frequency but there is no way to link freq so that other channel follow other channel freq adjust. This I have looked and try find but simply, there is no.
I miss this feature. (it was first note in my wish list "this I want".)
But it can live also without it.

Full modulation set fully separately for each channel is nice feature. And if this is not enough, then other channel can use for modulate other (From Ch out to Mod in)
Also one modulation I miss.. there is DSB-AM  (both sidebands without carrier)   but not SSB  (LSB or USB).

[nctnico]

Hi aghp

But I have question about feature I'm missing or just can't find.
Is it possible to "link" botch channels the way that we have output ie. square wave at both channels with the linked frequency but with different phases, amplitudes or duty cycles...
There is option to copy setting of one channel to second but not to link channel frequencies..

Afaik, there is not this feature.

That is pretty bad. The whole point of having two outputs is to be able to generate related signals. Since a few seconds I was planning to buy an SDG1000 series generator...

[nixxon]

I believe you can phase shift CH1 and CH2 in 1 degree increments (1/360 of full cycle). Or am I missing some part of the question here? I have only tested with sine waves though. Don't know about other waveforms.

I have testet sine wave phase shift with my SDG1020 and it seems to work  flawlessly.

nixxon

[aghp]

I believe you can phase shift CH1 and CH2 in 1 degree increments (1/360 of full cycle). Or am I missing some part of the question here? I have only tested with sine waves though. Don't know about other waveforms.

I have testet sine wave phase shift with my SDG1020 and it seems to work  flawlessly.

nixxon

Phase shift increment is 0.1degree.
There is some report old FW have 0.1degree increment but real step was 1degree.
Latest FW increment is 0.1degree and also real step resolution is 0.1degree.
(tested, not mutual)

[aghp]

That is pretty bad. The whole point of having two outputs is to be able to generate related signals. Since a few seconds I was planning to buy an SDG1000 series generator...

Of course you are able to generate related signals.

But manual frequency "sweeping" both channels together with one turn of knob, this feature there is not. But you can easy set both frequency same and even phase lock these. So for changin frequency you change it, then you push one button and change other channel freq. If you need then phase lock these, just press button. If you need 4MHz square and 4M sine.. just key in or turn 4M and push CH1/CH1 and key in or turn 4M.. you need 1.7 second time for this. if need phase lock, select it.
(how often need exatly same frequency all times manually variable  and so that signals are different, example square + sine or waveform a and waveform b.  Yes sometimes is is nice feature but in reality how often it need If really need this feature then it is better to buy some other generator what have this. Do you know good one in this price class and what also have this SDG all features? I want buy it.

I have not yet, as I have told, study features enough becouse more high priority have been signal quality reöated things. Making real lab tests is not "15 minute" work. If someone think these tests can do in few days he just have not any real experience about nothing.

So: Disclaimer. I know now somethng about signal quality and specifications but not nearly anything  about features in UI(exactly) . Just some basic things for generate signals for signal quality tests.  My answers may be wrong about some details. (good example was this misunderstooding this easywave software .csv automatic scaling to 14bit, independent of editing time voltage selections. (so softwares are more clever I first think)

This I know, both channels frequency adjust  can not directly lock to each others so that one channel follow other channel frequency changes in  real time automatically if other channel frequency is manually adjusted.
If control equipment with SCPI commands (using USB or GPIB) it can do both channels freq setting very fast and emulate locked freq between channels.
 
But there do not need many keys, it is just one key and all control change to other channel... if you have last changed freq, this selection is ready... so switching from channel to channel for freq change is fast.  They use same reference clock and they can also phase lock.

[Rosendorfer]

Hi Guys

I'm almost afraid to post anything, as seems that can be used against SDG1025....
Sorry aghp that was not my intention ... Actually opposite.. 

SDG1025 is not the best stuff possible.
But it there anything better for that price ??

...err... aghp ... attempt to measure "jitter" using Owon scope...... "one bridge too far"... . 


Regards
Rosendorfer

[aghp]

Hi Guys

I'm almost afraid to post anything, as seems that can be used against SDG1025....
Sorry aghp that was not my intention ... Actually opposite.. 

SDG1025 is not the best stuff possible.
But it there anything better for that price ??

...err... aghp ... attempt to measure "jitter" using Owon scope...... "one bridge too far"... . 


Regards
Rosendorfer

Yes, Owon is good (for me) for this becouse 1. I can get easy picture. 2. I can use 1GSa/s for 500us/div due to its full speed 10M sampling buffer for single shot realtime. I have not other scope now available what can do same. I can use my Tektronix but it have only 50k sampling buffer and maximum resolution is 2ns but Owon have 1ns. And my best old HP is even more poor for this, not with samplerate but there is less sampling buffer, so with low frequencies it have terrible low samplerate.
Of course this one example about jitter was only really one example and it is about "Pulse" function. Example for square wave function with SDG1000 is totally different case. Pulse function have only two possible placement hop for pulse timing.. 8ns time hop ot no time hop.  (one SDG1000 clock cycle) So you can see only two rising edges with 8ns time difference in picture. There is really not any kind of whole "picture" about all jitters. Sinewave phase noise/jitter is not bad, as can see in spectrum analyzer images. And tsts are just started... so more is coming. Also I try find how to use my Agilent time intervall counter for some measuremets.

[nctnico]

That is pretty bad. The whole point of having two outputs is to be able to generate related signals. Since a few seconds I was planning to buy an SDG1000 series generator...

Of course you are able to generate related signals.

But manual frequency "sweeping" both channels together with one turn of knob, this feature there is not. But you can easy set both frequency same and even phase lock these. So for changin frequency you change it, then you push one button and change other channel freq. If you need then phase lock these, just press button. If you need 4MHz square and 4M sine.. just key in or turn 4M and push CH1/CH1 and key in or turn 4M.. you need 1.7 second time for this. if need phase lock, select it.
(how often need exatly same frequency all times manually variable  and so that signals are different, example square + sine or waveform a and waveform b.  Yes sometimes is is nice feature but in reality how often it need If really need this feature then it is better to buy some other generator what have this. Do you know good one in this price class and what also have this SDG all features? I want buy it.

Sometimes I want to generate two arb signals which start at the same time (and have the same period). I have a Lecroy LWA420 (100MHz/400Ms/s) which can do that and it has a digital pattern output. I think I paid around $500 for it (second hand) including shipping.