Siglent SDG830 sync bug in sweep?

[regenfreak]

I have a new Siglent SDG830 signal generator. I have been struggling to get it to output the sync pulses in a sweep of a FM 10.7 MHz bandpass filter. On the front panel, there is a BNC connector marked "sync output".  By default, the sync pulse is turn OFF. You would have to go to page 2 of the Utility menu to turn it ON.

For sweep, the SDG830 will produce sync positive pulses of fixed width of 50ns at the start of the sweep, regardless of the sync frequency based on the datasheet. It produces sync pulses on my scopes (Siglent  SDS1202X-E and Rigol DS 1202E) no problem, as long as the sweep frequency is below 2MHz. I have done sweeps of 455kHz AM bandpass filters, no problem with getting sync pulses to display a steady waveform on my scope.

Now, the puzzling part  is that the SDG830 shows a warning message: Can't output the sync signal more than 2MHz when I do a sweep of 10.7MHz. It overwrites the sync ON and topples it to OFF whenever i try to sweep over 2MHz . I usually use 20ms sweep time. I look up the datasheet, it also states the maximum sync output frequency is 2MHz during sweep. Now you would think I am a dummy because 10.7MHz is bigger than the 2MHz sync frequency. BUT the sync pulse frequency should depends on the sweep time and NOT the sweep frequency:

the sweep time = 1/ sync pulse frequency

For example, the 100ms sweep time would generate a sweep response pattern across the screen of 10 divisions, the sync pulses would have a frequency of 1/100ms = 10Hz; For 20ms sweep, the sync pulses would have a frequency of 50Hz. I have checked the sync pulses on the scope and this is how it works.  It is nothing close to 2MHz of the synch output limit!

I know the Rigol DG1022Z also has a 2MHz maximum synch square waves as shown in the datasheet but it does not stop users sweeping at 10.7MHz. The difference between Siglent SDG800 series and Rigol DG1022Z is that:

1.Rigol DG1022Z produces sync square waves of variable wavelength that allows users to check the exact start, center or stop of the sweep pattern. So that you can see the whole response curve on channel 1 and the synch square wave on channel 2 to check the alignment on a scope. It does not stop users sweeping at frequency higher than 2MHz. Even the sync pulse not working, Ch2 can be used to generate a square waves in sync with Ch1 , as the phasing between channels can be matched.

2.The Siglent SDG830 produces a short fixed of 50ns (I measured it to be 10ns) at the start of the sweep and it stops users doing sweep frequency  of 10.7MHz. It is absolutely hopeless for FM radio alignment applications. I know it is entry level generator but it is a shocker that such limit is set quietly by Siglent for a 30MHz signal generator.

Now this is a serious limitation that the user manual does not mention. I am very disappointed with Siglent SDG830 and wish i bought the Rigol DG1022Z. I wish someone can give me more insight what to do. I believe there is a firmware bug in the Siglent SDG830  for sync output in a  sweep mode. I have spent hours and hours trying and hoping that I am wrong .

The second attachment is the picture of a 455KHz sweep of an AM fIF transformer with the Siglent SGG830 sync pulse displayed on Ch2



 

[regenfreak]

I have tried to find a workaround solution  to address the problem by using a cheap FeelTech signal generator. On the front panel of the Silgent SDG830, there is an external trigger. I feed sync square waves from the FeelTech to the Silgent external trigger and Ch2 of the scope via a T-BNC connector. It almost works but there is a very slow drifting of the sync waves relative to the sweep pattern on the scope.  I can set Siglent from internal to external trigger on the menu.  I speculate that the hardware (analogue to digital converter) of the external trigger introduces a small phase offset. The attached photo is 10.7MHz sweep with a sweep time of 20ms (50 Hz) sweep square waves.

BTW the external trigger button and light are not doing anything on the SDG830. They are probably for higher end model of the SDG800 series.

[uargo]

I'm trying to reproduce your problem on my Siglent SDG1032x and (SDS1104x-e), and the first thing is that I don't know how you can achieve an amplitude variation with a sweep? the sweep only varies frequency, not amplitude.
But hey:
1 - I have generated a 10.7Mhz sine wave
2 - I add a sweep, center frequency 10.7Mhz, Span 1Mhz, Source internal, direction up.
and there is a clear cut, there is no continuity.
But if I put Up_down in the direction, there is continuity.

See if that was the problem

[regenfreak]

The variation in amplitude is due to the resonance response of a LC bandpass filter. You cannot replicate it without a filter.

I believe Siglent SDG1032x  is free from this limiting problem with the SDG830.

I am going to try to do a synchronised sweep using my cheapo and ancient two channel FeelTech FY2200 on its own. If it can beat the SDG830 that is 6 six times  price of the FY2200, i throw the SDG830  out of the window (no, only joking).

[uargo]

The variation in amplitude is due to the resonance response of a LC bandpass filter. You cannot replicate it without a filter.

I believe Siglent SDG1032x  is free from this limiting problem with the SDG830.

I am going to try to do a synchronised sweep using my cheapo and ancient two channel FeelTech FY2200 on its own. If it can beat the SDG830 that is 6 six times  price of the FY2200, i throw the SDG830  out of the window (no, only joking).

I supose the filter

Are you put Up_down in the direction parameter of sweep???

put here various captures of all parameters of sweep please


PD:
My sweep menu have 2 pages, in second page have "direction" parameter, hve 3 options: Up, Down and Up_Down, put Up_Down.

Another thing, why do you use sweep if you can use BODE PLOT
Do you know how to use bode plot? The oscilloscope alone generates a frecuency filter response curve automatically, you only need to connect the SDG8300 via usb to the oscilloscope.


https://siglentna.com/application-note/bode-plot-filter-oscilloscope-generator/

[regenfreak]

Another thing, why do you use sweep if you can use BODE PLOT
Do you know how to use bode plot? The oscilloscope alone generates a frecuency filter response curve automatically, you only need to connect the SDG8300 via usb to the oscilloscope.

Both my Siglent  SDS1202X-E and Rigol DS 1202E do not have bode plot. Bode plot is only available in "posh" high-end scopes. However, poor man like me can create "pseuso bode plot" using log sweep with a function generator.

The parameters of sweep depends on  the frequency response, bandwidth and type of your filters. There is no one rule for all.

[regenfreak]

To investigate the small phase drift of my previous work around solution. I have reversed my set-up: my junkie FeelTech FY2200 as a sweep generator and the Siglent SDG830 as the sync pulse generator. The ancient and super cheap FY2200 can do sweep only up to 2MHz and the  shortest sweep time of 1s ( 1Hz  sync pulse)s. So I demonstrate the idea using a 455kHz AM IF filter sweep.  It locks the sweep response pattern absolutely dead on with no phase drift Cheap and cheerful it works!

The newer and low cost  FeelTech FY6900 can sweep up to 60MHz. So do i have to spend a lot of money to get FM frequency of 10MHz sweep done? I dont think so.

[regenfreak]

The final experiment is to see if the cheapo generator can do both sweep and self trigger sync using  its two channels. Now it struggles to keep the pattern stationary. I dont know why other than blaming it being so cheap. Note the less smooth response pattern, it is due to the fact that the FeelTech uses bigger frequency increment steps in a linear sweep.

[uargo]

Oh man sorry but you have bad lucky, read this:

SDS1000X-E User Manual
WWW.SIGLENT.COM   
Bode Plot II
The  bode  plot  application  for  the  SDS1000X-E  (only  4  channels  scope  supports  thisfunction)  controls  an  external  arbitrary  waveform  generator(only  Siglent’s   arbitrary   waveform  generator  can  be  supported)  to  sweep  a  sine  wave  across  a  range  of  frequencies while measuring the input to and output from a device under test (DUT). At each frequency point, gain and phase are measured and displayed on the Bode plot.

only 4  channels  scope  supports bode  plot 
and SDG1032X works fine doing sweep

PD:  SDG830 is suported in SDS1104X-E for bode plot

[regenfreak]

Oh man sorry but you have bad lucky, read this:

Not not much bad luck. I feel short changed with the Siglent SDG830. Seriously a 30MHz generator cannot output sync pulses of 10Hz if the sweep carrier frequency is above 2MHz? What a joke! 

 If you look at the specs of the cheap FY6800 or FY6900, it has much better specs and more features  than the SDG830 in many aspects. Ok the user interface and display of FY6800 sucks, I purchased the FY6900 and returned it to Amazon. Now I am thinking of buying the FY6900 again. 

[regenfreak]

Finally my workaround method described has enabled me to perform  perfectly synchronised with the cheapo FeelTech sweeping at 10.7MHz.

In the photo attached, the Feeltech FY2200 generates the sync square waves of 140ms wavelength or 7.14285Hz to fill up the screen of the Silgent which has 14 divisions. So the start, center and end frequencies are automatically aligned.

For a Rigol scope, a wavelength of 120ms or 8.3333Hz sync waves should be used for 12 divisions on the time base.

The 2MHz sweep sync limit seems to be a either a bug or a deliberate "dumb down" feature to make users to spend more money for high ended models.

[rf-loop]

Finally my workaround method described has enabled me to perform  perfectly synchronised with the cheapo FeelTech sweeping at 10.7MHz.

In the photo attached, the Feeltech FY2200 generates the sync square waves of 140ms wavelength or 7.14285Hz to fill up the screen of the Silgent which has 14 divisions. So the start, center and end frequencies are automatically aligned.

For a Rigol scope, a wavelength of 120ms or 8.3333Hz sync waves should be used for 12 divisions on the time base.

The 2MHz sweep sync limit seems to be a either a bug or a deliberate "dumb down" feature to make users to spend more money for high ended models.

Btw, why you need sync from sweep generator.
Oscilloscope have many trigger functions but of course it need bit knowledge abd experience how to use these.
I am now thousands of km far from my workshop so I can nolt demonstrate all possibilities. But I remember I have used it long time ago not exactly with just same what you do but somehow nearly. But I have done sync using scope when sweeper did not have sync out. And SDS1002X-E have really lot of different trigger and other features for this.

I can give one wink...  when your SDG sweep look very carefully this waveform and specially where it return back to start frequency and start new sweep. Look what all things in this change is what you can use for trigger using special trigger functions and also perhaps some time limits what there can use example for define just perfect slope and time what happen when sweep retract and start new sweep (for this you need use one direction sweep).

Perhaps yoy can show scope screen image just about sweep retract--- just this when sweep is end and it change to start freq. There happen things what can easy trig. One possible is find what slope there is most special for produce example voltage change speed defined trig ot some other. Use one direction sweep.

[regenfreak]

[Btw, why you need sync from sweep generator.

It is because of the problem described in my first post if you read it. I thought the title of this thread is self-explanatory.

Do i have too much time on my hands to create problem for myself? Shall i go out to pay more money to buy higher end signal generator if what i got is not working as expected? I am not a gearhead with a deep pocket.

SDS1002X-E have really lot of different trigger and other features for this

It can be done with manual adjustment by setting a high threshold level in order to stabilize the sweep pattern. It is time consuming and tedious to stablize sweep patterns without synch signal. I say it is  hit and miss. I dont know if you know about alignment of FM IF stages and discriminator valve tube / transistor radios. Every time, you tune the ferrite slugs, the peak amplltudes of the sweep response pattern varies and you have to vary the trigger threshold level.  I would describe it as fiddly. With S-curve for the ratio detector, this approach is even more fiddly if the ratio detector is badly out of alignment.

[uargo]

I really don't think you're understanding us.

You put the wave generator at 10.7Mhz, you activate the sweep function, you put the start and end frequencies (or center frequency and span, it's the same), you put the swwep time you want ...

AND IN THE DIRECTION PARAMETER BY DEFAULT IT PUTS UP, SO NO! YOU CHANGE IT TO: UP_DOWN and by doing this it synchronizes the wave itself, without doing ANYTHING ELSE. You dont need synchronizes nothing

when you put up_down the beginning and the end of the wave is synchronized automatically making a continuous smooth sweep without jumps.

a perfectly timed continuous repetition

[regenfreak]

I really don't think you're understanding us.

As I explained before your Siglent  SDS1202X-E has different features and menu from my Siglent SDG830.  It is like comparing an orange to an apple. I should stop beating the dead horse.

I think people assume I am clueless. I shall stop repeating myself

[TurboTom]

As long as there is enough amplitude at the beginning of the sweep, a single scope channel will do with any frequency generator that's capable of sweeping a sine wave. If the level is too low, you can use either a second channel or the external trigger input fed with the sweep signal from the FG before the DUT.

Then you select trigger type to "pulse" and "greater than". The time to dial in depends on the frequency range that you want to observe. If you want to analyze a 10.7 MHz filter, you may sweep from 9 to 12MHz. So the pulse duration to enter would be slightly shorter than half the period at the beginning of your sweep, i.e. (2*9E6 Hz)^-1 = 55 ns, so you may want to dial in 50ns. If you set the trigger level to the centerline of your sine wave, the scope should now trigger directly on the first wave of your sweep if the time base is configured correctly to match the sweep time.

Hope this explains the procedure half-way understandably... 

[uargo]

I really don't think you're understanding us.

As I explained before your Siglent  SDS1202X-E has different features and menu from my Siglent SDG830.  It is like comparing an orange to an apple. I should stop beating the dead horse.

I think people assume I am clueless. I shall stop repeating myself

ok I guess when you say SDS1202X-E you mean my SDG1030x generator, and I've been looking through the manual and it seems like you're right:

Your SDG830 has the "Direction" parameter but only the UP or Down options
http://www.siglenteu.com/wp-content/uploads/dlm_uploads/2017/10/SDG800_UserManual_UM02008-E02C-1.pdf

sorry man

[paul]

I have an SDG1020 which looks very similar and has the same menu layout as the SDG830, except you are missing the Trig Out option in the menu and the trig out BNC socket on the rear panel.
It looks like the SDG800 series is missing the Trig out feature, on my SDG1020 this gives the square wave out put you describe as on the Rigol generator in your first post, high at the start of the sweep and low at the mid point.
I also have a sync output which outputs a narrow pulse and is limited to 2MHz.

My first thought was that you dont have a bug just a missing feature, however it is a feature I am surprised would be missing, after reading the manual for the SDG800.

    “For a Sweep, when the sweep starts, the Sync Signal becomes TTL Level High. The Sync frequency equals the specific Sweep time.”

I agree this looks like a bug.

Also, if you look at the second picture in uargo last post you can see the `Trig Out` option, which I have on my generator, in this case its set to Off.

Paul.

[uargo]

Another thing, it can handle the SDG830 via (usb or LAN) with SCPI commands, if you know some programming you can handle the generator from your pc, making it work for you, and do what you tell it, to "simulate" a bode plot.
In fact, you can simultaneously control the SDS1202x-e oscilloscope also by (usb or LAN) and SCPI commands.

If you know how to program, you can create a program that, for example, does:

1 - SDG830: generates a 10.6Mhz signal at 4vpp
2 - SDS1202X-E: read the signal and tell me frequency and vpp
3 - PC: I read and store data
4 - SDG830: generates a 10.7Mhz signal at 4vpp
5 - SDS1202X-E: read the signal and tell me frequency and vpp
6 - PC: I read and store data
7 - SDG830: generates a 10.8Mhz 4vpp signal
8 - SDS1202X-E: read the signal and tell me frequency and vpp
9 - PC: I read and store data
etc etc
and when you have for example 1000 readings make a graph and show on the PC

SDG830 programming guide
https://www.siglenteu.com/wp-content/uploads/dlm_uploads/2019/12/SDG_Programming-Guide_PG02-E04A.pdf
includes examples in C++, VB6, MATLAB, LabVIEW and Python

SDS1202X-E programming guide
https://www.siglenteu.com/wp-content/uploads/2020/04/ProgrammingGuide_PG01-E02C.pdf
includes examples in C++, VB6, MATLAB, LabVIEW, c# and Python

I only give you information, maybe it is not what you need, or you do not know how to program, I do not know, but I will tell you I will give you this information in case it helps you.

[paul]

If you look on page 67 of the SDG800 manual you can see that the Sync output changes its signal depending on the application, so in sweep it should output a “frequency equals the specific Sweep time”. Which is exactly what the OP was expecting it to do.

This is a bug, we just need another SDG800 user to confirm this.

[regenfreak]

This is a bug, we just need another SDG800 user to confirm this.

Now you understand it . I contacted Siglent support and I got no reply.

Also I contacted the re-seller of Siglent here in the UK. They got confused the sync signal frequency with the main sweep carrier frequency in their reply. The irony is that they thought i was confused. 

[regenfreak]

I have repeated the sweep with the cheapo Feeltech as an ext sync using the Rigol  DS1202Z-E. You see the Siglent SDS-1200X-E sweep display looks much nicer I am not sure why there was a small phase drift the first i tried it. Now it is perfectly in sync. I have repeated this with 20ms sweep time, it is all fine and dandy.

Now Rigol  DS1202Z-E sounds like Sigent SDS-1200X-E? Is Rigol taking the Mickey of Siglent when they name this 200MHz model?

The next two photos are the alignment of the ratio detector of a FM transistor radio. Note the S-curve and the center of the screen is at 10.7MHz precisely. There is no need to use marker frequency.

[rf-loop]

[Btw, why you need sync from sweep generator.

It is because of the problem described in my first post if you read it. I thought the title of this thread is self-explanatory.

Do i have too much time on my hands to create problem for myself? Shall i go out to pay more money to buy higher end signal generator if what i got is not working as expected? I am not a gearhead with a deep pocket.

SDS1002X-E have really lot of different trigger and other features for this

It can be done with manual adjustment by setting a high threshold level in order to stabilize the sweep pattern. It is time consuming and tedious to stablize sweep patterns without synch signal. I say it is  hit and miss. I dont know if you know about alignment of FM IF stages and discriminator valve tube / transistor radios. Every time, you tune the ferrite slugs, the peak amplltudes of the sweep response pattern varies and you have to vary the trigger threshold level.  I would describe it as fiddly. With S-curve for the ratio detector, this approach is even more fiddly if the ratio detector is badly out of alignment.

Just this: "It can be done with manual adjustment by setting a high threshold level in order to stabilize the sweep pattern. It is time consuming and tedious to stablize sweep patterns without synch signal. I say it is  hit and miss."

Do you think scope really do not have any more sophisticated trigger functions. So I will write bit more.
If I am home I just make it for you with images and test setup and trigger setting guide.




Starting from 60's  I have aligned IF filters in many kind of radios, including also professional communication radios etc etc and lot of other things in hobby and also later in profession.  Of course in history and in work there was R&S SWOB etc tools and of course later more sophisticated tools.  In hobby side mostly analog scopes and analog commercial or even DIY sweep gens controlled by scope.
Old times with analog scope in this case it is important to have ext trigger from generator.  With analog scopes I have used method where analog scope generate ramp to control VCO what sweep syncronized with scope horizontal sweep ramp. Many Analg scopes have this output.

Also, ancient digital scopes need in this case trigger from sweep generator (in this case from SDG830 working as sweep gen) But you do not have ancient digital scope, even when is is not expensive mode.

Modern digital scope, like Siglent SDS1000X-E series (in home I have them many but I can not demonstrate to you because I am thousands of km far from home) have lot of different trigger functions what also works quite accurately because these scopes have true digital trigger engine (deeply inside heart getting data from ADC directly and before any decimation) instead of ancient digital scopes analog side pathway trigger system, what in this price class was mostly poor)

Perhaps you do not need trigger pulse from SDG830 at all.  It is fully possible you can set oscilloscope to sync to sweep output start. ( this depends  SDG830 waveform details when it goes to sweep upper freq and then jumps to start freq. And this was why I ask you take screenshot just from this moment when it retract so we can see this special waveform for help you to find right trigger method and adjust it and also there is reason why I ask to use single direction sweep as example just from down to up)

And when it start down to up you can also set oscilloscope horizontal reference position example to left of display or example 5 dif left from center. After then it is fixed there and not center so it do not move if you change t/div. 

But first work is adjust oscilloscope trigger so that it do enough perfect sync directly from 10M sweep output.
For this you need perfect image about signal what details happen just when it is reached upper limit and retract for start new sweep. There is key what you need for set scope to sync.  I do not know what there is but  sure there is fast frequency hop.

Now because your main signal go via filter under working it is very possible signal what is out from pass band is too weak for reliable trig for example to this frequency change instead of level change alone you need split signal from generator to two pathway. One where is constant level to Ch2 and other to DUT and then to Ch1.
Now you set all trigger things in Ch2 and signal for looking is Ch1.

Also for make this frequency step trigger bit more easy, specially if filer under work is very narrow, you can still sweep more wide and spread this image in scope. Trigger can be out from display.

Now there is one small "cosmetic" problem in this method. If it sweep, just example for thinking" one time in second oscilloscope do not trig exactly every second some times it keep 2 second pause, in worst case 3 second. Why.
Again, reason can find when looking this signal detail when frequency hop from end of sweep freq to start freq. It happen randomly what ever phase in signal. But when your sweep interval is more high and work is adjust filter it can do even when it some times drop out some single sweep.
You need learn what all trigger functions are and how these parameters are need adjust and how they affect.
Just normal trigger as rising edge and trigger level is not for this, not at all.

And again repeating, need know exactly this sweep retract waveform details in all cases also when frequency jump phase is moving. After then you can easy design what parameters you can use and set for trigger system.

This sync to sweep is fully oscilloscopes work and with advanced trigger functions you perhaps do not need at all separate trigger from sweep generator. 

I know learning is hard work...

If I have, say example normal IF filter for commercial wide FM reception and IF is this usual 10,7MHz  and I have SDS1x02X-E and SDG830  and normal other needed small things I can do normal center and width alignment etc.  It is not even difficult. Of course due to oscilloscope linear vertical scale there need forget some things where need high dynamic range at once on display. So stop band check and possible alignment is not for this. (With SDS1x04X-E models can do lot of more... but it is totally other story)

First break your locked thinking circle road and try other roads without hitting head again and again to some same rock what is in this case missing trig out  (sync) signal from generator even if it is bug or not.  Generator have sweep output where is signal what change. Why not use these changes for sync. Your scope have quite impressive amount of trig functions. One example is that you can define slope angle if there exist just special enough different angle in signal once when sweep retract.

And, do not use this signal to trig what goes thru filter under work to scope for display bode. Split signal after generator also to other channel what you use only for produce sync (trig) so you are not in troubles when signal is low due to out from filter pass band. This splitted signal you have more stable signal level for detect example signal some edge rising or falling slope, what is one possible method. (Use other main channel, not ext trig channel, it can nolt produce this sync)

One possible method. You know this trigger function where you set two levels and then time window. Then ignal signal need cross over these  points inside defined time window and direction. Of course it is more nice if there can AND / OR / XOR cascade these trig rules or more advanced method but there is not. But in this case I think one rule is enough. With previously told disadvantages.


Also for perhaps more nice bode like image you can use detector probe... no need tell to you because you work with radios.
It give more usable dynamic range for you for working without overload scope front end amplifier. (you can then set bode straight bottom (what is now your vertical center line) to display area bottom and then you have some more dynamic range (6dB).)


Because I can not demonstrate it now for give evidence I will stop here. If you still not understand perhaps some ones can help.

I am sure you need only SDG830 and SDS1x02X-E.
You do not need separate trig (sync) signal from generator. Just produce this sync inside your scope  using features what are there and all you need for this is generator sweeping output signal. It need some skills but no need any extra tools or investments.

[regenfreak]

quotFirst break your locked thinking circle road and try other roads without hitting head again and again to some same rock what is in this case missing trig out  (sync) signal from generator even if it is bug or not.  Generator have sweep output where is signal what change. Why not use these changes for sync. Your scope have quite impressive amount of trig functions. One example is that you can define slope angle if there exist just special enough different angle in signal once when sweep retract.e

I know how to trigger manually without sync signal in AM and FM radios. In any case, I have found a simple and reliable solution to my problem without spending more money.   But nobody has been able to answer my main question of this post  if SDG830 2MHz sweep limit with sync is a bug or a manufacturer's deliberate limit to make it more "useless" for sweep. Manufacturers want consumers to spend more money to buy models with more features and functions.

To trigger manually without external or internal sync, set the threshold level about 90% peak amplitude of the sweep pattern (top half of the response curve is bell shaped). It works certainly  but It can be time-consuming to get the threshold level right  if your IF filter is completely out of alignment. In addition, threshold has to be adjusted every time you adjust the IF transformers. If have a wonky or badly distorted response pattern to start with, it is not always working.  If you have ever done FM/AM alignment, or alignment of FM discriminator or ratio detector, you would know the difficulties. Without sync signal displayed on ch2, you do not know where is the alignment center with a reference point on the scope. In the old days, you use a marker frequency created out of a hetero beat. With digital scope, you use a reference square wave to see where is the start, center and end frequency of the scope.   

I built valve tube AM radios and I am building a FM stereo tube radio with solid state decoder now. I also repair vintage valve tube radios as a hobby. Therefore i know how analogue scope trigger works.

This guy is an expert on scopes. He works for Tek. He knows exactly what he is talking about. He should be someone to learn from:



This is how FM alignment should be done using sweep and sync square waves out of a Rigol DG1022 (old version of SG1022Z):

[regenfreak]

I am pleasantly surprised to receive an email from a Siglent support engineer. He has reproduced the same bug and reported the problem to their R&D team.