The Siglent SDG2042X Thread

[markone]

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.

I do not remember the exact numbers right now but i had the feeling that the displacement was sometime higher than theoretical DAC resolution step, sure not linear as a DAC scale.

[Performa01]

Very likely there is a separate DAC just for the offset, and its output is added to the main signal.

That way the main DAC can always have full resolution, i.e. all 14 (or whatever) bits can be dedicated to the signal waveshape and you still get a respectable offset range.

[uncle_bob]

Very likely there is a separate DAC just for the offset, and its output is added to the main signal.

That way the main DAC can always have full resolution, i.e. all 14 (or whatever) bits can be dedicated to the signal waveshape and you still get a respectable offset range.

Hi

Given that they have a very "affordable" instrument here and that DC performance is not a key spec, they may do something pretty simple (R2R on a 4 wire port). They also may ignore it completely. I suspect that every nickel (let alone dollar) gets watched on one of these designs. In this case, I'm glad they did. it's got a lot of bang for the buck.

Bob

[Performa01]

Hi Uncle Bob!

Well, from what I've seen so far (e.g. at the beginning of this thread) DC accuracy, while not stellar, is still pretty decent.

EDIT: Having said that, a waveform generator is just a waveform generator and not a process calibrator, so (DC) accuracy needs not be stellar in my book anyway

And you just need to find out what the resolution of the DC output is. Then if you know the max. output range (I think it is +/- 5V) you just need to divide 10V by the resolution and you get the number of steps required. Even if the resolution were just 10mV, you'd need a 10 bit DAC, so no way to do it with a 'homebrew' 4 wire R2R DAC

[markone]

Very likely there is a separate DAC just for the offset, and its output is added to the main signal.

That way the main DAC can always have full resolution, i.e. all 14 (or whatever) bits can be dedicated to the signal waveshape and you still get a respectable offset range.

Almost certainly it's made in that way, just out of curiosity i will watch again Dave's tear down in order to spot DACs chip marking.

Sure i agree, lot bang for the bucks, i'm using it on a daily basis, works good and it's pretty.

Not yet expanded, i will do soon 

[Performa01]

Sure i agree, lot bang for the bucks, i'm using it on a daily basis, works good and it's pretty.

It appears to be nice bit of kit indeed and I might be seriously tempted if I’d need one.

There are just two things I don’t like so much:

1)   There is no front connector for a sync signal. Yes, very likely there is one on the back, but I hate it when things that I actually use a lot are on the backside.
2)   The output amplifier is rather weak, as I understand it delivers full power only up to 25MHz.

Both are no huge issues, but still a pity…

[markone]

The output amplifier is rather weak, as I understand it delivers full power only up to 25MHz.

What would be your main use ?

[Performa01]

What would be your main use ?

Nothing special. Just as a general signal source for various tests and experiments.

When testing the maximum signal handling capability and overload behaviour of amplifiers/frontends I sometimes would have wished for even more than the standard ~3.5Vrms that I can get from 10Vpp into 50 ohms. And I would not want to have only half that at frequencies above 25MHz, even though I know that pretty much any signal generator will have a frequency limit where the output power drops. But 25MHz is a little on the low side.

[billfernandez]

There are just two things I don’t like so much:

1)   There is no front connector for a sync signal. Yes, very likely there is one on the back, but I hate it when things that I actually use a lot are on the backside.
2)   The output amplifier is rather weak, as I understand it delivers full power only up to 25MHz.

FYI:

o Yes there's a sync connector on the back.  Sync only works up to 1MHz (which seems really low to me).

o The output into 50Ohms is 10Vpp up to 20MHz, then 5Vpp above that (squarewaves go up to 25MHz, sinewaves go up to 40MHz).

o Into "hi-Z" the max output voltage is double the above.

[JohnG]

Update to my dual output pulse synchronization issues:

The issues I had earlier with output pulses temporarily losing the delay setting between channels while changing parameters seems to be fixed with the last firmware revision, so thanks, Siglent.

New problem discovered:
When enabling a channel, the DC offset appears to be enabled before the pulse output, and intermittently as well. This can be a big problem when you are starting up. I hope this can be fixed as well. Generator set at pulse output, low value zero, high value 10V (10Vpp, 5V offset), 3.33% duty cycle, 70 kHz, using Chan 1 out. The generator is driving a 50 ohm load, so one should see a 0-5V pulse.

The first waveform triggers when output is enable. There are a bunch of glitches, and note the amplitude of 2.5V (equal to the expected offset value into 50 ohms).

The second shows the same over a long time scale, showing that the output looks like the DC offset of 2.5V for almost 70 ms (!) before the signal kicks in.

The third waveform is just the signal.

John


As an aside, I think it is evidence that the dc offset is produced separately from the signal output.

John

[billfernandez]

On the Siglent SDG1000 and SDG800 thread user Deni mentioned that if you have both channels running, and make a frequency change on one channel, then a glitch appears on the other channel:

https://www.eevblog.com/forum/testgear/the-sdg1000-and-sdg800-thread/msg843225/#msg843225

User Paul explained an easy way to show this:

https://www.eevblog.com/forum/testgear/the-sdg1000-and-sdg800-thread/msg848349/#msg848349

I wondered if the SDG2042X had the same behavior (issue, bug, anomaly...) and it does!  Here's how to see it:
o Make channel A emit a 1Hz sinewave.
o Make channel B emit a 2Hz sinewave.
o Select the microhertz digit on the channel B frequency setting.
o Turn the knob 1 detent.
o The signals on both channels will immediately dive towards zero, emit two small sawtooth waves, then restart their sinewaves.

The first screenshot below shows the initial 1 and 2Hz sinewaves.

The second screenshot shows the glitch.

The third screenshot shows that if the two waveforms are 180 out of phase, then you do the glitch, they restart 180 degrees out of phase.

[Performa01]

o Yes there's a sync connector on the back.  Sync only works up to 1MHz (which seems really low to me).

Ouch! 1MHz isn't great indeed.
Does that mean the sync output is a train of 1µs pulses instead of a squarewave?

[billfernandez]

o Yes there's a sync connector on the back.  Sync only works up to 1MHz (which seems really low to me).

Ouch! 1MHz isn't great indeed.
Does that mean the sync output is a train of 1µs pulses instead of a squarewave?

It is a 1.5V, positive-going pulse, with a fixed pulsewidth of 500nS.  (so when the channel output gets up to 1MHz the sync signal is a squarewave.)

[uncle_bob]

o Yes there's a sync connector on the back.  Sync only works up to 1MHz (which seems really low to me).

Ouch! 1MHz isn't great indeed.
Does that mean the sync output is a train of 1µs pulses instead of a squarewave?

It is a 1.5V, positive-going pulse, with a fixed pulsewidth of 500nS.  (so when the channel output gets up to 1MHz the sync signal is a squarewave.)

Hi

Looking at it another way:

The sync comes out every thousand or so output samples. It comes out every hundred or so sine waves at max frequency. It may be an output from a MCU ( = jitter) or it could be from an FPGA ( = not real flexible). It is doubtful that they put in a dedicated sync channel with it's own giga sample sort of speeds just for the sync.

Yes, 10 or 20 MHz still seems like something you could easily do, even with those constraints.

Bob

[markone]

It is a 1.5V, positive-going pulse, with a fixed pulsewidth of 500nS.  (so when the channel output gets up to 1MHz the sync signal is a squarewave.)

Not in my case, amplitude is correct (0-5V) and with sweep function i get a 50% duty cycle square wave synced to  sweep ramp, see attachment.

[billfernandez]

It is a 1.5V, positive-going pulse, with a fixed pulsewidth of 500nS.  (so when the channel output gets up to 1MHz the sync signal is a squarewave.)

Not in my case, amplitude is correct (0-5V) and with sweep function i get a 50% duty cycle square wave synced to  sweep ramp, see attachment.

Sorry, I had a 50Ohm terminator on the scope input I used to view the sync waveform. 

[Siglent America]

Here is a video on generating a 4-phase signal using two SDG2000X generators:

[uncle_bob]

Here is a video on generating a 4-phase signal using two SDG2000X generators:

Hi

Nice demo !!

Ok, you have a sync out on the top generator and a trigger in on the bottom generator. I assume that there is a delay on the sync out and a delay on the trigger in as well. Not a big deal at 60 Hz. It likely would be a big deal if you did this same thing at 1 MHz.

So, what is the recommended method of taking care of the delay? One approach would be to trigger both generators off of a third. That seems a bit equipment intensive. I'm guessing there is another way to do this. Ideally it would not involve measuring and calibrating the phase on the second set of signals.

Thanks!

Bob

[nctnico]

60Hz isn't very demanding but I doubt you'll be able to do the same at 100kHz with these generators.

[neslekkim]

"Seems that the frequency counter leaks a 10 MHz signal to it's input."

We did identify a cross-talk issue that is causing this leakage to the frequency counter port. The factory tells me they are working on the problem.

Thanks for your comments.

Any information about fix for this?, I would assume this is an hardware issue and thus will only be fixed on newer units?
From which serialnumbers will this be fixed?

[uncle_bob]

"Seems that the frequency counter leaks a 10 MHz signal to it's input."

We did identify a cross-talk issue that is causing this leakage to the frequency counter port. The factory tells me they are working on the problem.

Thanks for your comments.

Any information about fix for this?, I would assume this is an hardware issue and thus will only be fixed on newer units?
From which serialnumbers will this be fixed?

Hi

Equally on the same topic:

Assuming it is a hardware fix (it's the ribbon cable from the back panel ... ) what happens with in warranty existing generators? Will they be updated / upgraded / fixed?

Yes, it is a bit early to start digging into this. Right now, it might not matter to most. To the few who might be about to cancel an order for one, it could matter quite a bit

Bob

[alank2]

I may be totally wrong in this, but a great deal of frequency counters are so sensitive that any strong signal around them will be enough to cause it to count.  Do we know that the 10M signal is indeed leaking into the counter input?  Or is the counter just sensitive and 10M is the strongest signal around?

[uncle_bob]

I may be totally wrong in this, but a great deal of frequency counters are so sensitive that any strong signal around them will be enough to cause it to count.  Do we know that the 10M signal is indeed leaking into the counter input?  Or is the counter just sensitive and 10M is the strongest signal around?

Hi

There are spectrum analyzer plots running around here showing the leakage. It's at a high enough level that it will mess up counting a 10 MHz signal. It also would give you a neat spur on the generator output.

Bob

[alank2]

There are spectrum analyzer plots running around here showing the leakage. It's at a high enough level that it will mess up counting a 10 MHz signal. It also would give you a neat spur on the generator output.

That answers it all right!

[Macman]

The fan in my unit started making a slight annoying buzzing noise, so I opened it up to see if I could manipulate the fan to get rid of the noise (which I was able to).

While I had it open I noticed it had the Siglent signature rust on the edges of some of the metalwork. It was only slight and am sure it won't be a problem but does show poor attention to detail.