Siglent SDS800X HD Review & Demonstration Thread

[awakephd]

There are more severe issues, and nobody really cares. Not even Siglent. So dont waste your time. 

Interesting.  My own experience suggests that, if truly no one cares about something that I find very problematic, it generally means that I am the only one who has the problem. Of course, YMMV.

[eTobey]

There are more severe issues, and nobody really cares. Not even Siglent. So dont waste your time. 

Interesting.  My own experience suggests that, if truly no one cares about something that I find very problematic, it generally means that I am the only one who has the problem. Of course, YMMV.

Indeed, i am a poor guy, that could only afford a Siglent SDS800X HD. But at the same time, i am in a fortunate position, that i can turn of a bunch of problems, with a simple button press.

Just yesterday, i used my DSO quad, to decode USART, and to view more than 64 bytes at once. I simply did not "turn on" those "problems" (the Siglent).  (The Siglent only shows like 7 bytes, and scrolling is a pain in the A. - The mouse scrollwheel has no function, and moving the scrollbar does not move the list at the same time.  ) BTW: those bytes on the DSO Quad are shown in ASCII and Hex at the same time. 

[crysti]

Is there any chance of frequency calibration? My Siglent SDS814X HD shows 10.00014 MHz GPSDO reference frequency of 10MHz, that is 14ppm. I know it is within specifications, but I would like more accuracy.
Another issue might be the input capacitance. I have measurement bridges with a precision of up to 0.001 pF, and I measured 19 pF on each input, with the Rigol DHO804 competitor, I measured 15pF. Impedance for 200MHz 15 pF > 53.05Ω  19pF > 42.58Ω.
  I would like to know what input capacitance others have observed.

[EvgenyG]

Is there any chance of frequency calibration? My Siglent SDS814X HD shows 10.00014 MHz GPSDO reference frequency of 10MHz, that is 14ppm. I know it is within specifications, but I would like more accuracy

Unless there is updated information, the answer is no it is not adjustable. I'd love to do the same.

[nanopico]

Is there any chance of frequency calibration? My Siglent SDS814X HD shows 10.00014 MHz GPSDO reference frequency of 10MHz, that is 14ppm. I know it is within specifications, but I would like more accuracy

Even if you could somehow do this calibration, which is only possible by changing the oscillator, the internal reference is not very accurate,
aging processes will lead to frequency changes.
The best method is to lock the oscilloscope to an external reference, a GPSDO for example.
Most equipment can use an external 10 MHz reference (seems to be standard these days), so it is a good idea to invest in something like that (e.g. Leo Bodnar GPSDO).

[crysti]

Is there any chance of frequency calibration? My Siglent SDS814X HD shows 10.00014 MHz GPSDO reference frequency of 10MHz, that is 14ppm. I know it is within specifications, but I would like more accuracy

Unless there is updated information, the answer is no it is not adjustable. I'd love to do the same.

I’ve now discovered why they’re all at + a few ppm: over time, the frequency decreases toward the nominal value due to aging.  In the first few years the drift is larger; thereafter the rate of change slows down (logarithmic trend).

[Martin72]

Hi,

I know it is within specifications, but I would like more accuracy

The question that arises for me is, what for?
All channels of the scope are “connected” to the same time base, so you will not initially experience any benefit from higher accuracy when displaying the respective signals.
I would rather perform frequency measurements/adjustments using a measuring device designed for this purpose (frequency counter).
Ultimately, it should not be forgotten that although this is a very capable scope, it is still limited in terms of price.

[crysti]

I’ve created a small Excel tool to help you calculate the actual frequency, provided you are able to measure the real frequency deviation. I reduced the read frequency by -2 ppm by coating the oscillators on the motherboard with DUE-CI V-66 Isolating Lacquer Spray.
Siglent only allows adjustment of the internal timebase offset on its top-end models (e.g., the 6000 or 8000 HD series), a feature that is very simply implemented via the PLL. In any case, the cost savings on the SDS800X are not significant, since the simple oscillator used costs €1, whereas a 0.3 ppm VCTCXO costs €8.
Likewise, on the lower-end models they have removed the physical fan-speed control, even though it remains available in the service menu.

[2N3055]

I’ve created a small Excel tool to help you calculate the actual frequency, provided you are able to measure the real frequency deviation. I reduced the read frequency by -2 ppm by coating the oscillators on the motherboard with DUE-CI V-66 Isolating Lacquer Spray.
Siglent only allows adjustment of the internal timebase offset on its top-end models (e.g., the 6000 or 8000 HD series), a feature that is very simply implemented via the PLL. In any case, the cost savings on the SDS800X are not significant, since the simple oscillator used costs €1, whereas a 0.3 ppm VCTCXO costs €8.
Likewise, on the lower-end models they have removed the physical fan-speed control, even though it remains available in the service menu.

So what is your point?
You would like also that 800xHD would cost 55 € and have OCXO, and active probe interface?
And also have 10" screen and also.... Yeah that would be great but not realistic.

There is a different made topic specifically about people expressing what they think might have been done differently. Your comments belong there, and it would be nice if you could move them there.

This is a topic to demonstrate use and what scope can and cannot do, not about how you think you would be more happy with a scope that was made differently.

At this moment there is no 8000 model in Siglent offering. There is 800xHD, 1000xHD, 3000xHD, 5000X, 6000A and 7000A.

[pezcat19]

Is there any chance of frequency calibration? My Siglent SDS814X HD shows 10.00014 MHz GPSDO reference frequency of 10MHz, that is 14ppm. I know it is within specifications, but I would like more accuracy

Unless there is updated information, the answer is no it is not adjustable. I'd love to do the same.

Me too! Without a 10MHz input on these scopes, a user adjustable reference calibration (or frequency display offset) would be great. Mine sits around the same 10.00014 and appears surprisingly stable. It just needs a tweak.

[crysti]

Has anyone else experienced acquisition memory limitations on the SDS824X HD with firmware version 1.1.6.5? When set to 10kpts, it drops to 5kpts at 5us/div, and the 100Mpts memory depth is only available for acquisition times of 5, 10, and 20ms/div. For other timebases, the available acquisition memory decreases. Is this an intentional limitation or a firmware bug? Thanks for any feedback!

[2N3055]

Has anyone else experienced acquisition memory limitations on the SDS824X HD with firmware version 1.1.6.5? When set to 10kpts, it drops to 5kpts at 5us/div, and the 100Mpts memory depth is only available for acquisition times of 5, 10, and 20ms/div. For other timebases, the available acquisition memory decreases. Is this an intentional limitation or a firmware bug? Thanks for any feedback!

Your scope is sampling at 100Mpt/s?
It is simple math.

If you are sampling at 100e6 point per second, how many points you get in 50µs (50e-6) worth of time?

[crysti]

Has anyone else experienced acquisition memory limitations on the SDS824X HD with firmware version 1.1.6.5? When set to 10kpts, it drops to 5kpts at 5us/div, and the 100Mpts memory depth is only available for acquisition times of 5, 10, and 20ms/div. For other timebases, the available acquisition memory decreases. Is this an intentional limitation or a firmware bug? Thanks for any feedback!

Your scope is sampling at 100Mpt/s?
It is simple math.

If you are sampling at 100e6 point per second, how many points you get in 50µs (50e-6) worth of time?

5,000 points are obtained in 50μs at a rate of 100MSa/s
But the oscilloscope I know only has 1200 points per refresh

[2N3055]

Has anyone else experienced acquisition memory limitations on the SDS824X HD with firmware version 1.1.6.5? When set to 10kpts, it drops to 5kpts at 5us/div, and the 100Mpts memory depth is only available for acquisition times of 5, 10, and 20ms/div. For other timebases, the available acquisition memory decreases. Is this an intentional limitation or a firmware bug? Thanks for any feedback!

Your scope is sampling at 100Mpt/s?
It is simple math.

If you are sampling at 100e6 point per second, how many points you get in 50µs (50e-6) worth of time?

5,000 points are obtained in 50μs at a rate of 100MSa/s
But the oscilloscope I know only has 1200 points per refresh

I don't understand, sorry.

[rf-loop]

Has anyone else experienced acquisition memory limitations on the SDS824X HD with firmware version 1.1.6.5? When set to 10kpts, it drops to 5kpts at 5us/div, and the 100Mpts memory depth is only available for acquisition times of 5, 10, and 20ms/div. For other timebases, the available acquisition memory decreases. Is this an intentional limitation or a firmware bug? Thanks for any feedback!

With memory management auto and one channel on, here can see how it use memory and use samplerate depending what maximum memory user have selected and what time base (t/div) is in use.
It works just as it is designed.

Yes there is some kind of exception if look this table. When step t/div down and samplerate drop after 500MSa/s it do not drop to 200MSa/s  but it drop to 100MSa/s  and because this, also memory need then reduce half for this step. Next step down in t/div speed do not drop samplerate, it stay 100MSa/s but it increase memory back to max selected.

Naturally when two or three and four channels are in use then many of these numbers are different. This table is just only for what ever one channel is in use alone.

[crysti]

Has anyone else experienced acquisition memory limitations on the SDS824X HD with firmware version 1.1.6.5? When set to 10kpts, it drops to 5kpts at 5us/div, and the 100Mpts memory depth is only available for acquisition times of 5, 10, and 20ms/div. For other timebases, the available acquisition memory decreases. Is this an intentional limitation or a firmware bug? Thanks for any feedback!

With memory management auto and one channel on, here can see how it use memory and use samplerate depending what maximum memory user have selected and what time base (t/div) is in use.
It works just as it is designed.

Yes there is some kind of exception if look this table. When step t/div down and samplerate drop after 500MSa/s it do not drop to 200MSa/s  but it drop to 100MSa/s  and because this, also memory need then reduce half for this step. Next step down in t/div speed do not drop samplerate, it stay 100MSa/s but it increase memory back to max selected.

Naturally when two or three and four channels are in use then many of these numbers are different. This table is just only for what ever one channel is in use alone.

Thank you!
The attached Excel calculation is correct or not?

[StevenMarshall]

I've upgraded my SDS-804X to 824X all worked well but I've noticed something unusual, can't remember if this was before the upgrade.

With nothing connected to the scope, at 2V setting I get something like a tiny impulse train on top of the waveform. At 1V I get a perfect flat waveform. Is this normal?

Here are the screenshots with the detailed settings:

[Performa01]

With nothing connected to the scope, at 2V setting I get something like a tiny impulse train on top of the waveform. At 1V I get a perfect flat waveform. Is this normal?

It might actually have to do with the scope bandwidth:

When switching from 1 V/div to 2 V/div you should hear an internal relay clicking. This is because the internal attenuator is switched from 20 dB (x10) to 40 dB (x100).

On the other hand, the 20 dB bandwidth limiter is directly before the ADC, yet it is 1st order only hence not very effective.

At 2 V/div the signal is only one tenth of that at 1 V/div, hence S/N ratio is 20 dB worse. That's only a concern for injected noise, entering the signal path after the attenuator, as it might happen in very noisy environments. The scope itself is electrically quiet, see my test results below (well, at least an original SDS824X HD is).

I've tested the SDS824X HD with a zero signal to rule out external noise as a factor, i.e. just a 50 ohm terminator at the input. The only difference between 1 V/div and 2 V/div is the noise level more than doubles at the less sensitive setting. Yet it's a uniform white noise without spikes. Take note of the different vertical gain settings in the zoom window!

[eTobey]

Has anyone else experienced acquisition memory limitations on the SDS824X HD with firmware version 1.1.6.5? When set to 10kpts, it drops to 5kpts at 5us/div, and the 100Mpts memory depth is only available for acquisition times of 5, 10, and 20ms/div. For other timebases, the available acquisition memory decreases. Is this an intentional limitation or a firmware bug? Thanks for any feedback!

There is something wrong with the acquisitions anyway. If you decrease the mem depth, then the trigger rate drops significantly, even in segment mode.

[eTobey]

With nothing connected to the scope, at 2V setting I get something like a tiny impulse train on top of the waveform. At 1V I get a perfect flat waveform. Is this normal?

This looks normal to be. The level is just at below the level where it goes from pixel height A to pixel height B. Put a 1 megaohm resistor on the input, and it should look ok.