Siglent SDS800X HD 12 bit DSO's

[matus]

Teardown photos of the 804 and 814 show two Samsung K4A4G16 chips (2*4 Gbit). I am not aware of a documented teardown of the 824 so far, but the 1000X HD series (which, like the 824, also has the doubled 100 MPt memory) comes with two  K4A8G16 chips (2*8 Gbit), on an otherwise very similar layout.

The assumption was that the doubled RAM capacity is required for the doubled memory depth -- but maybe it is not, and all the 800 series scopes get by with 2*4 Gbit?

Not sure about that, but 1000 might be built on different OS configuration, that might address your concern about the RAM discrepancy. Just compare the boot times.

https://youtu.be/NlCeWsD750w?si=hewGmu81Gw_AT0sf&t=348

[Martin72]

once someone really puts an upgraded 804 or 814 through its paces.

You, for example...
No risk no fun, if it doesn't work out you can always return it and buy the "right" 824.

[Martin72]

I don't think anyone has anything in stock at the moment.
Whether someone sets the marker to "green" and then delivers in 2 weeks, or whether someone else is "honest" and quotes 5-10 days...
Today was "only" the launch day on the part of siglent, but that doesn't necessarily mean that all eligible dealers have already been supplied.
If I'm lucky, my scope will arrive a week earlier than stated(mid march).

[Martin72]

siglent.eu

If it is actually on its way, congratulations, that would be something new, contrary to my experience with them and I mean that seriously, without any ulterior motive.

[baldurn]

baldurn had mentioned two days ago that he already received a tracking code -- i.e. siglent.eu had shipped the scope for which he had managed to sneak in an order. Guess he should receive it tomorrow or so and get his hacker-space team to work! 

They seem to have used the slowest delivery service ever. I have had stuff delivered faster from China. But now I got a confirmed delivery date tomorrow (thursday the 29th february). Unfortunately my business is closed tomorrow because I and everyone else are away. So will probably first pick it up friday.
 

[Bad_Driver]

I don't think anyone has anything in stock at the moment.
Whether someone sets the marker to "green" and then delivers in 2 weeks, or whether someone else is "honest" and quotes 5-10 days...
Today was "only" the launch day on the part of siglent, but that doesn't necessarily mean that all eligible dealers have already been supplied.
If I'm lucky, my scope will arrive a week earlier than stated(mid march).

My reseller told me yesterday, that I placed the first order for a SDS804x and that they will have the new models in stock beginning next week.

[Rudgas]

My 804X should arrive today, I might be able to answer some non-warranty voiding questions.

[BreakingOhmsLaw]

siglent.eu

If it is actually on its way, congratulations, that would be something new, contrary to my experience with them and I mean that seriously, without any ulterior motive.

Siglent.eu (or rather: JR Special Electronics, which is the actual name of the company) is not giving a delivery date yet. That's fine with me, it's not like there's a lack of scopes in my lab.
On the other hand, Batronix - which I know Martin you prefer, - is listing a wrong accessory for the 1000X-HD. In the logic analyzer license, they claim: "The hardware module SPL2016 is required!". AFAIK, that's wrong, the 1000X-HD only works with the SLA1016. Please correct me if that's not the case. That doesn't instill much confidence in me.

That said, i've never had a bad experience with neither Batronix nor JR. I've bought many times from both dealers and have had nothing but good experience with each of them.

[RAPo]

I was an early bidder at Ele-shop for an SDS81X.
Initially, it also displayed the wrong LA-sw/hw.
After an email it quickly was put in good order, things are brand new.
In the video of defpom, he was talking about a bundle: scope+la+awg. I haven't seen an offer for that yet.

[rf-loop]

On the other hand, Batronix - which I know Martin you prefer, - is listing a wrong accessory for the 1000X-HD. In the logic analyzer license, they claim: "The hardware module SPL2016 is required!". AFAIK, that's wrong, the 1000X-HD only works with the SLA1016. Please correct me if that's not the case.

Yes there is "copypaste error" in Batronix.  Of course "The hardware module SPL2016 is required!" is wrong and right one is SLA1016 (error is just in this one sentence in LA licence but if one buy this LA accessory (hardware)  for this model there is just right model.  Same text mistake is also with some other models).

[baldurn]

It arrived. Here is a photo of the yet unmodified SDS804X HD together with unmodified DHO804 and a fully unlocked SDS1204X HD.

[baldurn]

In the previous image we had a 70 MHz sine which both Siglents shows as approximately 500 mV. The Rigol however is down to 390 mV for some reason. At 1 MHz they are all the same. But my setup could be fighting reflections from input ports that are not quite 1M at 70 MHz and so this is just for fun.

Here are two more images with a square wave at 1 MHz and 500 mV overlaid with a sine of 50 MHz at 5 mV. I will do a single capture while zoomed out and then zoom after stopping. Only a true 12 bit scope would be able to show the sine with any detail. For some reason the Rigol fails this test completely. The trace looks nicer on the SDS804X HD compared to SDS2104X HD but I think that is because it has a lower bandwidth and does not capture higher frequency noise.

[Performa01]

It arrived. Here is a photo of the yet unmodified SDS804X HD together with unmodified DHO804 and a fully unlocked SDS1204X HD.
(Attachment Link)

The picture shows an SDS2104X HD (unlocked or not), which makes more sense anyway

EDIT: BTW, your setup really is just for fun and no conclusions can be drawn from this.

If you really want to feed all scopes with identical signals at the same time, then you need a resistive wideband power splitter and through terminators for those instruments that lack 50 ohms inputs - yes, BNC-Ts with end terminations will do as well.

[Martin72]

For some reason the Rigol fails this test completely.

That's pretty blatant.
Which generator did you use for the overlay?

[baldurn]

For some reason the Rigol fails this test completely.

That's pretty blatant.
Which generator did you use for the overlay?

It is from a SDG6022X unlocked. To figure out what was going on, I tried connecting only the Rigol. We then get the correct 500 mV signal, so that was just an artifact of trying to be too clever with connecting all three at the same time. However the Rigol still fails the 12 bit test.

So I noticed that the Rigol can show the sine if you are are in normal mode. It fails if you try to show it on stored data, when that data was stored before zooming in. You can hear relays clicking while zooming and it looks clipped. I am assuming this is a Rigol bug and they clip the stored data during the zooming. So you are unable to probably zoom and view your 12 bit data on the Rigol.

[baldurn]

Here is a picture with a proper setup. Single capture with a direct cable. We get more or less the same image on all three scopes.

This is the one case where I notice the two pixel thing on the SDS804X HD. The overlaid 50 MHz sine is just barely visible on all three, but on the SDS804X it is clearly just two values. Looks blocky compared to the other two. I don't think it matters much because you are going to zoom on that if you need it.

[baldurn]

But I would be really surprised if the Rigol could not zoom in properly on stored data and show you the high-frequency sine component. Could you share a closeup (or digitally captured) screenshot which shows what it looks like, with readable acquisition settings?

It happens if you allow the top of the data to go out of the screen while zooming. If you are careful to move the top line to the bottom of the screen before going to the next zoom level, such that the top line will still be within the screen after zooming, then it does not clip. Any mistake so that top line is temporary out of the screen will clip it and it stays clipped when you move it back down into the screen.

But this is the Siglent thread, not the Rigol bug thread (which I am sure is very long already).

[baldurn]

And now for the thing everyone wants: I upgraded the scope from 70 MHz to 200 MHz bandwidth. Used the script in the first post here: https://www.eevblog.com/forum/testgear/siglent-sds-sdg-hack-script/

Looks like this:

Code: [Select]

baldur@yakshi:~$ telnet 192.168.8.117 5024
Trying 192.168.8.117...
Connected to 192.168.8.117.
Escape character is '^]'.
Welcome to the SCPI instrument 'Siglent SDS804X HD'
>>SCOPEID?
MD5_SRLN?
0170e36144f12285
MD5_PR?
SDS800X-HD
PRBD?
70M
MCBD?
MCBD BQVPA2HNSPK3TEBG
MCBD 4DTUZNPN2N5M72H7
MCBD?
MCBD 4DTUZNPN2N5M72H7
MCBD I73W3KPT7EZA3MKD
MCBD?
MCBD I73W3KPT7EZA3MKD

baldur@yakshi:~$ telnet 192.168.8.117 5024
Trying 192.168.8.117...
Connected to 192.168.8.117.
Escape character is '^]'.
Welcome to the SCPI instrument 'Siglent SDS824X HD'

And now we have instrument model SDS824X HD on the screen.

I measured the 3 dB bandwidth to be 113 MHz before the upgrade and 286 MHz after entering the codes. Everything seems to be fine including 100Mpts. Did not even reboot.

[BillyO]

Nice!

[baldurn]

Ramp test with 100 Mpts.

[baldurn]

More ramp test. It may be hard to see on this screen shot but we are in dot mode and fully zoomed on that same 100 Mpts ramp test. I am counting 20 dots or 2 dots per div with 1 ns/div. This equals 100 million dots in 50 ms, which is the length of the capture. Seems to check out as far as I can tell.

[tautech]

And now for the thing everyone wants: I upgraded the scope from 70 MHz to 200 MHz bandwidth. Used the script in the first post here: https://www.eevblog.com/forum/testgear/siglent-sds-sdg-hack-script/

Looks like this:

Code: [Select]

baldur@yakshi:~$ telnet 192.168.8.117 5024
Trying 192.168.8.117...
Connected to 192.168.8.117.
Escape character is '^]'.
Welcome to the SCPI instrument 'Siglent SDS804X HD'
>>SCOPEID?
MD5_SRLN?
0170e36144f12285
MD5_PR?
SDS800X-HD
PRBD?
70M
MCBD?
MCBD BQVPA2HNSPK3TEBG
MCBD 4DTUZNPN2N5M72H7
MCBD?
MCBD 4DTUZNPN2N5M72H7
MCBD I73W3KPT7EZA3MKD
MCBD?
MCBD I73W3KPT7EZA3MKD

baldur@yakshi:~$ telnet 192.168.8.117 5024
Trying 192.168.8.117...
Connected to 192.168.8.117.
Escape character is '^]'.
Welcome to the SCPI instrument 'Siglent SDS824X HD'

And now we have instrument model SDS824X HD on the screen.

I measured the 3 dB bandwidth to be 113 MHz before the upgrade and 286 MHz after entering the codes. Everything seems to be fine including 100Mpts. Did not even reboot.

Why would you want to Telnet into the scope when you have a perfectly good SCPI command page in the web browser. I don't get it.

[matus]

Why would you want to Telnet into the scope when you have a perfectly good SCPI command page in the web browser. I don't get it.

Why not? You use your web interface, baldurn uses his terminal. What's the point here?

[lofe]

And now for the thing everyone wants: I upgraded the scope from 70 MHz to 200 MHz bandwidth. Used the script in the first post here: https://www.eevblog.com/forum/testgear/siglent-sds-sdg-hack-script/

Looks like this:

Code: [Select]

baldur@yakshi:~$ telnet 192.168.8.117 5024
Trying 192.168.8.117...
Connected to 192.168.8.117.
Escape character is '^]'.
Welcome to the SCPI instrument 'Siglent SDS804X HD'
>>SCOPEID?
MD5_SRLN?
0170e36144f12285
MD5_PR?
SDS800X-HD
PRBD?
70M
MCBD?
MCBD BQVPA2HNSPK3TEBG
MCBD 4DTUZNPN2N5M72H7
MCBD?
MCBD 4DTUZNPN2N5M72H7
MCBD I73W3KPT7EZA3MKD
MCBD?
MCBD I73W3KPT7EZA3MKD

baldur@yakshi:~$ telnet 192.168.8.117 5024
Trying 192.168.8.117...
Connected to 192.168.8.117.
Escape character is '^]'.
Welcome to the SCPI instrument 'Siglent SDS824X HD'

And now we have instrument model SDS824X HD on the screen.

I measured the 3 dB bandwidth to be 113 MHz before the upgrade and 286 MHz after entering the codes. Everything seems to be fine including 100Mpts. Did not even reboot.

Did you use the bundled probe to measure the bandwidth?

[Antonio90]

It's out at last. I intended to buy the 1000X-HD, expecting it to be double the price, around the €799-899 price point. Right now the SDS1104X-HD is closer to tripling the SDS804X-HD.
You can get the SDS804 plus the SDG2000 and still save €200, or the SDG1032 and save €400. The basic model is an incredible value TBH.
The real MVP of this release is the 2104X-HD though. It's € 1748.40 at batterfly.com. The 1000X-HD is in a somewhat strange price spot.