Siglent SDS2000X Plus - Bugs / Missing Features / Feature Requests

[Performa01]

I would only expect a fix improvement that is achievable thru firmware with existing hardware.  All of your issues are resolved if you just allow for two mode settings in the utility menu--SAFE and EXPERT or whatever.  In the SAFE mode the 50R terminator disconnects whenever the input is overloaded (even at the lower ranges) and AC50R is disabled.  EXPERT just gives you a warning and leaves you to fry your scope at will.  Or maybe even have a LOCK setting where you need a password to access the 50R menu...

Whenever the protection kicks in, the scope would exit the 50 ohms mode and revert back to 1 meg input impedance. By this, the amplitude gets doubled - if the frequency is low enough that transmission line effects don't come into play yet. Now adjusting for the correct vertical gain (which would usually result in an amplitude of more than 50% of the screen height in order to maximise dynamic range) could be very confusing. Of course, operators would have to get used to a workflow that starts at the lowest sensitivity (= 1 V/div) and only then connect the signal and increase sensitivity as required.

All in all, I'm predicting that this would not work. The average users would complain to no end, until they either RTFM or get advice in a (this?) forum, so they're finally able to turn this "foolproof mode" off. At the other hand, even though it's only "programmed hardware" (FPGA), this protection still requires resources, which the SDS2000x Plus might not even have - after all this model doesn't even have the otherwise indispensable "DVM". So it's certainly not very attractive to waste resources on something that in the end nobody will use anyway - because the way you have suggested (and the only feasible) is just too much hassle for the operator.

[bdunham7]

So it's certainly not very attractive to waste resources on something that in the end nobody will use anyway - because the way you have suggested (and the only feasible) is just too much hassle for the operator.

Of course I thoroughly disagree!  Its true that most people wouldn't 'use' it.  The SAFE mode would be the default and in the vast majority of circumstances, you'd never know the feature existed.  Some people might get slightly annoyed that it trips out of 50R if they exceed some low vertical limit---like the 10mV/div range--but they'd figure it out.  And dialing down from a higher range is always a good idea anyway.  What the SAFE mode would best prevent is people fat-fingering the 50R button while they are looking at PSU ripple on a 24V supply using AC coupling with their 1X probe. 

As far as it being a hassle, it's a pretty complex scope and the UI is, well, OK.  I doubt what I've proposed would really be a big deal either way.

[Martin72]

Hi,

Today I check the "50Ohm behaviour" of an much more expensive scope, in this case a WR9054 from lecroy, appx 12 times more expensive than sds2k+.
First, it got only DC50, not AC50.
Second, when turning to 50 ohms, voltage per division will be limited to 1V/div.
Third, carefully increased the (DC-)voltage above the mentioned maximum of 5V, nothing will happen, increased to 7.5Vdc.
No warning, no switchback to 1M.
FYI.

Martin

[Performa01]

Today I check the "50Ohm behaviour" of an much more expensive scope, in this case a WR9054 from lecroy, appx 12 times more expensive than sds2k+.
First, it got only DC50, not AC50.
Second, when turning to 50 ohms, voltage per division will be limited to 1V/div.
Third, carefully increased the (DC-)voltage above the mentioned maximum of 5V, nothing will happen, increased to 7.5Vdc.
No warning, no switchback to 1M.

I'm pretty sure, the LeCroy will have some thermal protection too. For this, you need to take into account that the internal termination actually can take a little more than the 500 mW suggested by the front panel message (5 Vrms). It will most likely be at least 1W, so there is some headroom for the overload protection. If you run the scope with 7.5V DC-input for a long time, the thermal protection should eventually kick in, but you might need an even higher voltage to achieve this.

[adam4521]

The instantaneous power in the resistor is V²/R. Would you guys be ok to lose a maths channel (ie use existing resources) to track this? Instead of having an artificial/compromise limit on the Y axis, have a limit on the time axis instead, eg 20ms total sample time? Then just put a small amount of logic in to automatically trip out if the mean V² on the input BNC exceeds a preset value. You could also have a V_peak threshold value if you wanted to, and trip on that too, just like a "real" circuit breaker.

While I appreciate that its not always possible to put comprehensive protection into high end instruments, because it can mess up the readings, it seems to me that having  protection 'coverage' in a general purpose bench oscilloscope is something that's worth spending effort on. With Siglent philosophy of putting as much function as possible while being thrifty with the hardware costs, the software approach would seem to be a good, achievable match.

In a lab setting, especially in a training/university lab I imagine it is reassuring for the demonstrator to be able to say -- "don't worry about experimenting with the controls, you can't break anything".

[2N3055]

The instantaneous power in the resistor is V²/R. Would you guys be ok to lose a maths channel (ie use existing resources) to track this? Instead of having an artificial/compromise limit on the Y axis, have a limit on the time axis instead, eg 20ms total sample time? Then just put a small amount of logic in to automatically trip out if the mean V² on the input BNC exceeds a preset value. You could also have a V_peak threshold value if you wanted to, and trip on that too, just like a "real" circuit breaker.

While I appreciate that its not always possible to put comprehensive protection into high end instruments, because it can mess up the readings, it seems to me that having  protection 'coverage' in a general purpose bench oscilloscope is something that's worth spending effort on. With Siglent philosophy of putting as much function as possible while being thrifty with the hardware costs, the software approach would seem to be a good, achievable match.

In a lab setting, especially in a training/university lab I imagine it is reassuring for the demonstrator to be able to say -- "don't worry about experimenting with the controls, you can't break anything".

If you put a channel in 1 mV/div, you overload ADC at 5mV. Software cutoff makes sense only if you are at highest range (1V/div) already and then ADC detects off scale.
Which might or not be the case. 
Putting anything "comprehensive" into any device always ends up much harder and more expensive than people think.
Even 500Mhz (and this is 500MHz scope) is high enough for this not to be trivial.
Simple example: everybody speaks about protection circuit. But you actually need 4, one for each channel. And then a connection back to CPU so it can know it scrambled. etc etc..
That is a mainboard redesign (provided current design has resources, that are not planned for other, more pressing stuff for the target users). And full characterisation. And you end up with a new product. With new name..etc... Then you have to support both...

And respectfully,  "" In a lab setting, especially in a training/university lab I imagine it is reassuring for the demonstrator to be able to say -- "don't worry about experimenting with the controls, you can't break anything"." is exactly the wrong thing to say to university students. That is kindergarten "positive reinforcement" attitude. That is preschool stuff. In university, they better be taught to know exactly what will happen if they turn that knob... They are not playing with legos, they are supposed to be engineers and scientists... They came to university to learn to think and understand...Not randomly try stuff until it either works or explodes...

[rf-loop]

And respectfully,  "" In a lab setting, especially in a training/university lab I imagine it is reassuring for the demonstrator to be able to say -- "don't worry about experimenting with the controls, you can't break anything"." is exactly the wrong thing to say to university students. That is kindergarten "positive reinforcement" attitude. That is preschool stuff. In university, they better be taught to know exactly what will happen if they turn that knob... They are not playing with legos, they are supposed to be engineers and scientists... They came to university to learn to think and understand...Not randomly try stuff until it either works or explodes...

 

[tv84]

They are not playing with legos, they are supposed to be engineers and scientists... They came to university to learn to think and understand...Not randomly try stuff until it either works or explodes...

Unless they are training for Quality Assurance!

I expect every life critical system to be tested in all work/explode scenario as practical possible.

[2N3055]

They are not playing with legos, they are supposed to be engineers and scientists... They came to university to learn to think and understand...Not randomly try stuff until it either works or explodes...

Unless they are training for Quality Assurance!

I expect every life critical system to be tested in all work/explode scenario as practical possible.

And that is almost as fun a job as working with explosives....

But it is even more dependent on predicting all outcomes up front...
In which case they designed scenario, simulated and calculated outcomes, and then made real life experiment where thing exploded exactly where calculation predicted. If it didn't, something was wrong and go back to the drawing board, until model aligns with real life testing. Only then we can have good certainty that product is safe by design....

[bdunham7]

They are not playing with legos, they are supposed to be engineers and scientists... They came to university to learn to think and understand...Not randomly try stuff until it either works or explodes...

We all make mistakes and errant button-presses.  Given the option, I'd rather have a device that doesn't blow up  with one errant touch of the screen during an otherwise perfectly normal operation (PSU ripple on AC1M, bump the 50R button and poof!)  It would have been easier to implement this all from the beginning, but the whole bang-for-buck thrifting process yielded what we have and any implementation now would necessarily be a bit clunky.  But I think it is perfectly doable, and for those users that don't like my SAFE mode, they can disable it and make their scope fry-able again.  If tripping the protection on the 1mV/div range is so terrible, perhaps they could introduce continuous autoranging, or at at least auto-UP-ranging when 50R is selected. 

However, I doubt we'll see this anytime soon.  I'm sure it isn't a priority and they will be putting their efforts elsewhere.  IMO the least they could do would be to put options in the setup menu to disable the AC50R mode  and perhaps the 50R as well so a user can proactively choose to protect themselves from their own thumbs.  So be careful with the buttons! 

[Martin72]

If you run the scope with 7.5V DC-input for a long time, the thermal protection should eventually kick in, but you might need an even higher voltage to achieve this.

You might be right, because:

Caution: The maximum input voltage depends on the input used. Limits are displayed on the body
of the instrument. Whenever the voltage exceeds this limit, the coupling mode automatically
switches to GROUND. You then have to manually reset the coupling to its previous state. While the
unit does provide this protection, damage can still occur if extreme voltages are applied.
Probe

This state I haven´t reached.....Mhh...Should I try it..

[2N3055]

.....
This state I haven´t reached.....Mhh...Should I try it..

Please don't...
Murphy never sleeps...

[Martin72]

 

There is something, that makes me a little bit confusing...
Either it was on the HDO6034A or on the WR9054 (although they got the same UI), but I could swear, one of them showed once a message while booting up, that there were a 50ohm coupling activated last time shutdown and gave me the try to change it before booting was complete.
But I couldn´t reproduce it in the last two days.
Maybe I´ve dreamed this..

[Plasmateur]

Mod + Burst

Is this ever going to happen?

[Martin72]

It takes some time to understand what you mean...
The siglent beta-testers here may corect me, but after 2yrs I don´t think they will change anything on the integrated awg.

[Plasmateur]

Yeah, probably not. Attempting to EasywaveX to do the same thing but the spectrum looks terrible.

[Martin72]

Maybe a bug:

While playing with the differential function for some reasons (see "original" thread),  I´ve found out that the universal knob won´t function when you want to change Dx although it´s activating.
You can change Dx by using the virtual keypad.
When you did this, then you can use the universal knob to change the value back - one time....

Martin

[Martin72]

Wanted features:

- Display split mode not only for FFT, it would be nice to have it for all math functions avaible.

- Additional print mode black&white.

Wanted "Polishes":

- Axis scaling : 4 digits after the comma is not useful and "steals" visibility. Better: Max. 1 digit after the comma.
- Channel colours are free selectable - Excellent. A little bit more nice would be when you can select between colours on screen and on print like lecroy scopes.
Example, channel 1 is yellow on the screen but will be "printed" in red or any other colour selected before.

[Martin72]

Hi there,

Not a bug, more a little wish:

We got several sds2104X+ at work and they will be daily used under "real conditions" (Pic shows scope while short circuit measuring)...
From time to time I go and ask the colleagues if they are satisfied with them.
They are....Mostly with all.
Here a little wish from the last asking.
On new lecroy scopes we hardly miss the free moveable measuring gate over the whole screen* - The siglent got it so all are happy.
To measure loadsteps on our inverters, we adjust the gate to say 2.5ms (for one 400Hz cycle) then turn the gate to the loadstep, measure what´s in this gate.
Then move the gate to the next cycle after loadstep, measure, move to the next and so on.
"Problem":
To move the gate you use the universal knob, so far so good.
But you must always go to the measure menu, go the Gate A-B, then universal knob is active and you can move - Otherwise the knob adjust the intensity..
For every measure-step we must repeat this instead having the universal-knob for this as long as we need it.
A more or less free alignment for this knob would be nice...(Must say, don´t looked at the manual after colleague asked me)
Then:
The measured value represent this what´s in the defined gate - As expected.
BUT: While you move the gate, the value won´t change - It changes only when you stop and wait a little bit.
Knowing the opposite from lecroy scopes that is uncommon, therefore the wish is to have actual changing value displaying.
Thanks in advance..
Martin

Edit: *) On elder lecroy models, you can define a gate and inbetween this gate it measures.
Today lecroys got it too, but not the "tracking function" of the elder ones, where you can move the before defined gate over the screen..

[TheDefpom]

So the SDS1204X-E has "Sample Logging" and "Measure Logging" built in, and NTP... I wonder if the 2000X Plus will ever get these ?

I found a small bug, there is a timezone setting option but it doesn't work ? I cannot get it to bring up any timezones.

[Martin72]

Hi,

Actually there is a "Timezone bug" present, I guess it´s "valid" for all touchscreen-models (2k+, 2k HD, 5k+, 6k A) because of more or less the same software.

[tautech]

So the SDS1204X-E has "Sample Logging" and "Measure Logging" built in, and NTP... I wonder if the 2000X Plus will ever get these ?

I found a small bug, there is a timezone setting option but it doesn't work ? I cannot get it to bring up any timezones.

Email sent.

[balnazzar]

So the SDS1204X-E has "Sample Logging" and "Measure Logging" built in, and NTP... I wonder if the 2000X Plus will ever get these ?

I found a small bug, there is a timezone setting option but it doesn't work ? I cannot get it to bring up any timezones.

Email sent.

If it was about sample/measure logging and/or ntp, please send it to me too

[bingo600]

Bug present in : 1.3.9R12

Bug Solved in in : 1.5.2R1 - Thank you Siglent 
Scope now renews on "half-time of lease", just as the "doctor (RFC) ordered"



***** TLDR *****
I just finished the "Options" on my new SDS2104X Plus SDS2504X Plus.
I haven't upgraded fw to latest yet, so i'm still on 1.3.9R12

I'm noticing that the above version have an "often occuring bug" on embedded systems...
The Scope DHCP Client does not renew the DHCP lease, on lease timeout.

Many users prob. won't detect that, as the scope just keeps using the DHCP IP it was "offered" when booting.
But in my setup i use my linux server with the ISC-DHCP server & the bind9 dns server setup for "dynamic dns add of DHCP clients".
The setup for my "measure vlan" is DHCP with a  1800s (30 min) lease time, and since the scope doesn't send a DHCP identifier i have just added one into my DHCP config file.
I have chosen : sds2104x
This means that when the scope accepts the DHCP offer, the linux DHCP server will automatically inform the DNS server that it should create a new "host entry" with the above name, and a TTL (lifetime) of 1800 sec.

I noticed the error, when i suddenly couldn't resolve the dns name anymore (more than 30 min since boot), and did a bit of DHCP/DNS log checkking.
Yepp ....
The scope does not respect my 1800s lease time, and for now i have only seen it request a DHCP address on boot, not ever seen a DHCP renew.

Here's a DHCP/DNS log snip from my linux server.

Note the :
Add new forward map, where it adds the scope dns name
And that exactly 30 min later the DHCP server "purges" the dns name , since it hasn't seen a new DHCP transaction from the scope.

Code: [Select]

Nov 19 12:32:00 frodo dhcpd[23534]: DHCPOFFER on xx.yy.70.130 to <Scope-MAC> via xx.yy.70.1
Nov 19 12:32:00 frodo dhcpd[23534]: DHCPREQUEST for xx.yy.70.130 (aa.bb.117.101) from <Scope-MAC> via xx.yy.70.1
Nov 19 12:32:00 frodo dhcpd[23534]: DHCPACK on xx.yy.70.130 to <Scope-MAC> via xx.yy.70.1
Nov 19 12:32:00 frodo dhcpd[23534]: Added new forward map from sds2104x.mydomain.org. to xx.yy.70.130
Nov 19 12:32:10 frodo dhcpd[23534]: reuse_lease: lease age 10 (secs) under 25% threshold, reply with unaltered, existing lease for xx.yy.70.130
Nov 19 12:32:10 frodo dhcpd[23534]: DHCPOFFER on xx.yy.70.130 to <Scope-MAC> via xx.yy.70.1
Nov 19 12:32:11 frodo dhcpd[23534]: reuse_lease: lease age 11 (secs) under 25% threshold, reply with unaltered, existing lease for xx.yy.70.130
Nov 19 12:32:11 frodo dhcpd[23534]: DHCPREQUEST for xx.yy.70.130 (aa.bb.117.101) from <Scope-MAC> via xx.yy.70.1
Nov 19 12:32:11 frodo dhcpd[23534]: DHCPACK on xx.yy.70.130 to <Scope-MAC> via xx.yy.70.1
Nov 19 13:02:00 frodo dhcpd[23534]: Removed forward map from sds2104x.mydomain.org. to xx.yy.70.130

I have been looking in the newer firmware 1.5.2R1 release notes , and nothing is mentioned about a DHCP Client fix.
So I suppose it's still there in the latest fw. Will upgrade soon, and will test again.

But i think it ought to be reported here.

/Bingo

[tautech]

I just finished the "Options" on my new SDS2104X Plus SDS2504X Plus.
I haven't upgraded fw to latest yet, so i'm still on 1.3.9R12

Why oh why are ppl so reluctant to upgrade FW ? 
Please just do it !