Products > Test Equipment
Siglent SDS2000X Plus - Bugs / Missing Features / Feature Requests
2N3055:
--- Quote from: ebastler on November 21, 2023, 03:33:50 pm ---Well, Siglent certainly has loyal customers/beta-testers/consultants here -- so they must be doing something right! :-+
Thanks for you long post, 2N3055. You make several valid points. I'll comment on just a few aspects where I don't concur:
--- Quote from: 2N3055 on November 21, 2023, 01:25:04 pm ---As to expectations, fact that Siglent is still keeping this scope in active development and keeps adding features long after release should not be confused with "Siglent promises to keep adding major features that will with time convert a 1000€ scope into same capability as 20000€ LeCroy".
Realistically, manufacturer will add features to keep scope competent contender in a marketplace where it belongs. They might add a bit more to keep it fresh and to keep certain strategic advantage. But mostly never it will grow so much to exit it's class and enter one above. For both strategic and technical reasons.
--- End quote ---
Can we please keep this discussion fair and level. I don't think I suggested anything unreasonable which would catapult the scope out of its class:
* The ability to "trim" a captured waveform and send it to the AWG -- such that a full period is played back, rather than some arbitrary chunk which happens to fit the screen width -- is standard on pretty much every DSO with a built-in AWG, to my knowledge.
* Filters (as a Math operation) are also standard in this class, e.g. on the RTB2000, DSOX1200, MSO5000. The SDS2000X+ is the odd one out here, since it apparently still does not have them. And while I get it that going beyond 2 math channels would stretch the computational resources too much, I can't see why simple FIR or IIR filters should require more computing power than other math operations.
* Being able to configure a WiFi dongle is not standard yet. But since Siglent have this feature in the lower-end scopes, and actually have the RTL8188 driver up and running on the X+, it seems like low-hanging fruit and a nice differentiator.So please don't give a distorted view of user expecations here.
--- Quote ---As for development and release schedules, that is also interesting and not so simple topic.
[...]
In addition to that, modern practice to develop software on fixed released schedules without and regard to quality or completeness of code has shown catastrophic results for quality of code.
Just look at Rigol which released 4 different new scopes (4 products) based solely on management decision to stick to aggressive timetables. [...]
This is what you get if you have to release software twice a year instead of when it's done..
--- End quote ---
Sure, I know about dependencies, platform approach etc.; I have been working in industrial instrument R&D for a long time. I specifically suggested that Siglent switch to a time-boxed approach once the product is in "sustaining" mode, with incremental feature additions and bug fixes -- not for new development. No need for shadow-boxing, please.
--- Quote ---None of the "big boys" will ever commit to any schedules or feature.
FW updates are in a way a strategic resource, nobody wants to open their cards..
I thing Siglent is actually doing a better job in this regard.
--- End quote ---
Again, commiting to features upfront is not what I suggested. The idea is to commit to a timeline cadence for updates -- so customers see that you are still committed to the product, and don't keep scratching at your door to ask when the next update will come. Keep the scope of the release open, to manage technical risk and keep some flexibility in the resources you can assign to firmware improvements.
Maybe this is not standard in the industry. I am not saying that Siglent is doing worse than other brands. But I think this would be a realistic approach how they could do better than the competition.
--- End quote ---
I do try to keep discussion fair and level. If I fail sometimes that is because of me being human. There will always be misunderstandings and failing to explain things in a way others can easily understand. Fact that most of us are communicating in second language does not help either.
I would just like to respond to you on few things I would like to point out.
I was talking about expectations in regards to capabilities as released/ few years later on the same scope. Not compared to other scopes in same class.
Those other scopes also have many shortcomings and architectural differences that benefit some features and are impeding others.
They work with smaller memories and decimated data.
As I said, a simplified Filter package just for 2000X+ probably could be made. But not very likely only for it.
If they find a way to run full filter package you will get it. Filtering also might be connected with manual control of sampling rate...
As for cadence of updates, in my opinion, those agile development processes are producing horrible code quality. Because they shift priority from quality or even functionality to regular releases. So big, chunky problems sometimes are not even solved because they don't fit into framework.. I have several friends working in very serious software development and they tell me horror stories all the time.
As I said, development of firmware is very interdependent. Sometimes one part waits for other.. Regular releases without substance but with full regression test nevertheless are huge resource hog. It is something Boeing does to US military because their contracts allow for it. And they have to give proof of progress all the time..
When you are doing most efficient internal development, on a stable and mature device, you obey your priorities. And you don't release just to throw dog a bone. I believe Siglent wouldn't do that to their customers, out of respect. When they have something of substance they release it. Otherwise they are not insulting customers with trivial distractions...
Just to make sure, I am not belittling your opinion but trying to give you more insight into process..
ebastler:
--- Quote from: 2N3055 on November 21, 2023, 06:11:48 pm ---So big, chunky problems sometimes are not even solved [in a time-boxed development process] because they don't fit into framework..
--- End quote ---
Looking back over the firmware release notes, what would be examples of big, chunky problems (or features), requiring major changes to the existing framework, which Siglent tackled in the 2000X+ after its release?
--- Quote ---[...] trying to give you more insight into process..
--- End quote ---
Thanks. As mentioned, I have done and led R&D for scientific instruments (which do of course include firmware and PC software these days) for a few decades, so I get the idea. ::)
I am still a firm believer in time-boxed releases, for the incremental features and bug fixes that typically get implemented after a product has shipped. I don't see why would not be entirely feasible, with good product quality, and with happy customers.
...
Heck, am I the only one who gets disappointed when tautech indicates that we will eventually get another firmware release (with a new major version number, and after more than a year of no user-facing updates) -- only to be told by Performa01 that it will not be a major update, but will comprise a single, small new feature? :-\
Oh -- I am the only one, at least in here? Never mind, I'll shut up then...
KungFuJosh:
--- Quote from: ebastler on November 21, 2023, 07:47:10 pm ---Heck, am I the only one who gets disappointed when tautech indicates that we will eventually get another firmware release (with a new major version number, and after more than a year of no user-facing updates) -- only to be told by Performa01 that it will not be a major update, but will comprise a single, small new feature? :-\
--- End quote ---
You made an assumption based on a version number without any other information. I miiiight have had a similar thought, but since nothing specific has been confirmed, I'm not going to get excited about it.
I want my scope to be and remain stable. That's what I care about. The biggest bug that bugs me is the popup not going away on its own after taking a screenshot. I'm pretty sure that's going to be taken care of in the next firmware update. Beyond that, I hope it keeps working forever until I get bored and get something shinier and newer.
My only big wish for the scope would be a more efficient bode plot that doesn't take as long. But I can live with it as it is if I need to.
tautech:
--- Quote from: ebastler on November 21, 2023, 07:47:10 pm ---Heck, am I the only one who gets disappointed when tautech indicates that we will eventually get another firmware release (with a new major version number, and after more than a year of no user-facing updates) -- only to be told by Performa01 that it will not be a major update, but will comprise a single, small new feature? :-\
--- End quote ---
As mentioned there will be improvements that won't be mentioned in release notes....there always is.
However full tests of this FW need be made and errors sorted.....
--- Quote from: 2N3055 on November 21, 2023, 01:25:04 pm ---Adding new features logically happens on platform level and then gets propagated to it's members, based on product placement, architecture and technical details.
--- End quote ---
This ^
One improvement hoping to be rolled into 2kX+ and gradually into all models is colored trace markers in Digital mode....a little but nice improvement I suggested some months back which was accepted by Siglent and endorsed by all beta testers.
Improvement is continual......just as it should be.
However development teams currently also have a lot on their plate......
Martin72:
--- Quote --- I'm pretty sure that's going to be taken care of in the next firmware update.
--- End quote ---
Me too. ;)
The Scope has been on the market for 3 years, here is an (incomplete) list of innovations and improvements that have been incorporated since then in addition to the bug fixes.
* Supported to hide analog traces
Added function expression information to the math box
Frequency counter supported to count serial trigger events
Supported Sign as a math function
New serial protocols supported (optional):
SENT, trigger & decode
Manchester, decode only
Measurement enhancement:
Supported user-defined thresholds (Upper, Middle and Lower): Measure | Config | Threshold
Added items: +Area@AC, -Area@AC, Area@AC, AbsArea@AC
Math: added new operator – Interpolate
Display:
Supported selectable color for traces: Display | Color Setting
Supported floating menu so that the waveform is not compressed horizontally when the right-side menu is displayed: Display | Menu Style
Supported to show bandwidth information on the channel descriptor box
Save/Recall:
Added option “Save all channel” for csv file
Supported to save math traces (except FFT)
Supported serial trigger as source of the frequency counter
Unlocked zoom in stop mode for Roll
Supported trigger Default or AutoSetup operations by pressing corresponding button twice
Measurement enhancement:
Supported cursors for measurement
Supported Track plot
Added parameters: Positive Slope and Negative Slope
Display:
Supported to display axis label
Auto hide menu setting increased to 10 seconds
Added “Hide” button in the fast menu of channel for quickly hiding trace
Save/Recall:
Supported network storage
Supported to Print only grid area
Supported to save FFT result
Updated the File Manager
Bode Plot:
The maximum generator amplitude changed from 6Vp-p to 24Vp-p
Supported single sweep
Horizontal: Time Zero strategy updated – time zero changed from center display to trigger position.
Supported setting Cursors by virtual keypad
* That's quite unique from one manufacturer, I don't know of any other, so we should "turn a blind eye" if there hasn't been another banger recently.... ;)
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version