Siglent SDS800X HD 12 bit DSO's

[eTobey]

This must be an outdated information then.

Even their manuals is a mess lol. As all of the rest isnt hard enough... The wildcard is not even mentioned in the CAN section. Couldnt find those in the other sections too.

[eTobey]

Quite obviously, there ia a request for wildcards in the equality condition for serial ID and/or Data triggers, like it is already implemented for SPI (see attachment).

It should be quite obvious, that on the CAN bus, there are not only three ID´s. So i would call a filter a *requirement*. I already worked many hours on a problem, and trying to figure out (learning) how to use this scope for that. Its a challenge with these limitations!

[2N3055]

Quite obviously, there ia a request for wildcards in the equality condition for serial ID and/or Data triggers, like it is already implemented for SPI (see attachment).

It should be quite obvious, that on the CAN bus, there are not only three ID´s. So i would call a filter a *requirement*. I already worked many hours on a problem, and trying to figure out (learning) how to use this scope for that. Its a challenge with these limitations!

Yes it is obvious to anybody that CAN supports more than one ID and Data. It is a communications protocol. If there was only one message being exchanged we would not need full communications protocol. Just toggle one pin on and off. It would be predetermined constant.

Could you please stop calling it a filter.
It is filter if you filter already captured full data.

In trigger it is a trigger mask, and function you are talking about is wildcards in a mask.

Your ambiguous usage of terminology was primary reason why I had to ask 5 follow up questions to understand what you are talking about.

[eTobey]

Your ambiguous usage of terminology was primary reason why I had to ask 5 follow up questions to understand what you are talking about.

I said: "filter for a range of IDs"
You asked: "Filter in decoded list or in trigger?"
I said: "Its the trigger"

From then on, i would say that it was pretty clear what my intentions were even though my terminology may not have been right.

I would call a "range" not exactly a mask. Try to mask the range of 1-2 !

[2N3055]

Your ambiguous usage of terminology was primary reason why I had to ask 5 follow up questions to understand what you are talking about.

I said: "filter for a range of IDs"
You asked: "Filter in decoded list or in trigger?"
I said: "Its the trigger"

From then on, i would say that it was pretty clear what my intentions were even though my terminology may not have been right.

I would call a "range" not exactly a mask. Try to mask the range of 1-2 !

That is exactly my point. I had to ask trice. Also me, asking  "Filter in decoded list or in trigger?" was not me validating your use of terminology but trying to ask you the way you seem to speak... But all that is beside point. We understand now.

My expensive Keysight have ID trigger mask, it is binary or hex.  There is no interval, unless it can be expressed by a single digit in a binary or hex mask.
It appears Keysight uses term filter by ID in CAN error triggers, but that is post error trigger filter. It is error trigger and then filter on ID as a second step.

Binary masks is what can be accomplished by binary logic running full speed in FPGA.
So in your case you could make a mask that works in range 0-3:
00xx

Could you please provide a link to Rigol document where it can be seen how it works there? I'm curious how it was implemented.

[eTobey]

Rigol used a X for a wildcard on binary and hex.

I think there is no documentation. It seems to just have been chucked into the 800 from the 900 recently.

If there is no range possible, it would be nice to be able to extract the decoded CAN (with frame time) and then beeing able to filter it via spreadsheet. Or even "filter" those in the CAN list that is below.

I tried to do it with sequence, but with that i only can capture like 2265 frames. If i would be able to set  a 5K max memory depth, then i could fit all my frames in there, and then get it out via stepping through the sequences with SCPI, which is all in all quite a task, that could otherwise easily be implemented into the scope (and webinterface!).

To bad that it is not open source. I would have loved to program that in!

[2N3055]

Could you please provide a link to Rigol document where it can be seen how it works there? I'm curious how it was implemented

https://int.rigol.com/file/DHO800%20User%20Manual.pdf

Chapter 8.7.16 describes the CAN trigger. For details regarding the mask entry (page 106), it refers to the description in section 8.7.14 (I²C trigger, data entry, page 99).

Wildcard entry is implemented just like for SPI data in the Siglent 800X HD: You get a keypad which has the digits for binary, decimal or hex entry enabled, plus a pseudo-digit "X" for don't-care digits.

Thank you!
So I was right, it is a single digit binary/hex mask with wildcards, not arbitrary intervals.
That makes sense.

[Martin72]

I just had the two "little ones" in operation, decoding.
I don't know which visual list display I would prefer.
I will make some comparisons between the two, at least until my new scope arrives.
If you want to know anything else, go ahead.

[eTobey]

Capture 100 frames, and export them in csv or what ever. But make a video about your struggles 

[Martin72]

I have now exported the list of 115 decoded frames(csv), if that is what you meant, it was no big deal.

[tautech]

I just had the two "little ones" in operation, decoding.
I don't know which visual list display I would prefer.

Will a mouse scrollwheel work on the HD list ? ....not tried it. 

[Martin72]

Hi Rob,

Just tried it out, yes, no problems, I could get used to the mouse control.
Oh yes, it's not in the data sheet, but just tested, maximum frames (CAN) are 2000.

[electronics hobbyist]

Rigol used a X for a wildcard on binary and hex.

I think there is no documentation. It seems to just have been chucked into the 800 from the 900 recently.

If there is no range possible, it would be nice to be able to extract the decoded CAN (with frame time) and then beeing able to filter it via spreadsheet. Or even "filter" those in the CAN list that is below.

I tried to do it with sequence, but with that i only can capture like 2265 frames. If i would be able to set  a 5K max memory depth, then i could fit all my frames in there, and then get it out via stepping through the sequences with SCPI, which is all in all quite a task, that could otherwise easily be implemented into the scope (and webinterface!).

To bad that it is not open source. I would have loved to program that in!

Each product has its own limitations, and it is impossible for a single product to fully meet the current and future needs of users. Your requirements are reasonable, and I believe Siglent will improve them based on your suggestions, including the request for wildcards in the equality condition for serial ID and/or Data triggers here and the can frame time saving function in another post.

[electronics hobbyist]

Hi Rob,

Just tried it out, yes, no problems, I could get used to the mouse control.
Oh yes, it's not in the data sheet, but just tested, maximum frames (CAN) are 2000.

The maximum frame limit for decoding should be written in the data sheet.

[Martin72]

The data sheets for the 800 and 1000X HD do not mention this, only from the 2000Xplus onwards.
I assume that they are the same for the 800/1000, I will check this as soon as possible.

[electronics hobbyist]

I suggest that when asking questions, please describe them in detail, preferably in the following form:

1. What is the signal I need to analyze and what is the expected result
2. How is one feature implemented in another product? I hope sds800x hd also supports it
3. Is there a certain function/is there a restriction on a certain function, which will affect whether I purchase sds800x hd
...

Everyone comes from all over the world, and there may be misunderstandings due to translation. It is best to include pictures and videos if possible.

[eTobey]

I have now exported the list of 115 decoded frames(csv), if that is what you meant, it was no big deal.

I have to laugh about myself. That task i gave you was too easy. 

Working a bit to much, i guess.
My problem is, that when i use 500k CAN speed, with 1s/div (max timebase to capture for that rate), i can only export 2000 entries, which are also corrupted (the shown data only), if you set the timebase above 100ms/div. Only 2000 frames is quite poor in my opinion, since we are talking here about ~ 20kb. And it should decode properly, what ever the zoom is, because one might think that this data is corrupted in reallity! That sends you far on the wrong track. A device for over 500 bucks should not show complete wrong data.



In other words:
I need to be able to capture and export 20.000 CAN frames (in 10s) with a range of IDs (they can be filtered out on export in this case, additionally to the trigger mask).

Another thing:
After a crash, i was was trying (thinking) to reload a waveform, thinking "ah no problem, i can just reload the waveform and decode it again". I was dissapointed. 
This would be a really nice feature, since sometimes you also hit the wrong button, and loose that waveform.

[2N3055]

I have now exported the list of 115 decoded frames(csv), if that is what you meant, it was no big deal.

I have to laugh about myself. That task i gave you was too easy. 

Working a bit to much, i guess.
My problem is, that when i use 500k CAN speed, with 1s/div (max timebase to capture for that rate), i can only export 2000 entries, which are also corrupted (the shown data only), if you set the timebase above 100ms/div. Only 2000 frames is quite poor in my opinion, since we are talking here about ~ 20kb. And it should decode properly, what ever the zoom is, because one might think that this data is corrupted in reallity! That sends you far on the wrong track. A device for over 500 bucks should not show complete wrong data.



In other words:
I need to be able to capture and export 20.000 CAN frames (in 10s) with a range of IDs (they can be filtered out on export in this case, additionally to the trigger mask).

Another thing:
After a crash, i was was trying (thinking) to reload a waveform, thinking "ah no problem, i can just reload the waveform and decode it again". I was dissapointed. 
This would be a really nice feature, since sometimes you also hit the wrong button, and loose that waveform.

I went and took a look at your post history.

Now it is clear to me that you are complete beginner in this whole oscilloscope thing.
I'm not going to spend time explaining why your expectations are unrealistic, just simply state that they are.

Let just put it simply: your expectations are unrealistic. Even on many scopes costing 50x more.
And 500€ scope is very inexpensive.  Until 10 years ago scopes that could to 10% of what it does used to start at 5000 USD price.

To put things in perspective, one (1, single) Keysight N2873A passive probe that came with my MSOX3104T has a retail price of 558 €. A single probe.
And no, MSOX3104T (a 12000€ MSRP scope when was new) can do maybe 1000 CAN packets at your specification.

And if you asked here what tool to buy for this CAN task you do, we would tell you to get some kind of CAN analyzer (something like Vector or Kvaser).
Or Picoscope, but not the cheapest one. I can do what you ask on a Pico 3406D that has 512Mpt memory, and cost €2405..
But that Picoscope is USB scope that is also missing other stuff, like any type of triggering on serial protocols etc.
But for this job of yours it would work.

[eTobey]

Now it is clear to me that you are complete beginner in this whole oscilloscope thing.

You should have though about this already when you said i had some bad terminology. But in fact this is not true.

I do not ask about things that are not implemented, im talking about a tiny bit more effort to have the funktionality perfectly useable.

About my specs:
If i havent misscalculated, those would mean a frame every 500us. In fact, i can capture and decode frames that come every 220us until the list is full with 2000 entries (~32kb is not so much memory nowadays). You may have misscalculated?

See the pictures: 1s/div looks good, 2s/div is the limit, or even a bit above. 10s with 250us frames would mean 40000 CAN frames. Even if it may not be possible to have 40000 in that list. It should be easily possible to export those!

[2N3055]

Now it is clear to me that you are complete beginner in this whole oscilloscope thing.

You should have though about this already when you said i had some bad terminology. But in fact this is not true.

I do not ask about things that are not implemented, im talking about a tiny bit more effort to have the funktionality perfectly useable.

About my specs:
If i havent misscalculated, those would mean a frame every 500us. In fact, i can capture and decode frames that come every 220us until the list is full with 2000 entries (~32kb is not so much memory nowadays). You may have misscalculated?

I haven't got miscalculated anything.

Let me guess.. You are software developer? Right?

[eTobey]

I haven't got miscalculated anything.

Let me guess.. You are software developer? Right?

Review my edited post.

Yes i do develop software.

[eTobey]

Trying to catch an error, i thought rolling would be perfect for CAN debug:

Got the error, hit stop, went to decode: "Function can not be used in roll mode". 

[2N3055]

I haven't got miscalculated anything.

Let me guess.. You are software developer? Right?

Review my edited post.

Yes i do develop software.

That explains things.

You live in virtual world where everything is possible. If your software needs 128TB of RAM and 256 core processor, you deadly seriously coldly "order" customer to buy better computer. And if that computer doesn't exist, it is not your problem.
And stupid hardware developers should develop "magical" computers that doesn't exist.

Because everything is possible in dream world..

I keep trying to explain some things about realities of real world devices and their limitations, thinking you need help.

But you are not here to ask for help.
You are here to explain how everybody else except you is stupid.
Good for you.

Wish you luck with your endeavors.

[electronics hobbyist]

Now it is clear to me that you are complete beginner in this whole oscilloscope thing.

You should have though about this already when you said i had some bad terminology. But in fact this is not true.

I do not ask about things that are not implemented, im talking about a tiny bit more effort to have the funktionality perfectly useable.

About my specs:
If i havent misscalculated, those would mean a frame every 500us. In fact, i can capture and decode frames that come every 220us until the list is full with 2000 entries (~32kb is not so much memory nowadays). You may have misscalculated?

See the pictures: 1s/div looks good, 2s/div is the limit, or even a bit above. 10s with 250us frames would mean 40000 CAN frames. Even if it may not be possible to have 40000 in that list. It should be easily possible to export those!

You're thinking too simply. Hardware resources are limited, and it's not just these results that require memory.
The calculation process will consume a lot of memory resources.
You also need to consider that users have opened many functions. If you write this code, you should be able to understand.
If you need 40000 frames, another person needs 80000 or 160000 frames, oscilloscope manufacturers can't meet this unlimited requirement.
Keysight can sell for $10000 with a storage depth of 4M, and it may not meet your needs.

[Martin72]

I have to laugh about myself. That task i gave you was too easy. 

I went back late in the evening and made the little video, and the only thing you say about it is that "my task" was too easy.
Thanks for nothing.
Others have already written about the rest.
When the first cheap scopes came out that could decode the basic protocols, we were happy and bought a few for our developers' desks - it was still cheaper than a dedicated device that could do nothing but that.
As always, it depends on the task as to when which device is sufficient and for us, the certainly not fully comprehensive functions of an oscilloscope are sufficient, as we do not develop software but hardware.
There is always something small that can be improved, but you certainly can't expect there to be a scope that can do everything and then for as little money as possible.
Nobody who knows anything about it expects that.