EEVblog Electronics Community Forum

Products => Test Equipment => Topic started by: snarkysparky on September 01, 2018, 02:57:08 pm

Title: DSO wanted feature.
Post by: snarkysparky on September 01, 2018, 02:57:08 pm
I need to watch for deviant conditions with my new SDS1104 X-E.   

Would love to be able to plug a flash drive into USB and set the scope to store all acquired waveforms directly to the stick while I am off doing something else.   Until memory is full or I stop it.

Is this a thing available.   And yes the events are spaced far enough apart so that a file could be written before the next likely event.

Title: Re: DSO wanted feature.
Post by: Fungus on September 01, 2018, 03:00:54 pm
Have you looked into the "Pass/Fail" test mode?
Title: Re: DSO wanted feature.
Post by: nctnico on September 01, 2018, 03:53:05 pm
I need to watch for deviant conditions with my new SDS1104 X-E.   

Would love to be able to plug a flash drive into USB and set the scope to store all acquired waveforms directly to the stick while I am off doing something else.   Until memory is full or I stop it.

Is this a thing available.   And yes the events are spaced far enough apart so that a file could be written before the next likely event.
That is called data logging. The scopes from GW Instek and Yokogawa have this feature.
Title: Re: DSO wanted feature.
Post by: wpwrak on September 03, 2018, 12:54:18 am
The 1000X-E does indeed seem to lack this. Even the beeper function is quite under-developed. (The beeper continuously alerts after the first failure, but with variable delay between beeps, which makes it sound as if the scope was complaining about new failures.)

One more item for the "alternative firmware" to do list, I guess ;-)
Title: Re: DSO wanted feature.
Post by: tautech on September 03, 2018, 02:21:53 am
The 1000X-E does indeed seem to lack this. Even the beeper function is quite under-developed. (The beeper continuously alerts after the first failure, but with variable delay between beeps, which makes it sound as if the scope was complaining about new failures.)

One more item for the "alternative firmware" to do list, I guess ;-)
I think you could offer better/more info.
Certainly Pass/Fail can do this but you need a consistent repetitive waveform so the Trigger suite is the first thing you'd use IF you know what your're looking for. Until you do use Infinite Persistence to get an idea of what the rogue waveform actually looks like then home in on it using the many available Trigger configurations.

Then later once the rogue waveform is known and can be triggered on you can set up the Search feature and capture each time it occurs to check if it's repetitive or something else is triggering/causing it. Then with any of the other 3 channels at your disposal check for correlation/s to the probable source.

You have a perfectly capable tool and all that is required is some thought as to how to best use it.  :)


Edit to add;
Search example:
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1370717/#msg1370717 (https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/msg1370717/#msg1370717)
Title: Re: DSO wanted feature.
Post by: wpwrak on September 03, 2018, 03:39:14 am
Certainly Pass/Fail can do this

I mean that the test/pass function can't record the offending waveform(s), or for that matter do anything beyond counting and (optionally) permabeeping.

If you can translate the excursion to a specific trigger, then you have more options. (Also, if you have a second scope, with segmented memory, you could use that as recorder.) But the case where you'd want recording most is probably when you don't understand the failure pattern just yet.

Quote
You have a perfectly capable tool and all that is required is some thought as to how to best use it.  :)

Oh don't worry. I have big plans for that little scope ;-)
Title: Re: DSO wanted feature.
Post by: tautech on September 03, 2018, 03:46:50 am
If you can translate the excursion to a specific trigger, then you have more options.
There are a few work around's, Runt and Period triggers spring to mind.
Once you know what the excursion is and how to accurately and repeatably trigger on it the battle is won.
Then we have the other channels to find an event correlation on signals that are the cause of it.
Title: Re: DSO wanted feature.
Post by: Fungus on September 03, 2018, 08:44:09 am
Would love to be able to plug a flash drive into USB and set the scope to store all acquired waveforms directly to the stick while I am off doing something else.   Until memory is full or I stop it.

At 1Gb/sec...?  :-DD

Title: Re: DSO wanted feature.
Post by: nctnico on September 03, 2018, 10:48:44 am
If you can translate the excursion to a specific trigger, then you have more options.
There are a few work around's, Runt and Period triggers spring to mind.
Once you know what the excursion is and how to accurately and repeatably trigger on it the battle is won.
Then we have the other channels to find an event correlation on signals that are the cause of it.
But the purpose of data logging is to do a very long term measurement to find a very rare event. Maybe even with power cycles in between. Storing data of each waveform with maximum details will be useful here because you can't reproduce it quickly let alone go back and adjust a trigger condition.
Title: Re: DSO wanted feature.
Post by: rf-loop on September 03, 2018, 12:54:38 pm
I need to watch for deviant conditions with my new SDS1104 X-E.   

Would love to be able to plug a flash drive into USB and set the scope to store all acquired waveforms directly to the stick while I am off doing something else.   Until memory is full or I stop it.

Is this a thing available.   And yes the events are spaced far enough apart so that a file could be written before the next likely event.

Can you explain more this signal and what kind of events (how long lasting, is this event what can generate trig and what is importantly needed sampling speed. and more if available)

I tell example.
Usually between 10 - 110s interval randomly happen some 2 - 5V spike what last max 200ns and I want full 1GSa/s and I want see every spike and also period and time from beginning of recording.
Outside of this spike I'm not interested what happend when level is under this 2V.

If average is 50s interval Sig can record every peak and total time then over 41 day if I do not interrupt it or power breaks.
Of course with other settings and signals all is different. Just only tip what history buffer or sequence acquisition can do.

But, it do not store it directly to USB. Result is in scope memory. Yes it can take out but not so simple that it only continuously save it to USB flash.

But, if you can give more information then can see what Siglent can perhaps do or not.
Title: Re: DSO wanted feature.
Post by: snarkysparky on September 07, 2018, 12:41:55 pm
The trigger event will happen maybe once an hour.  I want the scope to log the entire waveform.  The record length will be set to 10M points or whatever is practical.

I need to be able to review the stored waves to find a deviant event in product testing.  Sampling speed of 100k/sec would be about right.

I am reading GW instek manual 1000B series but I cannot confirm that the waveforms are stored to internal memory automatically.

Anybody know for sure of a scope to do this?
Title: Re: DSO wanted feature.
Post by: wpwrak on September 07, 2018, 01:49:08 pm
Since the event rate is low, perhaps you could set up waveform capture and test, poll by software over the SCPI interface, and download the waveform if the scope indicates a fail or trigger event.

Now, I don't know if any scope lets you have a regular capture running during a pass/fail test, and you'd have the additional problem that test and waveform capture may not want to use the same trigger. A way to turn pass/fail into a trigger would be to use a secondary scope for pass/fail and then use its pass/fail output signal to trigger the scope that captures the waveforms.

You'd then either use segmented memory to collect waveforms, or poll that scope from a PC, and, when detecting that it has triggered, download the waveform(s) and set up a new capture.
Title: Re: DSO wanted feature.
Post by: rf-loop on September 07, 2018, 02:00:58 pm
The trigger event will happen maybe once an hour.  I want the scope to log the entire waveform.  The record length will be set to 10M points or whatever is practical.

I need to be able to review the stored waves to find a deviant event in product testing.  Sampling speed of 100k/sec would be about right.

I am reading GW instek manual 1000B series but I cannot confirm that the waveforms are stored to internal memory automatically.

Anybody know for sure of a scope to do this?

Every "once/hour" record need be 10s or more length?

Siglent (example SDS1000X-E models) can not do this.
If look 100ks/s sampling speed. 100sec (140s in Siglent) acquisition it can keep only one. (14M)

If one  acquisition length is 14s and  100ks/s it can store 17 sequential trigged acquisitions. 17*1.4M
If one  acquisition length is 1.4s and  100ks/s it can store 188 sequential trigged acquisitions. 188* 140k
If one  acquisition length is 0.14s and  100ks/s it can store 1891sequential trigged acquisitions. 1891* 14k

After acquisition they all are in scope internal wfm buffer memory where from these need transfer to computer or analyze using scope functions.

Very old Siglent (some SDS1000CML etc  models) can do some kind of long time acquisition directly to USB flash but I do not remember tested working limits and samplerates what can use this feature.





Title: Re: DSO wanted feature.
Post by: MrW0lf on September 07, 2018, 02:02:28 pm
Now, I don't know if any scope lets you have a regular capture running during a pass/fail test,

https://www.youtube.com/watch?v=a65_EsTpAEQ (https://www.youtube.com/watch?v=a65_EsTpAEQ)

https://www.youtube.com/watch?v=xPGSQ0UaekA (https://www.youtube.com/watch?v=xPGSQ0UaekA)

Some combination of above may work but beware, this is all done in PC software so suitable only for relatively seldom events, probably not more 1Hz or so.
Title: Re: DSO wanted feature.
Post by: snarkysparky on September 07, 2018, 02:09:44 pm
Thanks for all input.

Would really like it to be all in the scope but I guess that doesn't exist.

Scope builders it would be a good feature....

Title: Re: DSO wanted feature.
Post by: nctnico on September 07, 2018, 02:53:49 pm
The trigger event will happen maybe once an hour.  I want the scope to log the entire waveform.  The record length will be set to 10M points or whatever is practical.

I need to be able to review the stored waves to find a deviant event in product testing.  Sampling speed of 100k/sec would be about right.

I am reading GW instek manual 1000B series but I cannot confirm that the waveforms are stored to internal memory automatically.

Anybody know for sure of a scope to do this?
I did some testing with my GDS-2204E and the data logging app. This seems to work but there is one issue though. You need to set a capture interval with a minimum time of 2 seconds. This interval seems to count down only when the oscilloscope is acquiring data. To capture each trigger event you have to make sure acquiring the waveform takes 2 seconds or more. A timebase setting of 200ms/div will do that.
Title: Re: DSO wanted feature.
Post by: 2N3055 on September 07, 2018, 07:32:39 pm
Picoscope was born for that. They have scopes with up to 2 Gsamples memory.
Is software you can setup event, that when buffers are full, it will flush buffers to disk. And then restart capturing...

And then in software you can open all files and inspect them, DeepMeasure them, apply math channels, compare them to ref channel....
Title: Re: DSO wanted feature.
Post by: tautech on September 07, 2018, 07:49:41 pm
Thanks for all input.

Would really like it to be all in the scope but I guess that doesn't exist.

Scope builders it would be a good feature....
Not once have you described the 'event' !

Do you know if you can trigger on it ?

If you can and have a slow timebase setting the 'event' will be in the center of the display and you'll have seconds of pre and post trigger info to investigate.
Requirements such as theses are just what many the trigger types and Single and Normal triggers are for ! Add sufficient memory like DSO's have and it's constantly being updated and flushed but when triggered all that info is available to inspect.
Such is the power of a DSO.
Title: Re: DSO wanted feature.
Post by: snarkysparky on September 07, 2018, 07:56:14 pm
tau

If i were designing the feature for my scope the 'event' is any acquisition.

Very simply.

1  wait for trigger.
2  after one triggered sweep, stop and write file to USB.
3  re arm trigger.
4  goto 1

Cmon,  get that feature implemented.  Would pave the way for scopes to cross over into data logging territory.

Title: Re: DSO wanted feature.
Post by: tautech on September 07, 2018, 08:04:48 pm
tau

If i were designing the feature for my scope the 'event' is any acquisition.

Very simply.

1  wait for trigger.
2  after one triggered sweep, stop and write file to USB.
3  re arm trigger.
4  goto 1

Cmon,  get that feature implemented.  Would pave the way for scopes to cross over into data logging territory.
SCPI commands are NOT my strength, if fact a weakness, however I'd be VERY surprised if this ^ cannot be done with the SCPI command set in most DSO's.
Now we know what you'd like to accomplish I'm sure you'll get the pointers you need to accomplish what you want.  :)
Title: Re: DSO wanted feature.
Post by: nctnico on September 07, 2018, 08:05:38 pm
tau

If i were designing the feature for my scope the 'event' is any acquisition.

Very simply.

1  wait for trigger.
2  after one triggered sweep, stop and write file to USB.
3  re arm trigger.
4  goto 1

Cmon,  get that feature implemented.  Would pave the way for scopes to cross over into data logging territory.
As I wrote before: Yokogawa scopes have this feature for sure. The GW Instek scopes can do this too with a simple workaround. I just checked and the data logger application is also available for the GDS1000B series.
Title: Re: DSO wanted feature.
Post by: Mechatrommer on September 07, 2018, 08:17:20 pm
tau

If i were designing the feature for my scope the 'event' is any acquisition.

Very simply.

1  wait for trigger.
2  after one triggered sweep, stop and write file to USB.
3  re arm trigger.
4  goto 1

Cmon,  get that feature implemented.  Would pave the way for scopes to cross over into data logging territory.
SCPI commands are NOT my strength, if fact a weakness, however I'd be VERY surprised if this ^ cannot be done with the SCPI command set in most DSO's.
he said he doesnt want any programming or any PC related mess, he wants "built-in to the scope". if 100 people with 100 different trigger/logging requirements, the scope should implement them all.  ::)
Title: Re: DSO wanted feature.
Post by: tautech on September 07, 2018, 08:49:49 pm
tau

If i were designing the feature for my scope the 'event' is any acquisition.

Very simply.

1  wait for trigger.
2  after one triggered sweep, stop and write file to USB.
3  re arm trigger.
4  goto 1

Cmon,  get that feature implemented.  Would pave the way for scopes to cross over into data logging territory.
SCPI commands are NOT my strength, if fact a weakness, however I'd be VERY surprised if this ^ cannot be done with the SCPI command set in most DSO's.
he said he doesnt want any programming or any PC related mess, he wants "built-in to the scope". if 100 people with 100 different trigger/logging requirements, the scope should implement them all.  ::)
Where ?
Did I miss something ?

Yes even the best featured entry level DSO the marketplace has ever seen does not have every thing on ppls wish list.  :horse:
But like most modern DSO's, the X-E can get the info snarkysparky wants if he uses some thought and work arounds to do it. He already has this DSO, he just needs to implement settings/procedures to capture his event/s.
Title: Re: DSO wanted feature.
Post by: 2N3055 on September 07, 2018, 08:54:39 pm
tau

If i were designing the feature for my scope the 'event' is any acquisition.

Very simply.

1  wait for trigger.
2  after one triggered sweep, stop and write file to USB.
3  re arm trigger.
4  goto 1

Cmon,  get that feature implemented.  Would pave the way for scopes to cross over into data logging territory.

Picoscope does that. Also Tek MDO3000 can.
Title: Re: DSO wanted feature.
Post by: tautech on September 07, 2018, 09:46:30 pm
I need to watch for deviant conditions with my new SDS1104 X-E.   

Would love to be able to plug a flash drive into USB and set the scope to store all acquired waveforms directly to the stick while I am off doing something else.   Until memory is full or I stop it.

Is this a thing available.   And yes the events are spaced far enough apart so that a file could be written before the next likely event.
Further thoughts......
You would love to but X-E DSO's are not the correct tool to do this. For one they have no RTC to give timestamps as when an event occurs. With X-E's a met trigger condition creates a zero point in time where before and after waveforms can be examined in times relative to the captured 'zero' time trigger point.

With X-E's to accomplish what you want to do, Trigger conditions to capture the event must be known/found and set to do so. Then X-E's have the tools to examine events in detail and then reset trigger conditions to find events on other bus's so the cause can be identified.

With external PC capture/logging there will be methods to apply real timestamps so the realtime cause for events can be investigated. Also the rear Trig Out BNC could be useful for instant notification (at logic levels) when an event has triggered the DSO.

All the tools are there, you just need find one that you can work with.
Title: Re: DSO wanted feature.
Post by: vk6zgo on September 08, 2018, 04:17:16 am
How fast are the "deviant conditions" when they occur?
If they are slow enough, you could go "retro" & use a chart recorder.

If it is an urgent problem, my other suggestion is forget about messing with your own 'scope & hire something which can do the job.
Title: Re: DSO wanted feature.
Post by: snarkysparky on September 08, 2018, 03:28:05 pm
Well this is for a work project.  I have three stations I need to monitor signals on for the errant condition. Asked boss if we can buy three budget scopes for this as this is the only way.  If I can't get a decently cheap scope with this ability ( Yokagawa will break budget ) I have another way.  The machines are each have a pc running a VB6 program.  I plan to add serial ports to each pc via USB and edit the code to output a byte  ( any byte will do ) if the event happens which will trigger a single capture on my cheap scopes because they will be monitoring the TX data line out of the serial port.

I just hoped for the lazy way.  It still seems like a simple feature to add to a scope.  I can do it manually by setting single shot aquisition and push file save after a capture.  Scope firmware just needs to hook into those functions.

Title: Re: DSO wanted feature.
Post by: 2N3055 on September 08, 2018, 04:02:08 pm
Again, Picoscope has the function to save whole waveform after trigger and than wait for another one. No need to program anything..

But you seem to be avoiding explaining what is exact signal you want to trigger on? Are you waiting for any signal, or you need to setup complex trigger condition?
Also how many signals do you need recorded at the time?

Title: Re: DSO wanted feature.
Post by: Mechatrommer on September 08, 2018, 04:23:40 pm
I have three stations I need to monitor signals on for the errant condition. Asked boss if we can buy three budget scopes for this as this is the only way.
whats the budget? if its $300 per scope then there is an offer in ebay right now... https://www.ebay.com.my/itm/RIGOL-DS1054Z-Digital-Oscilloscope/273448409163?hash=item3faacb984b:g:7skAAOSwFdpbMgGb (https://www.ebay.com.my/itm/RIGOL-DS1054Z-Digital-Oscilloscope/273448409163?hash=item3faacb984b:g:7skAAOSwFdpbMgGb) ;D

The machines are each have a pc running a VB6 program.
you are just one step close to what you want. there is programming guide for siglent, but if you dont want to invent a wheel, get a DS1054Z and someone or me may help you in programmming for a pack of coffee. ;)

I plan to add serial ports to each pc via USB and edit the code to output a byte  ( any byte will do ) if the event happens which will trigger a single capture on my cheap scopes because they will be monitoring the TX data line out of the serial port.
sounds like a nasty/poor hack, lets hope the cheap scope can send back the data or save it internally upon pc command.
Title: Re: DSO wanted feature.
Post by: wpwrak on September 09, 2018, 12:49:32 am
You would love to but X-E DSO's are not the correct tool to do this. For one they have no RTC to give timestamps as when an event occurs.

Just connect Ethernet to your 1000X-E, log in (as root), make sure it can reach the global Internet, then run
Code: [Select]
ntpd -n -d -p 201.217.3.85
and watch the miracle unfold :-)
Title: Re: DSO wanted feature.
Post by: wpwrak on September 09, 2018, 01:15:15 am
I plan to add serial ports to each pc via USB and edit the code to output a byte  ( any byte will do ) if the event happens which will trigger a single capture on my cheap scopes because they will be monitoring the TX data line out of the serial port.

Hmm, so it is the PC that detects the event, and not the scope ?

And you can normally just trigger the scope using an SCPI command.

If you need the scope only to archive waveforms at the time of the event, but the detection happens in the PC, things are really easy:
- set up the scope,
- trigger: set it such that the scope never triggers on its own
- stop capture, arm the trigger (the scope now captures the samples that will go before the trigger), wait for the event
- when the event happens, tell the scope to trigger
- wait until it has completed the capture (use a fixed time or, better but not really necessary in this case, poll its status)
- download and store the waveform
- stop capture, etc., again (and so on)

Some details would depend on the scope.
Title: Re: DSO wanted feature.
Post by: tautech on September 09, 2018, 02:35:00 am
You would love to but X-E DSO's are not the correct tool to do this. For one they have no RTC to give timestamps as when an event occurs.

Just connect Ethernet to your 1000X-E, log in (as root), make sure it can reach the global Internet, then run
Code: [Select]
ntpd -n -d -p 201.217.3.85
and watch the miracle unfold :-)
Can you please share all about this in the SDS1104/1204X-E thread ?
Some 'dummy guide' and usage examples would be much appreciated.  :)
https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/ (https://www.eevblog.com/forum/testgear/siglent-sds1204x-e-released-for-domestic-markets-in-china/)
Title: Re: DSO wanted feature.
Post by: wpwrak on September 09, 2018, 02:09:04 pm
Can you please share all about this in the SDS1104/1204X-E thread ?
Some 'dummy guide' and usage examples would be much appreciated.  :)

Okay, I'll add a few more details there. But I don't want to make it too detailed or too easy, since part of this is also the path to unlocking higher bandwidth, and I don't want to make Siglent feel threatened and pull a Sony on us.

My long-term goal is to have an easily applicable solution for running an alternative (Open Source) firmware (including the programmable logic in the FPGA) that does not depend on altering any of Siglent's code, and that can be modified/extended/customized at will. Hopefully that should make the platform popular enough that Siglent wouldn't feel compelled to make things difficult.