NEW 100MHz Keysight Scope Release on 1st March 2017

[daveshah]

£113 ex vat for I²C, SPI and UART decoding doesn't seem too bad

http://uk.farnell.com/keysight-technologies/dsox1embd/embedded-decodes-analysis-digital/dp/2706569?ost=DSOX1EMBD&selectedCategoryId=&categoryNameResp=All%2BCategories&searchView=table&iscrfnonsku=false

[nctnico]

Is this the one?

http://www.farnell.com/datasheets/2189548.pdf

Why the hell did they even bother to make this? 1Mpts on Keysight means 250kpts in real usage scenarios. No, segmented memory is far less useful than raw memory depth. Scrap the Toys it says in the datasheet. Well... why don't you guys at Keysight do that? FFS get real! This scope is clearly for the educational market where nobody gets fired for buying Tektronix or Keysight but other than that nobody with a sane mind would spend money on a scope like this.

 In Trumpish: Total Disaster. For Keysight's sake I hope it is a fake datasheet. 

[ebastler]

£113 ex vat for I²C, SPI and UART decoding doesn't seem too bad

http://uk.farnell.com/keysight-technologies/dsox1embd/embedded-decodes-analysis-digital/dp/2706569?ost=DSOX1EMBD&selectedCategoryId=&categoryNameResp=All%2BCategories&searchView=table&iscrfnonsku=false

So the commercial option is cheaper than the corresponding option for the EDU version which I had linked to above?!
OK, I will stop pretending that I understand Keysight's pricing structure then... 

[AndyC_772]

£113 is the price for DSOX1AUTO, DSOX1EMBD and EDUX1EMBD. All three are the same cost.

But:DSOX1EMBD offers I2C, SPI and UART, but EDUX1EMBD only offers I2C and UART, not SPI.

Clearly nobody needs to learn about SPI 

[mikeselectricstuff]

£113 is the price for DSOX1AUTO, DSOX1EMBD and EDUX1EMBD. All three are the same cost.

But:DSOX1EMBD offers I2C, SPI and UART, but EDUX1EMBD only offers I2C and UART, not SPI.

Clearly nobody needs to learn about SPI 

SPI on a 2-channel scope ?

[AndyC_772]

I'm just quoting from the data sheet 

[Neganur]

Why the hell did they even bother to make this? 1Mpts on Keysight means 250kpts in real usage scenarios.

Ahem, it's 2ch not 4 so it's probably 500k per channel and I see no MSO option (yet) to split the memory even further. If it really has the mega zoom 4 in it then the functionality will be pretty sweet compared to the rigol Imo. Sure, it can't be a feature galore but hey it's a 1000 series scope - look at the model it's replacing.

I imagine that people who want more memory buy scopes that provide it.

[nctnico]

No, with Keysight the memory is always split by 4. It is split across two channels and then split again for double buffering which on a 2 channel scope with 1MPts leaves you with 250kpts per channel.

[ebastler]

SPI on a 2-channel scope ?

Well, the data sheet mentions a "digital channel" in the specifications section, generously calling it "MSO" functionality. And next to the large, dual-page photo it explains that the trigger input can be used as a third, digital channel. That's actually a neat idea. (But still doesn't get me excited about the scope.)

[JPortici]

£113 is the price for DSOX1AUTO, DSOX1EMBD and EDUX1EMBD. All three are the same cost.

But:DSOX1EMBD offers I2C, SPI and UART, but EDUX1EMBD only offers I2C and UART, not SPI.

Clearly nobody needs to learn about SPI 

SPI on a 2-channel scope ?

2+ext. ext interrupt can be used for decoding. can't do full duplex (or maybe it can, without detecting for CS?)

[rf-loop]

If it decoded from acquisition memory and not from screen it would be a total winner over a comparable siglent.

Siglent decode whole acquisition length. Maximum acquisition length is limited to 1.4M when decode is in use.

--------

This Keysight model Segmented memory acquisition must be joke? max 50 segments. And more fun, 19us (!) minimum trigger rearm time in segmented mode - is it mistake in datasheet text or really true.  Siglent normal mode wfm history buffer can do it always in normal mode and when goes to Segmented mode (Sequence mode in Siglent) max speed is nearly ten times more even in Sigl 1000X.

 

[Neganur]

No, with Keysight the memory is always splits by 4. It is split across two channels and then split again for double buffering which on a 2 channel scope with 1MPts leaves you with 250kpts per channel.

But you still make it sound like the scope cannot capture 1M/500k 1ch/2ch which is not true. Single shot gives you all of that memory, doesn't it?

[Augustus]

It's a nice hobbyist scope, I wouldn't mind winning one 

[nctnico]

No, with Keysight the memory is always splits by 4. It is split across two channels and then split again for double buffering which on a 2 channel scope with 1MPts leaves you with 250kpts per channel.

But you still make it sound like the scope cannot capture 1M/500k 1ch/2ch which is not true. Single shot gives you all of that memory, doesn't it?

Yes, but you'd have to setup the scope specifically to do that which is a bit of a chore (I own an Agilent scope as well so I know all about it). On my GW Instek GDS2204E (for example) I don't have to bother myself with selecting the acquisition mode or which channels to use because it has enough memory for each channel to capture 10MPts no matter what. BTW the way Keysight counts memory the GDS2204E would have 40MPts.

[Neganur]

I was under the impression that all you have to do is to hit the Stop button on the InfiniiVision.

But the whole memory thing is not so important to me so I'm maybe the wrong person to discuss this.

[nctnico]

@Neganur: hitting the stop button isn't going to help because the scope is setup to acquire multiple sweeps. Only when in single mode you'll have double the memory because double buffering is disabled.

If it decoded from acquisition memory and not from screen it would be a total winner over a comparable siglent.

AFAIK Agilent/Keysight scopes decode from memory (the entire memory even when in segmented mode) and not just what is on screen. The only limit is the memory in which the decodes result is stored but you are into the many thousands of messages before you run into it. But that is judging from how my DSO7000 series works.

[PA0PBZ]

And more fun, 19us (!) minimum trigger rearm time in segmented mode - is it mistake in datasheet text or really true.

On page 17 it says:

Segmented memory optimizes available memory for data streams that have long dead times between activity. Maximum segments = 50. Re-arm time = 1 us (minimum time between trigger events)

So what is the truth?

[JPortici]

If it decoded from acquisition memory and not from screen it would be a total winner over a comparable siglent.

Siglent decode whole acquisition length. Maximum acquisition length is limited to 1.4M when decode is in use.

it does? couldn't find mentioned it in the manual for sds1000x series (putting 2GS/s aside, this is what i would use for comparison)
also i couldn't find out is the external trigger and the digital channels could be used as channels for serial decoder/trigger

[tautech]

If it decoded from acquisition memory and not from screen it would be a total winner over a comparable siglent.

Siglent decode whole acquisition length. Maximum acquisition length is limited to 1.4M when decode is in use.

it does? couldn't find mentioned it in the manual for sds1000x series (putting 2GS/s aside, this is what i would use for comparison)
also i couldn't find out is the external trigger and the digital channels could be used as channels for serial decoder/trigger

2 channels for decoding works perfectly fine, just not duplex and the Decoding Trigger suite in Siglents permits triggering on the packets or events of interest. I've tried to find a use case where 2 channels aren't enough to decode I2C, SPI, CAN, UART and LIN, there isn't one. 

[nctnico]

2 channels for decoding works perfectly fine, just not duplex and the Decoding Trigger suite in Siglents permits triggering on the packets or events of interest. I've tried to find a use case where 2 channels aren't enough to decode I2C, SPI, CAN, UART and LIN, there isn't one. 

Try and hunt for a (timing) problem when daisy chaining SPI devices. You won't be happy with just 2 channels! Looking at decoded data is one thing but usually there is other stuff happening in a circuit which has something to do with the decoded data and in many cases you'll want to see that as well.

[Fungus]

I've tried to find a use case where 2 channels aren't enough to decode I2C

 

SPI

Maybe ... but don't tell me it wouldn't be cool if you had the option to use ext-trigger for the CS line. 

[tautech]

Is this the 1990's?

No, it's the 50's:

Maximum input voltage 150 Vrms, 200 Vpk

And it comes with switchable probes too. 
 

[Fungus]

Maximum input voltage 150 Vrms, 200 Vpk

"Scrap the Toys, Get a Real Oscilloscope" 

[JPortici]

2 channels for decoding works perfectly fine, just not duplex and the Decoding Trigger suite in Siglents permits triggering on the packets or events of interest. I've tried to find a use case where 2 channels aren't enough to decode I2C, SPI, CAN, UART and LIN, there isn't one. 

so it doesn't. potentially one point for keysight

i *can* get around using 2 channels for spi (even for spi eeproms, it's not really full duplex anyway if MOSI is just there for shifting out data) and it's usually not a problem but having like analog on ch1 and data/clock on ch2/ext would be better (applies to other serial protocols as well)
no decoding on digital channels is borderline useless too, for my current use anyway

[tautech]

2 channels for decoding works perfectly fine, just not duplex and the Decoding Trigger suite in Siglents permits triggering on the packets or events of interest. I've tried to find a use case where 2 channels aren't enough to decode I2C, SPI, CAN, UART and LIN, there isn't one. 

so it doesn't. potentially one point for keysight

Where it appears they've chosen to use an Ext trig for CS, right ?
If so, it's a feature that can be added in FW for any DSO with a Ext trig in.