NEW 100MHz Keysight Scope Release on 1st March 2017

[Keysight DanielBogdanoff]

BTW does decoding in 1000 X have event table which covers full dataset captured? Lets say you captured 1M, but heavily zoomed in to some part. Will it decode only zoomed part or all 1M?

I don't recall seeing a lister anywhere, so I'm going to say no. I'll come back and edit if I'm wrong on that.

The decoding and measurements are performed on the screen data, but as soon as you scale out it'll re-calculate/decode without any lag or delay.

@rf loop, I'm looking into your question and should have an answer in the next 12 hours or so.

[Avacee]

Never knew www.aspen-test.com existed, thank you - had always used http://www.keysight.aspen-electronics.com/

£439 - http://www.aspen-test.com/dsox1102a
£488 - http://www.keysight.aspen-electronics.com/dsox1102a

Wonder why a discount on one site and not the other - are the sales that bad after only one day? :p

[ebastler]

Wonder why a discount on one site and not the other - are the sales that bad after only one day? :p

The more expensive aspen-electronics.com site is a "Keysight Premium Distributor".
Now you know what they mean by "premium"... 

[ebastler]

The decoding and measurements are performed on the screen data, [...]

Does that also imply that decoding only relies on the resolution currently used on the screen? I.e. if you zoom out to a time window where you can't resolve the individual bits on the screen anymore, does the decoding still work since it uses the high resolution data in main memory? Or does it stop working, like on the Rigol DS1000Z series?

[nctnico]

BTW does decoding in 1000 X have event table which covers full dataset captured? Lets say you captured 1M, but heavily zoomed in to some part. Will it decode only zoomed part or all 1M?

I don't recall seeing a lister anywhere, so I'm going to say no. I'll come back and edit if I'm wrong on that.

The decoding and measurements are performed on the screen data, but as soon as you scale out it'll re-calculate/decode without any lag or delay.

If I interpret the user manual right you can't enable decoding and decode a trace which has already been captured. You need to have decoding enabled before capturing data and I'm not sure you can change things like bits, bitrate, etc without making a new capture. Needless to say this can be annoying at times.

@ebastler: on (no longer) my DSO7000A series scope decoding could get wonky at lower samplerates while the signal itself was still recognisable as a good signal (I2C for example). It seems to me the decoded data is stored in a seperate part of the memory and it may be sampled at a lower rate. Then again the DSO7000A series had a 35MHz limit for SPI decoding but I can't find such a limit on the more modern Keysight scopes so it may have been improved.

[mikeselectricstuff]

The decoding and measurements are performed on the screen data, [...]

Does that also imply that decoding only relies on the resolution currently used on the screen? I.e. if you zoom out to a time window where you can't resolve the individual bits on the screen anymore, does the decoding still work since it uses the high resolution data in main memory? Or does it stop working, like on the Rigol DS1000Z series?

If you zoom out so you can't see the bits, you also won't be able to see the decoded data, as it scales the data 'envelope' to the length of the byte.
 
One problem with Keysight's serial decode is the size of font they use is too big, and the angled "envelope" onscreen clips the edges of characters so it can't display as many characters on one screen as it ought to be able to. The envelope can also sometimes cause ambiguities on certain characters when slightly clipped. e.g. 00 can look like 0C

<edit with right images this time!>

[ebastler]

Thanks Mike and Nico!

So it seems that there is a common Rigol and Keysight "school" regarding the implementation of serial decoding. Not surprising maybe, as Rigol has probably learnt from Keysight during their OEM relationship. But a bit disappointing that the more powerful hardware in the Keysight does not seem to provide much benefit for serial decoding.

[Keysight DanielBogdanoff]

BTW does decoding in 1000 X have event table which covers full dataset captured? Lets say you captured 1M, but heavily zoomed in to some part. Will it decode only zoomed part or all 1M?

I don't recall seeing a lister anywhere, so I'm going to say no. I'll come back and edit if I'm wrong on that.

The decoding and measurements are performed on the screen data, but as soon as you scale out it'll re-calculate/decode without any lag or delay.

If I interpret the user manual right you can't enable decoding and decode a trace which has already been captured. You need to have decoding enabled before capturing data and I'm not sure you can change things like bits, bitrate, etc without making a new capture. Needless to say this can be annoying at times.

That is correct, and news to me also.

[Keysight DanielBogdanoff]

@ Keysight_DanielBogdanoff

I repeat my question:

What is truth about 1000 X series segmented memory acquisition function?

http://literature.cdn.keysight.com/litweb/pdf/5992-1965EN.pdf?id=2832547

In data sheet text (page 3.):

Up to 50 segments can be captured on the 1000 X-Series models with a
minimum re-arm time of less than 19 us.

In data sheet specifications list (page 17.):

Maximum segments = 50.
Re-arm time = 1 us (minimum time between
trigger events)

If 1us is true, Keysight 1000X series maximum segments/s speed is 1Msegment/s.  Is it true?
Only max 50 segments, are there k missing or is it really only 50 segments?
Segment time stamps?
Single segment length, what give this high 1Msegment/s speed?

The 1us is for the normal trigger, but changes to 19us when in segmented memory mode.

50 is the max number of segments available, and they have time stamps. If you set the segment quantity to 1, you will get one "segment" that's uses the full memory of the scope.

I think that answers all your questions?

-D

[JPortici]

hi daniel, anything regarding the actual capabilities of EXT as a digital channel?
can it be used for something more than doing /CS in spi? (only time it's mentioned in either the videos and the manual)
other decode inputs?
i.e: can we do miso/mosi/sck in spi (and end packet with timeout. makes perfect sense)
i.e: can we do ANALOG CH1, ANALOG CH2/DECODE, EXT. DIGITAL/DECODE?

[mikeselectricstuff]

hi daniel, anything regarding the actual capabilities of EXT as a digital channel?
can it be used for something more than doing /CS in spi? (only time it's mentioned in either the videos and the manual)
other decode inputs?
i.e: can we do miso/mosi/sck in spi (and end packet with timeout. makes perfect sense)
i.e: can we do ANALOG CH1, ANALOG CH2/DECODE, EXT. DIGITAL/DECODE?

..or display it as a single digital channel ? That would be really useful.
Some old analogue 2-ch scopes had a "trig view" function to give 3 traces

[Keysight DanielBogdanoff]

I'm getting details, I don't want to mis-speak.

[rf-loop]

@ Keysight_DanielBogdanoff

I repeat my question:

What is truth about 1000 X series segmented memory acquisition function?

http://literature.cdn.keysight.com/litweb/pdf/5992-1965EN.pdf?id=2832547

In data sheet text (page 3.):

Up to 50 segments can be captured on the 1000 X-Series models with a
minimum re-arm time of less than 19 us.

In data sheet specifications list (page 17.):

Maximum segments = 50.
Re-arm time = 1 us (minimum time between
trigger events)

The 1us is for the normal trigger, but changes to 19us when in segmented memory mode.

This is weird! 

I like to see this 1us trig recovery in practice in normal acquisition mode with this model. In what circumstances and how it is visible in practice. Very nice if it can.

This I understand that this model segmented memory acq is very slow with minimum  19us trig recovery - just as in 2000 X  but up to 50kwfm/s scope (also same as in 2000 X)  have 1us trig recovery in normal mode, what it can do with it.  Let's hope we can see this demo.

[Someone]

On paper its competing directly against the DS1000Z on specs and price (assuming you don't unlock all the features of only one for free). Deep memory or fast update rates? pick one thats more important and now you know which scope is in your heart forever.

Not sure I understand. Regarding waveform update rates, the DS1000Z and the new Keysight are not worlds apart (30k/s vs. 50k/s). Regarding memory, the DS1000Z has lots of it, although with limited navigation capabilities; the Keysight has much less, apparently with a significant restriction on the number of segments.

So neither of the two criteria seems to massively favor either the one or the other scope I my view. Which one is "in your heart", and for what reason? (I know about my heart, and the answer has something to do with the price for a fully option-enabled scope... )

The waveform rates are very different when you compare them in the same memory depths, the Rigols can post high update rates only at certain settings with the memory depth at minimum. I've not seen a good table with all the points for a DS1000Z but the faster DS2000 still lags behind even when set for its peak performance:

So again instead of leaning on the banner specs, you might want to look at how it works in real world situations.

[Jay_Diddy_B]

BTW does decoding in 1000 X have event table which covers full dataset captured? Lets say you captured 1M, but heavily zoomed in to some part. Will it decode only zoomed part or all 1M?

I don't recall seeing a lister anywhere, so I'm going to say no. I'll come back and edit if I'm wrong on that.

The decoding and measurements are performed on the screen data, but as soon as you scale out it'll re-calculate/decode without any lag or delay.

If I interpret the user manual right you can't enable decoding and decode a trace which has already been captured. You need to have decoding enabled before capturing data and I'm not sure you can change things like bits, bitrate, etc without making a new capture. Needless to say this can be annoying at times.

That is correct, and news to me also.

This is true on the DSOX 3000A series also. The protocol has to be set and defined before the acquisition. I was trying decode a bus that was shared between SPI and I2C.

Regards,

Jay_Diddy_B

[nctnico]

I'll drop the curtain: decoding works that way in every Megazoom based Keysight scope.

[Keysight DanielBogdanoff]

BTW does decoding in 1000 X have event table which covers full dataset captured? Lets say you captured 1M, but heavily zoomed in to some part. Will it decode only zoomed part or all 1M?

I don't recall seeing a lister anywhere, so I'm going to say no. I'll come back and edit if I'm wrong on that.

The decoding and measurements are performed on the screen data, but as soon as you scale out it'll re-calculate/decode without any lag or delay.

If I interpret the user manual right you can't enable decoding and decode a trace which has already been captured. You need to have decoding enabled before capturing data and I'm not sure you can change things like bits, bitrate, etc without making a new capture. Needless to say this can be annoying at times.

That is correct, and news to me also.

This is true on the DSOX 3000A series also. The protocol has to be set and defined before the acquisition. I was trying decode a bus that was shared between SPI and I2C.

Double probe & use some digital channels.

[Keysight DanielBogdanoff]

hi daniel, anything regarding the actual capabilities of EXT as a digital channel?
can it be used for something more than doing /CS in spi? (only time it's mentioned in either the videos and the manual)
other decode inputs?
i.e: can we do miso/mosi/sck in spi (and end packet with timeout. makes perfect sense)
i.e: can we do ANALOG CH1, ANALOG CH2/DECODE, EXT. DIGITAL/DECODE?

..or display it as a single digital channel ? That would be really useful.
Some old analogue 2-ch scopes had a "trig view" function to give 3 traces

I did some playing around, and I'm able to assign the external trigger as a source for decoding. It basically acts as a bonus single digital channel.

[mikeselectricstuff]

BTW does decoding in 1000 X have event table which covers full dataset captured? Lets say you captured 1M, but heavily zoomed in to some part. Will it decode only zoomed part or all 1M?

I don't recall seeing a lister anywhere, so I'm going to say no. I'll come back and edit if I'm wrong on that.

The decoding and measurements are performed on the screen data, but as soon as you scale out it'll re-calculate/decode without any lag or delay.

If I interpret the user manual right you can't enable decoding and decode a trace which has already been captured. You need to have decoding enabled before capturing data and I'm not sure you can change things like bits, bitrate, etc without making a new capture. Needless to say this can be annoying at times.

That is correct, and news to me also.

This is true on the DSOX 3000A series also. The protocol has to be set and defined before the acquisition. I was trying decode a bus that was shared between SPI and I2C.

Double probe & use some digital channels.

On the 3000 you can do 2 simultaneous decodes of the same (e.g uart at 2 baudrates) or different protocols, I presume the 1000 won't?

[mikeselectricstuff]

hi daniel, anything regarding the actual capabilities of EXT as a digital channel?
can it be used for something more than doing /CS in spi? (only time it's mentioned in either the videos and the manual)
other decode inputs?
i.e: can we do miso/mosi/sck in spi (and end packet with timeout. makes perfect sense)
i.e: can we do ANALOG CH1, ANALOG CH2/DECODE, EXT. DIGITAL/DECODE?

..or display it as a single digital channel ? That would be really useful.
Some old analogue 2-ch scopes had a "trig view" function to give 3 traces

I did some playing around, and I'm able to assign the external trigger as a source for decoding. It basically acts as a bonus single digital channel.

Can it be used as a general-purpose extra trace ( that you trigger from, obviously) ?

[JPortici]

I did some playing around, and I'm able to assign the external trigger as a source for decoding. It basically acts as a bonus single digital channel.

which could be there to circumvent the 2 channel limitations on SPI and do full duplex.. but the manual doesn't seem to agree :/

[Cervisia]

Aimed 100% at the Tek TBS1000 series, to the point where IMO it puts them in the pricing valley of death as far as the hobbyist/Rigol DS1054Z market is concerned.

http://literature.cdn.keysight.com/litweb/pdf/5992-1999EN.pdf
"See how Keysight's InfiniiVision 1000 X-Series oscilloscopes compare against Rigol's 1000Z Series."

http://literature.cdn.keysight.com/litweb/pdf/5992-2000EN.pdf
"See how Keysight's InfiniiVision 1000 X-Series oscilloscopes compare against Tektronix's TBS1000B-EDU Series."

And I have to agree – the competitors' models look worse, with an empty screen … 

[EEVblog]

Aimed 100% at the Tek TBS1000 series, to the point where IMO it puts them in the pricing valley of death as far as the hobbyist/Rigol DS1054Z market is concerned.

http://literature.cdn.keysight.com/litweb/pdf/5992-1999EN.pdf
"See how Keysight's InfiniiVision 1000 X-Series oscilloscopes compare against Rigol's 1000Z Series."

http://literature.cdn.keysight.com/litweb/pdf/5992-2000EN.pdf
"See how Keysight's InfiniiVision 1000 X-Series oscilloscopes compare against Tektronix's TBS1000B-EDU Series."

And I have to agree – the competitors' models look worse, with an empty screen … 

And the standard technique of leaving out the competitors key benefit, which in the case of the Tek is the educational module.

[H.O]

Yeah, and it's "funny" (or pathetic) how they in the Keysight vs Tek document have included memory depth while conveniently leaving it out on the Keysight vs Rigol document. 

[EEVblog]

Yeah, and it's "funny" (or pathetic) how they in the Keysight vs Tek document have included memory depth while conveniently leaving it out on the Keysight vs Rigol document. 

Pathetic