Author Topic: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown  (Read 68965 times)

0 Members and 1 Guest are viewing this topic.

Offline Zbig

  • Frequent Contributor
  • **
  • Posts: 927
  • Country: pl
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #25 on: March 09, 2013, 08:51:55 pm »
You just signed up, your first posting is about telling us how good the LAs are. So, what is your relation with Saleae?

I'm also new here. So, you say it's not allowed to say you like a product in your first post? Good to know, I won't tell I have Logic16 and like it too, then. Oopsie daisy...
 

Offline Bored@Work

  • Super Contributor
  • ***
  • Posts: 3932
  • Country: 00
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #26 on: March 09, 2013, 08:59:40 pm »
Come on kids, three first time posters in a row. Pull the other one.
I delete PMs unread. If you have something to say, say it in public.
For all else: Profile->[Modify Profile]Buddies/Ignore List->Edit Ignore List
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37662
  • Country: au
    • EEVblog
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #27 on: March 09, 2013, 09:15:39 pm »
I don't have any relationship with them other than as a satisfied customer. I signed up just to make this post as I think Saleae has a great product and Dave's review wasn't super positive about it. (Nothing against Dave, generally I really agree with his reviews...)

All it had to do to get a thumbs up would have been a good search feature I suspect. I could maybe have overlooked the triggering (due to the almost unlimited buffer size) and free running mode. But missing all 3 is pretty inexcusable in such a mature product. And for a company that claims to: "tirelessly develop and refine the world's greatest [tiny] logic analyzer"

Dave.
 

Offline Zbig

  • Frequent Contributor
  • **
  • Posts: 927
  • Country: pl
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #28 on: March 09, 2013, 09:31:38 pm »
Come on kids, three first time posters in a row. Pull the other one.

Your level of paranoia is almost amusing :) I'd love to sit and wait for a FOURTH (imagine that...) positive post about Logic just to see your reaction if I hadn't have better things to do.
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37662
  • Country: au
    • EEVblog
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #29 on: March 09, 2013, 09:45:01 pm »
Come on kids, three first time posters in a row. Pull the other one.

Happens all the time.
You give a bad review to something people passionately like, and it triggers them to defend it.
(also happens in reverse)

Dave.
 

Offline mrflibble

  • Super Contributor
  • ***
  • Posts: 2051
  • Country: nl
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #30 on: March 09, 2013, 10:39:25 pm »
Lack of decent triggering seriously limits what you can do with a data pump like that...

I use a somewhat modified version of SUMP on a nexys2, and the first thing I did was to add some more triggering capabilities. Because not being able to trigger on for whatever it is you want to trigger on is so frigging annoying. You can argue about being able to do it post capture, but that just isn't the same.

If only because with post capture processing you don't have a handy TRIGGER output pin that you can use during your debugging.
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7799
  • Country: us
  • adieu
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #31 on: March 09, 2013, 11:40:25 pm »
The lack of proper triggering just lost them a sale. I was considering buying one but I refuse to fight with that.

It's a shame they seem to have put so much effort into wank factor when so much of the real development is unfinished. You could even consider the trigger limitation a "design choice", but there's no excuse for it choking on that UART stream...
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline David_AVD

  • Super Contributor
  • ***
  • Posts: 2797
  • Country: au
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #32 on: March 09, 2013, 11:56:55 pm »
I have a Logic ( 8 ) and found the software pretty good.  I did want to improve the DMX decoder plugin but can't seem to figure out how to use the SDK.   :(
 

Offline cwalex

  • Frequent Contributor
  • **
  • Posts: 299
  • Country: au
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #33 on: March 10, 2013, 12:19:08 am »
http://www.pctestinstruments.com

this is a usb logic analyser. 34 channels and $389

here is a list of bullet point features from their website:

               
               500MHz Timing-Mode (Internal Clock)
            200MHz State-Mode (External Clock)
             Advanced Multi-Level Triggering
       Real-Time Sample Compression
       +6V to -6V Adjustable Logic Threshold
       CAN, I2C, SPI, RS232, 1-Wire and more
       Built-in 300MHz Frequency Counter
       USB 1.1 and 2.0 Compatible


I'm not sure but I think dave may have done a teardown of one but it might have been a different person. So you don't have to spend huge money to get something with the features people are complaining about with the salae logic.

I'd be interested to hear if anyone thinks it is a good deal! seems like it to me but I'm a newb  :P

Cheers,

Alex
 

Offline free_electron

  • Super Contributor
  • ***
  • Posts: 8515
  • Country: us
    • SiliconValleyGarage
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #34 on: March 10, 2013, 12:22:46 am »
look , you can't really blame this thing for not having a trigger system. not at this price point. their software is pretty clever and they have a nice visualisation of many protocols. true, their are hampered by the hardware. but then again all these cypress based designs are in the same category ( Usbee , Salae , Logic and some others )

The ChronoVu http://www.chronovu.com/ for example is a different animal. that one uses a CPLD and on-board SDRAM to capture at up to 100Mhz. it does have triggering capabilities albeit not sequential triggering. Then again at that price point you still can't complain.
A real logic analyser with sequence triggering , fast , deep memory and advanced decoding ( like assembly ) is going to cost you an arm and a leg and there is much more involved in such a beast.

here is a VNC session connected to my logic analyser :


One of the 'blades'


This is a 16702B running HP-UX on a PA-Risc machine. The empty chassis ( so just the computer / display section ) of that thing was over 30K$ when new 10 years ago. The blades depend on their functionality but a blade holding 4 Meg and clocking in a 1Gsa/s pushes another 35K onto the machine

The blade above is a 500MHz 4 meg deep 68 channel acquisition card. It can do state and timing and hes sequentiel trigger.
I have another balde with 'advanced trigger' capabiltities. The user interface is like a flowchart. you can make really long trigger scripts.

1)- wait for pattern a to occur
2) if within time(x) pattern b occur then trigger
3) if within time y() from (1) pattern C does not occur then :
4) scan for pattern d , followed by pattern e ,f,g,h and if withing timestamp e from f there is no activity on input 4 then trigger with 70% pretrigger buffer.

for each step you can set masks and for each time stamp you can do shorter,longer , within or without.
you can do absolutely insane triggering with that machine. then again , take a look at the massive asics with heatsinks on their back ... that is not stuff you prop in a simple FPGA...

The machine can run cards in lockstep with each other , there is also a pattern generator card ( which i have too ) so you can generate stimuli as well , and the generator also has looping and sequencing capability , there is a scope card that basically lets you add 2 analog 1GHz channels. With its expansion chassis you can have 10 blades in this box. And if that is not enough there is the possibility to connect the analyser on the LAn and pair it remote with one or more infiniium oscilloscopes. Just plug in the IP addresses and the logic analyser goes and fetches the traces of the scope and displays it in time-step with whatever it captured.... there is a module to sync the scopes and analyser in time.
i hooked up my 54832D ( 1GHz 4 analog + 16 digital infiniium with 4Msa/s and 8 Meg memory) to this analyser and the i could simply scroll through the trace buffer while the analyser grabbed the analog data from the scope and visualised it on the same screen. everything was there : the 4 analog traces of the scope , the 180 or so digital traces of the anayser linked in various buses. it didn't even slow down.
pretty impressive given that this is a PA-risc running at barely 75MHz running a full bown Unix system with X-client

The analyser as it sits would have costed well over 200K$ when new. Pretty impressive stuff, even today, for a 15 year old machine... you should see the new machines like the 1690x ...

So all in all i think the little Salae does a good job for its price. I like the software , it reacts fast is intuitive and the packet decoders work well. ok there is no triggering but you can compensate using deep buffer memory and hope you catch everything in there.

Now, on the other hand a 'lister' like dave's 3000x scope has would be welcome.... That is especially handy for serial protocols. you can immediately see : here's a missing ack , here this got sent there etc. Clicking a line in the lister scrolls the traceview to that timepoint.

I got a demo yesterday of the new 4000x series scope. Pretty slick machine. I went to the boss afterwards and told him to buy me one.




Professional Electron Wrangler.
Any comments, or points of view expressed, are my own and not endorsed , induced or compensated by my employer(s).
 

alm

  • Guest
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #35 on: March 10, 2013, 12:24:24 am »
this is a usb logic analyser. 34 channels and $389

here is a list of bullet point features from their website:
[...]

They conveniently omit the memory depth, though. Development seems to be a bit stale, as far as I know they've been selling the same product for years. Users seem quite happy with it, though.
 

Offline David_AVD

  • Super Contributor
  • ***
  • Posts: 2797
  • Country: au
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #36 on: March 10, 2013, 12:37:53 am »
Just watched the review video to the end.  Dave seems unhappy with the software, but it's a breeze to use compared to the Bitscope unit I have sitting here as a paperweight.
 

Offline cwalex

  • Frequent Contributor
  • **
  • Posts: 299
  • Country: au
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #37 on: March 10, 2013, 12:45:11 am »
this is a usb logic analyser. 34 channels and $389

here is a list of bullet point features from their website:
[...]

They conveniently omit the memory depth, though. Development seems to be a bit stale, as far as I know they've been selling the same product for years. Users seem quite happy with it, though.

here is the specs list link: http://www.pctestinstruments.com/logicport/specifications.htm

Sample buffer: 34 x 2048 samples
Maximum sample compression: 2^33 to 1 (sample rates to 200MHz)
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37662
  • Country: au
    • EEVblog
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #38 on: March 10, 2013, 12:53:11 am »
Sample buffer: 34 x 2048 samples
Maximum sample compression: 2^33 to 1 (sample rates to 200MHz)

Data compression is great, until you need to capture a packet or two where the total number of transitions exceed the very limited memory.

Dave.
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7799
  • Country: us
  • adieu
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #39 on: March 10, 2013, 12:57:42 am »
ok there is no triggering but you can compensate using deep buffer memory and hope you catch everything in there.

Good point - you can capture a hell of a lot of data. Given that, and the lack of a "real" trigger, I think they should put some some effort into features for indexing and searching the captured data in software, as well as a simulated post-capture 'trigger' for those who don't want to learn a new way to do things. It would be a good selling point, I think - you could get pretty close to some real, professional triggering features, as long as you don't need any long "wait five minutes for this 100us packet" stuff.
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline cwalex

  • Frequent Contributor
  • **
  • Posts: 299
  • Country: au
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #40 on: March 10, 2013, 01:11:44 am »
Sample buffer: 34 x 2048 samples
Maximum sample compression: 2^33 to 1 (sample rates to 200MHz)

Data compression is great, until you need to capture a packet or two where the total number of transitions exceed the very limited memory.

Dave.

Thanks Dave :)
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37662
  • Country: au
    • EEVblog
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #41 on: March 10, 2013, 01:14:55 am »
Good point - you can capture a hell of a lot of data. Given that, and the lack of a "real" trigger, I think they should put some some effort into features for indexing and searching the captured data in software

Search is essential for such huge data buffers. In fact you could say it almost useless without it.
That's why I said if it had search it probably would have got a thumbs up.

Dave.
 

Offline c4757p

  • Super Contributor
  • ***
  • Posts: 7799
  • Country: us
  • adieu
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #42 on: March 10, 2013, 01:19:54 am »
Can it export data into a more or less "human readable" format? I don't mean some massive dump of the raw data, I mean something with all the decodings and whatnot as the software shows otherwise. I'd happily sift through and manipulate the data in Vim or something.
No longer active here - try the IRC channel if you just can't be without me :)
 

Offline JoeyP

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: us
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #43 on: March 10, 2013, 01:21:21 am »
Sample buffer: 34 x 2048 samples
Maximum sample compression: 2^33 to 1 (sample rates to 200MHz)

Data compression is great, until you need to capture a packet or two where the total number of transitions exceed the very limited memory.

Dave.

But that's where trigger capability becomes useful. You can just delay the trigger to a later event, thereby skipping the first n packets, and cause sampling to begin at the packet of interest. You then have the entire buffer available beginning at the point of interest.
 

Offline EEVblogTopic starter

  • Administrator
  • *****
  • Posts: 37662
  • Country: au
    • EEVblog
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #44 on: March 10, 2013, 01:30:58 am »
But that's where trigger capability becomes useful. You can just delay the trigger to a later event, thereby skipping the first n packets, and cause sampling to begin at the packet of interest. You then have the entire buffer available beginning at the point of interest.

Great, if you don't need to see any extensive pre-trigger data that may have caused the issue.
To do any sort of decent long packet data inspection requires more than a few KB of sample memory, that will fill up very quickly.
Murphy will ensure that your requirements will outstrip your logic analysers capability  >:D

Dave.
 

Offline TerminalJack505

  • Super Contributor
  • ***
  • Posts: 1310
  • Country: 00
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #45 on: March 10, 2013, 01:33:45 am »
The software looks pretty nice.  I have to wonder, though, if it was developed by people who don't actually use it on a day-to-day basis.

The developers might need to sit down with a domain expert and figure out how it should really work.  I can understand the software not having certain features due to the lack of hardware support but the "search the data" use case is something any domain expert would have known needed to be addressed.
 

Offline JoeyP

  • Frequent Contributor
  • **
  • Posts: 321
  • Country: us
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #46 on: March 10, 2013, 02:03:20 am »
Great, if you don't need to see any extensive pre-trigger data that may have caused the issue.
To do any sort of decent long packet data inspection requires more than a few KB of sample memory, that will fill up very quickly.
Murphy will ensure that your requirements will outstrip your logic analysers capability  >:D

Dave.

If your bug is repeatable, you don't need to see both at the same time. Capture the early information, then restart the sequence and capture the later information. I've yet to find a task that I couldn't complete using this method. No matter how deep your memory is, you'll eventually find its limit and need to use this sort of triggering.
 

Offline DavidJRobertson

  • Regular Contributor
  • *
  • Posts: 51
  • Country: scotland
    • davidr.me
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #47 on: March 10, 2013, 02:19:22 am »
About the async serial problem: in the analyzer settings dialog you typed in the correct baud rate, but didn't uncheck the autobaud option, so when you re-captured it attempted to determine the baud rate again (and failed).

About the lack of decoded data triggering: I posted on Saleae's Facebook page the other day asking about just this:

Me:
Will there ever be support for triggering on a certain pattern of analyzer data? I guess that won't work if the analyzer can't keep up with the incoming data though.

Saleae:
Hi David, thanks for asking! Yes, that is going to happen and it's a big deal. I'm sorry progress is so slow on that front. We've hired one additional software person who will start in a few months, and are looking for 2 more: http://www.saleae.com/jobs. I really hope to do something like "12 months, 12 major features" and be able to do that along with everything else so we're just constantly making the software better and better.


David.
 

Offline uprightsquire

  • Contributor
  • Posts: 45
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #48 on: March 10, 2013, 02:54:29 am »
Anyone have experience with the logic sniffer from dangerous prototypes?
 

Offline jpatten

  • Newbie
  • Posts: 8
Re: EEVblog #436 - Saleae USB Logic Analyser Review & Teardown
« Reply #49 on: March 10, 2013, 04:51:38 am »
Can it export data into a more or less "human readable" format? I don't mean some massive dump of the raw data, I mean something with all the decodings and whatnot as the software shows otherwise. I'd happily sift through and manipulate the data in Vim or something.

Yes, it can do this. The file includes timestamps too so it's easy to go back and forth.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf