EEVblog #436 - Saleae USB Logic Analyser Review & Teardown

[BartManInNZ]

If you have a STM32F4-Disc or TI-Launchpad Arm board these can also be used ...
http://jjmz.free.fr/?p=148
http://www.fischl.de/arm/sllogiclogger_logic_analyser_for_stellaris_launchpad/

/Bingo

Thank you Bingo - at last a project for my $15 launchpad eval board!

[cwalex]

Oops, sorry, those are solder types.
Genuine HP ones:
http://www.ebay.com.au/itm/HP-Agilent-5090-4833-Logic-Analyzer-Grabbers-Clips-Set-of-20-5090-4356-/321087386410

Dave.

Thanks for the link I think if I bought them at the price they are at now it would end up a couple of dollars per clip. The auction still has 5 days to go so no guarantee I would get them at that price.

Still looking for some recommendations for cheap alternative though I will put a bid on those items you linked just in case I win it at the current price

Hope I'm not derailing the thread. Just tell me off if anyone is annoyed by my questions!!!

[cwalex]

sorry to be a bit off topic but does anyone know where to get the grabber things that attach to the end of the leads? I don't care if they are cheap quality. I would just like to order a bunch for the cheapest price possible and then possibly buy some better quality ones later.

Just search for "mini grabber" on ebay.
$2.50 for 5 *delivered*
http://www.ebay.com.au/itm/5-Colours-Single-Hook-Clip-Mini-Grabber-Test-Probe-for-SMD-IC-Multimeter-DIY-NEW-/140898966082
or 20 for $6.95
http://www.ebay.com.au/itm/20-PCS-Mini-Grabber-SMD-IC-Test-Clip-Hook-Probe-Multimeter-Jumper-5-Color-P07-/200905035752

Dave.

I have bought these , and while the grabbing is ok.
The resistance/connections sucks ....

I intended to use them with my HP3488 Switch for the relay boards.

There's a brass ring that you solder the wire to , and it needs a good "bang" with a hammer and a tool , in order to make a decent connection with the steelplate.

If that's done they're ok , and don't wait until you have allready soldered 20  , so you have to desolder all those wires , remove the tin , and "bang" it ...  Do it upon receiving them. 

/Bingo

Thanks for the tip mate  I'm looking for the grabbers that have a pin sticking out of them that you push the logic probe female connector onto do you think I would be able to use these ones if I just solder a pin onto the connector after giving them a whack and checking the resistance?

[T4P]

Now REALLY i wonder why people buy this instead of the open logic sniffer 

because there is nothing good with ols, the hardware is "stolen" (yes, in my opinion if someone is selling OH/OS things with "own" name and zero innovation then this is stealing), the software partialy working - partialy not (yet, soon, maybe).
There is no innovation in simply copying other ppl work, giving funcy new name and making some stupid marketing
"open hardware/source" statements - buy here - we having so many cool things ...

Some ppl prefer to pay some more bucks to someone who really spend time on development.

Stolen? U srs?

[ju1ce]

Everytime i go onto their website i still think saleae's stuff is all about wank factor  It's like the apple of the logic analyzer 
Inadequate software, cheap hardware, cheap hooks and lastly ... barely any performance or whatsoever to back up the price tag.

Now, kids. A LA is going to be on a table pretty much all of it's life and usefulness isn't it? Well then what's the whole point of designing the damn case to be tough and making the software look oh-so-wanky 
Now REALLY i wonder why people buy this instead of the open logic sniffer 
You might say software support but have they done anything to fix the async serial problem? It's been known for quite a while now

I actually sold my OLS and bought a Saleae analyzer. I haven't regretted that move.

• The OLS software is a piece of crap, sometimes it worked and sometimes it didn't. The Saleae software, while pretty basic, is more polished. Moreover, you had to search the Dangerous Prototypes forum for updates, whereas Saleae updates are easily available on their website.
• The hooks that come with Saleae (sic!) are much nicer.
• I rather spend my time measuring instead of fighting with the measurement device. I think the price difference of $100 is less than two frustrations (on a student budget).
• I store my logic analyzer in a plastic storage box with a bunch of different meters, and a proper enclosure & case is a huge plus. I fitted my OLS in a generic ABS box, but had to store the grabbers in a bag...
• The performance may not be enough for demanding tasks, but as Dave pointed out, for SPI/I2C/UART it doesn't matter. Those cover 99% of hobbyist projects.
• Which async serial problem are you referring to? The autobaud may be broken, but it works a treat if you type the baud rate yourself.

Disclaimer: I am not affiliated with Saleae. I'm just their happy customer. I like Dangerous Prototypes too, but in my opinion their OLS is not too great.

[sagdahl]

I bought the 8 channel one year ago. Installed it in Win7 64-bit. Did not work! 
Contacted Saleae with the problem and got excellent help. They sent a "standalone-program" which worked without installation.
With the help of this little unit I suceeded in cracking the CAN-communication between the central unit in a Volvo
and the car-heater.
I do agree with Dave that more software functionality would be great. Killing the competition completely.
If possible do not buy "stolen" clones from China. Long term it will harm new products coming out.

/Roger

[jadew]

The problem with the software in general is that it appears there aren't many users for LAs, which is why most free software is incomplete or sucks. Lack of interest from the users results in lack of interest from programmers.

When I purchased my analyzer I both checked the software and compared the hardware features. For me the OLS one seemed the rational choice but, as others, I wasn't very happy with the default software so I wrote my own.

http://dumb.ro/lafront/

It currently does I2C, SPI, Async Serial, 1-Wire and uWire, however the protocol parsers are written in JavaScript so anyone can make their own fairly easy (the API is not documented tho, didn't find the time for it).

I even have a search function on the todo list and was planning to support more devices in the future, including the saleae.

Anyway, I stopped working on that project while still in alpha (after about a month of work) since not enough interest was shown, to make me go on, however I find it much more usable than many of the LA applications I tried, so it's clear that good software is not that hard to write, it just has a very small user base and it's not worth improving on.

On the other hand, companies that sell a complete package (hardware and software), fail at the software side either because of bad management (maybe too much marketing involved - the wank factor that was mentioned) or because of bad developers.

Either way, the fact is that LA software will suck no matter what you chose. And yes, this is the reason why people chose a product over another these days, because the current software for a particular product does something that the user wants and the other software doesn't and it's clear to everyone that the state of the software won't change anytime soon, if ever.


Sorry for the long post, have fun!

[David_AVD]

Has anyone here had a go with the Saleae SDK and either making new decoders or modifying existing ones?

[mrflibble]

because there is nothing good with ols, the hardware is "stolen" (yes, in my opinion if someone is selling OH/OS things with "own" name and zero innovation then this is stealing), the software partialy working - partialy not (yet, soon, maybe).
There is no innovation in simply copying other ppl work, giving funcy new name and making some stupid marketing
"open hardware/source" statements - buy here - we having so many cool things ...

I don't quite follow the "stolen" hardware logic there. AFAIK, and I could very well be wrong, they (dangerous prototypes) designed the pcb and then put the already existing SUMP firmware from That Other Guy on it. And while I would not call it innovation (I'd call it an incremental improvement) they did slap on the cheapo PIC to 1) connect it to the PC and more to the point 2) store the firmware image in PIC flash and as you no longer need a xilinx config prom.

The OLS software is a piece of crap, sometimes it worked and sometimes it didn't. The Saleae software, while pretty basic, is more polished. Moreover, you had to search the Dangerous Prototypes forum for updates, whereas Saleae updates are easily available on their website.

Now while the OLS software certainly has plenty of work to be done on it, AND while I am not a big fan of java for gui's (sluggish crap on average ), I wouldn't go so far as call the OLS software a piece of crap. I am assuming you refer to http://www.lxtreme.nl/ols/ this one?

A couple of things the OLS software could fix (regarding your "sometimes it works, sometimes it doesn't") is the java style RXTX lib. Now THAT is horrendous crap. IMO java sucks for both hardware integration and gui's. And the hardware integration does suck in this case and has issues. The gui side is doable here IMO. Not great, but doable. (which is why I use it ) And it exists, is free, and I can use it. And mostly ... the hardware doing the LA work in the back is definitely not a piece of crap IMO.

Incidentally, I don't own an open bench sniffer thingy, so no fanboism there I hope. One reason for not having bought one (to free up the nexys2) is it's still $50. That, and NO DIFFERENTIAL?!? O_O

Now to be fair, differential as a run time option is tricky. Right now I just have a modded sump with 24 single ended + 8 differential inputs. I would like to be able to switch between single ended and differential from user interface, but that is a bit tricky on the fpga side. At least, I have not figured out how to do that in an easy way. Yeah yeah, partial reconfiguration. Well, sod that. I said easy.

Anyways, lets not have this become a product X vs product Y thread. Both have strengths and weaknesses. You just pick whatever product is the best match for your needs. There's something to be said for a polished user interface. And there's also something to be said for decent triggering. And ... you can always buy a used proper LA on ebay if your budget allows. My budget doesn't allow for that, so I choose the tinkering option.

[mrflibble]

The problem with the software in general is that it appears there aren't many users for LAs, which is why most free software is incomplete or sucks. Lack of interest from the users results in lack of interest from programmers.

When I purchased my analyzer I both checked the software and compared the hardware features. For me the OLS one seemed the rational choice but, as others, I wasn't very happy with the default software so I wrote my own.

http://dumb.ro/lafront/

Neat!

You could see if you can join forces with http://ols.lxtreme.nl/ and double the development effort!

[andete]

The problem with the software in general is that it appears there aren't many users for LAs, which is why most free software is incomplete or sucks. Lack of interest from the users results in lack of interest from programmers.

When I purchased my analyzer I both checked the software and compared the hardware features. For me the OLS one seemed the rational choice but, as others, I wasn't very happy with the default software so I wrote my own.

http://dumb.ro/lafront/

Neat!

You could see if you can join forces with http://ols.lxtreme.nl/ and double the development effort!

I've used the dangerous prototypes openbench logic with the ols software with success to capture the behaviour of a vintage 80s microcomputer video chip (MSX2 V9938) with some friends.
It glitched three times on an entire day of usage and then a simple disconnect usb/connect usb did the trick. We ended up with tons of useful data though. The data can be saved to VCD which can then be further analyzed with the gtkwave [1] program.

Joost

[1] http://gtkwave.sourceforge.net/

[tinhead]

I don't quite follow the "stolen" hardware logic there. AFAIK, and I could very well be wrong, they (dangerous prototypes) designed the pcb and then put the already existing SUMP firmware from That Other Guy on it. And while I would not call it innovation (I'd call it an incremental improvement) they did slap on the cheapo PIC to 1) connect it to the PC and more to the point 2) store the firmware image in PIC flash and as you no longer need a xilinx config prom.

wow great, big improvement Seriously, i hate such zero innovation ppl, this PIC on ols have no relation to LA functionality.
This could be FX2LP, FTDI, CP21xx, CPLD bootlaoder, different µc .. what so ever .. it would be still the same LA with the same client - innovation was here the biiiiiiiig announcement to catch as much as possible "idiots" to sell them pseudo OH/OS project.

But when you wish to see real innovations, check this one:

http://www.bastli.ethz.ch/index.php?page=BitHoundEn

This is incremental improvement, OLS is not. But back to the topic, Saleae .. and i said, they working on search functionality,
they added already many things of which Dave spoke during review into the beta version, so they doing something.
I know as well that they working on FX3 solution, yes, this time with some protection against clones (nothing special, but at
least something to eliminate all these scrap based 5USD clones).

[jadew]

The problem with the software in general is that it appears there aren't many users for LAs, which is why most free software is incomplete or sucks. Lack of interest from the users results in lack of interest from programmers.

When I purchased my analyzer I both checked the software and compared the hardware features. For me the OLS one seemed the rational choice but, as others, I wasn't very happy with the default software so I wrote my own.

http://dumb.ro/lafront/

Neat!

You could see if you can join forces with http://ols.lxtreme.nl/ and double the development effort!

Thanks, small chances of that happening tho. First of all the Logic Sniffer is written in Java while my client is C++, I do know Java, but I don't really enjoy coding it.

To be honest, even if I'd keep working on the OLSFront, I'd still like to know there's the Logic Sniffer (jawi's client) under development, because (even tho I don't like Java) it DOES work on all platforms, while mine is Windows only.

[ErikTheNorwegian]

This must be a simular circuit, addet bonus as two channel scope.. clearly a copy both hardware and software.

http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=261158100499

[Denhart]

In fact, all you need is the dev board. The layout is damn near identical to the one in the product...

http://www.ebay.co.uk/itm/3-3V-5V-CY7C68013A-56-EZ-USB-FX2LP-USB-2-0-Develope-Board-Module-Logic-Analyzer-/281036187315

You don't even need any firmware - all you need to do is change the Product and Vendor IDs using this utility (google it) - all the firmware is transferred from the PC software.

I bought one, and I can confirm that it works fine.

Brilliant! Thanks for the advice and only $11, thats a bargain.

[free_electron]

It glitched three times on an entire day of usage

that alone would be enough for me to grab a shotgun and blast it to pieces.... i do not tolerate having to debug the debugger ! A piece of test equipment must work !

[frank26080115]

Dave, when you did the test with UART serial, you were using a inverted bit setting.

Does anybody have a really early version of the Saleae Logic? I got mine only a few months after it was released, it is solid aluminum, both front and back.

Personally I think the lack of advanced triggering isn't that big of a deal because of the massive sample memory and excellent export capabilities. I can do my own "advanced triggering" in a spreadsheet if I need to, and anything more advanced, I can just code some stuff up in minutes.

I'm really disappointed that Dave gave it a mediocre review because he already owns much better $1000+ hardware. I think Saleae is a key player in advancing the hobbyist scene.

My own opinion: http://frank-zhao.com/blog/index.php/opinion-low-cost-logic-analyzers

[EEVblog]

I'm really disappointed that Dave gave it a mediocre review because he already owns much better $1000+ hardware. I think Saleae is a key player in advancing the hobbyist scene.

I gave it a mediocre review because it deserved a mediocre review!
It has no search capability, no advanced trigger capability, and no free running mode. They are the 3 things that make a a good logic analyser (and you don't need all 3 in a low cost kit)
And this is for a unit that's been on the market for what, 3+ years, and has a company dedicated to "tirelessly develop and refine the world's greatest [tiny] logic analyzer".
It simply didn't live up to expectation.
If I had reviewed it 3 years ago when it came out I would have been kinder about missing basic software features.
 
Dave.

[AndyC_772]

What a shame there isn't even a CPLD sitting in parallel with the microcontroller to search for trigger patterns. I'm all for nice, simple hardware, but it does look like a software engineer's interpretation of what a hardware engineer might find useful - and lacking any development expertise which a hardware engineer might have been able to supply. Another £10 on the hardware could have made for a much more useful product

[amyk]

What a shame there isn't even a CPLD sitting in parallel with the microcontroller to search for trigger patterns. I'm all for nice, simple hardware, but it does look like a software engineer's interpretation of what a hardware engineer might find useful - and lacking any development expertise which a hardware engineer might have been able to supply. Another £10 on the hardware could have made for a much more useful product

At the speeds of current PCs, triggering could easily be done in software. Put the device in continuous streaming mode, stepping the trigger state machine on each input sample, and start recording once the state machine reaches its end state. For even more performance, compile the state machine to native code like what they do for packet filters, but a lot simpler. A slower PC might not be able to keep up with the full rate but that's better than not being able to do it at all. Pseudo-synchronous mode is also possible: whenever the bit designated as clock takes the desired transition, record a sample.

[Hypernova]

Anyone have experience with the logic sniffer from dangerous prototypes?

Much more limited memory but way faster (200MSPS if on 16ch) with more channel count (32 if you get the 2nd buffer board). I tend to use both, the Saleae for protocol and OBLS for multi channel timing data.

Disagree on free_electron's comment that MSO's built in LA are better, the one's I tested (Agilent 3000x) can only trigger on a single byte pattern. So if you want steady trigger on 'e' in a UART stream with "eee" you are shit out of luck, I got an mbed with a simple script to send a trigger pulse on a "sentence". Navigating the captured data is also worse than dedicated LA's. A scope interface simply isn't meant for such things.

[Noize]

The diodes used in the Saleae logic analyzer are these:

http://uk.farnell.com/stmicroelectronics/dviulc6-4sc6/diode-esd-array-bidir-sot23-6/dp/1751961RL

 %-B

[Noize]

Are the input protection resistors about 1 M ohm?

[amyk]

Are the input protection resistors about 1 M ohm?

They're around 4 orders of magnitude lower; IIRC Dave mentioned 100R in the video.

The specs say it has 1M input impedance but that's just the FX2LP's CMOS inputs.

[jesuscf]

In fact, all you need is the dev board. The layout is damn near identical to the one in the product...

http://www.ebay.co.uk/itm/3-3V-5V-CY7C68013A-56-EZ-USB-FX2LP-USB-2-0-Develope-Board-Module-Logic-Analyzer-/281036187315

You don't even need any firmware - all you need to do is change the Product and Vendor IDs using this utility (google it) - all the firmware is transferred from the PC software.

I bought one, and I can confirm that it works fine.

Brilliant! Thanks for the advice and only $11, thats a bargain.

Beware!  This general purpose development board comes with a 16-bit address EEPROM, the Atmel 24C128.  The latest Saleae software expects an 8-bit address EEPROM like the 24C02 or similar, which is what is installed into their board.   Therefore you'll need to change the EEPROM before even changing the Product and Vendor IDs!