Intronix LogicPort LA1034

Keith M

Just wanted to relate my experience so far with this product that arrived today.

A little while ago, I saw some brief discussion and a recommendation for the Intronix LogicPort logic analyzer.

It just so happens that I was searching for a better way to troubleshoot my long and drawn out floppy project, and my scope, while digital, had exactly two channels, really old, and so on.

I downloaded the demo software, which is the actually the full version (minus hardware, of course) and played around with it. Intronix advertises it was "very user friendly, intuitive, etc" and they do a pretty good job of it. There are some dialogs that are missing what I'd consider key buttons here or there, but overall, I picked it up quite readily. I can't help but feel it doesn't have the normal glitz of a new windows XP app, but the interface is clean enough.

Since I've never used a logic analyzer before, it took a little while to figure out the overall concepts of how to make the thing do what I want, but with hardware in hand, I was motoring right along.

My first goal I wanted to achieve was to find out why a particular bit was getting written to memory twice instead of once. This was happening very rarely (0.3%) but enough to screw up the integrity of my data. Because this event happened so rarely, and the fact that I only have two channels, I had no chance with scope of catching it. And even if I had, I may not have had enough data BEFORE the problem to figure out why the problem was occurring.

Suffice to say, I hooked up perhaps 4 pins, and setup a trigger that looked for the memory write line to be high for longer than normal, indicating that two of the same bits were written back to back, and set my pre-buffer to 50%. One acquisition and my problem was solved. Pretty cool considering how long I've been banging my head on this issue.......

For those who don't know it, the logicport is a 34-channel, 500mhz, 2k samples, PC USB-based logic analyzer that sells for $400.00.

I haven't made up my mind about the 2k sample size, which seems small, but they use sample compression(ie number of transitions are important, not time passing) and advanced triggering to make the most of it. They also have recurring acquisitions which seem pretty darn fast, although its not real-time. I like the fact that you can export raw data in files in the event that you need to interpret the data in other software, custom apps, etc.

Btw, there's really no probes to buy perse. You get this little 40-pin IDE-looking thing(they call it an input probe) that breaks out to 34-wires that have female header pin contacts. You can plug those directly into a board, or they sell these cute little ez-hook XKM grabbers that work really well. They are tiny, but work well, I've got to order more of them.

Oh yeah, it has these interpreters for I2C, SPI, and RS232. In addition to displaying the raw waveform, it decodes the data for those protocols and shows the decoded data on a separate line. Pretty convenient, though I didn't see any facility for making your own. I really think they need a plug-in architecture for that. It can interpret multiple parallel bits as a one byte, and display that instead of a bunch of 1's and 0's.

I haven't tried "state mode" yet which runs at 200mhz vs 500mhz in regular timing mode.

So these are my initial impressions, and perhaps I'll follow up in a little while to let everyone know how its worked out. My current project involves an SX28, some ramtron FRAM(which I really like too), a Parallax USB2SER, and a floppy disk drive.

Keith

P.S. I have absolutely no relation to Intronix other than as a happy customer(so far)

CCraig

Keith,

Thank you very much for the report. I have been looking at this unit for awhile. I need a replacement scope (it died). The only one I have now is the Parallax USB scope. My analog stuff is slow and I get by, but I'm moving up to the SX and maybe to the Prop.

Do you think someone could get by with just the LA? or Do I still need to see the dirty edges with a scope?

(I find it much easier to borrow a scope than a LA.)

Thanks, Chris

Keith M

Well I'm really an amateur when it comes to electronics, test and measurement stuff, etc, but it seems so far that the reason why I was using a scope was for digital stuff anyways. And I guess scopes don't really excel in that, especially when you are talking about STORING the actual data. What's nice is that the LA reports things as if a microcontroller or memory chip, etc would see them. The LogicPort has an adjustable Logic threshold expressed in volts, so you can define what's really considered a 0 or a 1. If you are using CMOS vs Schmitt Trigger levels, etc, you can configure it appropriately to match how the chip/device/etc would see it.

I bought an analog scope initially, when I first starting getting into this stuff, about 2 or 3 years ago. While it worked fine for some stuff(I was using it for PWM), this LA is obviously much more powerful. And I think cost has everything to do with it. A 500mhz digital scope is going to cost you some money, and may not give you the same feature set offered. This only costs slightly more than a dual channel slow analog scope, so I think bang for the buck is really good.

I've never been a fan of PC-attached stuff, to be honest, because it always seems that it's hardware that's expensive and software that is cheap. I like to spend my money on hardware........... While I think $399 yields the right price as an alternative to a much more expensive LA, it would much more palatable at say, $299 or something.

I will say this, and I know I mentioned this in the first post --- the ability to store the data in text-files(csv, I think) is really cool. When you deal with things that are encoded, (MFM in my case), the data needs processed before its useful, copying the data off the LA and then running it through some piece of software is nice. One other thing: it has "recurring acqusition mode" which is actually REALLY fast. Like maybe hundreds - 1,000 acquisitions per minute, so it does artificially inflate the buffer as well.

As far as dirty edges, I think they might be needed to diagnose some things. Like maybe slow-rising edges, noise issues, and so on. I think in the end, since some piece of hardware has to read it as a 1, or a 0, that if you have the LA configured to same specs, that knowing how that hardware is going to react to the same data is perhaps more important than an image of the "true signal."

Like I said, I'm not an expert, so don't rely strictly on my opinion......

HTH,

Keith

Paul Baker

LAs and OScopes each have thier place and function. If you only design digital systems you can probably get away with just owning a good LA. However, if you also design analog, you'll need an OScope too.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

kjennejohn

HI!
I too have one of those Intronix units. This thing is an incredible deal for the money. (No, I don't know the guy or own stock in his company. I do think highly of his product.) I had to T-shoot a 68 pin processor at work. I talked my employer into purchasing a diagnostic adapter where you plug the unit's base into the board's socket, then plug the processor into the adapter's socket. Festooned about the adapter are wirewrap pins to attach probes to. We installed this adapter, then connected a bunch of the LA's input probes. Worked like a charm! In no time we were tracking a watchdog set up problem back to its source.

I also have a USBEE ZX module. They retail for $895 but can be commonly found on Ebay for $495. This has a better display window, arguably superior trigger setups, and more functions, but only 8 inputs. It emulates several different instruments, like the Intronix LA does.
USBEE also sells the AX, which includes an oscilloscope. This is well over a grand, though, and I haven't seen any yet on Ebay.

Oh well, neat toys, one and all!
kenjj