Hi,
First, Thank you everyone for their answers.
Curious to hear what solution you end up with. Are you running the current Saleae software and have you asked them about the crashing issue?
Have you looked in to sigrok? Seems they support the original Saleae, and I'd guess are scriptable enough to help with your situation.
Regarding the input impedance; how about putting together a buffer circuit on some protoboard?
Is the problem apparent externally? In other words - could you record in 30 minute chunks, discarding them until you catch the problem?
Still no solution yet. Didn't contacted saleae yet , but i will. i tested with their latest 1.2.x family & 1.1.15.
About Sigrok: Yes , they support original Saleae , but their protocol analyzer is very slow compared to saleae , i'll test it capturing data and activating protocol analyzer later and see what happens.
Input impedance, i thought about it, but this device is an industrial device and installed in it's place, i can connect the usb LA to my work laptop, but if i want to put a buffer section, it needs power supply , ...etc. So if there is any instrument which eliminates the power supply, i'll be happy.
This is my last resort. but i have to log data in 30 minutes or less segments (~6 hours) and after that i have to analyze it, if there is no software crash.
My Saleae Logic Pro 16 (software version 1.2.9) definitely has compression. I told it to acquire a handful of channels at 500MS/s with nothing connected, it says it has collected 100 billion samples and that it's using a total of 0MB of memory. Windows shows it fluctuating between 105 and 107MB of memory but not increasing even as another 30 billion samples have been taken. I suspect what you're seeing is either a restriction on the older Saleae devices or a software bug. It's probably worth contacting them in any case.
I wish the software had segmented memory, but left and right trigger search in combination with data compression has been a workable substitute, although my requirements for long term capture haven't gotten close to yours yet.
Here is my Saleae (Logic Pro 16, as before) after recording 4 channels at 500MSa/s for 40 minutes, measuring the distance between a few bursts separated by 36 minutes, without exceeding 100MB of RAM (it was a bit higher by a constant amount during capture). The .logicdata save file is 136kB. Qualitatively, the display and scrolling are still butter smooth, and it's no hassle at all to zoom in until I can see the details of the "bursts" (i.e. the noise from brushing the logic probe against my power supply output).
Ok, it seems that new saleae pro series has some improvement over the old one. I'll contact them and ask them directly.
BTW, Could you please tell your computer specs? my computer is a Panasonic Industrial laptop , so it has not the latest CPU or massive RAM , but i don't see a cpu load on it too.(it doesn't have Anti-virus and firewall ,....etc too) only the required software and tools.
I installed linux on my laptop, and will try Saleae for linux and will see what happen. Usually linux is less intrusive and more stable for long runs.
If you are looking for some serious USB protocol analysis you might look at the products from Total Phase - www.totalphase.com . I've never used them, but I've heard good things.
There is nothing about USB Protocol Analysis. The only protocol analysis that i need is SPI.
Hi
You mention a problem after 1 to 5 hours. How long a capture are you trying to make?
If you are trying to log a few months worth of data, a USB device of any sort may not be the right tool to use.
Bob
According my knowledge of how this device works and current problem nature, I think that around 6 Hours data logging is enough.
You can try the Intronix software - full software with some demo files:
http://www.pctestinstruments.com/downloads.htm
Yes, i saw it and i tried it. the problem with Intronix software is , if you turn-on Repetitive Run to continuously trigger and Re-Arm itself, when a new segment of data comes in , it will replace the old segment of data and discard the old data completely!
BTW, I tried many demos of different logic analyzers and i have to say i like the Intronix much more than any other software. Their latest software is dated 2012 , but it works. if they implement a streamer mode in their logic analyzer (Even for low sample-rates like 4~5 MS/s) , i would buy their device immediately!