Everytime when I attempt to acquire data using SiGe GN3S Sampler v2, there is a message saying “Buffer overrun”. And only 0.015ms data is saved, which is about several kb only.
How could I fix it? Thanks in advance.
Several customers have said they have this problem, and we’re trying to track down a cause.
Which version of Windows are you running?
How much memory is in your PC?
Windows 7 +6GB memory
SFE-TS-Mike:
Several customers have said they have this problem, and we’re trying to track down a cause.Which version of Windows are you running?
How much memory is in your PC?
There is also something interesting. I switched to a computer with Windows XP sp3+2GB memory. It worked fine at first. But the same issue came out after a while. And then I tried another GN3S Sampler (I bought 2 of them) which worked pretty well on this computer. I guess I could safely conclude that previous malfunctioning Sampler is defected.
However, due to curiosity, I tried the ‘defected’ Sampler on a third computer with Windows XP and it still worked. So I’m confused. And I have no idea where could the problem lays.
High speed buffering is tricky.
We have a front end that uses USB here in the lab that I have been designing for the last year, and we have a pair of the SiGE front ends.
The SiGE needs to be ram hacked every time we connect it to disable the internal limit counter, but we also need to make sure to run it on a machine that has a fairly beefy processor and a good amount of ram. If we think about the 8051 in there, it only has something like 16kB of ram available as buffer, while it’s trying to move 16.384 MSample/Second or so. This means the hardware only has a ‘binary millisecond’ (1/1024 second) to unload a full buffer, not counting the data that will flow in during this operation.
While the front ends do work fine accros xp and vista machines, I’d try disabling everything you can on yours. Maybe even the anti-virus, visual themes, MSN, office, and anything else that can touch the drive.
My front is moving ~1.25x the data (20MB/sec) and I have troubles with it even when using six times the buffering (96kB), albeit with the finicky FT2232H chip.
Best of luck.
I have the first version of the SiGe device streaming arbitrarily large data sets using Ubuntu 8.10 and http://gn3s-nix.sourceforge.net/ I am also the admin for that project. As I said I just tested it with my device on a rather old (AMD single core 3000+ processor, 1.5 GB Ram) computer and it seems to run okay.
That said, it does not seem to like many platforms and setups. I blame this entirely on the very high data rate combined with poor software. I only recently inherited the project. Though we have limited resources, we plan on making this software work properly and well. It will take some time. Anyone with strong experience in C, Linux, and low level driver sorts of things is highly encouraged to join us in our quest. At the end, there might be cake.
Do you have any interest in testing on Linux? I won’t guarentee or even imply that it will work, but I’d surely appreciate your (or anyone’s) collaboration in making all of this better.
– Cole Christensen
University of Minnesota – Nanosat Flight Computer Team Lead.