I tried the new DLL but it also does not work. (just crashes matlab without warning).
Could I see the code you are testing the DLL with by any chance? Maybe I am doing something wrong.
Mark
Dave,
I attached a zip of my matlab code, you should be able to reverse engineer my steps. I think I did them correctly, It has been a while since I looked at it though.
http://dl.dropbox.com/u/147864/picoscope_test.zip
(The file size is too big, maybe because I included a pdf in there)
the _old.dll is the original dll that comes with the sdk, I was trying Martyn's fix just now.
Start with my_pico_test.m
I modified the .h file to change function pointers to void *. Shouldn't make any difference except for the fact that the matlab compiler won't complain.
Sorry for the lat reply, I was expecting an email notification when some1 replied.
The driver has now been fixed and will be released in SDK 10.4.2.2 shortly. The previous driver I posted was produced 1 hour before the bug was fixed so when looking at dates it seemed correct.
Martyn wrote:Take the current sdk and just drop in the dll from my previous post. If you are still getting errors after this let us know what they are.
Okay. I'm pretty rusty with C (more of a Java guy these days), so here's what I did.... I right-clicked on the project name in Visual Studio and did Add Existing Item, adding the (new) .dll and (old) .lib file. It compiles and runs, but I still get a crash. Is that correct?
I have a 6407 here. I'm trying to do a triggered block collection, nothing fancy. I suspect I'm doing something wrong with the timebase/# samples settings.... I started from the sample program and removed things not needed. I've attached the code.
By the way, I get a crash with your C demo program, the only changes being 62M samples and trigger on AUX instead of A. I can attach that file, if you like, but the one I attached before should be shorter and demonstrate the problem pretty well (I think). Thanks again.
Sure, I copied the new .dll into my Debug directory and ran the debug version from Visual Studio, but I'm still getting a crash in ps6000GetValues. I'm not seeing anything obviously wrong with my code....
Hi. I'm wondering if you've had a chance to investigate my crash? We were hoping to use your hardware to take some measurements next week, but this has slowed down my efforts.
I have seen the crash with your code as supplied, although if you select V and set one channel on and the rest off the crash disappears for me. I will add an update when I have more information.
Martyn wrote:I have seen the crash with your code as supplied, although if you select V and set one channel on and the rest off the crash disappears for me. I will add an update when I have more information.
Hi Martyn. Any news on this? We do need four channels, by the way.