ols
ols copied to clipboard
jawi
RLE problems
First I would like to thank you for such a wonderfull piece of software!
I just received my Openbench Logic Sniffer version 1.04, 4 days ago, and was using the software version 0.94 from you.
I have a problem with the RLE setting, the RLE setting is only usable when i included the 3'th channel group. However then the available memory is then only 50%. (i'm only needed the first channel).
When i'm not selecting the 3'th channel group, i will see a lot of data on the lines 2-3-4-5-6-7 of the first channel group (i'm not using these, and these were attached to the gnd). So I think this is a little bit strange.
Selecting the 3'th channel group and then perform a capturing, shows nice clean lines (for the ones, who are attached to the ground).
Just one hour ago, i installed your new version (0.95) but the problem still exists in my capture data. Could this be a bug.
If you need more information, I can give you more details (in dutch when needed).
I'm using Mac osX Lion for your information.
I tried the find out how to upgrade (or at least, read the OLS out to determine the different firmware from fpga and the pic). Unfortunately I could not do this within OSX. So i installed parallels and Windows 7. Had a lot of problems (opening com7/problem determine OLS ID). But finally i managed to do this.
I have upgraded both to the latest version of firmware/PIC software.
When I test the OLS again, a lot of problems where gone. Like the strange read signals on the channels which were connected to the ground. Also my IC2 decoding is working much better then before.
So this is verry good.
Now some test:
Capturing ic2 data at 5Mhz with only channel group 0 selected (Simple trigger) gives me 4.903ms of capture data to look at. ---> OK
Capturing ic2 data at 5Mhz, with only channel group 0 selected and RLE (Simple trigger) gives me 272.320 mS of capture data to look at. --> OK
but now the strange part!!!
Capturing ic2 data at 5Mhz, with only channel group 0 & 3 selected and RLE (Simple trigger) gives me 11.031 S of capture data to look at. --> OK
But when i select channel group 0 & 3 I'm able to capture 11.031 s (notice no mS but S!! ! ! ! !)
This is strange because in the second capture (only channel 0 & RLE enabled have twice as much memory (24K) as the last test (12k), but is much more captured data.
I have done this test now a few dozens of times, to make sure that my readings are correct.
With kind regards, Martin.
RLE works in combination with the channel groups that are enabled. So, 1 enabled group gives you a RLE count of 7 bits (the 8th bit is the RLE flag). With two channel groups, you already get 15 bits for the RLE count, allowing longer "dead" times in between samples. Might this be the cause of your issue?
mmh, could be, because the "dead" time is rather large in my test situation.
I will try with a more signal on the line, and find out the difference, and post the results off course.