Ever since I upgraded to the latest firmware I've noticed the display on the Classic occasionally goes to zero's and then it comes back. This is different than when the watts drops down quickly and come back up when it is doing a sweep. Other fields like battery voltage go to zero. It only happens for a split second and then everything is back to normal. It happens so fast that I can't look at all fields at the same time to see if they are all zero or just the field I was looking at but I think pretty much everything goes to zero for a split second. I don't ever remember seeing this before the last firmware. I have no idea if this happens at regular intervals or not but I've seen it roughly 5-10 times over the past month or so. Anyone else see this?
FW 1181 and I did do a VMM after updating the FW.
MNGP FW 1042
OK, I'll take a closer look. It may be because the Classic communications is a bit more busy nowadays.
I think I have noticed a wee bit of this. The text recall of logs data is slower to respond as well
lately but the reading of graphic data, which reads much more data at one time, is still
fairly fast.
Thanks for the input !
Hopefully we can speed that up a bit in the next update or two.
boB
Thanks for the response Bob. It's certainly not a problem but I just thought I would mention it in case it was only me seeing it. I had noticed the slowdown on the text logs recall as well and was wondering what happened there. Again, not a problem. I just wish I had more panels to give my Classic more of a workout. I added batteries this year so my panel additions will have to wait another year.
I have seen this and gotten zeroes while logging, just passed it off as it was "closing its eyes and thinking". I modified my logging script to reject zero values so it is all good.
Just FYI.
Tom
Quote from: TomW on January 22, 2013, 01:38:40 PM
I have seen this and gotten zeroes while logging, just passed it off as it was "closing its eyes and thinking". I modified my logging script to reject zero values so it is all good.
Just FYI.
Tom
Let me know if actually read zeros after a successful read. After a short timeout, or time without communication with
the Classic itself, it is supposed to show all zeros on the screen, rather than other wrong information.
i.e. I don't think you will actually read zeros like that through the modbus, but please let me know if
you do.
boB
Quote from: boB on January 23, 2013, 01:56:19 AM
Let me know if actually read zeros after a successful read. After a short timeout, or time without communication with
the Classic itself, it is supposed to show all zeros on the screen, rather than other wrong information.
i.e. I don't think you will actually read zeros like that through the modbus, but please let me know if
you do.
boB
I actually read the modbus data via the serial data dump. I have had no luck getting something that does modbus on the Arm architecture on the Pi. Mango is way too steep a learning curve and modpoll is not available for Arm. Modpoll works fine on X86 like my laptop but I want the low power solution of an Arm board.
It is possible my scripts misinterpreted something and I got zeroes not actually in the data? I did not get too deep into finding "why" but just fixed it so the zeroes were rejected. serial data on these Pi's seems to be overloaded by some "sharing" of resources with ethernet and I have seen occasional erroneous data. Maybe it is unrelated to the Classic itself.
Sorry if I misled anyone but I did not go too deep into the issue at the time. Still learning how this beast works.
I will run another unfiltered log and see what it shows.
Tom
I happened to see this occur tonight and it said Got Comm? in the lower right(not sure if it had a question mark or not, happens too fast) when the display went to zero's for the brief instant. Just providing some information.