That’s a good find!
I fully agree with you. Running a fan all the time never a good idea, but in this case, it will switch on only when the MPPT is starting to reduce the current.
So it should, SHOULD only run about 5-8 hours a day. About the same as a normal Desktop PC.
Just having a look around. Can i use my MP2 temp probe on the BMV702, or do i have to get one specific for the 702?
The temperature sensor for the BMV is a different kind. The one for the Multiplus is essentially an LM135. So you can’t use that one.
You can however attach the one from the Multi to the GX device, and there is an upcoming temperature relay feature in Venus 2.80 which will be released any time now.
I heard some good stuff and features with 2.80 but they don’t specify any ETA for it.
Do you know maybe, roughly when is it coming out?
Earliest estimate is early next week. If nothing serious is found. Don’t take this as εὐαγγέλιον just yet.
Ok, that sounds great…so in general Q1 2022 
I’ll hold my horses 
I assume you googled the cheeky Greek I stuck in there? 
Yep…definitely good news! 
1 Like
Been running the Beta 2.8…
Groetnis
Me too, till I had a problem where when there is LS, the BMS driver confuses the Carlo.
Had to revert back to 2.73.
Plonkster had a look, found the issues, and Louis made some changes to the driver, but I still ran into problems.
If 2.80 is released officially, I will have to try again. Eish. It is a mission to get the Carlo registered again.
NOTE: Par for the course running on DIY stuff. 
A tad more than that. There is a thing called serial-starter which has the task of detecting what hardware you have plugged in. What it does is try each “driver” in turn until one sticks. There is a particular protocol that must be followed by which serial-starter can see when a driver “sticks”. If your DIY driver crashes it is usually no problem: Serial starter tries the next driver, and the next, in a loop, until one sticks.
(Edit: So it will eventually come round circle and start the correct driver again).
If your driver doesn’t follow the protocol, then it either drops out of the loop, or it gets stuck inside the loop, and it cannot get to the correct driver.
And this is what happens. The BMS and the Carlo Gavazzi uses the same type of connection, so the system tries CG first, then serialBMS. So at bootup there is no indication that anything is not working as it should.
When the power fails, the CG goes away, so serial-starter tries the next driver in the list, eventually gets to serialBMS… and gets stuck there.
When the power returns, the CG does not come back, because the serialBMS driver is holding on to the port.
Drivers must terminate within a few seconds if they do not find the expected hardware on the port. I think this is currently not working in the serialBMS code, at least not for everything in there.
So technically the Carlo is not getting confused. It is simply never tried, because someone else has indicated that it’s handling that serial port.
Edit: Or in short: SerialBMS hangs when it tries to talk to a Carlo Gavazi. It should detect that it’s the wrong kind of hardware and terminate immediately.
1 Like
Thank you.
The driver works perfectly on 2.73, just not on Release Candidate 2.80.
BUT I’m quite aware that a whole lot has changed on the Cerbo/Venus OS, that everyone who made add-ons was given a heads up ahead of time, and I fully accept that one will experience issues. So I’m not worried at all … it is par for the course in my book.
Once @Louisvdw can get around to it, I’m sure the issue will be resolved, for not-my-business concern is that if more people upgrade to Rel 2.80, that they run into this problem, after LS events.
Can only hold thumbs.
Here in the Northern Cape it gets very hot. 40ºC+
Not good for Panels sitting in the sun the hole day.
One can clearly see when it is a bit cooler and nice sunshine. Production is 40% better.
Yonks ago there were panels in Europe that had water pipes at the back to cool them, the heat exchange to pre-heat water for geysers … sad it never came to SA.