Electrical puzzle

pessimist

Well-known member
Joined
7 May 2003
Messages
3,177
Location
Exmoor. Boat in Dartmuff.
Visit site
Is there anything connected to the Load terminals of the controller ? Bedouin might have something there; a :)

On Victron kit you can adjust the parameters for Load disconnect.
Yes, the RPi is conneted to the load but since it is measuring the current it obviously not switching off. Nice try though.
I'm using an Epever controller and this too allows me to adjust the load disconnect parameters (disconnect is set to 12.2volts) and the batteries are not getting close to that.
 

Daverw

Well-known member
Joined
2 Nov 2016
Messages
2,889
Location
Humber
Visit site
Is it worth moving the RPi off the load output just in case this effects the shutdown of the solar, ours appears to go to sleep when no input with just comms triggered and blue led just blipping every so often. I originally had a RPi running Venus OS but found current draw would vary from 0.3 to 1.2 amp for no reason, this was just the RPi and router running. Now have Cerbo and router and it’s steady
 

pessimist

Well-known member
Joined
7 May 2003
Messages
3,177
Location
Exmoor. Boat in Dartmuff.
Visit site
Is it worth moving the RPi off the load output just in case this effects the shutdown of the solar, ours appears to go to sleep when no input with just comms triggered and blue led just blipping every so often. I originally had a RPi running Venus OS but found current draw would vary from 0.3 to 1.2 amp for no reason, this was just the RPi and router running. Now have Cerbo and router and it’s steady
Thanks, worth a try. I'll switch it next time at the boat.
 

vas

Well-known member
Joined
21 Jun 2011
Messages
8,083
Location
Volos-Athens
Visit site
none mentioned the BMs, are you sure they are not to blame?
I'm on a fully victonised setup, never seen anything like that, but again I have quite a few things running 24/7 during the season.
inaccurate measuring due to V going slightly off spec on the BM chip, altering calcs and presenting slightly bogus A values is possible, ground loops all sorts of silly things can happen (well have happened to me at least ? )

good luck.

V.

PS. nice to see GHA back!
 
  • Like
Reactions: GHA

Thistle

Well-known member
Joined
2 Oct 2004
Messages
3,982
Location
Here
Visit site
For some time the local office of an international electronics company was puzzled by its alarm system going off at almost exactly the same time each morning. No amount of tinkering with the system would clear the fault and the alarm kept on going off each morning. A team was designated to stay in the building overnight and see what was happening: there was nothing going on in the building but still the alarm went off every day. Eventually someone looking out the window noticed that the alarm went off as a train passed on the adjacent line.
The moral of the story? When tracking down electrical faults, don't just look at what is connected to the circuit, there may be an external influence at work.
 

cbemoore

New member
Joined
10 Sep 2016
Messages
6
Visit site
As someone has previously mentioned, the load is dropping from 0.5A to 0.3A around midnight, so something is turning itself off, not on. The load increases again in the morning just before the solar panel starts charging. So I'm guessing your solar controller eventually goes to sleep if the panels don't produce any charge, and then wakes up as soon as the panels start charging.
 

GHA

Well-known member
Joined
26 Jun 2013
Messages
12,458
Location
Hopefully somewhere warm
Visit site
If you are familiar with Grafana you can view the live graphs here.
must check out skippers mate one day ?
You prob know this but maybe someone coming from a google search won't - you can get grafana to plot max data over a time period - useful for wind gusts. Group by time & select|selectors|max.
Prob won't help here but you never know, might pick up a weird spike or something.

1661940096325.png

If you really want to go to town for a few quid you can get a very accurate ammeter up to 3.2A (or more with a separate shunt) . Not actually tried it straight into a Pi ( I use esp32) but should work OK.
Signalk is just fantastic!!!! ?
signalk-raspberry-pi-ina219
 
  • Haha
Reactions: vas

pessimist

Well-known member
Joined
7 May 2003
Messages
3,177
Location
Exmoor. Boat in Dartmuff.
Visit site
Thanks everyone. I'll try to answer everyone in one post.

Can't guarantee it's not the BM without installing another monitor and comparing the results. Still not sure why it would have a nervous breakdown between 2400 and 0200 on a regular basis.


External influence has to be possible but I think it would have to be something on board - we're on a swinging mooring away from most things.

The time is consistent in as much that it is between 2400 and 0200. I don't have last month's data. Guess I can wait until next month:)

Yes, agree something must be switching itself off but I don't think it can be the solar controller as the specified self discharge for this is sub 15mA.

Didn't know you could do that with Grafana - I'm still a bit of a neophyte with it. Agree that SignalK is excellent. I love the way you can mix 0183 with N2K with home written routines. Wouldn't bother too much with SkippersMate - we're no longer supporting it for reasons I'm not allowed to mention here. All the sources and documentation are still avai;able to download, though, if you're curious.

Once again, thanks to all.
 
  • Like
Reactions: GHA

sarabande

Well-known member
Joined
6 May 2005
Messages
36,047
Visit site
I am casting flies wider and wider at the problem...

If you are on a swinger, does the PV panel input vary (especially for ~6 hours prior to dusk) according to the boat's ebb/ flow orientation ?

I am wondering the 0200 ish change is due to a good solar charge prior to 2000 (panel pointed at the sun) , and the 2400 change due to a weaker charge with the panel pointing indirectly.

If the controller is working properly , it might be a battery (or other sensed power source) running out of steam and its voltage runs below the parameter set in the controller for nighttime operation.


I have other weird hypotheses waiting ... :)
 

pessimist

Well-known member
Joined
7 May 2003
Messages
3,177
Location
Exmoor. Boat in Dartmuff.
Visit site
I am casting flies wider and wider at the problem...

If you are on a swinger, does the PV panel input vary (especially for ~6 hours prior to dusk) according to the boat's ebb/ flow orientation ?

I am wondering the 0200 ish change is due to a good solar charge prior to 2000 (panel pointed at the sun) , and the 2400 change due to a weaker charge with the panel pointing indirectly.

If the controller is working properly , it might be a battery (or other sensed power source) running out of steam and its voltage runs below the parameter set in the controller for nighttime operation.


I have other weird hypotheses waiting ... :)
The solar input certainly varies with the tide but the tide moving in doesn't seem to affect the timing of the "blip".
Don't think it can be the batteries (and they are the only power source) running out of steam as they've gone nowhere near the 12.2v set in the controller. Or have I misunderstood?
 
Top