Victron MPPT charging, Trojan T105RE, temp compensation and bulk voltage/duration Q

[vas]

morning all,

I vaguely recall discussing these issues before and various ppl contributing, noalex (IIRC) mentioning something about a substantial alteration of charging profile by Victron in a firmware sometime last year. I do have all devices in my system updated to latest firmware btw.

anyway, I've setup a raspberry pi3B+ with it's 7in touch screen running Victron OS (Venus) and wired my BMV700 and the MPPT30/100 on it. Not done the Multi inverter/charger but tbh I'm not using the boat right now so no big deal.
I've also temp fitted a wifi router and I'm "stealing" internet from two local places in port and as a result I have the raspberry sending it's logs to Victron cloud (VRM). There you can get lots of data on battery state (voltage/current from the battery monitor) as well as MPPT data (solar panel voltage and current). You also get a combined W consumed value (MPPT - BMV Amps).
About to complete the third month of full lockdown here you can imagine I have plenty of time to do nothing which I cannot stand. So ended up setting up another raspberry pi connected to the NMEA2000 bus and taking all sorts of values from there. This together with the Victron derived data gives a good overview of what's happening on board when I'm at home (admittedly less than 10min walk away but not allowed to visit for more than 2h and in theory only once a day...)

Enough intro!
have the thing logging at 1min intervals and checking visualisations with grafana.
There I first noticed two things (which are the point of the post!)
Bulk voltage was set as per Trojan manual at 29.6V
However MPPT was pumping 30.05V or around there!
checked settings to find that OK, there's a compensation for battery temp reducing bulk voltage by a set value for every 1C above 25C (I think). Didn't realise that when temps go down (was probably around 5-8C in the e/r at the time) V compensation goes up accordingly
just to be safe (and make sure I don't burn any other device onboard, I dropped bulk to 29, so now it only reaches up to 29.4-29.5 on a v.cold day, today fe is 29.45, yesterday was sunnier was 29.3
So first Q, is that OK/normal? is it acceptable to bulk charge them at 30+?
Mind batteries are doing bugger all atm, there's a fridge that turns on for half an hour once or twice a day (should probably leave the door open and it wont cycle at all ) the raspberry and an alarm/domotic system. Overall shunt reports a 0.4A consumption constantly throughout the day plus 2.0-2.5A when the fridge is on.

The other issue is that the last couple of days with clear skies and the Jan sun "shining" I do get the system to bulk charge at this 29.something for almost 2.5h daily.
Second Q, is that OK? seems excessive to me, no?
Actually having the batteries at 99% SOC at 8am when the solar start sort of working, means the 29.something starts at 10am till around 12:30 then hovers at float 27V till the sun sets then down to 25.5V and in the morning it's at 25.2V to start the new cycle.

Not complaining, just wondering if that's the acceptable/proper behaviour for this setup. Mind Trojans are not in a good state although only 3yo (and it's me to blame but I'll leave that for now) and I now regularly water them (got the watering kit fitted...)

cheers

V.

PS a couple of pics showing a typical day graph (grabbed five mins ago):

first from Victron VRM (just need a venus running device like a GX)




and next from grafana (raspberrypi-signalk-influxdb-grafana - that's the route of data to get there...)

 

[TernVI]

Victron seem to change their firmware quite a lot, quite often.
Leading me to think they were never happy with the previous edition, and why do we think this edition is any better?
The bulk of Victron's products don't go into yachts, so I have doubts that their firmware is ever really optimised with a typical yacht installation in mind.

The other thing is that the Trojan bulk voltage is based on a charging profile involving a lot of current early in the cycle. If you are not matching that current, then the voltage is no longer valid, you can't mix and match numbers from different charging profiles. (actually you can do what you like, they are your batteries, but don't blame anyone else for the charging profile if you don't follow all of it).

Thirdly, I never trust data loggng when there's SMPS involved, I'd want to have a proper look with a scope of bandwidth many times the switching frequency.

 

[GHA]

Does the BMV tlk to the regulator? Well chuffed with new smartshunt, tells the smartsolar what's going on at the batteries.

For the questions you'll no doubt loads of completely unsubstantiated armchair opinions coming up, for deeper cycling I'm all for charge them high, more often than not the only way you'll actually get to really really 100% a few days a week. On the hook I left float & bulk @ 14.9v and even then with 300W solar on 225Ah getting all the way back up was hit & miss. Your battery voltage must be a bit under the

Default settings of 4% tail current for fully charged on the smartshunt is way too high imho , Smartgauge and ammeter
What os interesting there is just how much the charging current drops off with tiny drops in voltage in bulk. And makes it more obvious how most systems will go to float before the batteries are actually back to 100%.

"Charge em high" is based on a large study unfortunately unavailable to the public but outlined here >
Need help with charging schedule for L16 interstate batteries

Plus trojan recommendations will very likely be a compromise based on what most of their batteries are used for - golf carts and other electric stuff put on charge most nights - not boats, we are on our own with next to no data to go on. Personally I'm thinking of dynamically controlling the mppt to drop off a little as charge acceptance gets down low, but lots data logging first.

Bottom line is no one knows, better to charge high and get to full at the cost of unknown slight damage from a little overcharging? Who knows, I like them charged. Though right now that's maybe not really a problem for you so you could try to back off the bulk voltage a bit and check tail current still levels off before going to float.
Nice graphs!!

 

[MapisM]

Vas, as you might remember I used to have the Multiplus 3kVA on my previous boat, and bought another at the FL sale to install on the current one, but that's still pending.
And I had/have the T105 (non-RE, though I never understood the difference, if any) on both boats, and both configured at 24V in blocks of 4 batteries.

Now, I can only contribute with two observations:

1) I can't recall to have ever seen 30V on my previous boat with the Multiplus.
I did see 29+V, for instance after a night at anchor, but the voltage progressively dropped (relatively fast) as the batteries were recharged.

2) These days, while docked/connected to shore power 24/7, and DC drawing anywhere between 3 and 30 Amps (just temporary peaks, like when a Tecma pump is running), the battery charger just "follows" the absorption, but the voltage remains constantly at 26+.
But in the last few weeks, the marina was doing some maintenance on shorepower stands, and had to cut power for up to half a day in a few occasions, so the domestic bank had to stretch its legs a bit.
And upon restart of AC supply, the charger was pumping current in bulk at 28+ Volts, but I've never seen a 9 after the 2, also right after turning it on, when sometimes it reached 60+ Amps (100A max being what it's rated for). And again, also the 28+ goes progressively down together with the Amps that the batteries get, usually in a matter of minutes.
The OEM charger is a Dolphin, and according to its specs, its output should be 28.8V with open lead batteries, but should be able to reach up to 30.8V with "Delphi" type batteries. Whatever that means - I never figured out how it can "see" what type of batteries he's connected to, since there is no way to adjust its working mode manually.

Bottom line, I'd be a bit nervous if I would see 30+ V in any condition.
Just my 2c, FWIW.

 

[jwfrary]

Are you using DVCC? - Needs to be switched on in Venus on the Pi. The you can use the BMV to tell the MPPT exactly what voltage to pump out. If you take a peek at the Victron help there's quite a detailed explanation of what DVCC will and wont do. (Mainly it cant appreciate that there is an alternator in the system!)

What you defo can't do is use smart ve networking at the same time.

To connect your multi plus you will need a Mk3 device to link it and the pi, you cant just ram in the ethernet :-(

Also its totally possible to do everything on the one pi, might save a bit of power, setting that up might occupy your mind in lockdown!

out of interest what are you using to read your n2k data?

 

[jwfrary]

Screen Shot for the Vessel Summary dashboard for my pi, signal k install. I'm using open plotter here but I have Venus and manage several large Victron installs on the boats at work (Data's just a load of mumbo jumbo in the screen shot but I thought I would share in case it was useful)

PM me anyone if you want the JSON file

Barometric pressures coming from a BME280 which also gives humidity and temperature and then I have 4 1 wire temp senders (Fridge, engine room. outside, inside)