question

hobohome avatar image
hobohome asked

DVCC Mode with 3 Smart Solar Controllers (and Venus GX)

Hi,

I have just replaced some [other brand] controllers with three shiny new Smart Solar controllers and a Venus GX (I already had a Victron 5kva MultiPlus). I have turned on DVCC mode and there is a Bluetooth temperature/voltage sensor feeding this info to the controllers.

I am struggling to understand the charge control process - can someone please explain it to me?

I was lead to believe that in this mode, all three controllers would operate together as one. With them all moving from bulk mode to asorb and then to float at the same time.

This is not what happens. They all seem to do their own thing. For example, for yesterday they each report...

Controller East
Bulk = 494min
Asorb = 19 min
Float = 297 min

Controller North
Bulk = 508min
Asorb = 205 min
Float = 99 min

Controller West
Bulk = 506min
Asorb = 194 min
Float = 109 min


In graphs showing the battery voltage, I would expect to very clearly see the three charging stages, this is not the case, the battery terminal voltage just seems to fluctuate between the programmed asorb voltage and the float voltage. I'm struggling to see how this can be a good thing for my poor battery bank. The battery manufacturer clearly states how the batteries should be charged, the voltages for each stage and the triggers for moving to the next stage. I have programmed both the MultiPlus and each of the Smart Solar controllers with these settings, but the result is a long way from what I expected to see.

Perhaps I'm missing something?


Hobohome - Australia

Venus GX - VGXDVCC
2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

hobohome avatar image hobohome commented ·

Just to add further information on my system to this question...

  • My system is off-grid. I have a backup generator.
  • The three controllers are attached to their own array each facing a different direction.
  • All three controllers feed into the same battery bank.
  • All Victron components (including the MultiPlus) are operating the most recent firmware
  • I have a BMV 702 - but this is NOT connected to the Venus (I may relocate it soon and then connect it to the Venus).
  • I do have a Smart Battery Sense connected by bluetooth to the controllers and I can see the temperature and terminal voltage displayed in the IOS app (but sadly that information is not fed back to the Venus).
  • I have ordered a temperature sensor to allow the Venus to feed battery temperature data back to my other systems via MQTT - this has not yet been installed.

The problem I face is that I do not fully understand what behaviour to expect from this setup - thus it is difficult for me to understand if it is performing as intended.

0 Likes 0 ·
hobohome avatar image hobohome commented ·

Just as some information to anyone else who finds this thread.

As I stated above, I have BOTH a Venus GX (now with a temperature sensor) AND a Smart Battery Sense bluetooth device. This seems to be a bad combination. I found that one of the Smart Solar Controllers changed mode about once every minute. In searching for information on this I found on this page https://www.victronenergy.com/live/victronconnect:ve-smart-networking Q6 of FAQ that having the Smart Battery Sense in a Venus controlled setup is NOT advisable.

Once I removed all controllers from the bluetooth network, this issue of the controller changing mode stopped.

0 Likes 0 ·
3 Answers
Peter Polz avatar image
Peter Polz answered ·

At the moment they are not working as one. The function of multiplie smart mppts beeing sync in loading stage is an announced feature. Actually only the good old canbus mppts have sync stages.

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

hobohome avatar image hobohome commented ·

Thanks Peter. I am very interested to read the announcement from Victron regarding planned feature of synchronised charging with Smart Solar Devices – can you please point me to that announcement?

0 Likes 0 ·
Peter Polz avatar image Peter Polz hobohome commented ·

Sorry i can't point to it at this moment because it was talked about in a disquse entry and it was definitly said that there will be no concret date of realizing this.

0 Likes 0 ·
JohnC avatar image
JohnC answered ·

Hi Hobo. (Peter's beaten me, but I'll post it anyway).. What you're experiencing is because your mppt's are not synchronised. Nor can they be, unless you have models with CAN (unlikely, but see here: https://www.victronenergy.com/blog/2013/11/15/synchronizing-multiple-mppt-15070-charge-controllers-2/ Note a few years old. If your supplier/installer has promised you synchronisation then perhaps call him to task.

I guess your panel orientations aren't helping either. You could try tuning the setpoints to bring them closer together, or even to manual setpoints. But frankly, I wouldn't do that until you get your head around why each mppt is doing what it does. You'll need to look at the Amps from each too, then nut it out.

Eg. In your figures above, the E array might kick off first in the day, go through Bulk all the way to Absorb. By then the N array also reaches Absorb. Call it manufacturing tolerance, wire length, whatever, but let's say the N array sees 14.40V (@12V) then, and the E array 14.41V. By then, overwhelmed by both the N/W arrays holding what it sees as 14.41V, it will wind back it's A quite quickly to a point where it sez "these batts are charged, off to float". Make sense?? Watch the A..

Complicating this is the Victron charge algorithms. Not published except for snippets (trade secret?, understood), but they are excellent. Smarter than I am to do manually, much smarter..

One published 'snippet' is the length of Absorb time. It will shorten it if you don't draw much from your batts overnight. 209min is 'up there', and suggests you may have pushed them a bit the night before??

Enough for now. The point being just get to understand why. The VRM portal graphs can explain it pretty well, invaluable stuff for analysis..

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

hobohome avatar image hobohome commented ·

Thanks for the response @JohnC. I have clearly never understood that even when connected to a Venus device, Smart Solar controllers just do their own thing.

Some things I still don’t understand…

  • If all three Smart Solar devices are getting their battery terminal voltage and battery temperature information from the Smart Battery Sense – how can they be selecting different charge parameters? Are they just ignoring the information they get from the Smart Battery Sense? Surly if the algorithm is being fed the same input data, you would expect the same behaviour from each of the three controllers?
  • DVCC mode – I understand that the Venus can limit the total battery current in this mode (that would be the current control part of the DVCC) – but what does the voltage control part of DVCC do? If the individual controllers are NOT being given a target voltage by the Venus device, it seems there is no Voltage Control in DVCC.
0 Likes 0 ·
hobohome avatar image hobohome commented ·

Just in response to the comment about the length of time the controllers spent in Asorb - the battery bank (flooded LA) is 675ah @ 48v = 32.4Kwh. My BMV 702 tells me that the SOC each morning is about 82% - 85% and I never allow them to discharge below 75% SOC. So I don't think that this quite fall into the category of "battery cruelty" :-)

The battery manufacturers do suggest an absorption time of up to 6hrs - so I'm not too concerned about the 209min.

My (probably poor) understanding of the asorb time algorithm is that the time is (in part?) based on the battery terminal voltage just before the array voltage was applied at sunrise in the morning. Again, all three controllers are being fed the same information - so why different outcomes?

0 Likes 0 ·
Mark avatar image
Mark answered ·

I believe that there are 2 contributing factors in the absorption time variation you have noted;

1- Even though all 3 MPPT's are provided with an identical battery voltage at any INSTANT point in time, they will NOT all 'wake up' at the same time (particularly since your arrays are oriented in different direction) - which is when the battery voltage is used to calculate absorption time. After 1 controller wakes up & starts charging this could have a significant difference in the voltage read by the remaining MPPT's when they eventually wake up.

2- The absorption time calculated is only a time LIMIT, the absorption phase can/will finish early for any individual MPPT if/when the charge current provided by any individual MPPT drops below a preset limit. As this is linked to the battery SOC / the current required to maintain the absorption voltage setpoint & you have 3x MPPT's charging in parallel, once the 1st MPPT drops off (goes into float phase) I would expect the remaing 2 MPPT's to then need to increase their output current to 'compensate' and accordingly the other 2 MPPT's may need to stay in absorption phase for quite a bit longer to drop below their individual current setpoint & switch to float. Same logic for the last MPPT remaining in absorption phase.

So in summary I believe that quite a bit of absorption time variation with a setup like yours is expected & normal.

Regards, Mark.

2 comments
2 |3000
Toggle Comment visibility. Current Visibility: Viewable by all users

Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

hobohome avatar image hobohome commented ·

Thanks @Mark. That makes good sense, my level of understanding has been raised (and my expectations lowered).

I look forward to seeing some information on the mode sync feature announced.

I wonder if the hold up on that feature relates to the problem of keeping the charging algorithm secret in an open project like the Venus? :-)

0 Likes 0 ·
JohnC avatar image JohnC ♦ hobohome commented ·

I haven't bothered to answer your queries that I prompted, as Mark seems to have covered them very ably. Even explains to me why I was a little tossed by your longish absorb time.

Re updates, I don't hold my breath expecting the features I'd like, but treat every download as if I'm opening a birthday present (they come more frequently too). Always surprises, and those hidden boffins do a helluva job. Must be a helluva job to do too, with the big boss so deep into IT development.

0 Likes 0 ·