question

I'm not understanding what your complaint is about the strategy... it is the 'safest' approach as you don't need to wait for the next bluetooth data packet to come across with the 'new' battery voltage. Also if there is a poor/intermittent Bluetooth signal everything keeps working relatively well.

If you have say 0.3v of voltage delta between the MPPT and SBS/BMV reading, then it will apply that amount of compensation to the MPPT measured voltage (resulting in the SAME voltage). But it will still use the MPPT measure voltage as the base for accurate millisecond to millisecond control (with the required amount of compensation applied).

If the charge current and accordingly voltage drop changes, then the amount of compensation applied will also change according - so again no issue.

This allows the best overall control system - you get direct / millisecond response to rapid voltage variations and also get to use the correct voltage as measured at the battery terminals.

An issue will only arrise if you have some kind intermittent wiring issue where the resistance can change drastically from one instant to the next. That is what I recommended for you to investigate. A constant resistance of Xohm will be totally accommodated for.

Also did you investigate why the charge current dropped significantly at the SAME time as the brief voltage spike? Did a large load turn off at that time?

Regarding #3 - The response dampening is reasonable to ensure proper operation during dynamic scenarios, again this is only for the amount of compensation applied that should not change drastically in an instant. The MPPT will react instantly (based on its own measured voltage) to any voltage changes with the predetermined amount of required compensation.

Also if you were to really have 1v of voltage drop with a 12v system, then the wiring is inadequate and probably unsafe.

Regarding #2 - Try it during the day when the MPPT is not in OFF state. There is no need for the MPPT to be compensating for voltage drop if it's not active/charging.

I don't see the issue again, you should NOT be connecting the SBS to a battery that is physically not connected to the MPPT (or that could be disconnected while other 'unknown' batteries are still connected) .

What did you expect it to do in your example?

It aims to achieve a voltage as measured by the SBS at the battery terminals.

The SBS is connected to a battery that is at 13.2v and the MPPT wants it to be at the 13.5v/13.8v target.

The MPPT will initially measure the voltage of the OTHER battery, but expect it to be the same battery and apply a compensation of whatever the delta is.

As nothing will be changing with the SBS measured voltage it will keep increasing it's output voltage in an attempt to achieve the 13.5v/13.8v target (as measured by the SBS), but it doesn't / can't because it's a different battery and not connected to the MPPT charge output.

The MPPT expects that the voltage delta is due to losses etc. and that a higher voltage at the MPPT output terminals is required to result in the CORRRCT lower target voltage at the battery terminals.

If it used the SBS voltage directly or any type of wired voltage sensing it would result in the exact same result in your example.

Hi Dean,

It seems that you were busy testing yesterday - nice report.

The independant baseline is good information, as are the recordings with SBS paired.

I agree that the data you provided suggests that an intermittent wiring issue in your setup is unlikely and also that there is a variance/lag between the voltage as reported/used by the MPPT (including compensation/correction from the SBS) and the voltage as reported by the SBS.

My comments are;

1- The examples that you provided are during conditions with rapidly and significantly fluctuating charge current, as the correction is based off voltage delta and has some integrated dampening - in these conditions it seems to not alter the compensation amount fast enough to 'keep up'.

As you would know, if the charge current goes from say 10A to 30A (3x) in an instant (or the same change in reverse), then even with a constant wiring resistance (no intermittent wiring issue) the voltage drop still increases by 3x.

Maybe there is some opportunity to reduce the amount of dampening to increase the voltage correction response speed, but generally these things are done for a good reason and after a lot of testing - so it may have negative repercussions elsewhere. But I will try to find out...

2- The PRIMARY and most IMPORTANT purpose of charge voltage compensation (in relation to the true voltage as measured at the battery terminals) is to ensure that the battery bank receives the correct/accurate CHARGE voltage.

This is specifically when the TARGET/SETPOINT charge voltage has been achieved and is being MAINTAINED. While charging at the correct charge voltage setpoint, the charge current will always inherently be relatively consistent.

It doesn't really matter (in relation to accurate battery charging) if a slightly different voltage is momentarily reported under highly dynamic conditions and when the battery voltage is BELOW the target charge voltage. During this time the MPPT is in 'max power point' mode and trying hard to increase the the battery voltage to the setpoint voltage.

3- Sorry to say but I am still worried about your wiring (and anything along its path, such as the BMS contactor and fusing/circuit breakers), due to the apparent high resistance and voltage drop.

I think that the wiring or something in the circuit is likely undersized or has an issue - it would be good to identify the primary contributor and upgrade/fix it. This IS proportionally related to the behavior you have seen.

Your baseline example shows that you have 0.56v of voltage drop at a charge current of 26A. This means that you would have >1.0v of voltage drop at the full MPPT charge current of 50A... thats over 8% loss for a 12v system... this is huge and potentially unsafe.

If you sort this out then any lag under highly dynamic charge current conditions will be proportionally reduced.

I've commented down-thread on one aspect of your suggestion/concern. And, I'm not here to defend Victron gear or the SBS remote sense implementation in particular. (I also don't have an SBS in my system.)

However, I feel compelled to note that despite your deep dive into the issue(s) you're having, it's still really hard to tell what's going on with the data you've presented.

One thing that makes it tough is that the graphs are not aligned, and it doesn't really even look like you are sampling your data points at identical times. Maybe you are, but it's impossible to tell as shown.

Another thing is that your spreadsheet columns don't match across the scenarios, and it's unclear why you wouldn't make the columns match. I'm not suggesting you're trying to hide something, I just can't follow them very well.

Another issue is that you aren't explicitly describing what mode the MPPT is in in each case. Are we in CC? CV? CV float?

If you're seeing the issues when in CC so far, I'd be very interested in understanding more about what happens when the MPPT is in CV (absorption). I can't see where you've written about this, but it's really important.

Lastly, I would suggest if you're up for it that you make a video of your first, most basic test. No need to do four runs or to do the second test. If I'm understanding your test correctly, it should only take a couple minutes to record a video. If you could show the wiring end to end, and if you could also take a good meter and make some independent readings at the same points (battery terminal, upstream at the BMS, MPPT output terminals) that would help a lot.

I suspect you're seeing an issue with your own setup and not a fundamental flaw in the SBS implementation, but if you have uncovered a real issue, I'm sure Victron will want to know about it and get it squared away. So far, though, it's still too vague to be actionable in my opinion.