question

@mvader (Victron Energy)

Thank you very much for your efforts!

You are saying "Multi RS: Fix switching from inverter only to passthrough (only applies to units with feed-in enabled)."

But from what I am reading here on the community board, these units, Multi RS, don't have yet the feed-in capability enabled... Am I wrong?

And because the ESS on these units is not as versatile like on the other units, can be implemented an option like "Don't feed-in in any condition"?

Because some of us have interdiction to feed-in and are subjects of fines if the provider's smart meter detects feed-in.

Installed , will give feedback if any issues from my end

Installed and looking good will feedback in due course.

Today, instead of forecasted mostly sunny (and shady) day, we have a continuously and evenly covered (by stratus) day, so no shades, but very regular illumination. I thought I couldn't report anything back...until...

However, with 1.16-beta02 I saw some strange "global optimum selections, on both MPPT RS 450/100's on string 1:

First graph is string 1 U (voltage) +I (current) curve of first MPPT 450/100 unit. Graphs 2/3 are first string from second MPPT 450/100 unit (graph 2 U+I curves, graph 3 P (solar power)).

You see regularly the U drop to around minimum MPPT U (48V system) for a minute (VRM doesn't log more than a minute, and unit unfortunately doesn't log U/I/P curves for when connecting with Victron Connect), so, without the I going up most of the time, except 1 time, that one compensating entirely the U drop in the P, but the others decisions of MPPT algorithm to choose a low voltage without any gain in current seem rather strange.

Tracker 2 of both units doesn't have this behavior, but they are in the opposite direction and have less panels on them, so a lower Voc, which may explain that they act differently.

It would really help to have a log of the P/U graphs of the global-optimum sweeps and a few hours of logs of second-level U and I curves inside the RS units, or on the GX, to understand what happened and why the unit made the decision to lower the U workpoint, while the current didn't compensate for the power.

But this looks to me like a new MPPT bug in this beta 02, as I didn't notice this in the previous days with fog in the beta 01.

I am trying the 1.16 beta firmware to see if it will correct an issue that I have only recently noticed with my 450/100.

I have two strings of panels. One facing South East, the other South West. Both are 2.2KwP and until recently the output was similar on sunny days, but with the weather recently and the low sun its been difficult to tell if the issue is just the weather/time of year.

Or is related to the partial shading algorithm. I am finding one tracker drops to very low output the other seems unaffected.

Unfortunately it has also coincided with me rewiring a DC isolator so potentially the connection is at fault and I will need to do more testing.

However the tracker 1 with the low output has peaked at over 1kW for several minutes whilst the normally better performing side tracker 2 fell away, so it's a bit puzzling.

I upgraded to v1.16-beta-02 (from beta-01) and it caused a short circuit alarm (Error 27) on the MPPT. I could not clear the alarm at all (including all power removed for 5 minutes)

Downgrading to beta-01 cleared the error. I will try again tomorrow. For now I need the power

Hi. Thank you very much! I just discovered the beta version, as I was concerned about the poor power output of my MPPT 450/100 RS. Partial shading occurs in the morning and afternoon, with small shadows passing over the upper string. I was surprised because these are 415W modules, yet four modules without shadow produced only 250W.

Initially, the update made a significant difference. I observed the current doubling immediately after installing the beta, but this change only occurred on tracker 2. Unfortunately, the logging system isn't fast enough to capture the impact. As far as I can tell, the behavior of tracker 1 hasn't changed at all.

Now, after a period without shadows, the previous behavior (at 14:19) returned, and it hasn't managed to reduce to a low voltage. I briefly observed the lowest voltage at 96V when only three modules were completely covered by the sun, which I assume is correct since my datasheet specifies Vmp as 31.45V per module. With about 3 1/2 modules covered by the sun, that should be around 94V. However, it consistently remains above 200V, whereas I believe it should be closer to 126V.

Consequently, I assume that the tracker is still not "searching deeply enough" for my lower string with four modules and for the eight modules when the lower four are shadowed.

v1.16-beta-02 is installed at 13:14.

Good day

We have an almost sunny day, with light clouds today

Here is a graph, thanks to @Guy Stewart (Victron Community Manager) for the Advanced Widget suggestion

I've done overnight discharge of 30kWh to SOC 30% and now we have a clear "do full charge as you could" situation in my setup

South facade - 90 degrees - a shading situation from neighbor's house

The west facade - 90 degrees - kicks in when the sun comes to direct light afternoon

My concerns are - flat horizontal bars, while King PV on a roof plays with MPPT, Victron stays on a steady level. Could be no clouds between the sun and the array, but for me it looks like Victron mppt is lazy in looking for a better power point or I could be just speculating here @mvader (Victron Energy) - Something wrong with VRM - it shows horisontal bars, while PV power is getting changed @Guy Stewart (Victron Community Manager)

Link to my setup https://betavrm.victronenergy.com/installation/370388/share/c0152a83

Report from yesterday https://community.victronenergy.com/comments/249721/view.html

Hi @mvader (Victron Energy) ,

Please help me with a way of reasoning..

In order for the MPPT to be precisely tested, you must draw a little more power above the maximum the tracker can give at a moment.

Of course the surplus needed will be compensated by the battery or grid, but you must draw above the tracker's limit, in order to see if the tracker is correctly following the maximum power point.

Why I am thinking this way?

Because, I believe, otherwise there can be a limitation imposed by the fact that the MPPT will limit the power generated, because that power doesn't have where to go, especially that now in the Multi RS case is not enabled to feed in. And the user has the impression that the MPPT doesn't generate the power as expected.

Am I talking nonsense?

Please go easy on me, I am a noob in this area...

I saw on VRM that yesterday (with 1.16 beta 02, and batteries with plenty of charging current reserve so that no tracker would have to limit its current) that one of the trackers of one of the 2 RS 450/100 was doing regularly strange drops in current and power, while others were not doing this, here the I&V and P curves of both RS 450/100s.

Both trackers 1 are low-angle South-West orientation, trackers 2 are low-angle slightly East-North.

Could that be a bad sampling timing of VRM data during swipes ? As mimimum interval for VRM being 1 minute, if that's the case, it would make sense to not send samples when scaning. On the other hand, I would expect at least power to be averaged over the sampling interval (1 minute in this case).

Second graph below shows on same South-West pane, in same interval, 3 trackers, the one in white is 2 panels on a Smart MPPT 100/20, while trackers 1 of RS 450/100 are 8 panels. The regular drops of one of the RS 450/100 tracker 1 are clearly visible, while the other trackers don't drop, so it's not a sunshine issue.

A notable thing that can be seen on this graph is that while 8 panels on 450/100 (especially on the orange tracker that is at same height than the white 100/20) should give 4 times the power of 2 panels on MPPT 100/20, they give around 3.5 times the power (around 13% less), when not dropping regularly. So it looks like there is still room for improvement in real conditions.

I've installed beta2 8 days ago, see my comment https://community.victronenergy.com/comments/251040/view.html

Dips I saw on old firmware from my Multi RS 6000 Dual tracker

Now they are seems like gone

Today we had a complex day, a lot of clouds/sun, can't complain

@mvader (Victron Energy) @Guy Stewart (Victron Community Manager)

Link to my setup where you can check results by yourself in Advanced graphs - https://betavrm.victronenergy.com/installation/370388/share/c0152a83

Thank you

Hi @mvader (Victron Energy) and @Guy Stewart (Victron Community Manager)

Firmware changed today...

A strange, or maybe normal, behavior as soon as the firmware was changed.

The first tracker, like in the case of @beat begin to manifest a saw-tooth behavior in power generation at 5 minutes interval.

See on the graph below, before (no saw-tooth) and after flashing (saw-tooth present).

As the sending data interval on VRM is set to 1 minute, I've watched closely the Remote Console for a second to second basis read-out.

Indeed, at 5 minutes intervals, it seems that the tracker is doing a full sweep, sweep that don't last more than 5 seconds, with a full voltage/current excursion. Then it seems to settle on the max found.

But on VRM, even if the drop is around 5 seconds, it is graphed like it was down for 1 minute. Probably the send data is somewhat synchronous with that power track/sweep and VRM records some of the low values during sweep.

Is this the intended behavior?

Probably if you move the send data to a different offset during time, it will not appear so obvious on the graph...

Thank you!

I've noticed some strange actions from panels on my RS450/100's.

Look at these charts and the red line. No shading affecting these panels (all roughly the same location). It should mimic the blue line as the panels are just in front of each other.

Also on the start of each day, the red tracker always shows this 'higher gain' than all other strings. This is the same every day,

I still think tracker 2 response time is poor , even with very minimal shading it stays on lower power for 1-2 hours more , during which if you reinstall or reboot the mppt, max power can be extracted from tracker 2 . I think time for seeking if shading is gone needs to be reduced further or a button for rescan should be introduced , problem solved

Hi. There were a few sunny days here and I like to update you about the performance of the bottom string at tracker 1 (see my pictures above).
The bottom string have most of the time at least two modules fully illuminated but the current stays low at around 25W. I guess it could produce 700-1000W with the partial string. I believe that the current of the lower string should be nearly the same as the upper string, while the voltage should drop. But the voltage stays high and the current drops. There are moments, where it seems fine.

I like to provide you with the information you need. Please tell me if I can do more to improve it :-)

Hi developers,

@Guy Stewart (Victron Community Manager) , @mvader (Victron Energy)

I've saw, at least monitoring my Multi RS Solar, that the (re)tracking process is done at 5 minutes (300 seconds) with an offset of 2.5 minutes (150 seconds) between Tracker1 and Tracker 2.

I mean something like: .... - retrack1 - 150s - retrack2 - 150s - retrack1 - 150s - retrack2 - 150s - ......

In order to get out of that sync with the VRM send data that generates that saw-tooth pattern sometime, can I respectfully suggest to choose the (re)tracking interval with a little bit of random?

Something like: retracking_seconds = 290 + random(20)

In that way the (re)tracking interval will be between 290 and 310 seconds, with an average on 300.

But for sure will be (out of sync) most of the time with the VRM send data.

Thanks for listening and again, no offense, it's just a suggestion...

Alex

Maybe I found a bug in the new v1.16 beta 2 algorithm:

When the optimum is just above Vmin, it works fine, but when the optimum is (just) below Vmin (48V system) the MPPT goes to Voc instead of staying at Vmin, which ruins the power output: That was the case at MPPT2 of both RS450/100 this morning as snow is melting, uncovering partly the panels:

In following graph: One line per MPPT:

Unbroken sunshine today and power is fluctuating up and down continually on one string whilst the other outputs the expected power. There is no rhyme or reason to it. If I could find version 1.14 I would go back to that as this is frankly unacceptable.