MPPT RS 450 shadow management?

Question

question

hotel-lima asked Sep 4, '23

MPPT RS 450 shadow management?

Hi,

I have 6 solar modules (east side of the roof) in series on one input of my RS450/200. I have a dormer which shades one solar module more than half during the morning and a chimney which shades another one partly. I added a screenshot of the MPPT input voltage/current curve. At some point the current drops significantly and the voltage goes up to the VOC. The power drops from about 1150 watts to about 100 watts. The other MPPT inputs keep working.

I think it's a strange behaviour. If modules are shaded partly I would expect a power reduction. But a loss of almost everything while the tracker stays at the VOC afterwards until the sun moved away from the east side of the roof looks like that there is a problem somewhere (MPPT tracker software or a diodes issue on the solar modules?).

What's going on here? Can anybody explain why the MPPT is behaving like this?

mppt_rs450_shadow_management.png

MPPT Controllers rs450

mppt-rs450-shadow-management.png (23.3 KiB)

2 |3000

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

Guy Stewart (Victron Community Manager) answered · Nov 08, 2023 at 09:07 AM

Hi @hotel-lima, @nesswill, @Broesel, @Madlight618, @admiralbulli, @beat, @Mo75, @Semlohnhoj

We have had some discussions about the best way to proceed here in light of all your feedback, and have decided to change our approach;

Please see here for the next step of beta testing - https://community.victronenergy.com/questions/241677/mppt-rs-shade-performance-testing-v116-beta.html

I would like to move any further discussion over to there please.

1

2 |3000

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

iammotorhomeless commented · Nov 08, 2023 at 09:23 AM

@Guy Stewart (Victron Community Manager) thinking I am a bit late...

I was just about to say I installed the update last week, and have not seen any real change

see you on the new thread

1 ·

118 People are following this question.

Modifications Space

Node Red Space

Answer 1 · 2023-09-04T13:16:47Z

seb71 answered · Sep 04, 2023 at 01:16 PM

You might be better of if you remove those modules (if you can't move them elsewhere). Especially the one you say is shaded more than half.

14

2 |3000

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

hotel-lima commented · Sep 04, 2023 at 01:37 PM

It is not shaded all the time. Nevertheless it should not pull down the whole string. Currently for me it looks like as if the MPPT does not try to find the correct power point. But why?

0 ·

Michelle Konzack hotel-lima commented · Sep 05, 2023 at 06:16 AM

If a SINGLE module is shaded or partly shaded, it pull down the WHOLE string and if the sun is shining trough the branches of a tree, then the shading can change so fast, that the tracker can not cope with it.

2 ·

pyrorider Michelle Konzack commented · Sep 28, 2023 at 04:00 PM

Thats the reason why you got up to 3 diodes on your module, to let the current "skip" the shaded part when the voltage of the diode is reached and not pull down the entire string

1 ·

hotel-lima Michelle Konzack commented · Sep 05, 2023 at 07:26 AM

From my understanding it is like this that a shaded module reduces the voltage of the string but due to the diodes in the module it should not block the other modules in the string.

0 ·

seb71 hotel-lima commented · Sep 05, 2023 at 07:34 AM

The important drop is in the current.

And in a series configuratiion, this means a current reduction for the entire string.

1 ·

nesswill seb71 commented · Sep 05, 2023 at 07:51 AM

I can confirm what being said above to be completed correct. I have x3 RS450/100 and have seen 2k form my panels on one string go to 200 to 300w because of a small bit of shade on just one panel.

1 ·

semlohnhoj nesswill commented · Sep 05, 2023 at 08:45 AM

Likewise, I have a Multi RS Solar that I used with 20 small 170w panels vertical on a rear wall at ground level with all sorts of varying shading issues at different times of the day, so worst case.

Even small amounts of shading on one panel (as long as it would affect all the diodes in the panel) would bring the whole string down.

Optimisers weren't cost effective with such a small set of panels, they are also NOT recommended by victron for use with their MPPTs so I switched to 5 separate victron 100/20 MPPTs with 4 panels each to split the problem up. Not recommending the approach!

My personal opinion with this having had plenty of experience of watching the effects of varying shadeis is that the diodes don't provide a reliable bypass to allow the MPPT to do it's work finding the correct power point in many conditions. This of course can be made worse by a particular type of panel and specific conditions.

You might want to work out if ditching the panels gives you a larger total annual output?

1 ·

herculess seb71 commented · Sep 05, 2023 at 07:49 AM

all modern modules has at minimum 3 bypass diodes included, which bypasses the current of the string - if only one module is shaded, the whole string should only reduce by the power of that panel. the mppt has to find the global mpp to "activate" the diodes - i solved that by adding optimizers to my shaded panels, but as far as i can tell rs450 isn't the best in shadow management.

Differences (which should also work without optimizers):

Without Optimizers:

Few Days later with Optimizers:

I use rs450/200 - which doesn't handle that very well, 2 strings (yellow and green) are affected by shading.

0 ·

shaded-wo-optimizer.png (153.0 KiB)

shaded-with-optimizer.png (184.2 KiB)

hotel-lima herculess commented · Sep 05, 2023 at 07:59 AM

I was also thinking about adding an optimizer but was not sure if it will really work. Which optimizers are you using?

Question to Victron would be, if the issues with the shadow management of the RS450 are known and if they plan to update the software in this regards?

1 ·

herculess hotel-lima commented · Sep 05, 2023 at 08:02 AM

is use tigo ts4-a-o as standalone solution, without any further tigo equipment and only on the shaded panels.

1 ·

hotel-lima herculess commented · Sep 05, 2023 at 08:35 AM

Thanks for the tip.

0 ·

Michelle Konzack herculess commented · Sep 05, 2023 at 10:56 AM

But with 60€ including shipping in Estonia quiet expensive.

0 ·

Michelle Konzack hotel-lima commented · Sep 05, 2023 at 10:49 AM

This was my understanding too for the last 30 years.

But it seems, the diodes are installed in a way, that ifvyou cover one section in the solarpanel, the diodes are working, but if you cover the whole panel, the whole solarstrinh goes down.

I can proove this on my own solarpanels with

1) SmartSolar 150/70 and 3s2p (310Wp modules)

2) SmartSolar 250/100 and 5s (410Wp modules)

3) SmartSolar 250/100 10s3p (120Wp modules)

0 ·

semlohnhoj Michelle Konzack commented · Sep 11, 2023 at 07:19 AM

Hence why people use the Tigo units but unfortunately your not supposed to use those with Victron systems for some reason because it messes with the MPPT algorithm?

0 ·

Answer 2 · 2023-09-05T12:02:18Z

hotel-lima answered · Sep 05, 2023 at 12:02 PM

@Guy Stewart (Victron Community Manager)

Is it known at Victron that the shadow management of the MPPT RS 450 does not behave very well on partly shaded modules and are there plans to provide a firmware update that improves the performance in this case?

Thanks in advance for your feedback.

2 |3000

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

Answer 3 · 2023-09-10T21:43:22Z

beat answered · Sep 10, 2023 at 09:43 PM

Today, I had two very strange situations with MPPT of RS 450/100.

MPPT RS 450/100 (firmware up to date) with 2 strings of 10 panels each, in series, same orientation, same sunshine. Modern 415 Watt panels with 3 by-pass diodes.

Morning with a 20% shadow on one panel of MPPT2:

Switched the MPPT RS 450/100 on after the rest of the system was on (CerboGX, Multiplus II 5000, 4x Pylontech US3000C batteries charged 30%).

MPPT1 was working fine at around 1000 Watts

MPPT2 was quickly oscillating between 200-250 and 0 Watt, and the MPPT RS 450/100 was making a discrete fast clicking noise at same speed then the power changes, for a long time (several minutes), before settling slightly above only 200 Watts.

It then finally settled way later at:

MPPT 1 without shadows: 1500 Watts

MPPT 2 with only 1 panel 20% shadowed (shadow of an open roof window): only 1000 Watts (one would have expected 90% of MPPT1 at least).

Middle of morning without shadows, both were producing fine at MPPT 450/100 peak power, each string at 2500 Watts, maxing the 5000 Watts output of 100 Amps at 50 volts.

Mid-day, when 2 panels were shadowed and a third one 10% on MPPT2, and no panels got shadowed on MPPT1, I had:

MPPT 1 still producing 2500 Watts

MPPT 2 producing 300 Watts only

and total was only 2800 Watts.

I was expecting MPPT 2 being at at >70% of MPPT 1, and MPPT 1 increasing power production (maximum in datasheet is 4000W/MPPT).

Does this look like a bug ?

My RS450/100 production date is 2222.

Is this a known problem ? is it software or hardware ? do this years model solve it ?

I really need to have this working to specs (this larger install has a second MPPT 450/100 that will be wired shortly, and there are shadows, and panels mounting wiring is done, so replacing MPTT RS 450/100 by many smaller ones is not an option, and so isn't an option to add TIGO optimizers (not really needed with correct MPPT, not to mention that Victron is not compatible with those nor with others).

Here the daily graph from VRM (at end of day, battery was charged, so the need was little except for a small consumption peak and that's why the curve falls steeply):

Any help really appreciated!

And a reply from Victron too ;-) @Guy Stewart (Victron Community Manager)

screenshot-from-2023-09-10-23-29-42.png (156.8 KiB)

21

2 |3000

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

seb71 commented · Sep 11, 2023 at 08:34 AM

replacing MPTT RS 450/100 by many smaller ones is not an option

Using many smaller MPPTs only works because you remove the shaded PV panel from the series connection (you remove the shaded PV panel from the string).

And the MPPT/PV panel with shade will barely produce power when the PV is shaded.

0 ·

beat seb71 commented · Sep 11, 2023 at 09:11 AM

Maybe the smaller and more mature Smart MPPT 100/20 and such have a better hardware/algorithm ? I strongly hope so, because in my design that I'm wiring, I have small roof orientations with 2 or 3 panels going to 3 Smart MPPT 100/20 and to one Smart MPPT 150/30, and those will have partial shades in winter.

MPPT means Maximum Power Point Tracker. That's quite clear, and Victron's position seems to be "no optimizers needed with MPPT", so that means their MPPT works!

Anyone with smaller Victron Smart MPPT charges having bad MPPT issues with little shadings ? (except that the MPPT scans seems to be done every 30 minutes instead of every 5 or 10 minutes with AC inverters, according to another thread).

0 ·

seb71 beat commented · Sep 11, 2023 at 09:20 AM

That's quite clear, and Victron's position seems to be "no optimizers needed with MPPT", so that means their MPPT works!

I think it's actually "optimizers are not compatible with Victron MPPTs".

You risk destroying the MPPTs.

0 ·

beat seb71 commented · Sep 11, 2023 at 09:34 AM

Exactly, Tigo told me they destroyed a Victron MPPT by trying a compatibility test on their own.

But the gain of using optimizers vs string MPPT, even if noticeable, is not huge, even according to a technical note from the optimizers-king:

https://knowledge-center.solaredge.com/sites/kc/files/se_technical_bypass_diode_effect_in_shading.pdf

0 ·

herculess beat commented · Sep 19, 2023 at 02:37 PM

If you look to my pictures above, there's a huge difference with/without optimizers - but sure, there shouldn't be because of the functionality of the bypass diodes. i hope victron finds a solution for that issue.

0 ·

seb71 beat commented · Sep 11, 2023 at 09:27 AM

Maybe the smaller and more mature Smart MPPT 100/20

The "Smart" you emphasized in bold does not mean what you think it means.

For Victron devices, "Smart" merely denotes built-in Bluetooth.

Unfortunately they already used "Blue" as a prefix in their product names before Bluetooth was used by them, so another prefix was needed to differentiate between mostly identical products, but one with and one without built-in Bluetooth.

0 ·

beat seb71 commented · Sep 11, 2023 at 08:50 PM

Indeed (and I was aware of this), it's extremely frustrating (to say the least) that the MPPT function of the SmartSolar MPPT RS 450/100 doesn't perform as expected, like a typical MPPT, in finding the global maximum power point. My primary question regarding a bug fix for this significant MPPT issue with the MPPT RS 450/100 remains unanswered. Additionally, my secondary question remains open: Do smaller and older SmartSolar MPPT chargers like 100/20 or 150/35 also suffer from the same issue of not adjusting correctly in the presence of partial shades on arrays of panels with bypass diodes?

0 ·

seb71 beat commented · Sep 12, 2023 at 06:24 AM

Small MPPTs can have only 2-3 PV panels in a string (I am talking about "regular" 60-cell or 72-cell PV panels, not very small PV panels). And usually you have 2-3 strings in parallel on such MPPTs.

If you have a few PV panels which are shaded in a part of the day, and you insist on using such areas, at least you should not mix them in a string with PV panels which never get shaded.

If you have areas which are always in shade throughout the day, never place any PV panel there.

1 ·

beat seb71 commented · Sep 12, 2023 at 08:14 PM

Thank you for your advices regarding shading. However in inhabited urban and suburban areas with trees and neighbourgs, shading during parts of days or some months of years, is a fact of panels life, and most MPPT inverters handle temporary shadings well.

However, instead of speaking of shading avoidance, I would like to keep this thread focussed about this major bug in the MPPT trackers of SmartSolar RS 450/100.

2 ·

nesswill beat commented · Sep 12, 2023 at 06:55 AM

I had a 150/70 on 6 pv's in 3x2 (only option for the MPPT because of the pv voltage) and then moved them to a 450/100 in 3x2 then tried x6 in series to see what worked best for their location, there was some shading on them moving across from right to left, the 3x2 looked like it worked best because of the shading (when 1 set was in full sun) on the 450/100 but because of the higher voltage with x6 in series this worked out better (on days with not so much sun but not by a lot overall) because of the long days. (shading only lasted for about 2h in the summer).

With both MPPT in 3x2 i did not notice any real difference, not say there wasn't any but i just did not notice it.

I suspect though i might move back to 3x2 for the winter (but this will be their first year so might not get any shade in the winter because of where the sun comes up), but time will tell.

NOTE - I have x4 other arrays so i am not dependent on the above x6 panels and this might be why i did not notice too much.

Dave.

0 ·

beat nesswill commented · Sep 12, 2023 at 08:21 PM

Thank you very much for your reply. I didn't notice the issue immediately either, as there was enough power left, except for that morning where the customer needed full power. But when we start to export/sell PV excess, and also in winter months when there is less energy for self-consumption (and more shading due to lower sun), this will become critical, and I hope that the issue will be solved soon.

To notice it, if interested, you can go to the VRM portal, and in advanced mode, plot the voltage, amperage and power output of the trackers with partial shading.

1 ·

nesswill beat commented · Sep 13, 2023 at 04:42 AM

To notice it, if interested, you can go to the VRM portal, and in advanced mode, plot the voltage, amperage and power output of the trackers with partial shading.

Thank you and good luck.

Dave.

0 ·

seb71 commented · Sep 10, 2023 at 10:29 PM

The solution is simple: do not mount PV panels in shaded areas.

-1 ·

hotel-lima seb71 commented · Sep 11, 2023 at 07:55 AM

@Seb71 It is rather a workaround than a solution. The real solution would be a firmware update with an improved handling in case of shaded PV panels.

1 ·

seb71 hotel-lima commented · Sep 11, 2023 at 08:08 AM

PV panels need Sun.

You can't workaround that. No firmware update will remove the shadows.

Basic rule #1:

If there is no Sun, do not place PV panels there.

0 ·

hotel-lima seb71 commented · Sep 11, 2023 at 08:23 AM

You will always have to deal with shadows. As the sun moves throughout the day, shadows will move over your PV field (chimneys, dormers, etc). It is no solution to say that you should not install PV panels on an area where you have partly shadows during the day (we're not talking about areas that are totally shaded the whole day).

The firmware update cannot remove the shadows but it can set the working point in a way that the bypass diodes of shaded panels open and that the nonshaded panels continue to work.

0 ·

seb71 hotel-lima commented · Sep 11, 2023 at 08:31 AM

The MPPT firmware can't control PV diodes.

1 ·

beat seb71 commented · Sep 11, 2023 at 09:30 AM

MPPT is exactly made to make good use of by-pass diodes by doing Maximum Power Point Tracking, this means that there is a regular scan (or even triggered by sudden changes in power at the working point), to find the best working voltage giving the most power.

1 ·

hotel-lima beat commented · Sep 11, 2023 at 09:36 AM

Exactly. And with shaded modules in a string you will get more than one maximum: Local Maximum Power Point (LMPP) and Global Maximum Power Point (GMPP). The job of a good tracker is to find the GMPP and not to operate on the LMMP.

2 ·

hotel-lima seb71 commented · Sep 11, 2023 at 08:46 AM

Currently, I doubt that.

The MPPT can control the current. In my case I see that the voltage goes up to VOC and the current goes down. It even does not try to find a working point where it can gain more power.

0 ·

semlohnhoj seb71 commented · Sep 11, 2023 at 07:21 AM

I'm justifying that comment with a response purely for the irony :)

0 ·

Answer 4 · 2023-09-12T20:31:37Z

semlohnhoj answered · Sep 12, 2023 at 08:31 PM

I hate to say this here but I have NodeRed code on my Cerbo that I wrote because my MPPTs randomly go to zero (probably due to intermittent shading) and don't recover for anything up to hours for no apparent reason and if I reset them they start generating power again. I've never got to the bottom of it and just live with it now. If one MPPT does it and not the others I assume it's having a mental moment and I reset it.

4

2 |3000

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

seb71 commented · Sep 13, 2023 at 04:44 AM

If reseting the MPPT fixes the issue, I would say it is a different situation.

This is about power drop when shadow is present, not after shadow was gone.

0 ·

semlohnhoj seb71 commented · Sep 13, 2023 at 07:21 AM

I agree but I initially thought it was just the shading until I checked the panels and realised that on some occasions this was still happening after the shading had passed and would then continue for random periods. I'm just making the point that everything isn't always as simple as it seems. I'm also with you on the point that no amount of firmware updating will resolve an issue with diodes blocking due to shading assuming the software is already correct.

0 ·

beat semlohnhoj commented · Sep 13, 2023 at 07:31 AM

Thank you for your return of experience and extreme idea to reset, better than nothing, but still a hint that there are major bugs in the MPPT algorithm of Victron.

Other brands' (AC) inverters' MPPT do find the global maximum power point, and scan every 5 or 10 minutes to be sure to still be at maximum. And shading, handled by a working MPPT is not an issue. If out of 10 pannels, 2 are shaded, you get 80% of the power, that's quite simple, and perfectly ok. The issue here is that when 1 out of 10 panels is shaded, Victron SmartSolar RS 450/100 gives 66% (1000W/1500W) instead of 90%, and with 2 shaded, gives 12% (300W/2500W) instead of 80% like it should.

I would really appreciate a reply from Victron on this, @Guy Stewart (Victron Community Manager) ;-)

-1 ·

kimmo-paukku beat commented · Sep 13, 2023 at 08:17 AM

You can think of a string of panels as something like a piece of pipe, and the solar power is like water flowing through that pipe. In conventional solar panel strings, shade is something that blocks that flow. If, for example, shade from a tree or a chimney is cast on even one of the panels in the string, the output of the entire string will be reduced to virtually zero for as long as the shadow sits there. If there is a separate, unshaded string, however, this string will continue to produce power as per usual.

1 ·

mppt-solar-pv-system.jpg (30.9 KiB)

Answer 5 · 2023-09-13T07:30:00Z

semlohnhoj answered · Sep 13, 2023 at 07:30 AM

Just a thought, when the panel(s) are shaded and hence the diodes are essentially removing those panels from the string, is the series voltage of the remaining panels high enough for the RS450 to operate?

If this happened though I'm pretty sure it would show up as in 'Off' rather than 'bulk' mode.

If you've got the details of the panels you could check if the system is in spec with the reduced quantity of panels using the Victron MPPT calculator to give you an idea:

https://www.victronenergy.com/mppt-calculator

2 |3000

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

Answer 6 · 2023-09-13T12:02:34Z

Guy Stewart (Victron Community Manager) answered · Sep 13, 2023 at 12:02 PM

Hi,

There is a setting in VictronConnect to enable or disable the RS MPPT algorithm for shaded panels.

Does enabling, or disabling this setting any effect?

cleanshot-2023-09-13-at-220147.jpg (99.6 KiB)

3

2 |3000

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

deonkorb commented · Sep 13, 2023 at 01:01 PM

This setting had no effect (positive or negative) on my installation.

1 ·

hotel-lima commented · Sep 13, 2023 at 01:09 PM

This setting has no effect on my side either.

1 ·

beat commented · Sep 14, 2023 at 11:50 AM

No effect either, here the result with "Partial shade detection" OFF with:

MPPT1: 10 panels (Trina Vertex S TSM-415NEG9.28), no shade

MPPT2: 10 panels (same panels, same orientation), 2.5 ones temporarily shaded:

And, sometimes, the MPPT works better, like one would expect it, the working voltage has been dropping and around 75% of the power is harvested, here with "Partial shade detection" ON and less sun (partly cloudy varying today):

I have the feeling that some things trigger the MPPT not working right, e.g. when maximum power of around 5000 Watts is reached, but at other times too.

0 ·

screenshot-from-2023-09-14-13-30-41.png (32.2 KiB)

screenshot-from-2023-09-14-13-47-48.png (36.1 KiB)

Answer 7 · 2023-09-13T13:55:16Z

nesswill answered · Sep 13, 2023 at 01:55 PM

Out of interest are your panels split cell or full cell perc panels

2 |3000

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

Answer 8 · 2023-09-14T08:30:41Z

Guy Stewart (Victron Community Manager) answered · Sep 14, 2023 at 08:30 AM

Hi All,

Thanks for all your info and reports here.

I've just learned myself that there was a new shade detection algorithm introduced with the RS series compared to the previous MPPT range.

The RS developers are now aware of this thread and looking to see how things can be improved.

22

2 |3000

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

Guy Stewart (Victron Community Manager) ♦♦ commented · Sep 28, 2023 at 04:53 AM

A quick update, the developers have made some changes that they believe should address the issues raised here. They are now in internal testing, once complete they will be publicly tested in an upcoming VictronConnect Beta version.

7 ·

semlohnhoj Guy Stewart (Victron Community Manager) ♦♦ commented · Sep 28, 2023 at 07:24 AM

Would this apply to the Multi RS Solar as well?

I'd be interested in the improvement if it did. I'd be very happy to use the tracker on the RS rather than several separate MPPTs to break up the string like I'm doing at present if it handled shading better.

2 ·

Guy Stewart (Victron Community Manager) ♦♦ semlohnhoj commented · Oct 18, 2023 at 02:49 AM

Yes, it will also apply to other RS solar products.

1 ·

beat commented · Oct 09, 2023 at 02:31 PM

Hi @Guy Stewart (Victron Community Manager) , any news from the bughunt ? :-)

0 ·

Guy Stewart (Victron Community Manager) ♦♦ beat commented · Oct 10, 2023 at 06:06 AM

Hi @beat,

No bugs yet afaik from the RS team.

It looks like the change will also add an additional setting to force (or disable) the periodic sweep, instead of depending on the detection of shading.

It's now being worked on by the VictronConnect team to include in the UI. VictronConnect has lots of moving parts and a new product (Victron Energy Meter) and may need to wait for inclusion as part of the development cycle.

0 ·

broesel Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 17, 2023 at 04:31 AM

Hi,

Great News that an improvement is upcomming.

Andy Chance to get the Beta Firmware File to Test it while waiting for the VictronConnect Team?

Best regards

Daniel

1 ·

mo75 Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 10, 2023 at 08:02 AM

I can't imagine how shadow detection works at all without a sweep.

Hopefully the sweep is implemented better/faster than with a 150/45 that takes 45 sec every 600 sec to perform 3 (!) sweeps in a row. This is 7.5% of the time.

0 ·

Guy Stewart (Victron Community Manager) ♦♦ mo75 commented · Oct 10, 2023 at 10:34 AM

The new algorithm tried to detect shading by a looking for changes in the voltage or current output of the string, and then used that as a trigger for the sweep.

This algorithm method is being improved; and if successful will be the most efficient method.

However we will also be re-introducing the timer based sweep method as a custom configuration option for when the algorithm isn’t getting triggered for some reason. So it will be at least as efficient as the rest of the range.

If you are certain there is no shading on your array, it will also be possible to disable the sweep entirely.

I think that gives the best range of options possible.

None of that is finalised either yet, so subject to change before release, but appreciate people in this thread want to know we are on the case.

3 ·

beat Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 13, 2023 at 10:52 AM

Thank you for the update @Guy Stewart (Victron Community Manager) !

The worked-on solution seems fine to me if the algorithm becomes "smarter", and if sweeps are very fast (talking about miliseconds or a couple of seconds maximum) like on Fronius or Huawei, which if I saw it right do a first very low-resolution sweep based on number of panels, then a second one based on diodes per pannel (typically 3), and then a third around the 1-2 local maximums if needed.

If you need an alpha/beta tester (or remote or VRM access to the installation), please PM or reply.

For this larger installation (75x415W panels, 12 strings, 5 orientations) which is still in testing, I will have to decide what to do in the next weeks before going into full production, as the two RS 450/100 are barely usable, especially in winter months, loosing around 80% of the solar energy. I love Victron equipments' unmatched flexibility, open-source mind (at least for VenusOS, btw if the RS 450/100 algorithm was open-source, I would already have done a review and probably made a suggestion with a PR) and community-openess, so would really love to be able to make it work instead of having to return the Victron equipment to the distributor.

Btw if you would like to consider my proposal for testing on site, I'm, among other things, a senior electrical systems and system software engineer, well tooled, and also Victron reseller and installer, with completed successfully all relevant Victron online trainings (your own ones being particularly well done and well packed with information (proof: no way to view them at more than 1x speed lol)), so I don't really need fancy configuration interfaces to try things out. I've installer and root access to the Cerbo GX turned on, and have ofc command-line tools if needed ;-).

3 ·

mo75 Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 17, 2023 at 09:27 AM

Dear Victron people, sorry for this rudeness, but you really should come around the corner with at least a beta version to fix this slowly embarrassing situation. Or may I send you an invoice for lost revenue?

1 ·

beat mo75 commented · Oct 17, 2023 at 10:22 AM

@Guy Stewart (Victron Community Manager) sorry to ping you, but Urgent Action is needed here too: Please provide us with a download link for a alpha/beta build with new algorithm and/or periodic full scan simply turned on!

This "MPPT" feature is currently a joke to stay polite: 450/100 are idling most of the time at Voc with 100-200W instead of 2000-3000W at GMPP.

Our customer is very unhappy not only to not get (almost) anything back from his heavy investment, but to have to pay real money on heavy grid cost after a heavy investment in plenty enough solar panels and well-dimensioned and duly installed Victron gear.

Thank you very much for your understanding, hope to hear back real soon.

2 ·

Guy Stewart (Victron Community Manager) ♦♦ beat commented · Oct 18, 2023 at 10:36 AM

Hi @beat, @Mo75 & @Broesel,

We aren't ready for a public beta test yet, but I can share a private beta version with you, as you've all expressed interest in trying it out.

It is still not yet feature complete because it depends on a VictronConnect update to enable the periodic sweep mode.

The improved detection sweep algorithm is included now though, and perhaps that is enough.

I will send you each an email with the file.

3 ·

beat Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 18, 2023 at 03:18 PM

Thank you very much @Guy Stewart (Victron Community Manager) for making this happen!

I installed it here fine too, and replied to your email with first observations of the PV voltage curves (cloudy here now). Now waiting for sunshine for shades.

2 ·

herculess beat commented · Nov 02, 2023 at 12:23 PM

is there already an outcome, is the new algorithm better than the old one or are there no changes?

0 ·

beat herculess commented · Nov 02, 2023 at 03:37 PM

Here, the new beta seems significantly better. We had only very little direct sunshine here over the last week, so it is hard to tell if it's perfect, we did have only a few hours of sun, not even an hour at a time, so it was difficult to compare with a sunny day with old version, but at the same time it's challenging for MPPTs to switch between the GMPP of sunshine to the GMPP of cloudy weather with clouds coming and going every few minutes.

But from what I can tell, it is already significantly better. I'll be sending privately more detailed feedbacks and questions to Victron shortly.

0 ·

mo75 Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 18, 2023 at 10:43 AM

Hi Guy, that sounds really good to me. Thank you for making this happen!

I'm looking forward to testing the new beta and hope for the expected improvements. Unfortunately, there is no good weather in the near future, but what the heck...

Best regards, Mario

1 ·

mo75 Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 18, 2023 at 11:09 AM

That was really quick, thank you very much. The beta has been successfully installed and is working perfectly for the time being. I'm curious to see what the result will be over the next few days, when there will hopefully be enough sun (and partial shade).

1 ·

getacca Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 18, 2023 at 03:25 PM

Dear Guy, could I also receive the beta file by email. I have the same problems. I am happy about the feedback.

1 ·

hotel-lima Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 18, 2023 at 04:28 PM

Hello @Guy Stewart (Victron Community Manager)

I would also like to test the beta firmware. Please also send me the file.

Thanks.

1 ·

leonbed Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 31, 2023 at 12:49 PM

Hi,

Ive just installed a MPPT RS 450/200 a few days ago and its performing much worse than cheap chinese MPPTs.It starts to track the MPPT and then it just loses it and power falls off of a cliff. I have also noticed that, even when only indirect light is available, the voltage still is too high to achieve reasonable power output.

I would really appreciate if I could receive the beta software as well, because I would rather have a device that might not work, than a device that is guaranteed to not work.
Thank you very much
Leon.

Edit: Just received it, I will have to see if it will do what it claims, but it very likely will.

0 ·

mppt-rs-bad-solar-charge-algorithm.png (121.6 KiB)

Guy Stewart (Victron Community Manager) ♦♦ mo75 commented · Oct 18, 2023 at 02:42 AM

I've just asked if a public beta test can be made available, but there are no promises for all the reasons already mentioned above.

Some things like this just take as long as they take for each team in the stack to do their part (with all the other competing changes in the works and testing that is required even to get to public beta).

I am pretty hopeful that shade performance can be improved with these changes, but less so that I can force the process to make the change required happen any faster.

2 ·

mo75 Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 10, 2023 at 11:01 AM

Thanks for this info.

The new algorithm seems to have problems at least when a module array is only released piece by piece from the shadow. So if the shadow goes slowly instead of coming. Unfortunately I can only observe this case at my installation (see house in the picture of the simulation).

It seems that the triggered sweep tends to look up (to higher voltages), with the better power point being below.

It would be good if the option to select the sweep could be selected for each tracker individually, as different shading situations may exist (as in my case).

I hope that the changes will be implemented soon, because a lot of important energy can not be used at this (upcoming) time of year.

0 ·

Answer 9 · 2023-09-14T11:26:32Z

beat answered · Sep 14, 2023 at 11:26 AM

Thank you very much @Guy Stewart (Victron Community Manager) !

Btw, is there a way to get/see the I/V (and while at it P/V) curves of the strings, colledcted during MPPT scans of the chargers ?

Some solar AC inverters provide that possibility, and even to trigger an I/V scan, which is really useful to debug such situations, and to optimize strings. E.g. SMA explained here: https://manuals.sma.de/STPxx-US-41/en-US/8639465483.html SOFAR there: https://www.sofarsolar.com.au/why-is-i-v-curve-scan-function-so-important/

Such curves are really useful to identify loss of PV module/array defect, shading, dust accumulation, temperature rises. They can provide series/parallel adaption loss calculation. They can also help detect modules mismatch, and with some computation even bypass diode faults

And any MPPT charger has it anyway during its MPPT scan....

Moreover, the I/V curve of the strings can give precious indications about the security of the string and provide early detection of contact faults, even before an arc happens in the string (btw, any arc detection in the plans at Victron?).

A I/V + P/V curb visualization, like on page 6 here https://solar.huawei.com/en-GB/download?p=%2F-%2Fmedia%2FSolar%2Fnews%2Fwhitepaper%2FIV-Curve-whitepaper.pdf in VictronConnect would look awesome marketingwise, even for small MPPT chargers, not to speak about RS.

In the mean time, is there any possibility (Modbus-TCP, MQTT, NodeRed) to get today the last I/V measurement points of last MPPT scan of the MPPT RS 450/100 or of a MPPT 100/20 ?

Or, if not, are the NodeRED inputs fast enough to collect the I (amps) and V (volts) string data during MPPT scans ? Or if not, the MQTT ones ?

2 |3000

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

Answer 10 · 2023-09-15T12:29:56Z

beat answered · Sep 15, 2023 at 12:29 PM

Some further really strange behavior here: I made a small NodeRED flow to plot the MPPT voltage and power of MPPT2 over time on the Cerbo GX, and I saw following: Total power went up to 5000W, and then dropped around 4200W for around 5 minutes with no reasons, while there was sunshine without shadows on both strings, and the 48V battery's charge limit was 148 amps, the MPPT started doing a strange adjustment (while being in MPPT operation mode 1 as PV solar power available maxed the charger):

If that's the MPPT algorithm, then there is something really strange or buggy happening.

screenshot-from-2023-09-15-14-23-28.png (32.7 KiB)

2 |3000

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

Answer 11 · 2023-09-28T06:55:22Z

mo75 answered · Sep 28, 2023 at 06:55 AM

Hello all,

I have exactly the same problem with the RS450/100.

MPPT1 is on 9 modules on a dormer with south orientation, about 20 ° inclination and only in the morning some shade by a chimney. This tracker makes little problems.

MPPT2 is on 7 modules, which are located west of the dormer on the roof, also oriented exactly to the south with about 40 ° inclination. The modules are very close to the dormer and thus they get full sun only after noon.

Most of the time at MPPT2 the power over a few hours in the morning is much lower than the expected power, if the number of modules exposed to sunlight is taken into account. It seems that the selected MPP voltage is much too high. This is because it is at the level of the open circuit voltage of the complete string for a very long time. Then, at some point, the moment comes when the voltage is lowered and the power jumps up many times. However, this does not mean that this is the global MPP, the voltage is still too high. Often, some time later, a much higher voltage is simply selected again and the power collapses again. This goes back and forth a few times over the morning until all modules are in the sun from noon.

Sometime a restart of the charger help to get a better MPP.

The shadow management of the RS450 is simply a disaster.

I have already opened a ticket at Victron, but was then only referred to my dealer. Since I can not do without the device at the moment, I hope very much for a firmware update.

Here a screenshot of the both trackers:

And here the corresponding Voltage/Power diagram for Tracker 2:

This is the mounting situation:

unbenannt.png (39.5 KiB)

unbenannt2.png (64.1 KiB)

bild-2023-08-12-135324155.png (1.3 MiB)

2

2 |3000

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

beat commented · Sep 28, 2023 at 07:45 AM

Yup, very same as the other reported issues and mine too. Glad that Victron devs are handling it quickly now (saw reply this morning), as this MPPT is otherwise unusable (and turning "shadow management" on or off doesn't change the results fundamentally, prooving that the algorithm is currently buggy).

1 ·

Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 18, 2023 at 03:05 AM

Hi @Mo75

We have an almost uncanny similarity in our roof alignments :)

0 ·

cleanshot-2023-10-18-at-140416.jpg (3.1 MiB)

Answer 12 · 2023-09-28T07:16:48Z

hotel-lima answered · Sep 28, 2023 at 07:16 AM

This looks exactly the same as on my side.
Guy from Victron posted today that they have made some changes to the firmware and that they're planning to release a beta version soon. I'm confident that this will make things better.

Here is the link to Guy's comment in case you missed it

https://community.victronenergy.com/comments/233847/view.html

1

2 |3000

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

mo75 commented · Sep 28, 2023 at 07:18 AM

Thanks!

Yea, I read it.

Looking forward for better things to come :-)

1 ·

Answer 13 · 2023-09-28T15:05:19Z

janvi answered · Sep 28, 2023 at 03:05 PM

Can confirm this for my RS450-200 with a partly shaded string. Power goes down frequently to only 150 Watts with most modules illuminated. Up on now, I had no time to examine the problem in detail. I plan to install a camera to the roof what allows assignment of power drops to shaded areas.

Already recognized another marvelous thing. My modules have 400 Watt, Voc 37 Volt and Vmpp 31 Volt with 3/6 bypass diodes. With 7 modules in series I had about 228Volt typical operation voltage. Assuming the voltage is too low for activating bypass diodes, I added 3 more modules of same model in series. Typical mppt operation voltage went up to about 250 Volt. With a total of 12 modules the typical mppt voltage is only 280 Volt. With other words: The increase of voltage does not show my expected linearity with the number of modules why I assume the mppt found is a resaonable one for typical situations but but not a optimum one in many non typical situations.

5

2 |3000

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

janvi commented · Oct 11, 2023 at 02:22 PM

After examination of my modules I have probably amend my complaints above.

There is a really unfortunate shadow of 5,5 modules while 3,5 modules are illuminated well. This results in a string voltage of 270 Volt with 0,6 Amp @160 Watt while neighbour strings with same direction has about 5 Amp at the same time. Assume activation of bypass diodes for the shadow modules, voltage from 3,5 modules would probably drop slightly below the minimum activation threshold of 120 Volt. Therefore the operation point of 3,5 Modules * 5 Amp = 31V*3,5 * 5 Amp =540 Watt is rejected.

Either I have to try adding a 10th module to the illuminated group what lifts voltage above 120Volt activation threshold or playing around with external optimizers to allow 5 Amp through the shadow modules.

For this situation, it would be helpful to publish what the MPPT algorithm is doing exactly if switched on. In contrary therefore, the RS450 manual only mentions no use of external otimizers while the optimizer manufacturers claim it will work with any MPPT.

Please, no discussion if external optimizers are economic - they are probably not. I simply wants to understand where the power diffrence comes from.

0 ·

leonbed janvi commented · Oct 31, 2023 at 01:02 PM

The algorithm of the MPPT RS just sucks right now, and they do state 80-450V MPPT tracking, so once the voltage has reached 120v once it should even track below that.

It does that on my unit, but only for a while, then it just lets the voltage rise way too much until sometime later on it decides to lower the voltage again, but by then much of the solar energy was already lost.

I really hope they can provide a fixed software soon (beta or not, I dont care, as long as it works better than it does now), because I didnt spend good money to upgrade my charge controllers only to receive much less energy than before.

0 ·

Guy Stewart (Victron Community Manager) ♦♦ leonbed commented · Oct 31, 2023 at 01:10 PM

Hi @leonbed,

We have a private beta going on right now with an improved algorithm. I will send you an email to try it if you like. Otherwise you can wait until the next VictronConnect release.

1 ·

leonbed Guy Stewart (Victron Community Manager) ♦♦ commented · Oct 31, 2023 at 01:22 PM

Thanks a lot, its already installing.

It would be less work for you if you would just release the beta with tons of disclaimers, but this works too.
I hope the charging algorithm will only get better, so I will be able to recommend victron to other people.

0 ·

Dierk Grossfeld Guy Stewart (Victron Community Manager) ♦♦ commented · Nov 02, 2023 at 11:32 AM

Hey @Guy Stewart (Victron Community Manager) ,

I can also try it if there is some more actual input needed.
Having two MPPT RS 450/200 with 7 partially shaded strings attached.
All Strings contain the same modules and string sizes.

Greetings,

Dierk

0 ·

Answer 14 · 2023-09-28T20:45:46Z

beat answered · Sep 28, 2023 at 08:45 PM

@Guy Stewart (Victron Community Manager) Thanks for the update! Looking forward! While at it, partial quick scans could be made more often and complete scans every 5 or 10 minutes ?, especially been re-triggered when there is a change in only one of the 2 (or 4 with the 450/200 models, or even in the networked chargers, with e.g. 2 MPPT 450/100 and soon 3 100/20 and 1 150/35). This would typically mean that there is a moving shading, and not clouds. Ideally, the regular scan interval and the sensitivity to changes could be an advanced setting per string ?

Fronius for instance, to my knowledge, does a quick scan every 5 minutes, and more often in cases where the U/I behaviour changes faster than a rate.

Also trying to keep the current constant, changing voltage seems like a better strategy to follow the MPPT locked-on all the time than keeping the voltage constant. See video below at around 7 minutes in.

Diodes in solar panels work wonders, but only with fast and effficient MPPTs with shadow management.

This very well done video with experiment and theory, and simulations, convinced me to not put optimizers (and another brand!):

https://www.youtube.com/watch?v=E9NVLB_OYSU

it's a well worth watch until the end (and this guy has many other nice videos):

1

2 |3000

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

mo75 commented · Sep 29, 2023 at 07:59 AM

It's the old discussion use vs. use not optimizers.

A very eye opening and professional scientific view is in this (german language) video:

https://www.youtube.com/watch?v=yKz-zbxijFU

He also has some english videos an his channel.

This was for me the reason not to buy optimizers and go just with a modern String inverter. Sadly, the RS450 could not hold his promise until now. I hope that good things will come.

1 ·

Answer 15 · 2023-10-16T19:56:57Z

admiralbulli answered · Oct 16, 2023 at 07:56 PM

I have very similar problems with my 450/100. A little shadow from a tree on my 2x 7 string on tracker 2 and the power drops massively. As the light intensity reduces due to the afternoon sun lowering, obviously the MPPT decides to scan in lower voltages. Once the voltage is low enough power raises by a factor between 4 and 5.

@Broesel mentioned this here, too:

https://community.victronenergy.com/questions/233050/mppt-450200-poor-shadow-management.html

2 |3000

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

Answer 16 · 2023-11-02T22:14:46Z

Guy Stewart (Victron Community Manager) answered · Nov 02, 2023 at 10:14 PM

Hi All,

We have just released a new RS firmware version with improvements to the MPPT shade detection algorithm.

You can do a manual update with the attached files.

This firmware version will also get rolled out and released in the near future via a VictronConnect update, and the VRM menu.

If you've had issues with solar underperformance due to shading (as described in this thread) please perform this update and report back.

Thanks to everyone who reported this shade detection issue, and your patience in waiting for us to improve it. Also special thanks to those who reached out to participate in the private beta.

Multi_RS_48-6000-100-Solar_A443_v01.14.xup

Multi_RS_48-6000-100-Solar_A442_v01.14.xup

MPPT_RS_450-200A_A111_v01.14.xup

MPPT_RS_450-100A_A110_v01.14.xup

Inverter_RS_48-6000-Smart-Solar_A401_v01.14.xup

multi-rs-48-6000-100-solar-a443-v0114.xup (644.2 KiB)

multi-rs-48-6000-100-solar-a442-v0114.xup (638.3 KiB)

mppt-rs-450-200a-a111-v0114.xup (392.2 KiB)

mppt-rs-450-100a-a110-v0114.xup (380.4 KiB)

inverter-rs-48-6000-smart-solar-a401-v0114.xup (583.3 KiB)

9

2 |3000

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

beat commented · Nov 02, 2023 at 11:40 PM

Awesome! Thanks a Lot, @Guy Stewart (Victron Community Manager) and Victron Team!

Already updated both MPPT RS 450/100, will see in the next days.

Are there any differences/improvements in the MPPT algorithm compared to the private beta ?

Sun (and thus shadings) was really scarce in the last two weeks, only an hour or less at a time, with clouds passing, so MPPT had a lot to do. It looked a lot better, but was hard to compare with v1.10. I was about to write you privately a short report of my findings with power/voltage graphs. But if algorithm did improve since beta, then there is no point to write the short report, better to re-observe in the next days/weeks with v1.14 stable.

1 ·

Guy Stewart (Victron Community Manager) ♦♦ beat commented · Nov 03, 2023 at 01:40 AM

Hi @beat,

Afaik there has been no change since the 1.14b2 - enough testing has now been done on that release to make it public.

Your feedback would still be welcomed, and has been very helpful so far.

1 ·

Dierk Grossfeld commented · Nov 03, 2023 at 02:07 PM

Thanks a lot also from me @Guy Stewart (Victron Community Manager) !

I upgrade both of my MPPT RS 450/200 this morning with the firmware files provided.
Just a quick report after almost one production day :)

Both MPPT are producing substantially more power that before.
As already mentioned above:
Both are holding strings of the same amount and type of modules, so are directly comparable.

On the first tracker of each MPPT is a string containing 6 trina modules with 425 watts each and those are facing the same orientation and azimuth.

The only difference is the shading.

In the morning one string is stronger since the other one is shaded by extends of the roof.

In the afternoon its the other way around due to shading from the roof of the neighbour building.

Before I had 2000 Watts agains 300 Watts.
Since roughly a third of one string is shaded i would expect two thirds of power on the shaded string.
Now it is like this.

Had 1200 Watts to 1800 Watts a few minutes ago with 1/3 of shading.

Perfect !

Can only say:
Well done !
Thank you very much for the fix !

Dierk

1 ·

nesswill commented · Nov 03, 2023 at 07:58 AM

@Guy Stewart (Victron Community Manager)

Guy this looks good on my 450/100 where you are using both trackers but on one of my 450/100 i only use one tracker, the Pannels are now in full sun and the output is about x20 down on what i would expect, see below...

Dave.

0 ·

just-one.jpg (20.5 KiB)

nesswill nesswill commented · Nov 03, 2023 at 08:01 AM

can i downgrade to 1.11 to test?

Dave.

0 ·

nesswill nesswill commented · Nov 03, 2023 at 08:38 AM

Not going above 36w regardless of sun/shade or clouds.

Dave.

0 ·

nesswill nesswill commented · Nov 03, 2023 at 08:45 AM

Ok fixed it by disabling the charger then turning it back on again.

Dave.

0 ·

screenshot-20231103-084614-chrome.jpg (25.1 KiB)

Guy Stewart (Victron Community Manager) ♦♦ nesswill commented · Nov 05, 2023 at 02:06 PM

This (of course) should not be necessary.

I'll pass on the report to the engineers.

1 ·

beat Guy Stewart (Victron Community Manager) ♦♦ commented · Nov 05, 2023 at 03:53 PM

@Guy Stewart (Victron Community Manager) would it be thinkable to open-source your MPPT chargers codes, like you guys brilliantly did for Venus OS (btw open-sourcing Venus OS was one of the main reasons I chose Victron) ?

My feeling is that your MPPT tracking is not as ripe as the best ones in the industry. So getting community review and help in optimizing it could help you vastely. And there are already several open-source MPPT chargers codes and optimized algorithms on github as well, so it's not really like disclosing stellar algorithms or company secrets, but it could benefit the community and Victron. Just a Sunday thought.

3 ·

Answer 17 · 2023-11-05T07:10:31Z

madlight618 answered · Nov 05, 2023 at 07:10 AM

hallo, since uploaded the new firmware i have to restart the MPPT every morning for working correctly to deliver more than 30 W ...? Anyone with same problem ?

vielen Dank

Michael

screenshot-249.png (167.1 KiB)

7

2 |3000

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

nesswill commented · Nov 05, 2023 at 08:48 AM

@Guy Stewart (Victron Community Manager)

Possible bug in v1.14

Yes me too on my 450/100 i have just one tracker connected on this 450/100 but on my there two 450/100's with two tracker connected there are fine.

It will start ok then go to 35w, i have to stop and restart the charger to get it to work correctly.

1 ·

450-100-single.jpg (24.4 KiB)

madlight618 nesswill commented · Nov 05, 2023 at 12:42 PM

yes, like me, one tracker 8S2P ( 380W Modules)

0 ·

beat commented · Nov 05, 2023 at 03:34 PM

I have both 450/100 with both trackers connected, and haven't noticed that either, but I am just below maximum voltage on each (8-10 PV in series on each) and do not have parallel strings.

But With v1.10 I had to turn switch off and on again sometimes, otherwise it could take up to 30 minutes for the MPPT to react to changes.

Generally speaking, I believe the reactivity to tracking the local MPP could be improved vastly (separately of whole swipes, they could constantly "explore" how the current varies slightly below and slightly above current voltage. And that remark is not only for the MPPT RS, but also for "classic" smaller MPPTs: When a cloud goes away, power gets up (like from 400W to 2000W), but it takes time (15-60 seconds) to get the maximum power (like 3500W) while sunshine irradiation hasn't changed.

0 ·

admiralbulli commented · Nov 05, 2023 at 08:35 PM

Same for me.

450/100 with both trackers connected to 2x7 330W panels each (9.24 kWp total).

On 1.14 since yesterday (updated via VRM, NOT via the provided files above!)

Really bad day today from a radiation point of view, but perfect for comparing the 2 trackers that would be behaving somewhat identical - normally.

First Tracker (T0) shows the "approx 30W" - performance. Second Tracker (T1) is somewhat "active":

@Guy Stewart (Victron Community Manager) : Ready to perform some private beta testing if helpful for you guys.

Device is of HQ2250 series and "partial shading detection" is switched ON.

0 ·

1699216216397.png (362.1 KiB)

Guy Stewart (Victron Community Manager) ♦♦ admiralbulli commented · Nov 05, 2023 at 11:56 PM

Hi @admiralbulli, @beat, @nesswill, @Madlight618

I will pass the feedback on to the RS developers.

I expect this issue will be resolved with the upcoming update to VictronConnect that includes this firmware and enables the periodic scan, and not just relying on the (now improved) detection method.

1 ·

admiralbulli Guy Stewart (Victron Community Manager) ♦♦ commented · Nov 06, 2023 at 06:12 AM

Hi @Guy Stewart (Victron Community Manager) :

"updated via VRM, NOT via the provided files above!"

Just to make sure... Looks like 1.14 IS already rolled out to everybody since a couple of days. You might want to stop this...

0 ·

Guy Stewart (Victron Community Manager) ♦♦ admiralbulli commented · Nov 06, 2023 at 11:20 AM

Hi @admiralbulli,

The rollout of 1.14 is now being replaced with 1.15 that includes the essential fan control changes, but reverts the shade detection for now.

1 ·

Answer 18 · 2023-11-05T10:36:23Z

michi answered · Nov 05, 2023 at 10:36 AM

Hallo i have v1.14 on my 450/100 and one tracker and it worke fine. But now sun todey.

1

2 |3000

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

madlight618 commented · Nov 05, 2023 at 12:37 PM

This is very peculiar, same MPPT, dfferent behavior...

0 ·

Answer 19 · 2023-11-06T08:57:26Z

Guy Stewart (Victron Community Manager) answered · Nov 06, 2023 at 08:57 AM

Hi All who have been watching this issue.

It appears that the despite the changes the automatic shade detection feature in the RS range is still causing issues for some people.

We have another more urgent change that needs to be made to the RS fan control, and this will now be implemented in a new mandatory update v1.15.

This v1.15 release only contains the updated fan and dimming temperature profiles. The partial shading changes from v1.14 are reverted.

We will continue to work on the partial shading solution with another release coming shortly.

7

2 |3000

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

broesel commented · Nov 06, 2023 at 11:53 AM

Ok, thank you for the status.

What if involved in the Beta 1.14?
Can i stay there with MPPT 450/200 or do you recommend updateing to 1.15?
Then i loose the shadow Improvements again... Or will there be a new Beta including Fan and Shadow changes?

0 ·

nesswill commented · Nov 06, 2023 at 12:50 PM

@Guy Stewart (Victron Community Manager)

Hi Guy

Reading the release note on Victron Professional the update is for the MPPT so not relevant for the Inverter RS 48/6000 Smart as there in no MPPT in it? If this is so can it be corrected please.

From the release note...

Changes applying to all RS models with a built-in MPPT:

Dave.

0 ·

Guy Stewart (Victron Community Manager) ♦♦ nesswill commented · Nov 07, 2023 at 07:12 AM

@nesswill
- Thanks for the clarification, This release was made for all units, even the Inverter RS without MPPT. I'll find out about the discrepancy between Vic Pro news and release notes.

1 ·

Guy Stewart (Victron Community Manager) ♦♦ nesswill commented · Nov 08, 2023 at 09:29 AM

Hi @nesswill,

The v1.15 was only made mandatory for the models with the built-in MPPT.

So the Inverter RS without Solar doesn't have to be updated if you don't want to. It will still remain available, just so that the firmware versions are consistent across the range.

Testing is now going on for v1.16 beta here - but for now that release is exclusive to MPPT RS.

MPPT RS shade performance testing, v1.16 beta

Once we are certain that it is performing as expected, then other betas will be available for the Inverter RS Solar, and Multi RS as well.

1 ·

broesel nesswill commented · Nov 06, 2023 at 03:23 PM

Not quite the answer.

Changelog for 1.14 beta 02 said:

Changes:
Fix automatic partial shading algorithm: fix interval timer and increase scan depth and jump to the highest peak.
Add periodic partial shading mechanism: do a (fast) broad scan and jump to the highest peak. Interval is 5 or 10 minutes depending on the Vmpp vs Vocv
Update fan limits to prevent overheating the MPPT boost coil under certain conditions.

So am i right that the fan fix was included in 1.14 already?

Or can you send me the latest Beta?

Broesel

0 ·

Guy Stewart (Victron Community Manager) ♦♦ broesel commented · Nov 08, 2023 at 09:29 AM

Hi @Broesel

Please see here for the replacement beta v1.16 - MPPT RS shade performance testing, v1.16 beta

0 ·

broesel Guy Stewart (Victron Community Manager) ♦♦ commented · Nov 08, 2023 at 10:53 AM

Great, thank you :)

0 ·

Answer 20 · 2023-11-06T12:32:57Z

hotel-lima answered · Nov 06, 2023 at 12:32 PM

I have the MPPT 450/200 and I'm using 3 out of 4 trackers. Until now the 1.14-beta2 as well as the 1.14 seem to work fine on my side. Full comparison to the old firmware is not that easy as we have a lot of bad weather currently and conditions changed since beginning of September. But so far it looks good. Updated fan control seem to be included since 1.14-beta2 (fan is much more often active and with higher speed than before). Can I stay on 1.14 then?

2

2 |3000

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

Guy Stewart (Victron Community Manager) ♦♦ commented · Nov 08, 2023 at 09:23 AM

Hi @hotel-lima - and anyone still on v1.14

I would suggest moving away from it now, and instead trying the v1.16 beta available here - MPPT RS shade performance testing, v1.16 beta

Once you update VictronConnect you will be forced to do so. This isn't anything personal just a side effect of how the development rollout happened.

That fan control fix is very important, and needed to be made mandatory immediately.

However the shade detection adjustments that were also included in v1.14 were not ready for wide scale roll out, they worked better for some (such as yourself) but worse for others. So we had to remove them, and then force the fan control change without the shade changes.

It's all a bit muddled now, and apologies for that, we were excited to see the early results and for some customers (even in this thread) were getting quite frustrated to not see the progress released, but as nearly always with these things, pushing faster than the normal process reminds us why we have the normal processes in place.

Hopefully now with the v1.16 beta method we can return to some steady progress.

0 ·

hotel-lima Guy Stewart (Victron Community Manager) ♦♦ commented · Nov 08, 2023 at 09:45 AM

Alright then. Will have a look at 1.16 beta.

Anyhow I appreciate that you picked up the topic and working on it. I'm confident that you will come up with a decent and robust solution at the end.

Thanks so far.

2 ·

question