question

StraightUp avatar image

When is Victron going to fix the crippling bugs in their MPPT controllers?

Six months ago, following multiple user complaints, I posted this, outlining how Victron MPPT controllers consistently fail to fully charge batteries in intermittent sun, due to a bug in their "smart" Absorption timer algorithm.

"Fast" forward to last week, and finally a supposed solution is given in terms of an update to Firmware 1.42.

Except it's not.








Nothing has changed with v1.42 at all. The controller is still resetting to its shortest absorption period every time a cloud goes over, and dropping to float when the batteries are desperate for charge. And yes - I manually ensured the battery V was under 12V every sunrise, and I did isolate all loads during the day for these tests - that difference in Amps between the app and the Trimetric is just Victron's apparently poor accuracy (another problem in its own right).


This isn't some obscure, niche problem. This is core function - reliably controlling the absorption stage length is the fundamental point of a controller. It's been reported again, and again, and again, and again, and again, and again. We have a suspected case where it's ruined someone's batteries as a result. For it to take six months for this to be addressed, while adding pointless gimmick charts to the app instead, was poor enough. For it to now turn out to be a non-solution is unacceptable.

I'm addressing this squarely at you @mvader (Victron Energy Staff). You promised this was being fixed when I first bought it up. Your response here determines whether I take this unit back for a refund.

MPPT - Solar Charge Controllerabsorption
2.jpg (2.0 MiB)
1.png (106.8 KiB)
13 comments
10 |3000 characters needed characters left characters exceeded

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

"Nothing has changed with v1.42 at all. The controller is still resetting to its shortest absorption period every time a cloud goes over, and dropping to float "

I have been running v1.42 on a smart 100/50 since 25/6 and it seems to be behaving properly during the absorption phase, as per description in the changelog.

⦁ Stop counting down on the absorption timer when there is (temporarily) not enough PV power available to maintain the voltage at the absorption level. Previously, the absorption time kept counting; even though the voltage was not at the absorption level. Note that the history module might show a larger absorption time value compared to the maximum absorption time setting, as we extend the absorption phase as long as the absorption voltage cannot be maintained.

I dial in 15 minutes of absorption for my batts, and that's what I get, unless the sun hides behind a cloud.

What model mppt are you using?

SmartSolar 100|50. But I think youre confusing the issue here. It might be worth reading the previous thread.

The problem is that when the sun is blocked by a something to make the PV voltage drop <5V over the batt voltage, the controller is sensing it as a 'sun down' event. When the shade passes, the controller perceives it at a 'sunrise' and re-calculates the Absorption time, based on an inflated (usually 13v+) battery voltage. It interprets this as a 'full' battery and applies the minimum Abs time multiplier...even if the battery ACTUALLY started the day as quite flat, and needing a lengthy absorb period.

"The problem is that when the sun is blocked by a something to make the PV voltage drop <5V over the batt voltage, the controller is sensing it as a 'sun down' event."

I am not seeing that (how dark does it get for the pv voltage to drop to batt volts?). Maybe someone else is seeing your problem? This should no longer be an issue, it has been written out of the firmware (?).

⦁ Fixed bug that caused the charge algorithm to reset when the voltage on the PV terminals dropped 2V below the battery voltage. In other words, a situation where there was totally no power any more on the array, for example when mounted on an RV and driving under an over-pass. The reset could lead to a too short absorption time. This has been fixed.

"I am not seeing that (how dark does it get for the pv voltage to drop to batt volts?). "

A decent rain cloud does it. You can see it on the monitor - charge amps drops to 0, voltage drops to ~13v ie panel are inputting nothing. Its a totally normal phenomenon and all my other controllers cope just fine.


" Maybe someone else is seeing your problem?"

Did you read the link in my post? I gave 6 examples off the bat. Search and youll find dozens. Its been a problem for ages.

"This should no longer be an issue, it has been written out of the firmware (?)."

Exactly my point - it ISN'T fixed, despite what the 1.42 notes say.

Even you own chart appears to show that although you say "you dont 'see it", I suspect it's happening to you too.

Look at all those days where the absorption periods was tiny, and the majority of your charging is occurring in float (specifically 2, 3 5 and 6 days ago). Unless you basically used NO power at all on the nights preceeding that (which I sspect not, because the total kWh yeild is consistent across most days), you should be getting a good several hours of absorb each day, or you're killing your batteries with undercharging. On the highlighted day where you DID get a good 3hr 27min Absorb, look at how much larger the total kWh yield was.

I seem to be experiencing exactly the same problems at my lodge with my Victron system (i.e. "low battery" was coming on far too soon and battery never seemed to fully charge). This is one reply I received from my battery supplier after my installer pointed to a problem with the battery BMS;

"Your assessment is completely incorrect in this regard. Your installer set the solar charger to charge to 56.2V and it only charged to 55.9V. This can be verified by a multi-meter. He then had to increase the setting on the solar charger to 56.7V to get it to charge to 56.4V. You can see this on the recorded data. We have a special instruction in our manual to ALWAYS verify that the chargers charge to the correct set point as we’ve seen some of them not to do so. The Victron system has a function called DVCC (Distributed Voltage and Current Control) which allows the battery to control the voltage set-point. This is not enabled on your 8kWh battery.

Your belief that the BMS is unable to communicate with the Victron system is unfounded. I have investigated each and every parameter between the battery and the CCGX and these are communicated correctly."

As you can see our installer fiddled with the Victron charger to get the system to work. I really don't understand what DVCC means and whether or not it is wise to use this setting, or whether it will even solve the problem. Any advice will be highly appreciated. I will post this in the main-stream as well since I am now experiencing the same problems at my home having just bought new lithium batteries (interestingly we did not have any such problems with our lead-acid batteries).

EDIT: I updated this message; my earlier explanation was over simplified.

Hi Gordon, I'm not sure that what you're seeing is related.'ll keep an eye on your question to make sure its cleared up:

https://community.victronenergy.com/questions/27255/dvcc-constant-low-battery-warnings.html


sorry to disturb. where is this new version of firmware. 1.42

i dont seem to have it. i thought that the connect app should always be upto date?

how do i get it

cheers rob

Hi @Robin, no problem! Indeed its not (yet) in the latest official VictronConnect version. There are two places where you can find it:

1. by using a beta version of VictronConnect, see here https://www.victronenergy.com/live/victronconnect:beta. That version includes more settings such as configurable tail current, and choosing between fixed and variable absorption time.

2. By downloading the file from https://professional.victronenergy.com.


Lastly, in case you have the MPPTs connected to VRM on the internet, then you can use the Remote Firmware update function.


Or, wait a little (at least a few weeks) for us to finalise that new VictronConnect version.


Robin avatar image Robin mvader (Victron Energy) ♦♦ ·

Thanks Vader.

I must be doing something wrong as can not see v1.42 in either place, can you advise.

Also can you confirm this will solve the shadding issue and not course the problems v1.41 did. Thanks rob

Hi @Robin, it might be that for your model, v1.41 is the latest version. v1.42 fixes a few issues specific to a few models, and was thus only released for those models. see the change log in this folder:

https://www.dropbox.com/sh/6cky0413xkwa7a2/AADMbZSnRVR3wgfjcwNp30W3a?dl=0

(Same content is on victron professional btw - this dropbox and VicProf are the same system).

Let me know if there is still something that I can help with.

Hm. I’m mixing things up, sorry. v1.42 is latest version for most; and for a few there is v1.43.

Anyway, see that folder; you’ll find them there.

4 Answers
StraightUp avatar image
StraightUp answered ·

And it's returned.

A week a passed, and the only response from Victron is denial and deflection. Not even an acknowledgement of the (continued) existence of the fault.

Disappointing product, support and company attitude as a whole.

Share
10 |3000 characters needed characters left characters exceeded

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

OffgridSA avatar image
OffgridSA answered ·

Straightup, I appreciate your post and it's refreshing to see some one back up issues with hard data to support what they claim

Most end users (customers) will be quite reasonable when it comes down to ironing out bugs or flaws in a products software even after its released to the public.

That said if you are end user who has a above average level of knowledge and confidants in you analysis of a issue and measuring equipment to back it all up. It can be VERY frustrating when you feel like you are being dismissed by the by the people who should be taking your issues seriously. If they really do want to improve there product's.

Especially when the shortcomings relate to core functionality of the product.

1 comment Share
10 |3000 characters needed characters left characters exceeded

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

Thanks.

I think I've been more than reasonable and does basically everything that can be (justfiably) asked in terms of helping fix this. I'm absolutely staggered by how Victron staff seem to think it's acceptable to basically fob the whole thing off. This isn't some issue with bluetooth or some minor measurement anomaly - it's a completely device-breaking bug. It's going to kill peoples batteries. Really poor company form.

mvader (Victron Energy) avatar image
mvader (Victron Energy) answered ·

Good morning @StraightUp,

Indeed we fixed the issue. Here is the entry from the change log:

Fixed bug that caused the charge algorithm to reset (and re-evaluate battery voltage to determine the absorption time) almost immediately when the voltage on the PV terminals dropped 2V below the battery voltage. In other words, a situation where there was totally no power any more on the array, for example when mounted on an RV and driving under an over-pass. The reset could lead to a too short absorption time. This has been fixed, by adding a delay of one hour: the charger restarts (and redetermines the absorption time based on battery voltage) only after the PV has dropped 2V below the battery during one hour.

(From the v1.42 entry in the change log. Note that I expanded that item in the change log this morning a bit to by putting in more details; thanks for pointing to that @klim8skeptic).


Also we fixed issues in the daily history logging.


I don't yet understand what happened exactly during your test; I'll send you an email, please respond to that; thanks!


18 comments Share
10 |3000 characters needed characters left characters exceeded

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

ps. still emails are rejected. Please let me know how I can reach you; you can use a private message here on the forum.

StraightUp avatar image StraightUp mvader (Victron Energy) ♦♦ ·

I think I explained this once before? I used a throwaway email to create my account, after seeing that other moderators previously lifted such personal information to initiate unwanted contact.

You can, and already are, reaching me here. I see no reason not to keep this discussion public?

There are plenty of reasons to take part of the debugging offline. You are so smart that I can't believe you can't think of what any of those might be.

You can always come back here with any conclusions, results, or whatever.

At some point you may, in fact, have to identify yourself as an individual, specific human here on this planet in order to get the support you deserve.

At the same time, there is no need to be so ill-mannered in this community. No matter how frustrated you are, this is not the place.

Again, I'm yet to hear a good reason as to why this 'debugging' can't take place where all the people affected (paying customers) can have visibility. Surely YOU can see how this saves time, increases the intellectual pool, and cuts down on rumour and misinformation?

There is absolutely ZERO need to disclose personal information to solve this matter. It's irrelevant, and Victron has already demonstrated they can't be trusted to respect members privacy. Any relevant technical discussion can be had as easily here as with private email, and any information can and should be assessed on its own merit, not the name attached to it.

If I wanted to embark on some collaborative technical marathon, I could have tried to build my own controller, and take the process to half a dozen other (anonymous) forums without a commercial self-interest for actual advice. But I didn't. Because there are only 24 hours in a day, and mine are full. So I paid $500 for a product that claimed to work off the shelf. I didn't pay $500 for the privilege of having to then hand over my personal information and time, at my cost, to help a company do what they should have in the first place, for their profits.

As for being ill-mannered - I reject that entirely. I'll concede blunt, but after tolerating this for six months, I don't think that's unreasonable.

What is unreasonable, @ben, is making snarky, sarcastic posts with no technical content relevant to the issue, complaining about how other people aren't 'taking part' and are being 'ill-mannered'.

If your game, and accept the risk, try a beta version of victronconnect. https://www.victronenergy.com/live/victronconnect:beta

Tip download onto a secondary device, as updating device firmware may not be possible with a beta version.

Your tone is unacceptable in this good-faith community. The community guidelines explain how to receive support, if you require it, directly from Victron.

The rest of us want you to be polite. There is no snark or sarcasm in anything I wrote. (And I wouldn't write at all, if you were not at risk of damaging the community with your inappropriate tone. This has nothing to do with technology.)

For your first question, I continue to state from personal experience: certain types of assistance are not rendered via this community. You cannot avail yourself all of it if you are not willing to make a personal contact -- over the internet, no less -- with someone.

@ben You honestly just seem to be trying your hardest to derail this away from the technical matter at hand. I've got no interest in playing that game,

@klim8skeptic When you say "secondary device", do you mean another controller? I only have one, and I'm sure not going to go buy another.

Show more comments

We also don't want to hear your whining about dollars and time. Many of us have spent well over an order of magnitude more than you with this company, and we have products with issues as well, and yet we are not in here ranting and raving like young children.

Instead, we reach out via the appropriate channels -- which go well beyond this forum and include our dealers and/or the customer support mechanism -- to seek resolution.

Finally, if we don't get the result we want, we return the product and find something else that does what we need instead. Which is what I highly encourage you to do, instead of threatening that you might do it to the Victron CEO on this site intended for sharing and support. It truly does not seem like your time is worth wasting any further on this, @StraightUp.

@ben , I suggest you step back from the keyboard, and calm down a little.

Show more comments
StraightUp avatar image StraightUp mvader (Victron Energy) ♦♦ ·

I also cannot for the life of me find any "Private Message" function on this forum.

Yes, I read the change log myself. Hence my making this post - the words written in that changelog seem not to be reflected in reality.

I don't know what other details there are that I can provide, over the top of what has already been shown. Cloud cover is resetting the controller. The issue isn't fixed.

@mvader (Victron Energy Staff)

....so, that's it? "Nah we like totally fixed that, honest...".


Are you, or are you not, going to resolve this very clearly unfixed problem? Or are you now just going down the denial route?

And don't give some weak excuse about need to email me privately. Whatever email you're going to send, you can put it here. Trust me, I'll read it just the same.

Oh @StraightUp you know better than this. Ofcourse I’m keen to find out what is happening on your system.

But the way these discussions between you and me tend to go, not here in public.


StraightUp avatar image StraightUp mvader (Victron Energy) ♦♦ ·

@mvader (Victron Energy Staff) .

I could have already just made another temporary email (and redistributed any correspondance....), if communication medium was really the issue here. But we both know better that that, don't we?

I could have returned this unit in February. Instead, I did the testing and write-up that I did, and have patiently persisted with the problem for six months since. Because, believe it or not, I am actually keen to help with a solution. Within reason - namely, a respect for personal privacy and transparency .

You're the one obfuscating this with conditions. You're the one who's turning this into more than just a technical issue, but a case study in Victron's company values.

You say you're keen to figure out what's happening. So am I. Right here, for all your other customers to benefit from.

Your move.


StraightUp avatar image StraightUp mvader (Victron Energy) ♦♦ ·

@mvader (Victron Energy Staff) your silence has been an admission in its own. Extremely disappointing.

Frederic.mora avatar image
Frederic.mora answered ·

First of all, what are we expecting of Absorption time vs Bulk time? I read on a Pheonix Charger FAQ doc that for FLA batteries Absorption time should be 5 times bulk time? can someone confirm? If that is the case, even when adjusting Absorption time to max 24h, I never ever get 5 times bulk time in Absorption. The only way I can get enough Absorption is to use the 1/6th rule of max Absorption time, still true with new 1.42 version and boost the max Absorption time to 36 hours.

Also under slight load (ie 60W fridge), with more than 250W of available solar,(440AH AGM bank) , MPPT doesn't keep up with absorption voltage. So we can't really talk about constant voltage while still being in Absorption step. So even in Absorption step (counting in the total Absorption duration), batteries don't see a constant voltage at the set Absorption Voltage. Assuming that there is enough Solar power available, I can't quite understand why Voltage is not held at Absorption voltage.

If we add this low voltage Absorption phase issue to the shorter Absorption reseted time after getting into short bulk step during the day, batteries are never really under a quality Absorption voltage and time... The never get charged properly.

This needs to be addressed quickly and seriously.




14 comments Share
10 |3000 characters needed characters left characters exceeded

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

Regarding absorb time:

Absorb is currently caculated using this method:

So, in theory, to get your require 5 hrs absorbtion (which I'd agree with for most batts), you SHOULD be able to set the max absorb time to 5 hrs, and then providing your battery voltage is under 11.9v at sunrise, the controller will calculate to hold the full 5hrs.

Except it doesn't, if you have ANY clouding or intermittent shading. Should the panels drop voltage during the day, the controller turns off, and turns back on when the sun comes out. As it does this, it recalculates the absorption time for that day based on the same chart above - except because the batteries are mid-bulk, the V s usually over 12.6, so it THEN applies the 1/6th rule. Meaning in most real world conditions, the controller just winds up setting you absorb at 1/6th whatever you have it set to, resulting in undercharging.

Of course the problem is that if you did/could set the max time to 36hrs and bank on the 1/6th fact to bring it back to 5, then if you get a clear, sunny patch, the controller could indeed try and hold absorb for 36hrs, resulting in overcharging. Same if you set float to= absorb. You're out of luck either way.

This patch claims to have set the controller on a 1-hr delay (ie you need shade over an hour long to reset the controller) but it hasnt fixed the problem, because a) a 1hr storm front (super common in my part of the world) triggers the same issue again (ask @mvader (Victron Energy Staff) why they arbitrarily picked 1 hrs and not say, 5 hrs?) and b) it doesnt seem to be effective anyway, my controller still resets with just minutes of sun loss.

1217-websiteabs.png (152.6 KiB)

Regarding absorb Voltage:


I've noticed this too - for some reason the controller is slow to up the power to compensate for loads . I suspect it's related to voltage drop from controller to battery . The controller is just providing enough current to hold a constant voltage at its own terminals, and has no visibility of loads 'diverting' that current when they cut in, causing actual battery v drop. When I see it, the controller keeps showing the same V and A output, but i get slight V (~0.1 - 0.2v) and net A input drop at the battery terminals from the independent monitor. If the load increases above about 2-3 amps, the controller then registers a V drop at its own terminals, and ups the power until voltage rises to setpoint again.


Are you measuring your Abs voltage at the battery or the controller? What sort of load size do you need to get the controller to increase power and get back to the set Abs V?


It's of less concern to me. So long as you're in the ballpark of the batts abs V, youll be pushing the ions back in the right direction - it'll just take longer. Of course with the issue with the absorb time being cut short....

Like you I m reading both (at the battery with a BMV which match my measurements with the DVM (Digital Voltage Meter) , and on the App for the MPPT. Even at low Amps (3 or 4 Amps) , there is a 0.1 to 0.2V difference, so it cannot be a voltage drop due to too small gauge wire. I believe there is a problem here too, so I ordered a voltage (and temp) sensing unit. Here again, I consider this as a design flow not to have a way to physically connect a Voltage sensing wire as default (not with add on equipment). Until now I boosted my Absorption Voltage to 14.6V in lieu of 14.4V

In Absorption phase , what I feel is that when I start to see a voltage drop when load is applied (6 to 8 Amps) , voltage doesn't come back up until load is gone. And this happens despite available power from solar panel. I have 350W installed and on a sunny day get 260W off of them. That is plenty enough to cover the 6 to 8 Amps load and maintain voltage.

I will double check this and report back.


Thanks for answering my previous questions.

So I tested a few more things this morning.

Unlike I was saying previously, when load is applied to the system and assuming there is enough solar power available, the voltage at the MPPT remains constant. But as the controller compensate for that extra current consumption, extra current is sent to battery increasing voltage drop at battery terminals. That should be fixed with voltage sending unit.

BUT, I realized that the temperature compensation was way off with the built temp sensor. Controller temp is between 38C and 35C. So for a 14.4V Absorption Voltage and - 24mV/C compensation coefficient, I should have at least 14.09V and I only get 14.01V. ambiance température in battery compartment is 32C, so I m actually targeting 14.23V.


I can't find a way to calibrate the built in temp. Sensor withe app, so.... That is another inaccuracy that undercharges batteries. Since I m now in a tropical environment, I can disable the temp compensation feature until I receive the temp sensing unit.

These are two good important issues that can only be solved with the temp and voltage sensing unit.

I had an issue with the V 1.42 yesterday. I had a 15 A @ 12 V load connected to the battery and the solar panel setup was providing around 120W @ 24V. Then I turned off the load and battery reached the floating charge somehow, even though I have consumed quite a bit of energy from the battery and would expect further charging. Then I turned the load back on, but the system was still in floating mode, even though I was pulling 15 A again. The battery voltage was stable at the floating voltage. I restarted the system and it was instantly in bulking charge mode and provided the 120 W again. Seems buggy to me.

Hi Simon.

"Seems buggy to me."

Without quantifying the details properly, it would be impossible to give advice or explain what is happening with your system.

"I had a 15 A @ 12 V load connected to the battery and the solar panel setup was providing around 120W @ 24V. Then I turned off the load and battery reached the floating charge somehow, even though I have consumed quite a bit of energy from the battery and would expect further charging."

How long was your load connected, a few seconds, or a couple of hours? What charge stage (and battery voltage) was the controller in when you turned the load on?

"Then I turned the load back on, but the system was still in floating mode, even though I was pulling 15 A again. The battery voltage was stable at the floating voltage. I restarted the system and it was instantly in bulking charge mode and provided the 120 W again."

If your battery was still at the float voltage, why did you need to restart the mppt?

Hi klim8skeptic,

it's an 18 Ah battery and I was using it together with the panels (2x100W in series) for about 15-20 minutes already. Since the Victron mppt (75/15) was able to provide around 8,5 amps @ around 14,X V max at the time, I would expect constant discharging over that period. Hence, I wonder how the system (read from the app) could reach the floating charge state and voltage, when it has been constantly discharged for that time with around 7 amps until I turned the load (trolling motor) off.

Hi Simon.

"it's an 18 Ah battery"

You have an extremely high load/charge for such a small battery.

Have you checked the battery manufacturers specs?

I checked those I could find for the charging voltage (up to 15 V) and minimum voltage (11,89 V). There was no limit regarding the (dis)-charging amperage described, but I contacted the distributor asking for it. However, you are most certainly right, that using the battery basically as a buffer has to be considered as improper treatment. Still, when I reconnected the mppt, the actual voltage was 1 V less (around 12,9V) and it started charging again, which tells me that it didn't recognize the consumption of 15 A before and I guess that the actual battery voltage was disguised by the floating voltage achieved through the solar panels. Do you see what I mean?

@mvader (Victron Energy Staff): May you tell us if this is being investigated further? I can second, that the controller shows the described behavior with being unaware of the actual consumption at the battery in the charging modes above bulk. Maybe it would help if it cut off the solar voltage from the terminal from time to time, but long enough to read the actual battery voltage?

Hi Simon, there can be many different reasons for what you are seeing; from what you are have posted I don't know if you have an issue or not.

Please post a new question, include all details of what you have seen as well as type of battery, used configuration (end-of-absorption-by-tail-current on or off for example?), preferably by screenshot.

Thanks!

I see the same issue. Absorb stage (with ample pv and sun avallable) when looking at my voltage logs the graph should look like a dead flat line where battery voltage is held constantly at the absorb voltage set point. As I have always experienced with other charge controllers. I dont see this with Victron. It does a very bad job of managing a steady constant voltage absorption stage. I can provide all data to back this up if needed. Along with comparative data to other controllers.

Hi OffgridSA, please post the details is a new question rather than in this thread that started about length of absorption; I’ll make sure its answered; thanks. Matthijs

Do you have a Battery voltage sensing unit? if not I believe it is voltage drop through the cable.