Victron solar charge controller for LFP?

[nebster]

Just to answer the very specific question tanglewood keeps posing but that is not really answered:

Today, on the current firmware, with DVCC enabled on a Venus device and Shared Voltage Sense turned on:

1) If there is a BMV or BMS online, the voltage from that device will be sent to all chargers (Quattro, Multiplus, MPPT) on the network.

2) If there is not a BMV or BMS, each charger will use its own voltage readings. For the big inverter-chargers, that means they will fall back to their Vsense pair first. For the MPPTs, they don't have a dedicate sense pair, so you just get the single battery reading.

With SVS, it will NOT send the Vsense from the Quattro over to the MPPTs, in case 2. That data is however exposed on the underlying databus, so it should be easy to do. It might be something Victron would add one day, or someone sufficiently motivated could perhaps do it themselves since the CCGX/Venus stack is quite open.

For what it's worth, my experience with my two Quattros is that their Vsense is not quite as accurate as the BMV (although it is far better than using the Quattro's main DC leads).

[tanglewood]

Originally Posted by nebster

Just to answer the very specific question tanglewood keeps posing but that is not really answered:

Once again, thank you.


I have a love hate relationship with Victron. They are well made, very capable products. But also a very well kept secret. I'm a techie, and actually download and read manuals to see how things work, and what they are capable of. But you can't figure any of this out from Victron's web site or published manuals. It's very frustrating.


See more below..

Originally Posted by nebster

Today, on the current firmware, with DVCC enabled on a Venus device and Shared Voltage Sense turned on:

1) If there is a BMV or BMS online, the voltage from that device will be sent to all chargers (Quattro, Multiplus, MPPT) on the network.

I see this now. I was reading a version of the CCGX/Venus manual from about a year ago, pre-DVCC.

Originally Posted by nebster

2) If there is not a BMV or BMS, each charger will use its own voltage readings. For the big inverter-chargers, that means they will fall back to their Vsense pair first. For the MPPTs, they don't have a dedicate sense pair, so you just get the single battery reading.

The manual also says it will fall back to the battery voltage reported by the VE.bus system. I don't have all the combinations of which products use VE.Bus, VE.Direct, and VE.CAN straight in my head, so will have to research which products might report this, and whether they actually do.


I would also expect each charger & inverter manual to explain what sources it uses fro voltage sense, and under what conditions it changes.

Originally Posted by nebster

With SVS, it will NOT send the Vsense from the Quattro over to the MPPTs, in case 2. That data is however exposed on the underlying databus, so it should be easy to do. It might be something Victron would add one day, or someone sufficiently motivated could perhaps do it themselves since the CCGX/Venus stack is quite open.

So how does this relate to their statement about using VE.Bus voltage data? This is a good example of the well kept secret mystery of Victron. Just out of curiosity, how have you come to know this?

Originally Posted by nebster

For what it's worth, my experience with my two Quattros is that their Vsense is not quite as accurate as the BMV (although it is far better than using the Quattro's main DC leads).

I hadn't planned to use a BMV, but it sounds like it's the only shared voltage sense available.

[nebster]

Originally Posted by tanglewood

I have a love hate relationship with Victron. They are well made, very capable products. But also a very well kept secret. I'm a techie, and actually download and read manuals to see how things work, and what they are capable of. But you can't figure any of this out from Victron's web site or published manuals. It's very frustrating.

Yes, they are understaffed relative to how much gear they produce, and they seem to count on that dealer network to hold all the "institutional knowledge" in their heads, I suppose.

The good news is that they are aggressive on the software side in terms of involving the community, and Matthijs is directly involved with many of the points of engagement there. So I think this will improve.

In fact, if you figure out more that should be documented, it is a wiki after all, and I'm sure they'd be happy for you to edit the master document.

The manual also says it will fall back to the battery voltage reported by the VE.bus system. I don't have all the combinations of which products use VE.Bus, VE.Direct, and VE.CAN straight in my head, so will have to research which products might report this, and whether they actually do.

I actually think the wording there ("VE.bus system") is ambiguous. And I would expect the MPPTs to use the Vsense on the Quattro in the absence of a BMV or BMS-reported voltage, but that's not what I'm observing when I yank my BMV offline and watch to see what voltage the MPPT picks up. I am actually a couple versions back on the Venus OS, so it's possible that this has changed in the last few months.

So how does this relate to their statement about using VE.Bus voltage data? This is a good example of the well kept secret mystery of Victron. Just out of curiosity, how have you come to know this?

I just have the system running right here, so I messed with it to see what it would do.

One thing you could do is post your question to the new community.victron site and hopefully get an answer that clarifies what is supposed to happen.

I hadn't planned to use a BMV, but it sounds like it's the only shared voltage sense available.

I think you're going to want one anyway, because I don't know why you would want to start your generator at a specific voltage e.g.. (Unless you are somehow computing a load-compensation offset?) You'll probably be happier autostarting at an SOC, if you can.

But it should be possible to pick up the Vsense line. Maybe that's coming.

[tanglewood]

Originally Posted by nebster

Yes, they are understaffed relative to how much gear they produce, and they seem to count on that dealer network to hold all the "institutional knowledge" in their heads, I suppose.

The good news is that they are aggressive on the software side in terms of involving the community, and Matthijs is directly involved with many of the points of engagement there. So I think this will improve.

In fact, if you figure out more that should be documented, it is a wiki after all, and I'm sure they'd be happy for you to edit the master document.



I actually think the wording there ("VE.bus system") is ambiguous. And I would expect the MPPTs to use the Vsense on the Quattro in the absence of a BMV or BMS-reported voltage, but that's not what I'm observing when I yank my BMV offline and watch to see what voltage the MPPT picks up. I am actually a couple versions back on the Venus OS, so it's possible that this has changed in the last few months.



I just have the system running right here, so I messed with it to see what it would do.

One thing you could do is post your question to the new community.victron site and hopefully get an answer that clarifies what is supposed to happen.



I think you're going to want one anyway, because I don't know why you would want to start your generator at a specific voltage e.g.. (Unless you are somehow computing a load-compensation offset?) You'll probably be happier autostarting at an SOC, if you can.

But it should be possible to pick up the Vsense line. Maybe that's coming.

Thanks, very helpful.


It does seem that Victron relies on their dealer network for all this, and in my experience distributing, selling, and supporting tech products in Europe, that's the norm. But in the US it's a giant fail, for better or worse. 99% of the time if I ask a dealer a questions, I find that I know more about the product than they do. And it's not because I'm some genius. All I've done is read the manuals. In the US, one goes to the manufacturer to get anything other than trivial questions answered, and it's expected that they are ready and waiting for the call. No right or wrong - it's just different.


Right now all the Victron equipment for my boat is on paper only, so I really appreciate the hands-on experience.

[CatNewBee]

Hi, nebster,

I understand your urge, but it is much simpler than you may think.

MPPT controller simply run from time to time the curve to find the MPPT point on the panels, yelding during this time very little current, so the controller itself has a pretty good idea of the battery voltage on the poles without communication - it is the best bet if you not have a BMV.

Another point is, the on board system is an integrated one with various charge sources and loads at the same time, even the BMV has no clear picture at all time what causes the current measurements, but it is closest to the battery and the only device that knows if there is a current flowing, in what direction it goes, it can also get an idea by differential measuting of the battery resistance and resulting voltage drop to correct the voltage measured to the current current flowing to calculate a virtual cell voltage and resulting SOC.

No other device is able to do this, because all they sit on fractional parts of the installation and have no idea about the others.

[noelex 77]

Originally Posted by CatNewBee

Hi, nebster,

I understand your urge, but it is much simpler than you may think.

MPPT controller simply run from time to time the curve to find the MPPT point on the panels, yelding during this time very little current, so the controller itself has a pretty good idea of the battery voltage on the poles without communication - it is the best bet if you not have a BMV.

Some of the old MPPT controllers did this, and a few of the very cheap “no name” models still do, but it is a very wasteful way to establish the MPP.

The goal of any MPPT controller is to spend the maximum time at the MPP. The controller has to be particularly wary of spending any time at voltages above the MPP, as the output will be severely reduced. On occasion the voltage/output relationship can become multi-lobed, which can be particularly challenging to track.

The better solar controllers have a very sophisticated algorithm for establishing and tracking the MPP. The success of this tracking algorithm has a significant impact on the output. We really need independent tests using different solar controllers. Unfortunately, most consumers incorrectly assume all MPPT will have a similar performance.

[CatNewBee]

This is another can of worms. What I can say, my Victron MPPT 150/100 Smart solar gets every now and then a quick full cycle through the spectrum and I can see the current drop and the search for the best point on the display, but most of the time it sticks and fluctuates at a narrow range around the MPP he has found during the last full search, i can track all values and states either on the wired (optional) MPPT controller screen or on the Victron BT App.

Because the title of this thread is explicitly to Victron controllers and I assume they perform similar, I gues my statement is valid so far for most of their gear.

Anyway, what I have is one of the "high-end" things, that communicates via BT to the BMV 712, I also have installed his display (and I guess BT sender) in the salon at starboard, while the MPPT controller is in the port hull under the bunk almost near the outer hull - so approx. 6m away, with structures like floor, hull, the freezer and walls and stuff in the locker in between, communication is impressively stable so.

The Voltage readings jump between what the controller is measuring life and the Voltage from the BMV, that is updated every minute or so.

[noelex 77]

Originally Posted by CatNewBee

This is another can of worms. What I can say, my Victron MPPT 150/100 Smart solar gets every now and then a quick full cycle through the spectrum

Interesting.

The actual MPP is always fluctuating and the tracking algorithms can only establish the MPP by altering the input voltage. So some variation will occur and is normal.

The tracking algorithms also need to incorporate occasional wider excursions to reduce the risk of becoming locked onto an incorrect MPP in the case of a muti-lobe voltage/output relationship, but the type of full cycle scan you are describing (if I am understanding correctly) is incredably wasteful and this not normal regular behaviour for the better controllers.

Is there any chance that you able to post a video showing the behaviour you describe?

[CatNewBee]

I am at the moment not on board, I will be back in May.
But yes, it is doable I guess.

Full range does not mean down to zero, it just drops current and starts a Voc 62V and scans down to 49V
52V is usually the MPP he locks in on my panels (5 x 96 cells 330Wp AUO BenQ Solar SunForte PM096B00)

[noelex 77]

Originally Posted by CatNewBee

I am at the moment not on board, I will be back in May.
But yes, it is doable I guess.

Full range does not mean down to zero, it just drops current and starts a Voc 62V and scans down to 49V
52V is usually the MPP he locks in on my panels

That voltage range is much more consistent with the voltage I would expect so I don’t think your unit is unusual or has a problem.

Your description of a “full cycle through the spectrum”, as well as your earlier post suggesting that the controller knows the true battery voltage because it regularly adjusts the input voltage to a level where there is very little output current (which should not happen BTW), left me with the impression of larger input voltage range.

Thanks for clarifying.

[nebster]

Originally Posted by CatNewBee

I understand your urge, but it is much simpler than you may think.

Sorry, I'm not sure what you're referring to?

MPPT controller simply run from time to time the curve to find the MPPT point on the panels, yelding during this time very little current, so the controller itself has a pretty good idea of the battery voltage on the poles without communication - it is the best bet if you not have a BMV.

That's interesting. You're suggesting that it applies current compensation to its ongoing readings, using a baseline voltage obtained during its sweep?

Another point is, the on board system is an integrated one with various charge sources and loads at the same time, even the BMV has no clear picture at all time what causes the current measurements, but it is closest to the battery and the only device that knows if there is a current flowing, in what direction it goes, it can also get an idea by differential measuting of the battery resistance and resulting voltage drop to correct the voltage measured to the current current flowing to calculate a virtual cell voltage and resulting SOC.

I'm not quite sure I follow your point. Are you suggesting that this is something someone could do in the future, or are you saying that some part of a Victron system already does this? Perhaps you could explain further.

[CatNewBee]

The current drops almost to zero when scan starts at 62V (Voc), so there is a window where the controller can measure the voltage at the battery by neglecting the power drop along the connection - like a voltmeter.

I do not know of course if they actually do it in their firmware, but this is a way to overcome the systematic voltage error when measuring at the controller under load and I would use this if I would programm that thing.

[notz]

Originally Posted by tanglewood

Once again, thank you.

I have a love hate relationship with Victron. They are well made, very capable products. But also a very well kept secret. I'm a techie, and actually download and read manuals to see how things work, and what they are capable of. But you can't figure any of this out from Victron's web site or published manuals. It's very frustrating.

I hadn't planned to use a BMV, but it sounds like it's the only shared voltage sense available.

I can wholeheartedly agree with the above around a love/hate relationship with Victron. I am extremely technical as well, but finding definitive information about how the product works with another is like pulling teeth.

On my last boat I installed a complete LiFePO4 system from Victron including a MultiPlus, MPPT, CCGX, and more. Prior to that, I posted many comments on their "community" website asking how X component worked with Y device, and bit by bit figured out what worked/didn't work.

Peter Kennedy at PKYS also helped with some of my questions, as he had to reverse engineer some of it himself given that he's a large Victron reseller in the US.

DVCC was the key for me, and was released only a bit over a year ago if I remember correctly, and has next to nothing about how it works in the manuals even today.

Still, I would prefer a company that releases new features, new software, and continues to innovate but has bad manuals over one that doesn't!

[drcat]

Ever thought of something other than Victron? Take a look at EPever Triron series with it's optional wired monitor. At $170 and $35 each respectively for 40A version and more for higher Amp version, it might be an alternative. My understanding is board is conformal (sp) coated so should be suitable for marine use. It has some standard Lithium settings but also an "other" setting allowing user settings. The wired remote controller comes with just a 6' cable but people have used longer Ethernet cables . Youtube has a number of EPever solar controller reviews. Just a thought.