Talk me into a Cerbo?

[Capt.Don]

Originally Posted by CaptainRivet

Get a raspi and run Venus OS on it. Cerbo doesn‘t even have a real hardware reset and can be bricked. The raspi you simply format the SD and newly install Venus os. Also touch screens are much cheaper, non Victron don‘t work well with cerbo but no problem with raspi.
On the raspi you can then also run home assistant where you can much easier and cheaper integrate sensors. For one ruuvi tag I can get 6 equal sensors for home assistant. Also turn your analog sensors wireless with an ESP32 into home assistant. And from home assistant you can then get all into Venus OS or vice versa.

Do you have separate RPi's for HA and Venus.OS -or- are your running both on a single RPi?

[Plumbean]

Originally Posted by s/v Jedi

That’s not how lte/4G works. You don’t connect periodically and you don’t pay for time online. It’s just like a smartphone, always connected to the cell tower and the Cerbo will send data periodically. It’s very little data use and it doesn’t change the amount of data until it is offline too long and you loose data.

Victron indeed sells everything you need. I believe they even have a LoRa option, check that out.

Thanks, very helpful.


FYI, it does not look like the LoRaWan device works with the Cerbo

[CaptainRivet]

Originally Posted by Capt.Don

Do you have separate RPi's for HA and Venus.OS -or- are your running both on a single RPi?

I have a cerbo and an older Intel Nuc with SSD running home assistant and ordered a Raspi 4 that will run graphana and orca navigation (easier to setup, the raspi will have a backup installation of venus OS). I don't need a touchscreen for venus and my system is running stable but if I would I switch to raspi.
When I got cerbo 2 years ago I got told raspi and cerbo is the same which in detail is not true. So I got a cerbo as thats cleaner install but wouldn't anymore.
And the NUC I had lying around and is much more powerfull then a raspi so I can run home assistant voice steering offline.
It depends on what you wanna do with HA if you can run both venus and HA on one raspi or need each a seperate one.
In case you don't know better install each on a single raspi as it's pita to have HA set up on the same and then you realize you need eg. HACS and therefor need a Home assistant core which needs his own raspi and you have to start all over on a new raspi again.

[sailingharry]

Well, it seems that I am fundmentally blocked from using a Cerbo. Rats. Not sure I actually WANT to, but I surely wish it were an OPTION! Here's the problem.


I am using a DIY battery and BMS. While the charge sources should be set at voltages that will never invoke BMS protection, the BMS has the ability to shut down charge sources, and this is an important feature. The way the modest MPPT controllers (100/50 and 75/15) implement this feature is to repurpose the VE.Direct port. Unfortunately, this prevents using this as a VE.Direct port! An alternative way would be to insert a Battery Protect into the output of the solar panels (they do combine in a fuse block, with a single line to the electrical system, so this would work). With nearly 100A of solar output, the 220A unit for $100 would solve the problem (and eliminate the need for 4 non-inverting VE.Direct cables at $20 each). But I've heard that MPPT don't like to be disconnected from the battery while still having solar input.



So, the question. Do I control the MPPT as designed (and use the VE.Direct port, preventing a Cerbo). Do I put in a BP, which interrupts the MPPT output, perhaps with undesirable results. Or, best answer, do I set up a Cerbo and there is some way for the BMS to tell the Cerbo to shut off all charge sources (4 MPPT and a MultiPlus).

[s/v Jedi]

Originally Posted by sailingharry

Well, it seems that I am fundmentally blocked from using a Cerbo. Rats. Not sure I actually WANT to, but I surely wish it were an OPTION! Here's the problem.

I am using a DIY battery and BMS. While the charge sources should be set at voltages that will never invoke BMS protection, the BMS has the ability to shut down charge sources, and this is an important feature. The way the modest MPPT controllers (100/50 and 75/15) implement this feature is to repurpose the VE.Direct port. Unfortunately, this prevents using this as a VE.Direct port! An alternative way would be to insert a Battery Protect into the output of the solar panels (they do combine in a fuse block, with a single line to the electrical system, so this would work). With nearly 100A of solar output, the 220A unit for $100 would solve the problem (and eliminate the need for 4 non-inverting VE.Direct cables at $20 each). But I've heard that MPPT don't like to be disconnected from the battery while still having solar input.

So, the question. Do I control the MPPT as designed (and use the VE.Direct port, preventing a Cerbo). Do I put in a BP, which interrupts the MPPT output, perhaps with undesirable results. Or, best answer, do I set up a Cerbo and there is some way for the BMS to tell the Cerbo to shut off all charge sources (4 MPPT and a MultiPlus).

You have several options:

- select the MPPT controller models that have a separate remote on/off connector.

- alternatively, select a MPPT controller model that has a Ve.Can port and use that to connect to the Cerbo instead of Ve.Direct. This is the best and thus most expensive option.

[sailingharry]

Originally Posted by s/v Jedi

You have several options:

- select the MPPT controller models that have a separate remote on/off connector.

- alternatively, select a MPPT controller model that has a Ve.Can port and use that to connect to the Cerbo instead of Ve.Direct. This is the best and thus most expensive option.

Yeah, that crossed my mind. My MPPT controllers are in and working brilliantly, so any new controller is fairly pricey. Also, I have 2 100/50, each controlling a 400W panel and 2 75/15 each controlling a 100W panel (and all controllers are already oversized). I think the ones you are recommending are much larger than my needs.


Another option I've found is choosing a BMS that speaks Cerbo (CAN does it, from what I can tell). I haven't bought a BMS yet, but the one I'm looking at speaks TTL (actually, a controllable 50ma FET, basically a dry contact) that works brilliantly on the MultiPlus and the Balmar regulators, and the MPPT if I repurpose the VE.Direct. It would be wonderful if the Cerbo had an input that could be programmed as a "Charge Enable" signal to all devices.


Interoperability is a real problem in this space. It's even a problem in Victron, with all their communications protocol. My MultiPlus, bone stock, requires TWO separate dongles for my installation, one to program the unit and one to make it BlueTooth -- and they can't be run at the same time. Why Victron runs VE.Direct, VE.CAN, VE.Bus, VE.Smart, and who knows what else, such that Victron can't work with Victron, amazes me.

[CaptainRivet]

Originally Posted by sailingharry

Well, it seems that I am fundmentally blocked from using a Cerbo. Rats. Not sure I actually WANT to, but I surely wish it were an OPTION! Here's the problem.


I am using a DIY battery and BMS. While the charge sources should be set at voltages that will never invoke BMS protection, the BMS has the ability to shut down charge sources, and this is an important feature. The way the modest MPPT controllers (100/50 and 75/15) implement this feature is to repurpose the VE.Direct port. Unfortunately, this prevents using this as a VE.Direct port! An alternative way would be to insert a Battery Protect into the output of the solar panels (they do combine in a fuse block, with a single line to the electrical system, so this would work). With nearly 100A of solar output, the 220A unit for $100 would solve the problem (and eliminate the need for 4 non-inverting VE.Direct cables at $20 each). But I've heard that MPPT don't like to be disconnected from the battery while still having solar input.



So, the question. Do I control the MPPT as designed (and use the VE.Direct port, preventing a Cerbo). Do I put in a BP, which interrupts the MPPT output, perhaps with undesirable results. Or, best answer, do I set up a Cerbo and there is some way for the BMS to tell the Cerbo to shut off all charge sources (4 MPPT and a MultiPlus).

I actually first had this problem.
1) you don't want the cerbo to shut down your MPPTs, that's job of the BMS and not the cerbo.
2) MPPT: one of the most common way victron MPPT die is letting the panel voltage through and go open current with the remote port unable to shut them off.
3) the victron Ve-direct remote cable for the 75 and 100V MPPTs is hot when its off, means if the remote cable gets disconnected or cut in anyway the MPPTs switch on (absolute EE noGo of Victron).
So remote port of MPPT is not sufficent to do the job doesn't matter which victron MPPT you use. Due to security you need to cut the hot output side and done right this has no undesired issues.

The solution is:
1) you need the smart solar MPPTs and the BT dongle for the multi to build a BT network to sync charge.
2) You need a BMV 702 or 712 battery monitor (which a lot have as independent last resort anyhow) or a victron battery sense on the bank terminals to deliver via Bluetooth the reference voltage of the bank to the BT network. Sync of charge sources is done by BT network. Victron fanboys claim steering via DVCC and cerbo is much better, I saw no differences at all but DVCC made a lot of other issues.
3) a BP220 smart that cuts all victron MPPTs due to security(!) On the output. So the MPPTs don't go haywire as well as the BP when you cut the ouput BP and MPPT receives the reference voltage via BT network and MPPT only raise voltage to +2V above absorption voltage without no connection. With 13.8V +2V is 15.8V with no load. No issues.
4) now for VE.DIRECT ports of MPPTs use the Duppa 4 port VE-DIRECT to USB C adapter and connect 4 devices to one USB at the cerbo. Much cheaper then 4 Ve-direct cable and better to route one USB c to A cable then 4 VE-direct cable with their flimsy connector. The multi with a rj45 into cerbo.
5) with cerbo you get now all the data from MPPTs via VE.Direct and Multi via VE-can and steer other sources like water heater....whatever you wanna do but don't use DVCC.
6) the remote of the BP220smart goes to your BMS to cut charge for emergency. Via BT connect app you can simply switch off the MPPTs via the BP220smart for eg maintenance purposes.
7) if you have an issue with eg one MPPT diying and going open voltage the BP220smart shuts off at 16,0V cutting the MPPT of the charge bus and avoid like this a complete shutdown of the system due to overvoltage (i had this case). The BMS has a time delay for overvoltage so that eg balancing doesn't cause an unwanted disconnect so the BP is faster cutting the MPPTs off in this case then the BMS does an emergency cut off. Also a great 2nd line of defense in case BMS fails.
8) your BMS delivers the cell voltages and SOC to cerbo or if that doesn't have communication the BMV 702 or 712 delivers via VE.DIRECT the same reference voltage to cerbo then to BT network for charge sync and optimization.

Like this you have all data in cerbo doesn't matter which MPPT, have additional security and use that for automation plus all kinds of warnings but stay away from cerbo running the whole show via DVCC. You cannot run DVCC and BT connect simultaneously, stands nowhere in the documentation and you only find that on forums, experienced user telling you and get all sorts of gremlins making you think it works and suddenly it's not.

[Dockhead]

Originally Posted by s/v Jedi

. . . This is the best and thus most expensive option.

ROTFLMAO.

[sailingharry]

Originally Posted by CaptainRivet

3) a BP220 smart that cuts all victron MPPTs due to security(!) On the output.

A question on this (and it actually plays out in several applications). The BP is unidirectional. Power FROM the MPPT TO the battery. They are pretty strong on saying no reverse flow. But in many cases, including the MPPT, there is a need for reverse flow. The MPPT look for a battery source for various intel. That current is in mA, not A, but it exists.

From what I understand, the "No reverse current" comes with an unmentioned exception -- that a 220A BP can easily handle fractional Amp reverse current. Is that your understanding as well?


My Watt&Sea charge controller similarly is lit up all the time with little LED's even when the unit is out of the water. I think my D400 Wind Generator had similar reverse flows.

[sailingharry]

Originally Posted by sailingharry

A question on this (and it actually plays out in several applications). The BP is unidirectional. Power FROM the MPPT TO the battery. They are pretty strong on saying no reverse flow. But in many cases, including the MPPT, there is a need for reverse flow. The MPPT look for a battery source for various intel. That current is in mA, not A, but it exists.

From what I understand, the "No reverse current" comes with an unmentioned exception -- that a 220A BP can easily handle fractional Amp reverse current. Is that your understanding as well?


My Watt&Sea charge controller similarly is lit up all the time with little LED's even when the unit is out of the water. I think my D400 Wind Generator had similar reverse flows.

Answered my own question. Victron in several places (doing some searching/browsing) repeatedly states that the self consumption of a MPPT is so low (mA) that the reverse current causes no concerns (they mention heating as the (non)issue). While other devices than Victron solar controllers are not mentioned, I would assume similar devices would be similarly safe to use.

[s/v Jedi]

You do not want a device that is triggered by a warning level from the BMS to interrupt high current circuits like the output from a charger. Don’t do that, it is the wrong thing to do. The BMS will cut the high current circuit when needed, don’t replicate it.

You want to use alarm level triggers to control chargers, inverters, regulators, controllers so that they stop charging or discharging when the battery gets too full/empty.

If you can’t afford the equipment that supports this then simply configure it to stay away from extremes (lower the bulk voltage a bit or set the low voltage value for an inverter higher).

If you want to get this right the first time, ask here before buying stuff. An installation needs to be designed with these aspects in mind before buying the parts used.

[CaptainRivet]

Originally Posted by s/v Jedi

You do not want a device that is triggered by a warning level from the BMS to interrupt high current circuits like the output from a charger. Don’t do that, it is the wrong thing to do. The BMS will cut the high current circuit when needed, don’t replicate it.

You want to use alarm level triggers to control chargers, inverters, regulators, controllers so that they stop charging or discharging when the battery gets too full/empty.

If you can’t afford the equipment that supports this then simply configure it to stay away from extremes (lower the bulk voltage a bit or set the low voltage value for an inverter higher).

If you want to get this right the first time, ask here before buying stuff. An installation needs to be designed with these aspects in mind before buying the parts used.

Well then victron should design the MPPTs savely that under all conditions switching off by remote is guaranteed to work (it's not and that is for all series 75 till 250) and also if the remote line is interrupted in any way charger is off (thats only 75/100 seies). Both is without a doubt not the case so you need the best way to do it.
So the only way is to cut their MPPT save is to disconnect the high output side.
Victron is doing that the same way in some reference diagrams eg for the smart and the lynx BMS.
What I wrote above is standing again nowhere how to avoid the MPPT going haywire if the output is cut doing it the way victron shows in these reference diagrams.
Done right this even has several advantages I described above to compared to cutting off by remote.

[s/v Jedi]

Originally Posted by CaptainRivet

Well then victron should design the MPPTs savely that under all conditions switching off by remote is guaranteed to work and also if the remote line is interrupted in any way charger is off. Both is without a doubt not the case so you need the best way to do it.
So the only way is to cut their MPPT save is to disconnect the high side.
Victron is doing that the same way in some reference diagrams eg for the smart and the lynx BMS.
What I wrote above is standing again nowhere how to avoid the MPPT going haywire if the output is cut.

This is how they work. When you receive the unit, there is a wire bridge between the contacts. Without it, it doesn’t work.

A second way is to provide a positive voltage to the + terminal of the remote on/off but that is a secondary option: switching between both contacts is the primary way and exactly what is needed.

[CaptainRivet]

Originally Posted by s/v Jedi

This is how they work. When you receive the unit, there is a wire bridge between the contacts. Without it, it doesn’t work.

A second way is to provide a positive voltage to the + terminal of the remote on/off but that is a secondary option: switching between both contacts is the primary way and exactly what is needed.

I must learn that the most common way a victron MPPT dies (all series) is that it let's full panel voltage through to output and gets open current and the remote function is not working anymore=unable to shut the MPPT off.
So the remote port is not suitable as cut off for the BMS.
Additionally the 75/100 series the only remote cable offered via ve-direct port needs hot=voltage appllied to be off, means if the remote cable gets cut in any way the MPPT is switched on. So it's suitable for steering to optimize charge but not for security purposes which is clearly the prio1 task of BMS.
On the other side its not necessary and not beneficial that the whole charge bus is cut when one MPPT dies so cutting the MPPT in groups by a battery protect smart with sufficent current rating is the best option and victron is doing that in some reference diagrams too. As also wrote above the BPs adding another security level which is always welcome. And I had lately exact this defect and BP done it's job cutting the MPPTs and prevent an overvoltage chargebus disconnect when alternators were running.

That's different on multi and Quattro where the remote on/off for load and charge working under all condition the unit is on. Either the remote or the internal fuses cut the DC side sufficiently. So here the remote is sufficent to be used by BMS to cut it off for security purposes.

[sailingharry]

Originally Posted by s/v Jedi

You do not want a device that is triggered by a warning level from the BMS to interrupt high current circuits like the output from a charger. Don’t do that, it is the wrong thing to do. The BMS will cut the high current circuit when needed, don’t replicate it.

You want to use alarm level triggers to control chargers, inverters, regulators, controllers so that they stop charging or discharging when the battery gets too full/empty.

If you can’t afford the equipment that supports this then simply configure it to stay away from extremes (lower the bulk voltage a bit or set the low voltage value for an inverter higher).

If you want to get this right the first time, ask here before buying stuff. An installation needs to be designed with these aspects in mind before buying the parts used.

Doing MPPT controllers that safely work as you describe is, well, expensive. I have 4 panels, each with their own controller. Two at 100W using a 75/15, two at 400W using 100/50. All 4 are well oversized. Today at PKYS, the 75/15 are $67, the 100/50 are $184. These all use the backwards VE.Direct system, where either the VE.Direct network or the adapter for BMS control both are default "ON" (meaning any loose wire leaves the MPPT uncontrolled). As Rivet has pointed out repeatedly, and other sources on the web agree (some sources recommend gluing the VE.Direct plug into the socket as a work-around), that's kind of pointless (either control is pointless in the first place, or an OFF control that fails ON is pointless). This system, as installed, is $502 spread over several years. The cheapest MPPT Victron sells that has a credible control system (fails off) is the 150/70 on a $70 discount at PKYS for $382, for a system price (if I had planned ahead in 2021) of $1528, or over a grand extra for a control wire. And that assumes I could fit those massively oversized (2.5-7x the current and 3x the voltage) controllers in the space I had.



I do intend to set my various controllers(*) at a point that they should never be tripped by the BMS. In addition, the BMS will trip all of them(*) gracefully (control wires, not disconnect) before any HVC ever occurs. (*)The one exception is the solar, for two reasons. First, it's an always on source that can with a fair amount of finesse get me into the knee for balancing and to reset the SOC meter. It is also the only device that apparently cannot be reliably controlled for remote shutdown. An out-of-line cell can trigger a BMS call for charge shutdown, and my inverter and alternators can shut down -- but there is no way to do that with solar. This applies to an off-brand BMS, a BMV, or a Cerbo, since all of them depend on illogical control wiring.


Using a BP to shut off charge from a Victron MPPT is specifically endorsed by Victron. Not directly in the specs, but by various reference drawings and employee comments, such as these (note, the drawing has the BMV tripping the BP in the MPPT output): https://communityarchive.victronener...y-protect.html https://communityarchive.victronener...ng-diagram.png



The fail-on configuration of SmartSolar below 150/70 size is a big concern. In theory, I could design a system that barring an anomaly would never require the BMS to trip a SmartSolar. However, that same logic would suggest that a BMS contactor (or using a BMV, or any other manner to trip a LFP off line on an errant cell) isn't needed either. Allowing the BMS to trip charge sources including solar (reliably!) is how you can absolutely rule out a HC Cutout -- if the BMS is aware enough to do a HVC, it would have already cut off all charging.