Victron Cerbo GX (RPi) HELP...

[Raul]

Out of curiosity, what firmware you running on RPI? I'm running 3.33 and I heard that later version is not stable for PI.

 

[AdrianChen]

So it's wired correctly then, thats good.

So here's where I am, the system has been installed for 3 months and has been working perfectly, I have been monitoring it (locally via victron connect bluetooth) when we have been away, both on EHU and Off-Grid.

The lowest SOC I have seen was 85% after a long night off-grid with everything lit up like an Xmas tree, the telly going and the inverter charging my E-bike (highest draw was around 14Amps). Battrry was back to 100% within a few hours of the sun coming up the next day (via 435w panel & Victron MPPT).

When on sites with EHU, the battery never fell below 100%, as the IP22 was covering the load.

When parked up on our drive it is always on EHU and the IP22 eventually drops into "storage mode" where the battery sits at 13.6v, with the battery master happily keeping the starter battery circa 0.5v below the leisure battery.

I then decided I wanted to have a play with RPi GX and Venus, all went well. Put the RPi in M/H and connected it to VRM via the M/H wifi router. Left it sitting disconnected from any VE hardware while I waited for the VE Direct to USB cables to arrive. The draw shown on the shunt locally via Connect was less than 0.5A with the Pi and router being the only things running and the IP22/MPPT were both in storage mode, just maintaining the battery at 100%.

VE Direct Cables arrived yesterday so I thought I would start with the shunt. Connected that to the RPi, was pleased that it showed up on the touch screen & via VRM, but suddenly I was now seeing a 5amp draw and the SOC dropping (even with the IP22 running).

Left it overnight last night, I now have an SOC of 79% and the 5Amp draw is still there (still on EHU).

It is as if connecting the GX has knocked out the IP22 and MPPT and created a 5amp draw from somewhere (which is very odd).

Am going to disconnect the Rpi GX this AM and revert to monitoring Connect locally to see if that then gets me back to where I was previously (inc the IP22 and MPPT charging the battery again).

Does seem very odd - interesting to try the VE Direct cable plugged into the orion XS and see if you get similar results, but disconnectings a good start and re-booting the Pi?

 

[Two on Tour]

Have just triple checked against VE instructions and it's the right way round. Which would make sense as its been fine the last 3 months, happily reporting via Connect app & Bluetooth.

This only started when I plugged the shunt into GX via VE Direct cable.

Are you using the correct USB to RS232 TTL UART cable, as only the PL2303HX type will work ?

 

[CueBall]

Are you using the correct USB to RS232 TTL UART cable, as only the PL2303HX type will work ?

I bought these: https://www.amazon.co.uk/dp/B0B2RRMG9T?tag=mhf04-21

And then these: https://thepihut.com/products/jst-ph-4-pin-to-male-header-cable-i2c-stemma-cable-200mm

Then pinned them together according to a couple of You-Tube videos (as although the cable colours match, the pin order on the JST PH doesn't).

I got my TX & RX around the wrong way the first time, which resulted in no connection, once that was rectified the shunt showed up on the GX.

 

[Two on Tour]

I bought these: https://www.amazon.co.uk/dp/B0B2RRMG9T?tag=mhf04-21

And then these: https://thepihut.com/products/jst-ph-4-pin-to-male-header-cable-i2c-stemma-cable-200mm

Then pinned them together according to a couple of You-Tube videos (as although the cable colours match, the pin order on the JST PH doesn't).

I got my TX & RX around the wrong way the first time, which resulted in no connection, once that was rectified the shunt showed up on the GX.

Did you use all 4 wires or only the 3, not using the positive as the Pi supplies 5v as the Victron mppt's and shunt's run on 3.3v.

^{_{Subscribers  do not see these advertisements}}

 

[CueBall]

Did you use all 4 wires or only the 3, not using the positive as the Pi supplies 5v as the Victron mppt's and shunt's run on 3.3v.

All x4 wires are pinned together according to the video I watched.



Albeit his cable colours on the JST connector were different to mine that I got from the PiHut, I just crossed referenced the location on the JST connector itself.

 

[CueBall]

Just been out and turned off the 12v router & unplugged the Pi.

The current draw has dropped to 0.5A (which I know is the battery master, as I can now see the starter battery voltage increasing).

The only odd thing is that the IP22 isn't putting any current into the battery, not even covering the discharge load from the battery. It is in float mode (as is the MPPT as they are wirelessly networked together. The MPPT only seems to do anything when I unplug the IP22 from the mains, as if Victron decided mains charging should take priority over solar charging).

I've re-checked my IP22 settings and they are as they were previously (I think).

 

[Two on Tour]

All x4 wires are pinned together according to the video I watched.



Albeit his cable colours on the JST connector were different to mine that I got from the PiHut, I just crossed referenced the location on the JST connector itself.

I disagree with using the 5v positive, but it's up to you, but I would never put 5v into something that runs on 3.3v, I only use the 5v positive on the isolated Duppa Isolated 4 port USB to TTL UART FTDI, as it is needed for that.

You would better off following this video.

 

[AdrianChen]

Just been out and turned off the 12v router & unplugged the Pi.

The current draw has dropped to 0.5A (which I know is the battery master, as I can now see the starter battery voltage increasing).

The only odd thing is that the IP22 isn't putting any current into the battery, not even covering the discharge load from the battery. It is in float mode (as is the MPPT as they are wirelessly networked together. The MPPT only seems to do anything when I unplug the IP22 from the mains, as if Victron decided mains charging should take priority over solar charging).

I've re-checked my IP22 settings and they are as they were previously (I think).

View attachment 1059758

OK could you pop up the charger battery settings - voltage may not have dropped enough for it to start a new cycle.

If the IP22 goes to the same busbars surprised it is at a different voltage to the rest, are you back on the smart BT network?

 

[Scottish Scrutineer]

I have not used the red 5v feed. Also I had to change the wires over at the junction between the JST wires

^{_{Subscribers  do not see these advertisements}}

 

[Lenny HB]

Out of curiosity, what firmware you running on RPI? I'm running 3.33 and I heard that later version is not stable for PI.

I thought I had problems with 3.54 then found out it was user error now running 3.55 without any problems.

 

[CueBall]

OK could you pop up the charger battery settings - voltage may not have dropped enough for it to start a new cycle.

The battery is a 280Ah LiFePO4 from Eco-Worthy (which has been great so far, touch wood).

Here are the IP22 settings.

If the IP22 goes to the same busbars surprised it is at a different voltage to the rest, are you back on the smart BT network?

Yes, it's on the same busbars and everything is within 50cm of each other (apart from the MPPT, which is about 4m away from the bus bars).

 

[Two on Tour]

I have not used the red 5v feed. Also I had to change the wires over at the junction between the JST wires View attachment 1059760

View attachment 1059766

View attachment 1059762

It's quite easy to de-pin the JST connect by lifting the little plastic tab that locks the barb on the pin and pulling out the wire, if you like me and like colour to colour.

 

[CueBall]

I disagree with using the 5v positive, but it's up to you, but I would never put 5v into something that runs on 3.3v, I only use the 5v positive on the isolated Duppa Isolated 4 port USB to TTL UART FTDI, as it is needed for that.

Will give that a go, have just grabbed the cable from the M/H and will re-pin it as per that video (not using the 21V+ cable).

Hopefully I get a lower draw than 5A.

Still need to work out why my IP22 no longer covers any load and keep the battery at 100% as it used to though.

 

[AdrianChen]

The battery is a 280Ah LiFePO4 from Eco-Worthy (which has been great so far, touch wood).

Here are the IP22 settings.



Yes, it's on the same busbars and everything is within 50cm of each other (apart from the MPPT, which is about 4m away from the bus bars).

OK so if the battery is at 13.9v and float is set to 13.65v (quite high for LiFePo4) there won't be any charge, so seems right.

The SOC is 79% which if the battery is rested, albeit with 0.4a draw, and 13.9v I would think it is full and the SOC is out a bit, sorry to be a pain, do you have the shunt battery settings handy?

^{_{Subscribers  do not see these advertisements}}

 

[CueBall]

Have just re-pinned the cable as per video above, taking out the red 12V+ completely.

Powered up the Pi and am back to a circa 5A draw and discharging again.

 

[CueBall]

sorry to be a pain, do you have the shunt battery settings handy?

No problem at all, pleased to have the assistance from everyone...

 

[AdrianChen]

No problem at all, pleased to have the assistance from everyone...

View attachment 1059784

The only tweaks I would make are minor, not likely to make any difference to your issue though. Charged voltage normally set to 0.2v below the absorption figure in your chargers so 14.2 - and Peukert - but as I say wont make much difference.

Can you connect the VE/USB cable into the Solar or B2B to see what that gives?

 

[Raul]

Have just re-pinned the cable as per video above, taking out the red 12V+ completely.

Powered up the Pi and am back to a circa 5A draw and discharging again.

View attachment 1059782

Why the solar does not appear on the GX display?
Did you try my advice with the negatives of the shunt load side? You have to eliminate any hardware error one by one. Even the comms cables, swap them around and see if any difference. I remember you mentioned you got a victron VE-USB cable.

 

[CueBall]

Why the solar does not appear on the GX display?

That's because I ran a VE direct cable from the MPPT down into the main installation area (thinking I would buy a genuine Cerbo GX). But now decided to go the RPi route. I am just waiting on receiving a JST socket, so I can break out the VE direct cable from the MPPT to USB - https://thepihut.com/products/jst-ph-4-pin-socket-to-color-coded-cable-200mm

Can you connect the VE/USB cable into the Solar or B2B to see what that gives?

Have now plugged in the B2B plus the VE 12-1200 Inverter and now have the following on the portal.

Obviously the portal can't see an AC input as the IP22 doesn't have VE Direct.

What is interesting though is that the load has now dropped to 15W (sub 1A).

Battery voltage is now 13.65v, with an SOC of 77.3% (whereas Eco-Worthy suggest 100% SOC on a resting battery is 13.46v).

Perhaps my charger settings are wrong somewhere?

I have also just found some settings in the GX console that allows me to turn on/off solar support to maintain SOC, which would be nice - Obviously can't do that until I get the MPPT hooked up.

^{_{Subscribers  do not see these advertisements}}

 

[Lenny HB]

Perhaps my charger settings are wrong somewhere?

They are a bit out but it won't affect your problem.
I would sethetabsorbgion voltage to 14.2v and float & storage to 13.5v.
With the float at 13.5v the battery won't accept a charge unless there is a load drawing current.

 

[CueBall]

I would sethetabsorbgion voltage to 14.2v and float & storage to 13.5v.
With the float at 13.5v the battery won't accept a charge unless there is a load drawing current.

Ok, done that, interestingly the battery "time remaining" halved when I did that, from 6+ days to 3 days.

It seems the charger being in Float at 13.65v was holding the battery at that voltage.

By dropping it to 13.5v it allowed the battery to continue discharging again as the charger stopped sending any current.

Out of interest, should I re-set the SOC, if so, at what voltage and under what conditions should I do that?

 

[Lenny HB]

Shunt setting are OK you could make a couple of slight changes, the charged voltage to 14.1v & peurkant to 1.02.

 

[AdrianChen]

Ok, done that, interestingly the battery "time remaining" halved when I did that, from 6+ days to 3 days.

It seems the charger being in Float at 13.65v was holding the battery at that voltage.

By dropping it to 13.5v it allowed the battery to continue discharging again as the charger stopped sending any current.

Out of interest, should I re-set the SOC, if so, at what voltage and under what conditions should I do that?

I would reset the SOC on your shunt when the battery is at 100% for certain! IE let it charge right through the bulk stage, and when it clicks over to absorption, then you could reset to 100% That's really the only time you can be sure of the SOC....

 

[CueBall]

Having put the IP22 float & storage voltage down to 13.5v, the battery discharged to that voltage and is now sitting there. With the SOC at 74.2% (which is probably wrong and not sync'd properly to be fair).

Reading some Victron forums, they suggest trying the default Li-ion setting as it turns off float mode (as it isn't healthy for Li-ion batteries apparently).

That seems to be the issue I now have, it's now stuck holding the battery at the float voltage and won't go back to bulk/absorb mode.

^{_{Subscribers  do not see these advertisements}}

 

[CueBall]

That's interesting, I switched the IP22 to stock Li-ion mode and it immediately went into bulk mode for around 5 mins before dropping into absorb mode, not only that, the MPPT followed suit and started supplementing the IP22 charger with additional solar power (which I have never seen before and was beginning to wonder if it was even possible).

Battery now charging at almost 30A, with 20A from the IP22 and an additional 10A from the MPPT.

Now to see what happens when the battery is fully charged again.

 

[Lenny HB]

Reading some Victron forums, they suggest trying the default Li-ion setting as it turns off float mode (as it isn't healthy for Li-ion batteries apparently).

Victron default Li-ion is 13.5v float at which the battery shouldn't take a charge.

 

[AdrianChen]

Having put the IP22 float & storage voltage down to 13.5v, the battery discharged to that voltage and is now sitting there. With the SOC at 74.2% (which is probably wrong and not sync'd properly to be fair).

Reading some Victron forums, they suggest trying the default Li-ion setting as it turns off float mode (as it isn't healthy for Li-ion batteries apparently).

That seems to be the issue I now have, it's now stuck holding the battery at the float voltage and won't go back to bulk/absorb mode.

I use float, so it supplies the van - just my preference though - mine will only start a new cycle if switched off/on, or the voltage drops to the re-bulk setting. There must be some logic to that, but I'm happy that the battery is kept pretty full on EHU so in practice it's fine.

 

[AdrianChen]

That's interesting, I switched the IP22 to stock Li-ion mode and it immediately went into bulk mode for around 5 mins before dropping into absorb mode, not only that, the MPPT followed suit and started supplementing the IP22 charger with additional solar power (which I have never seen before and was beginning to wonder if it was even possible).

Battery now charging at almost 30A, with 20A from the IP22 and an additional 10A from the MPPT.

Now to see what happens when the battery is fully charged again.

I guess when you changed the profile, it restarted a new cycle so all well - you can reset the SOC when it gets to Absorption if it doesn't do it for some reason.

Have you set a max charge figure in the Pi (DVCC) if not all chargers should give what they can (up to the battery BMS max), including the B2B if you start the engine.

The Solar should do that if the profiles are set the same, and you have BT Smart networking on although you can I think give preference to Solar, but I don't use that on the basis that if I have EHU I've probably paid for it....

 

[CueBall]

I use float, so it supplies the van - just my preference though - mine will only start a new cycle if switched off/on, or the voltage drops to the re-bulk setting. There must be some logic to that, but I'm happy that the battery is kept pretty full on EHU so in practice it's fine.

Based on setting the float to 13.5v (where Eco-Worthy suggest 100% SOC is 13.46v) what would be the harm in resetting the shunt to 100% SOC when it goes into float mode and holds the voltage at 13.5v? Whereas people suggest re-setting the SOC at the higher Absorption voltage which would result in a drop in SOC % as soon as the charger drops down into float (or storage, when you have float turned off).

Have you set a max charge figure in the Pi (DVCC) if not all chargers should give what they can (up to the battery BMS max), including the B2B if you start the engine.

No, as the Victron 20A IP22 isn't seen by the Pi as it doesn't have VE direct. Should I still do that if that were the case and given the GX can see the B2B and the MPPT (or at least it will see the MPPT once I sort a cable).

Will likely figure out how to prioritise solar once I get the MPPT into the GX, as it's plugged into my house when we aren't using it, so best to take advantage of the sun rather than the price of electricity lol.

^{_{Subscribers  do not see these advertisements}}