question

Michael Riley avatar image

VE.Bus System [261] - #2

I have the same/similar error as with Harvey-ricky at

https://community.victronenergy.com/questions/6045/vebus-system-261-on-vrm.html and as answered by @WKirby.

System: 2 x 24 V 200 AH LFP, Venus, Ve.Bus BMS, BP, 24/5000 Multi, 150/70 MPPT +..(on a sailing catamaran).

At around 0800 (AEST) I noticed a low voltage alarm on the digital remote control (and on the Multi). Logged into local.venus Battery SoC 74% and same warning. BlueTooth to both batteries with the following result

With the above battery cells voltage results I decided to let the discharge continue to my floor of 50% SoC. The system is relatively new (few weeks) and I am still trying to test the various limits etc. of the system. Also trying to utilise the battery bank between 95%-50% SoC. At around 09:36 both Battery Protects disconnected and the Multi, Venus and BP went down - ship went dark so as to speak. Very lucky we were not underway.

Connected AC at 09:40ish and started charging. Logged into VRM.

On VRM found the above info so went looking on the Community and found the Harvey-ricky entry. Then went looking for my Device ID and found that in my system the 24/5000 Mulit is device 261.

Ignore the first three (3) alarms as these are valid - at that point I disarmed and disabled the data and alarm monitoring (around middle of March).

So it appears that the BMS is working; the SoC is seriously wrong, the Battery cell voltages are ??, and perhaps some assistants are not showing up or loaded. I was there when the AC Detect assistant was loaded in the installer workshop but have left a voicemail for him to call.

Can anyone shed further light on this error and why only now (battery SoC and other conditions have been lower in the past weeks and there was no heavy load (in fact almost no load and the 24/5000 was switched off over night).

VRMVE.Bus
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.

9 Answers
ben avatar image

Your SOC meter looks completely messed up to me. Before you go into all the other issues and questions, you need to get it set up correctly. (It will never be completely accurate, but it will be close enough for the goals you've described.)

At 3.04V per cell, with no load on the battery, your cells are at somewhere around 3% SOC, a.k.a. completely discharged.

At 0.25C charging, when your pack is at 26.1V, your cells are probably at about 3.26V, and that corresponds to about 8% SOC, a.k.a. still almost empty.

When you figure out why your SOC meter is so far off, other stuff will come into clarity.

I sympathize with your observation that distributors have trouble keeping up with all the tech. But yours is a very basic, glaring issue that has little to do with fancy software. You need to find someone who understands LFP batteries and get some help.

3 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.

@ben As with others - big thanks for putting in the time to respond. I agree with all you say except the last bit - this is ALL to do with 'fancy software' as SoC is a calculation not a measurement and that is in software - Surely?

The technology is ALL Victron - except the wiring and some fuses and solar panels. The 'lithium' chemistry and BMS seems to be working - why do you refer to getting someone who understands LFP? Do you have a specific thinking?

Thanks for the metrics - will work on this however noting that these are 'results' (i.e. the voltage is now XX, the current is now YY) and I think, as you eloquently put it ' the SOC is completely messed up - because of some 'input' (not result) and unless I have a 'leak' that is being missed by the shunt then it points towards the result being 'posted' by the Shunt.

Note for clarity - there is NOTHING on the Lynx Power In except the batteries and nothing connected directly to the batteries other than the Lynx. the Lynx Shunt is immediately after the Power In and then the Lynx Distributor. From this (in parallel) the MultiPlus, MPPT (see earlier post about where to connect this), VE.Bus BMS and 2 x Battery Protects. In this configuration I believe that when the BMS signals a 'low voltage' shutdown it turns off the Multi, the MPPT and BP's. There is absolutely no load on the battery(s) other than the BMS (<10mA). So I think that the 'input' to the Lynx SoC calculation is OK (i.e. physical leak or other 'load' is not it. Therefore the SOC moves to the 'calculation' (current in, current out referenced to 'its zero'). Other than getting into the 'software calculation error' (which would show up on everyone else's installs) then this is where I am at. See below as this relates to Boekel comments also.

Would appreciate your thinking/feedback on the above.

@ben

It is me who needs to understand Lithium and look at the results right in front of me (aka that BlueTooth Battery figures above).

Completely missed that; to my excuse I had a yacht that was fast becoming out of control... no excuse I need to pay attention to the numbers I am being presented with - I am building 'trust' in the new LiPo install and this is not without its issue.

Thank you for the numbers.

Yes, I think you're on the right track. Those cell-level values show us the ultimate truth about the battery. You need to compensate for any current into or out of them, but you can look up curves showing SOC versus {voltage, current} for various currents as a percentage of pack capacity.

Once you know how your battery chemistry works and what's really going on in there, then you can start tracing your way up the chain to figure out why the SOC is so far off. Maybe a shunt is mis-wired, or maybe something is simply misconfigured. Victron systems can have multiple SOCs coming from different estimators, some of which are less accurate than others. The problem could indeed simply be in the software.

Your displayed SOC is so far off that it should be "easy" to sleuth out what is going on, because it's egregiously wrong.

Good luck!

Michael Riley avatar image

@Boekel, Thank you for the response. For anyone else reading this entry this is what we now understand (6 days later); SUMMARY

a. First up - my fault in inadvertently letting the battery voltage get critical. As a newly installed system this was a test that I was intending to do at some stage, but anyone using LiFePO4 really need to understand the battery voltage(s)/cell. As stated above - this is one of the only 'truths'.

b. the VE.Bus error is 99% likely because the MultiPlus shut down. Will need to look into the sequencing as to what shut down first (if possible) as I would have understood that the VE.Bus BMS initiated this (correctly protecting the battery) but should have also shut down the Venus at the same time - so how did it record the error?

c. Assistants - only one needed. The LiFePO4 assistant - for the functions of just 'running' the batteries. I will also load the Start/Stop Generator assistant now as this functionality can be used have a more sophisticated battery charging arrangement when on shore power/MPPT etc. (Tks Boekel for that on another post).

d. SOC. Whilst there are many who say you should manage LiFePO4 on the cell voltage [only] - this is not practical in my opinion for our operation. I need something that has more predictive capability. In addition PROVIDED you can get the 'ZERO Current' accurate then the SOC should be more accurate. Given that there is still a 'discharge' across the shunt, which is what one is 'calibrating', by the very equipment that are 'setting the zero-point' then the SoC will always be a little optimistic. That is to say that the battery actual SOC will be lower than the measured one - for any one cycle - being until the next 100% resync and you should do this, say, monthly to balance the cells in any case. I disconnected every load that was practical to do so and then pressed the space bar (in the Venus display) as well as using the mouse to press the button. No visual or aural response noticed. As of this time no-one from Victron has deemed to respond to the above query about button press 'feedback', which in my opinion is important to have.

Now monitoring battery cell voltage regularly; SOC seems great but have not had the same load on the system as previously and the solar is topping up early - have lost faith in the system a little and can re-iterate that there are 'gaps' in the information available (just look at the posts on this site) for any contemporary configuration. Thanks to all above who responded.

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.

Michael Riley avatar image

@Boekel. Thanks very much for your time to respond. The 'reply' button was not visible to your post so I respond here. I have separated your comments and itemised - trust this is acceptable.

[Boekel] You don't seem to have the VE.bus assistant installed in the Multi, that's one of the errors.

[Riley] - incorrect. See attached. Assistant is loaded.

For a system with Lynx Shunt, Mutiplus, MPPT, VE.Bus BMS, and Venus with DVCC enabled:

  1. Q1 How many assistants in the Multi should there be? If more than the VE.Bus assistant what is the title as seen in VE.Configure?
  2. Q2. Is there anything in DVCC that may add to this problem. I note a few comments ( incl from Victron) about the solar currents and voltage. What are your thoughts in respect of the DVCC settings. Can you give me a brief list.

[Boekel] What is set in VRM as 'battery monitor'?

[Riley]. The Lynx Shunt. See attached. I am not sure why you refer to the VRM site. Let me explain - to be specific there are 'results' I see on VRM which are total wrong (i.e. my generator production hours showing up as 'solar' generation) so there can be a difference between VRM and 'reality'. The 'Battery Monitor' logo on VRM also shows up as a BMV?

This is also an area with a few options. See two screen shots below from my Venus - Settings - System Setup - Battery Monitor / sub options are: Automatic, MultiPlus, No battery monitor, Lynx Shunt VE.can on VE.can. I have selected the last option - the Lynx Shunt. This is shown on the screen prior and on VRM.

[Boekel] Comment "About SOC: remember that at 24V your batteries are almost completely empty ('running on fumes')".

[Riley] Understood. The System Overview - DC System graph is included above. For brevity I have expanded this to the last few days and included below. In any case the lowest 'voltage' I have seen is 25.21 V and this graph covers the period of the 'black out'. Why do you mention 24Volts? Am I missing your point?

[ADDENDUM] Went looking for the reason for Boekel's comments re 24V. I now see that the Bluetooth login result for both batteries shows the 24 V reference and have added this addendum. This value was missed by me and it does not show up anywhere else. Is this another issue or a clue?

Note: On VRM I still do not see any 'low voltage' alarms in the 'Battery History' so I do not entirely rely on VRM as a source of facts.

[Boekel] Comment "Also I wouldn't worry too much about the amount of SOC you use with LiFePO4 batteries, I'd use 80% of SOC without any worries, they have a very different lifespan compared to 'regular' Li-ion batteries (those last longest when not fully charged for example)"

[Riley]. A little confusing. My interpretation. You use from 100% down to 20% SOC (i.e. you use 100-20 = 80% of bank capacity) without worries.

Happy that you are not worried but what is your expected life cycle value. i.e. are you looking for 5,000 cycles? Are you saying that the Victron LiFePo4 (new format 24 v 200 AH) is different from 'other batteries'. I was using the Victron datasheet for these batteries to get my 'cycle' information.

What do yo recommend for my batteries as the SoC range you would use if you had them (or do you - have the same?).

I also am 'happy' to use a significant amount of capacity - but this is becoming the key aspect of this post I believe - i.e. we are both referring to SoC and this is now what appears to be the problem I need to resolve. Meaning getting a more accurate reading. Given your incorrect assumption about the assistant above and the reference to the SoC therefore being wrong - what would be your next suggestion be as to the cause. I believe it is something to do with the 'Zero current' button.

On this issue.

  1. How do you get 'zero' current when the device you are using to set it is part of a system that is consuming electricity and which includes (to the best of my knowledge); Lynx Shunt, BMS, 2 x BP and Venus still as 'consumers'. I am reasonably confident that I can remove all other 'loads' being: Multi (turned off using Digital Remote connected to BMS and switch on front of Multi also off), MPPT and panels (isolated), and Loads (DC Load isolators turned off). Note that the Multi and MPPT are still physical wired to the Lynx Distributor.
  2. Do I need to 'remove fuses' in the Distributor to get this achieved?
    <Note in my past I was working as a test engineer for Hughes satellite power supplies at one time (Morebank - Sydney-AU) and we were measuring very small amounts in clean rooms etc. As the head of Philips Test and Measurement supplying equipment to the Hughes Laboratory , I am used to seeing '0' current as 0 not something close by - but I acknowledge that this is a different application> I believe the above questions will assist me in getting some context in this application.

Once again thank you for your time so far.


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.

I'll try to make it short :)

[Boekel] You don't seem to have the VE.bus assistant installed in the Multi, that's one of the errors.

[Riley] - incorrect. See attached. Assistant is loaded.
Ok, than the errors might have come up due to the Multi being shut down?

  1. Q1 How many assistants in the Multi should there be? If more than the VE.Bus assistant what is the title as seen in VE.Configure?
  2. Q2. Is there anything in DVCC that may add to this problem. I note a few comments ( incl from Victron) about the solar currents and voltage. What are your thoughts in respect of the DVCC settings. Can you give me a brief list.

I think DVCC should work just fine here, and no further assistants needed (except if you want to change other behavior)


[Boekel] What is set in VRM as 'battery monitor'?

[Riley]. The Lynx Shunt.

Ok, than this is where the SOC comes from and this is where the settings are incorrect: giving an incorrect SOC.
(and my apologies: I should have typed 'Venus' where I typed 'VRM)


on SOC:

Riley]. A little confusing. My interpretation. You use from 100% down to 20% SOC (i.e. you use 100-20 = 80% of bank capacity) without worries.

I think Victron is a little conservative on the cycle life, it is common to use LFP li-ion batteries for deeper discharge. But: if you calculate 50% DOD for 5000 cycles vs 20% DOD - 2500 cycles, you don't gain much on total used capacity: you can use almost twice as much (or need half the batteries) and it lasts half as long.

If you really use them for (an average of) 50% every day, you still get 2500 cycles out of them, that's almost 7 years of continuous cycling. I'd say in 7 years there are better / cheaper batteries available so that's why I don't worry about using them to 80% (most cycles will be shallower)

So the issue with this system seems to be an incorrect programming of the battery monitor (Lynx shunt)

1554966759203.png (256.5 KiB)
Boekel avatar image

Hi @Michael Riley

sorry I can't comment on all issues at once, but some observations:

You don't seem to have the VE.bus assistant installed in the Multi, that's one of the errors.

This can also explain the incorrect SOC, what is set in VRM as 'battery monitor'?

About SOC: remember that at 24V your batteries are almost completely empty ('running on fumes')

Also I wouldn't worry too much about the amount of SOC you use with LiFePO4 batteries, I'd use 80% of SOC without any worries, they have a very different lifespan compared to 'regular' Li-ion batteries (those last longest when not fully charged for example)

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.

Michael Riley avatar image

With respect to the Venus.local and Lynx Shunt - Battery settings as shown below.

Can someone give guidance as Sync SOC. When the charger has been on float for how many minutes/hours do you consider that LiPo are 100%?

How often would you 'sync' the SOC?

Reference 'Calibrate Zero current. I have switched off all loads except the VE.Bus BMS and Venus. I press the 'Press to set to 0' button and also on the keypad I press the 'space' bar. There does not appear to be any reaction; no resetting or other indication.

Can someone advise if there is any 'feedback'?

How often would you do this action?

Can it be confirmed that if this is done in error (i.e. pressed when indeed there is a current (load) that this will affect the SoC calculation?



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.

I would consider the battery 100% when the battery switches to Float or some time after provided a charge current is still maintaining the Float state.
I like to check periodically that my switch to Fload and my 100% charged event occur at roughly the same time. If things start to skew due to a string of cloudy days and partial charges then I'll perform a manual sync. Consult you VRM graphs periodically to ensure everything is running in synchronicity.

Calibrating zero current when current is indeed flowing will certainly affect SOC calculations.
Example: If 1A is flowing out of the battery and you tell the system to ignore this current by telling it that zero current is flowing, then the SOC will remain at 100% whilst 1A continues to flow out of the battery - discharging it by 24Ah (AKA 6% for a 400Ah battery) every 24 hours.
It gets worse, the battery monitor might measure 10A of charge current but in fact only 9A is actually going into the battery because 1A of discharge has been discounted. The SOC will hit 100% prematurely and continue to get worse with each cycle. Indeed, the battery monitor could stay pegged at 100% and the battery keeps charging by the sun, but if you decide to stop a generator based on the false 100% figure then you are undercharged.

Michael Riley avatar image

With respect to our 2 x 200 AH (total 400 AH) 24 V LiFePO4 batteries using a Multi 24/5000 with a 4.5 KVA generator to charge them; can someone suggest what settings they would use in their system.

Can someone also advise what they understand is meant by 'charged voltage' @ 28.4 v in this instance. (Lynx Shunt 1000 settings).

With respect to the following Venus settings - is there anything that stands out as wrong or needs attention?

Thanks in advance.


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.

Michael Riley avatar image

@Paul B, thanks for the lesson in Ohms Law and it is refreshing to see the 'hose pipe' example used to explain this. I think I first heard that at university (Canterbury Engineering School - Christchurch) in 1975 when doing my airforce degree (day 1, lesson one in DC Theory I recall).

I understand all you have said - that being said I do not necessarily agree with the interpretation of the simplicity.

Note the first reply exceeded the 2000 characters limit so this is an abridged version. Excuse the brevity.

Point 1a.

extract from Victron documentation. Note comments in pink are my PDF edits.

How do you double the life of your battery if you cannot rely on this SoC. Victron themselves refer to the SOC (Depth of Discharge) in their guidance on the 'Cycle Life'. From an operational perspective it is almost impossible to 'manage' [DC] systems on a variable [voltage] that changes extremely quickly, as you correctly say. SoC is at least manageable, if accurate.

How do you manage your longevity (battery cycles)?

Point 1b. Reference LiPo charge voltage.

You appear to be unhappy with 28.6 V; see Victron battery documentation. At 28.6 V this is within the tolerance given by Victron).

Why do you suggest 28.2-28.4 V?

By being 'too high' what do you understand I am doing or missing?

Point 1c. By charging 'fully' are you actually stressing the chemistry and reducing battery life. What is your understanding? Perhaps you can advise me what settings you have on the 'Charger' page (Ve.Configure) of your Multi?

Point 2. For clarity I am trying to get a clear (absolute if possible) understanding of how our system will perform under stress because we are on a blue-water sailboat and I must be able to understand how the system will perform in the worst of weather (stormy and raining for days) - which leads directly to high use of autopilot hydraulic pumps and sail winches combined with little solar and absolute necessity of keeping the navigation equipment up and running.

I have regularly charged the batteries so with the batteries at >90% for long periods including 100% for numerous hours on numerous days as shown by the VRM output below;

Further I had previously seen the SoC down to 6% as per following screen shot on 16 March.

Assuming for a moment that it is not just the 'get your battery fully charged' scenario); What goes into the 'setup' that makes the SoC (sample of two occasions) so variable?

[Note. What is not shown is that shortly after the 6% the BMS also gave a low voltage shutdown (sample 1) - a deliberate system test to see if the VE.Bus BMS was actually configured and wired correctly].

Point 3. Given the above history of charge state combined with previous SoC levels - Why would I not be able to rely on it now (ie. the difference between 74% and 6% is a significant 'discrepancy' and clearly at 6% (an actual reading) they system was 'working'.

I am at a loss to understand what is actually different (i.e. no deliberate setting changes). Further VRM reports (nil) battery low voltage alarms???

Point 4. Your suggestion about another 'installer' is based on what? I am happy to provide any necessary detail as to the configuration - perhaps also your assumption is that this is the 1st installer - (not correct); and that he is on his own (also not correct); ALL are Victron 'resellers' by the way - hence this post.


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.

Michael Riley avatar image

@Guy Stewart (Victron Energy Staff); @WKirby; @Boekel; @Margreet Leeftink (Victron Energy Staff)

Hi Victron Staff, There is a sense of increasing concern and even borderline disillusionment on many of the sailing vessels that I am in close contact with and many of them have Victron equipment on board. This includes sailing cats (catamarans) from countries as far away as South Africa.

That is not to say that Victron equipment is necessarily at fault but as my reply above indicates there are a number of us who are having a range of 'difficulties' getting (or keeping) our systems operational and this in part relates to getting 'definitive' answers in respect of the software configuration aspects. I understand the variations can make this 'complex'. I am a little stressed as this is not a motor-home or other installation where I can simply park-up or ring and have someone visit to get a solution; our vessel safety and navigation systems entirely rely on DC power and indeed our lives can be in jeopardy if loss of DC occurs at an inopportune time. I tried writing up this query but this site now seems to have 'lost' its carriage return capability and I can only write on one line (string). Can someone contact me so that I can give a detailed query; In particular I need to know what the 'Muti' settings, Lynx Shunt 1000 settings; Venus with DVCC enabled settings should be for our system. Finally I have spoken to dozens of boat owners and they (we) almost all relay on State of Charge as a basis for making decision on charging (generator usually) and load management (discretionary dump loads) etc. and wish to operate our battery on the basis of 'extending the life' (for LFP); I need some guidance from Victron as to WHY my system when DOWN with 74% SoC? Thanks.

2 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.

@Guy Stewart (Victron Energy Staff), @WKirby, @Boekel, @Margreet Leeftink (Victron Energy Staff)

The answer to the above - is simply - Because I let the batteries deplete to the point where the BMS stopped the discharge and protected the battery(s).

Correct: the displayed SOC was off, so you didn't know the batteries were empty.

One of the good things about a BMS: it really safes your batteries :)

Paul B avatar image

Best that you contact your Installer and he will explain it all to you and answer your questions, as he has setup the system for you.


2 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.

Paul B,.... how do I respond? I know my installer very well AND I am a qualified avionics engineer with a design engineering background. My installer; and many of the 'experts' offering advice are WAY BEHIND the knowledge curve when it comes to Victron and its complexities. They may know 'electricity' but these 'blue-boxes' are now extremely complex systems which are outstripping most of them; at least here in QLD (Australia).

What is it you think he would 'explain'!!!!! I have a operational vessel which needs its DC to run and be safe (navigation equipment, bilge pumps, autopilot, sail winches etc.... ad infinitum).


To me off the cuff your batteries are flat and or are not being refilled fully.

it does not matter what your SOC says the voltage is always correct, work out why your SOC metering is wrong and correct it

so get your Batteries upto 100% by voltage 28.2 is full and 24 is flat

SOC is just a flow meter if its not set correctly then the flow through will not reed correctly and thus it will drift out of balance to the real charge level

Voltage is Pressure

Amps is Flow


Note that with Lithium they will basically charge at 26.1 for a LONG time and then raise quite quickly to the set level of 28.6, (Which I feel is to high but its your system 28.2 to 28.4)

and the other suggestion is to maybe contract in another installer to have a look, a remote overview is always possible service@octopusasia.com