Cheap LiFePO4 BMS?

[Dr Bob]

13 minutes ago, ivan&alice said:

Ah yes, sorry Bob I remember now. When the Li's are disconnected you run your domestics from LA.

 

Yes, the alternator side is definitely the tricky bit. @Richard10002 what are you using to disconnect your alternator? I'm going to try to live domestically on solar alone for the first few months while I figure out how to tackle that monster. But leaning towards a Tyco relay, opened on high voltage by the GWL BMS, that will isolate the alternator/lead acid/engine side from the solar/lithium/domestic side.

Ivan, the GWL BMS looks very interesting (I think Tom said so as well). Report back on how it goes. The fact it works with the Tyco is a big plus and negates the price of the BEP if you are trying to save money.

Dunno what I would do without my BMV 712 though. I use a mobile phone to view the volts as I trundle along so always know where I am wrt to SoC of the Li's. It will all depend how anal you are as to keeping a track of what is going on. I am!!!?

[jetzi]

13 minutes ago, Dr Bob said:

Dunno what I would do without my BMV 712 though. I use a mobile phone to view the volts as I trundle along so always know where I am wrt to SoC of the Li's. It will all depend how anal you are as to keeping a track of what is going on. I am!!!?

Yeah perhaps I need one of these too. Going to see what I can get out of the GWL BMS before I drop the 170 GBP+ on the BMV 712! The main advantage seems to be to give you an easy-to-understand "fuel gauge" so you know your SoC rather than volts and amps.

[Tom and Bex]

18 hours ago, ivan&alice said:

Am I to understand that you're able to modify the voltage output of your alternator?

. 

Only using the Arduino based alternator controller that Dr Bob alluded to earlier. This controls all alternator charging of the lithiums including switching to float when fully charged (based on voltage and tail current). Only available commercially now at c£500, but as originally open source, design and code still available with some searching (pm if interested). Not given much thought to how I'd do alternator charging now, but probably be based on a lower (or adjustable) voltage regulator if available, or simply on cut charge when set voltage is reached using something like the Amazon voltage board linked to. Nick's design looks promising as well

 

18 hours ago, ivan&alice said:

4 in series = 800A? (to get up to 12V - with each 3.2V cell being 3 cells in parallel?)

In series the current stays the same! 3 cells in parallel would give up to 600A @3.2v, 4 of those in series would give up to 600A at 12v (nominal)

 

18 hours ago, ivan&alice said:

If I got the 3kVA inverter, that would be 2400W using Peter's 80% rule and 240A using Nick's 10V rule for losses. I can check with Jeremy if he means 200A per cell or per battery.

Jeremy would bre talking about 200A per cell. EV use demands very high peak discharge rates, and the problem with some cells as they age is the voltage drops too much under heavy loads >200A. If used at much lower charge and discharge rates voltage drop is negligible, and capacity remains close to badge capacity. Jeremy said my cells had exactly the same issue. We've had no problem with them in use, and have used a measured 280Ah which is not bad from 160ah cells in 2p4s configuration. Probably could have got more if pushed to the limit of charge and discharge, bit don't see the point of pushing boundaries for the sake of it. 

 

18 hours ago, ivan&alice said:

 

What would happen if I tried to pull more than 200A out of the batteries, would something bad happen? Would the inverter just dip the voltage too low and crash the system?

As above,  cell voltage would dip below bms cut off and (hopefully!) bms would trigger the low voltage relay.

18 hours ago, ivan&alice said:

 

Oh dear. Monitoring is an aspect I seem to have overlooked!! I thought I'd be able to integrate it with the Victron monitoring (MPPT and inverter) somehow, but I haven't given proper thought to this. The instructions of the BMS say:

Most of that nothing to me. I'm a software engineer by trade what I might do in the medium term is write my own interface and run it on a Raspberry Pi or something. Do you have any recommendations on monitoring?

I think most on here are using the Victron BMV for monitoring.  I know @Dr Bob does, although he doesn't seem to think the SOC is very accurate. We use it, and with the settings I've used the soc seems to remain fairly accurate until we re-sync it around every 3 months or so. Maybe slightly pessimistic, but not enough to cause problems over a 3 month period, and would rather that than give opportunistic readings.

 

18 hours ago, ivan&alice said:

I've sited the washing machine next to the cauliflower for this reason - well insulated pipes will feed the machine with hot water as needed. Washing machine has a "cold wash" button that completely disables the heater rather than simply setting to 30 degrees. The mixer tap next to the machine will then be like the "temperature dial" on your fully automatic. And will keep the same temperature for the whole cycle - rinses as well as washes. Old fashioned I know but I'm a big believer in hot water for cleaning everything.

On our machine you can also set to cold on the temperature dial,  which performs no heating at all. We have the mixer set to hot for the wash cycle, but cold for the rinse to reduce creases. 

 

18 hours ago, ivan&alice said:

I'm excited to report back, so I'll definitely supply a full review in a couple of months! The system voltage cutoff idea is really just a belt and braces thing, but yes I can just rely on the BMS. I do plan on getting four really cheap low voltage buzzers to put on each cell just in case the BMS doesn't work. Would love to have high-voltage buzzers too, but couldn't find those with a cursory look on Amazon.

Personally I think the bms will be more reliable than the low voltage buzzers,  I've seen very mixed reviews about their reliability and accuracy. If you want additional safeguard personally I'd recommend the ISDT bc-8s if you can still get it. Basic display, but reliable and simple to use, with individual cell voltages, and (loud!) customisable high and low cell voltage alarm. This has been supercded with the bg-8s which despite bigger full colour display and more features, doesn't seem to be as suitable (think @Dr Bob has both - maybe he can advise?).

18 hours ago, ivan&alice said:

 

Not just for the SoC monitor, but initially I want to rebalance my cells manually every 2-3 months just to see how they are working. Whether I choose to top balance or bottom balance, I still plan to check the other (top or bottom) to help inform whether 80/20 is a reasonable rule in my case. If I top balance, I think it would make sense to follow something like 85/20 or 90/25 and if I bottom balance 80/15 or 75/10. Initially though I will be extremely conservative - 70/30 rather than 80/20!

Top balancing is definitely the recommended way for domestic use. Not possible to fo both. No need to be too conservative. Wer usually run ours between 20% and 90%, with occasional charging to 100%. Generally if cruising we seem to charge to around 95%, general opinion seems to be charging to 100% and then using them is not really an issue, it's the storing at 100% that causes long term deterioration. Problem comes with determining when 100% is reached, and ensuring all cells reach it at the same time without one of them going into the upper voltage knee (hence top balancing). Bottom balancing tends to find favour in ev use due to the high currents used, and thr need to ensure all cells are performing equally when drawing high current at low soc, thus ensuring maximum range is achievable. Ideally when supplying domestic loads, you'll not be pushing the bottom limit!

 

18 hours ago, ivan&alice said:

So yes - no memory effect for me, even if it is a real thing.

Still not convinced on this memory effect. I've read anecdotal reports of it, but no explanation for it, and not seem any actual data to back it up. 

 

18 hours ago, ivan&alice said:

Can anyone think what else I might want to power one day off that inverter that I should think about? How much power does an electric welding machine use?

I think you'd really struggle to run any welder from the typical inverter found on boats. 

 

1 hour ago, ivan&alice said:

Sorry I may have confused things by starting to talk about Richards system. Mine is going to be quite different. For one, I don't have a BMV. I'll only set high and low voltage cutoffs on my BMS and hook them up to 2 separate Tyco relays. Does mean I will have no idea how many ants are in my batteries.

As above, you'll find the BMV very useful tool. We use ours purely for monitoring, and the bms for providing the high and low cell voltage protection. The fact your bms provides support for the Tyco relays directly is a definite bonus!

 

1 hour ago, ivan&alice said:

I mentioned that I'm planning not to connect the alternator at first, so on the charge side, I would use the MPPT's voltage cutoff to switch to float (i.e. off). This would happen on high voltage OR on low temperature. On the load side I'd use the GWL BMS board hooked up to a Tyco 190A relay.

I'd still recommend using the bms to provide a high voltage cut as an emergency back up in case of fault. How does the MPPT work to provide a low temp cut? Make sure you're not confusing voltage temperature compensation for low temp protection, LA's need higher voltage at low temperatures - the exact opposite of what you need for lithiums! t's only charging to be concerned about low temp, but if inside like ours are, I don't think it very likely anyway. 

 

1 hour ago, ivan&alice said:

When I connect the alternator, I will parallelise the LAs and use a second Tyco relay to isolate the Li's. The GWL BMS can connect up to 3 latching relays! The third opens when 0.3V higher or lower than the min/max values you set - I'll use this for an alarm I think. But yes, the charge-isolating relay will isolate the Li's from the LAs.

You shouldn't really rely on the bms for everyday charge control. Bms should be last ditch protection if the other systems fail! You could use the voltage board, linked to a relay for providing charge control. If the relay was only powered when ignition on then no problem using normal relay as no ongoing power drain. Alternatively use a relay to cut the alternator regulator directly, (very easy to do with a127 alternator) then no need for lead acid batteries at all!

 

1 hour ago, ivan&alice said:

I know, but that's not what we were asking. I wanted to know if Richard could drive a load-disconnect relay to replace his Victron Battery Protect that's between his battery and inverter - can you connect a load-disconnect relay to the Valence internal BMS to trigger on low voltage? I'm guessing not. But @Dr Bob does mention that you can get "a cheap Amazon voltage board (£10 ish) for the low voltage disconnect on the load side."

 

I found this: https://www.amazon.co.uk/KKmoon-Voltage-Detection-Charging-Discharge/dp/B019YWUI5G but it doesn't look like you can connect a latching relay, only a powered one. Any idea what to search for?

That's the type of board I've referred to above. Very difficult to find any devices giving direct control of Tyco relay, you'd need an additional control circuit @MoominPapa posted about a simple one he designed I think. 

 

1 hour ago, ivan&alice said:

Ah - I interpreted what you said as a BEP switch controlled by the BMV on the charging side and a Tyco controlled by a voltage board on the load side. What do you use to switch the voltage on the load side, if not a latching relay?

I use a BEP motorised switch on the load side controlled by the bms, and a Tyco relay in the charge side, again controlled by the bms,  but via a control circuit as my bms does not support direct control of the Tyco relays. 

 

1 hour ago, Dr Bob said:

No, on my system the BEP disconnect isolates the Li's the LAs so stops both charging and loads. I have only one disconnect. That is the simple way of doing it and it has worked great for 12 months. Tom has both the BEP and a tyco i.e. 2 disconnects, one for charging and one for loads. I think he uses a BMV on the BEP to do the loads but for charging he has an alternator controller that does the latching tyco relay. That alternator controller is now not available - it's gone commercial and costs £500.

My bms controls both. The BMV is used purely for monitoring, and the alternator controller is purely for controlling alternator charging, including controlling split charge relay for connecting both alternators to lithium bank during bulk charging. 

 

38 minutes ago, ivan&alice said:

 

Yes, the alternator side is definitely the tricky bit. @Richard10002 what are you using to disconnect your alternator? I'm going to try to live domestically on solar alone for the first few months while I figure out how to tackle that monster. But leaning towards a Tyco relay, opened on high voltage by the GWL BMS, that will isolate the alternator/lead acid/engine side from the solar/lithium/domestic side.

As mentioned above, don't recommend using bms as main charging control. Leaning towards voltage control board for cutting charging at set voltage point, say 14.0v. This would probably result in charge terminating at around 80-90%, but would depend on alternator output, and wouldn't be exactly the same each time. Might require experimenting to find best voltage point for your system, and probably an override to enable manual (carefully monitored!) charging to 100% on occasions. Alternatively the BMV could be used to control a relay based on soc, so cutting charging when 100% is reached based on voltage and tail current settings in BMV. Actually, thinking about it, thiis might be the best option for alternator charge control, particularly if used with a lower voltage regulator. 

[cuthound]

On 02/04/2020 at 22:20, peterboat said:

No Nick its not confusion its practical. I was repairing generators when I was 17 they were rated in KVA but we always used .8 for conversion to watts load, its like the industry standard and Victron have done just that. I spoke to John V after I posted and he as an electrician that owned his own company. he said the same, his last big job was installing two 700 KVA and one 900 KVA  gennies and he  used the same conversion system, theory is one thing but in practice its less 400 watts in this case.

 

You are correct that the industry standard for generators is at 0.8 lagging power factor.

 

The limiting factor with a generator set is the power of the engine (kWe). The alternator, being an electrical device could produce more current with a unity power factor, but the engine isn't capable of turning the alternator with that much load on it continuously. 

 

I often specified generators at 0.9 pf, and in every case, the generator set came with a larger engine to meet the additional load.

 

Generator sets are usually specified as either "continuous" (100% rated load @ 0.8pf continuously, with 1 hour in 12 at 110% @ 0.8 pf) or "fuel stop" (up to 125% rated load for a maximum annual running time of 250 hours per year).

[jetzi]

12 minutes ago, Tom and Bex said:

Only using the Arduino based alternator controller that Dr Bob alluded to earlier. This controls all alternator charging of the lithiums including switching to float when fully charged (based on voltage and tail current).

This does sound ideal, and I'll get in touch when I am ready to think about alternator charging, thanks!

 

13 minutes ago, Tom and Bex said:

In series the current stays the same! 3 cells in parallel would give up to 600A @3.2v, 4 of those in series would give up to 600A at 12v (nominal)

Makes sense! You said series in your original post, but yes each cell would be made up of 3 units in parallel, giving 600A. Either way it's plenty.

 

14 minutes ago, Tom and Bex said:

I think most on here are using the Victron BMV for monitoring.  I know @Dr Bob does, although he doesn't seem to think the SOC is very accurate. We use it, and with the settings I've used the soc seems to remain fairly accurate until we re-sync it around every 3 months or so. Maybe slightly pessimistic, but not enough to cause problems over a 3 month period, and would rather that than give opportunistic readings.

OK. Does the Victron BMV integrate with the MPPT and inverter? I guess I will need something like this eventually. At 170+ GBP though it really seems like a hell of a lot for what it is - especially if I am already measuring the needed inputs with the BMS anyway. Smile and pay?

 

21 minutes ago, Tom and Bex said:

Personally I think the bms will be more reliable than the low voltage buzzers,  I've seen very mixed reviews about their reliability and accuracy. If you want additional safeguard personally I'd recommend the ISDT bc-8s if you can still get it. Basic display, but reliable and simple to use, with individual cell voltages, and (loud!) customisable high and low cell voltage alarm. This has been supercded with the bg-8s which despite bigger full colour display and more features, doesn't seem to be as suitable (think @Dr Bob has both - maybe he can advise?).

Thanks for the advice! I just looked for the cheapest most basic one I could find, since it's a backup. Cursory look on Amazon and eBay says you're right and only the BG-8S is available. I'll get one of these instead - @Dr Bob would you recommend?

 

24 minutes ago, Tom and Bex said:

going into the upper voltage knee

I've seen this term all through the thread but I don't know what it means... the "voltage knee"?

 

28 minutes ago, Tom and Bex said:

How does the MPPT work to provide a low temp cut? Make sure you're not confusing voltage temperature compensation for low temp protection, LA's need higher voltage at low temperatures - the exact opposite of what you need for lithiums!

According to the Victron MPPT SmartSolar manual:

 

"Low temperature cut-off: This setting is available when a VE.Smart network is set-up and the battery temperature is available (e.g. by for example using a Smart Battery Sense or a Smart BMV with the optional temperature sensor). It can be used to disable charging at low temperatures as required by Lithium batteries. For Lithium Iron Phosphate batteries this setting is preset at 5 degrees Celsius, for the other battery types it is disabled. When creating a user defined battery the cut-off temperature level can be adjusted manually."

 

Will use with the Victron SmartSense temperature sensor on the battery.

 

31 minutes ago, Tom and Bex said:

You shouldn't really rely on the bms for everyday charge control. Bms should be last ditch protection if the other systems fail! You could use the voltage board, linked to a relay for providing charge control.

Now I'm really confused!!
 

I thought the whole point of the BMS was to control when charging and load was allowed!! Hence why it is so useful to have one that supports the Tyco latching/bistable relays. The GWL BMS actually supports three relays - overvoltage disconnecting charge (battery full), undervoltage disconnecting load (battery empty) and emergency (over or under +-0.3V), which could either drive an alarm or - perhaps - a third Tyco relay isolating both charge and load from the negative side??

 

Having a separate voltage board was supposed to be the "emergency" backup, just sounding alarms?

 

What should you use to control charging, if not the BMS - can't use those cheap voltage board because they don't support latching relays?

(This is maybe specific to the alternator - for everyday charge control from solar, can use the high voltage from the MPPT).

 

Also you say you use "a Tyco relay in the charge side, again controlled by the bms,  but via a control circuit as my bms does not support direct control of the Tyco relays.  " - which sounds like it contradicts not using the BMS for charge control?

 

I just thought I had all my understanding ducks in a row and now I feel quite lost again, please help!

[peterboat]

5 minutes ago, ivan&alice said:

This does sound ideal, and I'll get in touch when I am ready to think about alternator charging, thanks!

 

Makes sense! You said series in your original post, but yes each cell would be made up of 3 units in parallel, giving 600A. Either way it's plenty.

 

OK. Does the Victron BMV integrate with the MPPT and inverter? I guess I will need something like this eventually. At 170+ GBP though it really seems like a hell of a lot for what it is - especially if I am already measuring the needed inputs with the BMS anyway. Smile and pay?

 

Thanks for the advice! I just looked for the cheapest most basic one I could find, since it's a backup. Cursory look on Amazon and eBay says you're right and only the BG-8S is available. I'll get one of these instead - @Dr Bob would you recommend?

 

I've seen this term all through the thread but I don't know what it means... the "voltage knee"?

 

According to the Victron MPPT SmartSolar manual:

 

"Low temperature cut-off: This setting is available when a VE.Smart network is set-up and the battery temperature is available (e.g. by for example using a Smart Battery Sense or a Smart BMV with the optional temperature sensor). It can be used to disable charging at low temperatures as required by Lithium batteries. For Lithium Iron Phosphate batteries this setting is preset at 5 degrees Celsius, for the other battery types it is disabled. When creating a user defined battery the cut-off temperature level can be adjusted manually."

 

Will use with the Victron SmartSense temperature sensor on the battery.

 

Now I'm really confused!!
 

I thought the whole point of the BMS was to control when charging and load was allowed!! Hence why it is so useful to have one that supports the Tyco latching/bistable relays. The GWL BMS actually supports three relays - overvoltage disconnecting charge (battery full), undervoltage disconnecting load (battery empty) and emergency (over or under +-0.3V), which could either drive an alarm or - perhaps - a third Tyco relay isolating both charge and load from the negative side??

 

Having a separate voltage board was supposed to be the "emergency" backup, just sounding alarms?

 

What should you use to control charging, if not the BMS - can't use those cheap voltage board because they don't support latching relays?

(This is maybe specific to the alternator - for everyday charge control from solar, can use the high voltage from the MPPT).

 

Also you say you use "a Tyco relay in the charge side, again controlled by the bms,  but via a control circuit as my bms does not support direct control of the Tyco relays.  " - which sounds like it contradicts not using the BMS for charge control?

 

I just thought I had all my understanding ducks in a row and now I feel quite lost again, please help!

How about just forget about the charging issues and fit one of these comes in other sizes https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=10&cad=rja&uact=8&ved=2ahUKEwiWlr-X6s7oAhXzrnEKHe2UDPYQFjAJegQIAhAB&url=https%3A%2F%2Fwww.12voltplanet.co.uk%2Fsterling-pro-batt-ultra-waterproof-battery-to-battery-charger-12v12v-120a.html&usg=AOvVaw27R-xCc90KTxyABvnpuult

[peterboat]

35 minutes ago, cuthound said:

 

You are correct that the industry standard for generators is at 0.8 lagging power factor.

 

The limiting factor with a generator set is the power of the engine (kWe). The alternator, being an electrical device could produce more current with a unity power factor, but the engine isn't capable of turning the alternator with that much load on it continuously. 

 

I often specified generators at 0.9 pf, and in every case, the generator set came with a larger engine to meet the additional load.

 

Generator sets are usually specified as either "continuous" (100% rated load @ 0.8pf continuously, with 1 hour in 12 at 110% @ 0.8 pf) or "fuel stop" (up to 125% rated load for a maximum annual running time of 250 hours per year).

On exercise we knew how much power each vehicle consumed so we placed them so that we never overloaded the gennies, Signal regiments have sophisticated repair equipment in box bodies, they are heated and cooled plus air filtered all driven by electric, and then their is the test and repair kit onboard to account for, war isnt just fought by soldiers in a trench anymore. They have large UPS with inverters on board but thats only for a short time on the filter equipment, heating and cooling its basically for power, lights and repair

equipment.

They worked on an over pressure, woe betide anybody that forgot to put the door chain on it could knock the visitor off the steps !!

[cuthound]

23 hours ago, WotEver said:

And this is why electricity companies would love to charge you by kVAh instead of kWh, but they’re not allowed

 

Or even kVArh ?

Edited April 4, 2020 by cuthound

[Dr Bob]

40 minutes ago, ivan&alice said:

 

Thanks for the advice! I just looked for the cheapest most basic one I could find, since it's a backup. Cursory look on Amazon and eBay says you're right and only the BG-8S is available. I'll get one of these instead - @Dr Bob would you recommend?

 

 

I think the BG-8S is great. The numbers on the display are bigger than the Bc-8s so easier to read. I got a BG8s first and started to use it - then discovered it didnt have the high voltage alarm (IIRC it has the low voltage alarm only). I therefore found a BC-8s and bought one but never replaced the BG8S with it as the display is smaller. I would defo recommend the BG8S as the way to see all the cell voltages in one  simple display.

[Dr Bob]

43 minutes ago, ivan&alice said:

 

Now I'm really confused!!
 

I thought the whole point of the BMS was to control when charging and load was allowed!! Hence why it is so useful to have one that supports the Tyco latching/bistable relays. The GWL BMS actually supports three relays - overvoltage disconnecting charge (battery full), undervoltage disconnecting load (battery empty) and emergency (over or under +-0.3V), which could either drive an alarm or - perhaps - a third Tyco relay isolating both charge and load from the negative side??

 

 

Tom and I both think the same way here.

You control the charge sources so they back off charging when the batteries get full...ie the solar goes into a lower voltage float mode or the alternator controller turns down the voltage. The BMS is then only an emergency shut down and triggers the autodisconect in an emergency. This means the Tyco or BEP switch is not activating on a daily basis but the solar and alternator look after themselves. This is the way my system has worked for 12 months. My BEP switch has never worked in anger....other than my testing of it. I think the same is true of Toms system.

I dont think Peter ever takes his system to the limits that his BMS shuts the systme down.

I think MP runs his system differently and routinely runs it so the tyco relay activates to shut down the Li's when charging to full. That could me the tyco relay switching every day. I would prefer to have a system where the charge sources themselves did the shut down and then the BMS is only for emergency use. Its up to the user. There is no right or wrong. I did a comprehensive risk review on mine (ie what happens if something goes wrong.......and work through all the eventualities and what the response of the system could be).

[jetzi]

OK I think I understand - we're back to the three options that I summarised a few pages back.

  

On 02/04/2020 at 14:38, ivan&alice said:

• Parallel LA with the LiFePOs as a load dump to protect the alternator from rapid disconnection. No need for split charge relay - the LiFePOs will only charge once the LAs are full due to their higher nominal voltage. At 80% SoC, the voltage monitor sends a pulse to an (expensive) battery isolation switch or high-current latching relay, which isolates the lithium portion of the bank from the alternator. The spike from disconnecting the Lithium load is dumped into the fully charged LAs. This has the advantage of using your existing LAs (but doesn't sound particularly good for the alternator to me?)
• Modify the alternator to have an externally switchable field. At 80% SoC, the voltage monitor turns off the field, shutting down the output more gracefully. This involves dismantling the alternator to access the field wire. Would suggest a backup abrupt overvoltage cutoff like (1), so perhaps a LA load dump is still a good idea.
• Control the voltage output of your alternator, so that it behaves like a smart charge controller. At 80% SoC, the alternator is switched to a lower "float" voltage (i.e., charging off). With this you could also monitor alternator temperature and shut off charging if the alternator gets too hot.

@Tom and Bex and @Dr Bob are advocating options 2 or 3. I agree that controlling the charging source is probably the correct way to do it, but those things are much harder to do than option 1.

 

I won't connect the alternator initially until I have a better idea of how it works.

 

Please could I ask you guys to have a look at this circuit diagram and see if I've overlooked anything?

[WotEver]

59 minutes ago, ivan&alice said:

The spike from disconnecting the Lithium load is dumped into the fully charged LAs. This has the advantage of using your existing LAs (but doesn't sound particularly good for the alternator to me?)

Because the LAs are connected there is no spike. It therefore doesn’t stress the alternator. 

[Dr Bob]

1 hour ago, ivan&alice said:

OK I think I understand - we're back to the three options that I summarised a few pages back.

  

@Tom and Bex and @Dr Bob are advocating options 2 or 3. I agree that controlling the charging source is probably the correct way to do it, but those things are much harder to do than option 1.

 

 

 

...but option 1 is the back up if 2 or 3 go wrong.

There is another way ....if you have a big enough alternator then you could use a B to B charger between the LAs and the Li's.

We are in the process of selling our current boat and buying a new (longer) one with a 240A alternator. I am current torn between using Nick's solution (but that means taking the back off the alternator) or using a Sterling 120A BtoB which would restrict the charge to the Li's to 120A so the alternator wont overheat.

[jetzi]

1 minute ago, Dr Bob said:

if you have a big enough alternator then you could use a B to B charger between the LAs and the Li's.

My alternator is only 90A or less. I have measured it putting out a maximum of 50A with everything on a very flat LA.

[Dr Bob]

1 minute ago, ivan&alice said:

My alternator is only 90A or less. I have measured it putting out a maximum of 50A with everything on a very flat LA.

Ok, forget the BtoB method then!

[Tom and Bex]

3 hours ago, ivan&alice said:

This does sound ideal, and I'll get in touch when I am ready to think about alternator charging, thanks!

Controller I use now only available for around £500 I'm afraid. 

 

Quote

You said series in your original post, 

Whoops my mistake! Surprised no-one else picked up on that!

 

Quote

OK. Does the Victron BMV integrate with the MPPT and inverter? I guess I will need something like this eventually. At 170+ GBP though it really seems like a hell of a lot for what it is - especially if I am already measuring the needed inputs with the BMS anyway. Smile and pay?

I think it integrates with you're other victron stuff, but as it's the only Victron item I own I can't confirm!

 

Quote

I've seen this term all through the thread but I don't know what it means... the "voltage knee"?

Look at the graph below. You can clearly see the voltage is only rising very slowly until right at the end of charge, when it climbs very quickly. The point it starts to climb is called the knee. This is why balancing is important. Imagine you're charging at 14.0v. If not balanced, then one cell could easily climb to 4.1v, whilst the others are still at 3.3v. That's why  monitoring cell voltages is so important (as your bms does) and why just monitoring total voltage is not really enough. Avoiding the knee on both charge and discharge, reduces this risk considerably. 

Quote

According to the Victron MPPT SmartSolar manual:

 

"Low temperature cut-off: This setting is available when a VE.Smart network is set-up and the battery temperature is available (e.g. by for example using a Smart Battery Sense or a Smart BMV with the optional temperature sensor). It can be used to disable charging at low temperatures as required by Lithium batteries. For Lithium Iron Phosphate batteries this setting is preset at 5 degrees Celsius, for the other battery types it is disabled. When creating a user defined battery the cut-off temperature level can be adjusted manually."

 

Will use with the Victron SmartSense temperature sensor on the battery.

That sounds fine. Not being familiar with Victron kit I wasn't aware their MPPT controllers had that option. Just a word of warning - be very careful of default lifepo4 settings. Most of the default charge voltage settings I've seen are much to high. 

Quote

Now I'm really confused!!
 

I thought the whole point of the BMS was to control when charging and load was allowed!! Hence why it is so useful to have one that supports the Tyco latching/bistable relays. The GWL BMS actually supports three relays - overvoltage disconnecting charge (battery full), undervoltage disconnecting load (battery empty) and emergency (over or under +-0.3V), which could either drive an alarm or - perhaps - a third Tyco relay isolating both charge and load from the negative side??

 

Having a separate voltage board was supposed to be the "emergency" backup, just sounding alarms?

 

What should you use to control charging, if not the BMS - can't use those cheap voltage board because they don't support latching relays?

(This is maybe specific to the alternator - for everyday charge control from solar, can use the high voltage from the MPPT).

As you say, the MPPT can have custom settings such that the bms never has to activate the emergency disconnect. If relying on the bms then you'll be pushing the cells into their upper limit each time you charge. Charge sources should be set not to overcharge under normal operating conditions. The bms high voltage cut is to prevent an overcharge from occurring if the charge source regulation fails (which can happen - particularly with alternators).

 

Not something that needs sorting immediately, but need to have an idea how you plan to manage alternator charging going forward. 

 

No need for voltage board to control latching relay, keep Tyco relay for bms emergency control, and use normal 180A relay for alternator control, linked to voltage board. If voltage board connected to lead acid side, then once set voltage point is reached, it will cut parallel relay and voltage will rise until engine switched off again. 

 

Quote

Also you say you use "a Tyco relay in the charge side, again controlled by the bms,  but via a control circuit as my bms does not support direct control of the Tyco relays.  " - which sounds like it contradicts not using the BMS for charge control?

 

I just thought I had all my understanding ducks in a row and now I feel quite lost again, please help!

As above, Tyco relay is in charge side, but never activates in normal use as not a control relay, only there to prevent damage if charge regulation fails. 

 

3 hours ago, peterboat said:

How about just forget about the charging issues and fit one of these comes in other sizes https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=10&cad=rja&uact=8&ved=2ahUKEwiWlr-X6s7oAhXzrnEKHe2UDPYQFjAJegQIAhAB&url=https%3A%2F%2Fwww.12voltplanet.co.uk%2Fsterling-pro-batt-ultra-waterproof-battery-to-battery-charger-12v12v-120a.html&usg=AOvVaw27R-xCc90KTxyABvnpuult

That's one option. Ot a standalone regulator (disabling internal alternator regulator). Not difficult on a127 and probably fairly easy on others, just need access to one of the brushes and to disconnect/disable internal regulator. 

 

1 hour ago, ivan&alice said:

OK I think I understand - we're back to the three options that I summarised a few pages back.

  

1 hour ago, ivan&alice said:

• Parallel LA with the LiFePOs as a load dump to protect the alternator from rapid disconnection. No need for split charge relay - the LiFePOs will only charge once the LAs are full due to their higher nominal voltage. At 80% SoC, the voltage monitor sends a pulse to an (expensive) battery isolation switch or high-current latching relay, which isolates the lithium portion of the bank from the alternator. The spike from disconnecting the Lithium load is dumped into the fully charged LAs. This has the advantage of using your existing LAs (but doesn't sound particularly good for the alternator to me?)

 

No problem at all for alternator. Also no need for expensive latching relay or motorised switch, normal split charge relay more than adequate.

 

1 hour ago, ivan&alice said:

1 hour ago, ivan&alice said:

• Modify the alternator to have an externally switchable field. At 80% SoC, the voltage monitor turns off the field, shutting down the output more gracefully. This involves dismantling the alternator to access the field wire. Would suggest a backup abrupt overvoltage cutoff like (1), so perhaps a LA load dump is still a good idea.

Probably not as difficult as you think to modify alternator to switch regulator off (depending on alternator type). Definitely the more elegant way rather than using lead acid dump load.

 

Your bms would then provide for the emergency backup disconnect. Yes of it activated you risk alternator damage, but by it's nature, something would have already failed to get to that point anyway

. 

1 hour ago, ivan&alice said:

1 hour ago, ivan&alice said:

• Control the voltage output of your alternator, so that it behaves like a smart charge controller. At 80% SoC, the alternator is switched to a lower "float" voltage (i.e., charging off). With this you could also monitor alternator temperature and shut off charging if the alternator gets too hot.

You're looking at modifying the alternator to use external regulator again by this point. 

 

1 hour ago, ivan&alice said:

 

 

@Tom and Bex and @Dr Bob are advocating options 2 or 3. I agree that controlling the charging source is probably the correct way to do it, but those things are much harder to do than option 1.

 

I won't connect the alternator initially until I have a better idea of how it works.

You know it makes sense!

 

 

[Tom and Bex]

16 minutes ago, Dr Bob said:

...but option 1 is the back up if 2 or 3 go wrong.

There is another way ....if you have a big enough alternator then you could use a B to B charger between the LAs and the Li's.

We are in the process of selling our current boat and buying a new (longer) one with a 240A alternator. I am current torn between using Nick's solution (but that means taking the back off the alternator) or using a Sterling 120A BtoB which would restrict the charge to the Li's to 120A so the alternator wont overheat.

Nick's option is much a more elegant solution in my mind. Sounds similar to what our alternator controller does, but more integrated with the rest of his system. 

 

13 minutes ago, ivan&alice said:

My alternator is only 90A or less. I have measured it putting out a maximum of 50A with everything on a very flat LA.

We have twin 70A alternators. With very flat lead acid we never saw more than 50A. With the lithiums that could easily be 100A (and burnt out alternator in a few months use!).

[Dr Bob]

2 hours ago, ivan&alice said:

OK I think I understand - we're back to the three options that I summarised a few pages back.

  

@Tom and Bex and @Dr Bob are advocating options 2 or 3. I agree that controlling the charging source is probably the correct way to do it, but those things are much harder to do than option 1.

 

I won't connect the alternator initially until I have a better idea of how it works.

 

Please could I ask you guys to have a look at this circuit diagram and see if I've overlooked anything?

On the diagram.

It is not in great detail ie no fuses, BMV shunts etc.

The thing I would be worried about is control of the alternator. So, you are using option 1 to isolate the alternator on high SoC so that sorts the overcharging, but you will have an issue with the LI's demanding 60A+ from you alternator which will fry it. You may need to put a temp sensor from the BMS on the alternator to cut power when it gets to 90C. YOu could use a BMV to do this but better alternator controll is the answer. I use a Sterling AtoB to control my alternator. Best to not wire the alternator in and just try it.

Edited April 4, 2020 by Dr Bob

[nicknorman]

I’ve got all the individual bits of my regulator working as far as I can do without actually testing it on an alternator (but the actual regulating bit should just “work” because it’s handled by the chip). So being able to set the regulated voltage, set the field current limit, read the rpm and actual field current, measure voltage, measure remote temperature (alternator temperature), receive CANBUS data from the Mastervolt system, control the alternator warning light etc. Next step is to create the “state machine” ie what behaviour actually arises from the different sensors etc. 

 

But it occurs to me that a simpler and intermediate step might be to have a digital input that simply selects between two regulated voltage levels, eg 14.2v and 13.4v, the latter being a “float” voltage that would allow the alternator to continue to supply boat loads without further charging the batteries. You folks using BMVs could use one of its alarm outputs to switch over to the float voltage at the set SoC or cell voltage alarm.

[jetzi]

14 minutes ago, Dr Bob said:

It is not in great detail ie no fuses, BMV shunts etc.

I don't have a BMV, so no shunts.

 

But yes I'll add a 200A fuse directly between the positive terminal of the battery and the ammeters, should have put that in.

 

Is there anything else I'm missing?

 

 

22 minutes ago, Tom and Bex said:

If relying on the bms then you'll be pushing the cells into their upper limit each time you charge.

The BMS voltages are configurable and I would ideally set them to be voltages representing 20% SoC and 80% SoC. Although there are only four options - 2.8V, 2.9V, 3.0V, 3.1V for the lower bound and 3.5V, 3.6V, 3.7V, 3.8V for the upper. There is then an "emergency" that will trigger 0.3V above/below what you set for the max/min. That could be used to isolate both charge sources and load at once. So using the BMS doesn't have to imply pushing the limits of the batteries?

 

22 minutes ago, Tom and Bex said:

No problem at all for alternator. Also no need for expensive latching relay or motorised switch, normal split charge relay more than adequate.

Option 1 seems like the simplest and cheapest option then. Modifying the alternator seems more elegant. But why complicate things if there's no downside? As I understand it the Li's are parallelised with the LAs, no split charge is needed since the LAs run at a lower voltage. It's OK (beneficial actually) for the LA to be taken to 100% while the Li's are still charging.

 

14 minutes ago, Tom and Bex said:

We have twin 70A alternators. With very flat lead acid we never saw more than 50A. With the lithiums that could easily be 100A (and burnt out alternator in a few months use!).

Is this the downside? I would have to tackle the problem of the overheating alternator somehow, and a simple dump load won't help with that. Does anyone here actually use the dump load in routine use, or is it only in the case that the charging is shut off in an emergency?

 

 

[peterboat]

17 minutes ago, ivan&alice said:

I don't have a BMV, so no shunts.

 

But yes I'll add a 200A fuse directly between the positive terminal of the battery and the ammeters, should have put that in.

 

Is there anything else I'm missing?

 

 

The BMS voltages are configurable and I would ideally set them to be voltages representing 20% SoC and 80% SoC. Although there are only four options - 2.8V, 2.9V, 3.0V, 3.1V for the lower bound and 3.5V, 3.6V, 3.7V, 3.8V for the upper. There is then an "emergency" that will trigger 0.3V above/below what you set for the max/min. That could be used to isolate both charge sources and load at once. So using the BMS doesn't have to imply pushing the limits of the batteries?

 

Option 1 seems like the simplest and cheapest option then. Modifying the alternator seems more elegant. But why complicate things if there's no downside? As I understand it the Li's are parallelised with the LAs, no split charge is needed since the LAs run at a lower voltage. It's OK (beneficial actually) for the LA to be taken to 100% while the Li's are still charging.

 

Is this the downside? I would have to tackle the problem of the overheating alternator somehow, and a simple dump load won't help with that. Does anyone here actually use the dump load in routine use, or is it only in the case that the charging is shut off in an emergency?

 

 

What you have to remember is whilst you have a smallish alternator your solar will be running as well cruising on a sunny day could mean upwards of a 100 amps going in to the LifePo4s which they will handle with ease and be charged in a record time. On Bobs boat he has fitted a bilge fan to blow cold air at the alternator [I use the same to force cold air through the electric motor as the manual fan does nothing at low RPM] so factor one of these into the equation

[Dr Bob]

18 minutes ago, ivan&alice said:

 

Option 1 seems like the simplest and cheapest option then. Modifying the alternator seems more elegant. But why complicate things if there's no downside?

For me there is a downside.

I have 5 levels of safety on my system for most of the troubling stuff. So for instance take over charging. I manage mine in 5 steps

1) the charge sources cut back when the SoC is high enough.

if that goes wrong then

2) Alarms sound on the BMV to show over voltage

If that fails - or I dont hear it

3) the auto disconnect then disconnects at a higher voltage from the BMV

if that fails then

4) the BMS sounds an alarm for cell over voltage

If that fails  - or I dont hear it

5) the BMS activates the auto disconnect based on high cell voltage.

Mine has never gone wrong at level 1.

 

If you rely just on the BMS for auto disconnect then you are significantly reducing the safety net and you are reliant on the tyco switch to work all the time - and not just in emergency.

 

25 minutes ago, ivan&alice said:

 I would have to tackle the problem of the overheating alternator somehow, and a simple dump load won't help with that. Does anyone here actually use the dump load in routine use, or is it only in the case that the charging is shut off in an emergency?

 

 

I think @MoominPapa uses the dump load in routine use. I will start using the dump load this way on my new boat until I work out the best way to do it as I am doing it now.

[Col_T]

Quote

I don't have a BMV, so no shunts.

 

But yes I'll add a 200A fuse directly between the positive terminal of the battery and the ammeters, should have put that in.

 

Don't all ammeters have shunts?

 

[MoominPapa]

1 hour ago, Dr Bob said:

I think @MoominPapa uses the dump load in routine use. I will start using the dump load this way on my new boat until I work out the best way to do it as I am doing it now.

Indeed he does.

 

MP.

 

[nicknorman]

21 minutes ago, Col_T said:

 

Don't all ammeters have shunts?

 

No, moving iron ones pass all the current themselves. Generally not a particularly good idea for a boat with a large alternator/inverter etc.