Cheap LiFePO4 BMS?

[Dre]

Hi Dr Bob, 

I was gonna buy the Renogy 40a DC DC but it's out of stock at the moment (Simone from Renogy actually emailed me back to say they'll be in stock end of Feb). I would think the Sterling is a better option as it has a few more features. I don't think I would charge at more than 60a (and never more than half that bank rating) so I'm not worried about the balancing. I'd do an initial balance and monitor but don't expect to have problems with balancing at those charging or discharging C ratings. Moreover those batteries from that vendor seem to be very well matched. 

As to charging parameters I have seen reviews on the Sterling with Lifepo4s as being optimal in use. I'm also gonna monitor the charging while using the engine as I hope it won't be needed as much once I install a kW worth of solar. Also the bms is fully programmable and that will be a failsafe if the Sterling can't be optimally programmed. For the money I think either the Sterling or the Renogy are better and easier than anything else at charging lifepo4s via an alternator and they both have been used and reviewed extensively. 

As for splitting the battery bank I doubt it would make a difference as I'd have to balance 8 cells anyway as I'll be doing it manually and hopefully not for a long time after the first balance. I can parallel two battery banks with a bms each or parallel the two bms onto one battery but it's the same. Two bms will give me double the current regardless. 

I think there's a lot of technical and not so technical talk online about lifepo4 installations on the net but with the kit available nowadays and at those prices it all seem pretty straightforward and I don't expect any problems apart from maybe quality control issues from some of the cheap components. Lifepo4s are super safe and are successfully used in much more demanding applications, both electrically and mechanically speaking. I really don't think a few point volts difference in charging parameters will make much of a difference to the amount of cycles you get out of them and if I had to sacrifice a few hundred cycles of life or a few amps of capacity I am not too bothered as I can just add a few more cells at those prices.

I'll definitely keep you posted but I guess it might take a while still before I start building as I expect quite a delay for the batteries and bms from China. Poor souls have worse problems right now! 

 

https://battlebornbatteries.com/sterling-pro-batt-ultra-battery-battery-charger/

 

https://faroutride.com/b2b-review/

 

https://marinehowto.com/understanding-the-sterling-power-pro-batt-ultra-battery-to-battery-charger/

 

 

[Dre]

1 hour ago, peterboat said:

Very good glad to see someone else going the lifePo4 route 

Cheers ?

[nicknorman]

Whilst we’re on I’ll report my progress. Sorry if it’s not interesting and or too technical! I have the battery monitoring board working from the point of view that I can sample the cell voltages, receive data over CANbus from the Mastershunt and Combi, operate the real-time clock and show some stuff on a display. Drivers for the latching relays also work. Not sure I will keep the display because everything should be automatic and if I really want to know the cell voltages, I can always plug in a BG-8S. The display is just a drain on power, which I’ve been trying to keep to a minimum as this will be on 24/7.
 

When idling my circuit uses about 0.5mA, and I don’t see a point in sampling the cells very often. I think it only needs to be sampled once every several seconds, or less, unless charging is in progress  (when power consumption isn’t important) or perhaps when close to a critically low SoC. When the measuring chip fires up the current consumption goes up to about 10mA, but it only takes about 0.25seconds to read and process the data into floating point numbers. So the average power drain will only be marginally above 0.5mA.

 

Here’s a short video of it in action, measuring some AA LiFePO4 cells. There is a little blinking red LED on the board which is on when the measuring chip is working and off when it is sleeping. Current sample rate is every 2 seconds but I think it could be much slower, since SoC, temperature etc can’t change very quickly. Oh and as you can see, I made a major design cockup on the £2 PCB (red wires on the board) due to misreading the 800 page microcontroller datasheet. Never mind, new boards on order from China. £8 this time, as the introductory £6 discount was a one-off.

Quite pleased with the stability of the voltage measurement, the resolution is 1mV and that seems to be the maximum jitter, despite no signal averaging going on in the software.

 

Edited February 20, 2020 by nicknorman

[Dr Bob]

1 hour ago, Dre said:

I'll definitely keep you posted but I guess it might take a while still before I start building as I expect quite a delay for the batteries and bms from China. Poor souls have worse problems right now! 

 

 

 

 

DRE. All sounds good.

Just one other thought. The BtoB will likely charge the Li's at 60A - but I've not read reviews on that particular model - so it is going to pull 60A out of the alternator. How big is your alternator? My 90A alternator overheats ie gets to 100°C if I pull 60A continuously but is ok at around 45A. With bilge cooling it is better but 60A might still be pushing it. If your alternator is bigger than 90A you will probably be ok, but if not you may have problems with it overheating.

Best of luck and keep us all posted.

[Dr Bob]

Great progress Nick.

Looking forward for the first unit to test!!! Lots of options to control the 240A alternator I will end up with but at the moment a big BtoB or your bespoke system looks the best.

[nicknorman]

7 minutes ago, Dr Bob said:

Great progress Nick.

Looking forward for the first unit to test!!! Lots of options to control the 240A alternator I will end up with but at the moment a big BtoB or your bespoke system looks the best.

Hopefully I’ll send off for the alternator controller board soon. Having made a major cockup on the monitor board I thought I’d wait until I’d got the software to drive the various elements working properly, in case there were any other hidden issues! Of course there are currently issues in China with getting stuff made and delivered, but the factory is back at work albeit with reduced people (some are in quarantine) so I think it will be a bit slower.

 

Just following on from our private discussion, I should mention that any alternator controller, including mine, needs to be wired into the alternator internally. I recall you were worried about the guarantee on your new alternator. However with my Iskra, I’ve just bought a replacement brush/regulator module (about £20 from eBay) which I will hack in order to connect the regulator. Thus in extremis, the old brush/regulator module could be replaced leaving no trace of intervention. It’s just held in with 3 screws, so a quick change is easy (also have to remove the plastic cover, 2 nuts).

 

If you go for a B2B I think you will struggle to get one big enough to make good use of the alternator. And at nearly £500, even the 120A is very expensive!

[Dre]

1 hour ago, Dr Bob said:

DRE. All sounds good.

Just one other thought. The BtoB will likely charge the Li's at 60A - but I've not read reviews on that particular model - so it is going to pull 60A out of the alternator. How big is your alternator? My 90A alternator overheats ie gets to 100°C if I pull 60A continuously but is ok at around 45A. With bilge cooling it is better but 60A might still be pushing it. If your alternator is bigger than 90A you will probably be ok, but if not you may have problems with it overheating.

Best of luck and keep us all posted.

170A alternator I'm sure will be able to cope

[Dr Bob]

1 hour ago, nicknorman said:

 

 

Just following on from our private discussion, I should mention that any alternator controller, including mine, needs to be wired into the alternator internally. I recall you were worried about the guarantee on your new alternator. However with my Iskra, I’ve just bought a replacement brush/regulator module (about £20 from eBay) which I will hack in order to connect the regulator. Thus in extremis, the old brush/regulator module could be replaced leaving no trace of intervention. It’s just held in with 3 screws, so a quick change is easy (also have to remove the plastic cover, 2 nuts).

 

 

Yes I was looking at the Prestolite brush replacement stuff

http://www.prestolite.com/literature/tech/alts/TSB-1148_AVI160J_P_Reg_Brush_Change_Procedure.pdf

and it does just look like you could do it without trace of intervention. It just looks like a plastic box with the 2 brushes. Do you take the regulator out and just connect to the brushes (ie hack it as you say)? I guess that may be the way to go. Few months away yet from getting the new boat so will keep you posted.

 

3 minutes ago, Dre said:

170A alternator I'm sure will be able to cope

Oh, that should be fine. No probs of over heating.

[nicknorman]

42 minutes ago, Dr Bob said:

Yes I was looking at the Prestolite brush replacement stuff

http://www.prestolite.com/literature/tech/alts/TSB-1148_AVI160J_P_Reg_Brush_Change_Procedure.pdf

and it does just look like you could do it without trace of intervention. It just looks like a plastic box with the 2 brushes. Do you take the regulator out and just connect to the brushes (ie hack it as you say)? I guess that may be the way to go. 

Yes my device is high side switched so one brush to ground, the other to the device. I think most integral regulators are low side switched, ie one brush to +ve one to the regulator. But anyway, the only issue is that ordinary soldering with lead solder melts at about 185C and alternator internal gubbins can get hotter than that on a bad day, so silver solder is probably a good idea.

[nicknorman]

Tried my monitor on the boat today, connected to the Masterbus (CANbus). Works really well and I can access any parameter from the Mastershunt or the Combi. Display (which is really just a development tool) here showing some parameters from the Mastershunt. I must say, now I’ve got it working, I really like CANbus - seems very robust- and with all the hard work done by the dedicated hardware within the microcontroller, the software is very easy, once set up properly.

 

Cell voltages are of course from my AA sized LiFePO4 batteries, which is why they don’t sum to the Mastershunt voltage (ie LA house batteries).

 

PCBs for monitor mark 2 (without the cockup!) and the alternator controller and now winging their way back from China, eta 2 weeks or so.

[Dr Bob]

3 minutes ago, nicknorman said:

Tried my monitor on the boat today, connected to the Masterbus (CANbus). Works really well and I can access any parameter from the Mastershunt or the Combi. Display (which is really just a development tool) here showing some parameters from the Mastershunt. I must say, now I’ve got it working, I really like CANbus - seems very robust- and with all the hard work done by the dedicated hardware within the microcontroller, the software is very easy, once set up properly.

 

Cell voltages are of course from my AA sized LiFePO4 batteries, which is why they don’t sum to the Mastershunt voltage (ie LA house batteries).

 

PCBs for monitor mark 2 (without the cockup!) and the alternator controller and now winging their way back from China, eta 2 weeks or so.

Nick, will this work from a Victron Combi....or can it be connected to a victron combi/BMV 712?

[nicknorman]

2 minutes ago, Dr Bob said:

Nick, will this work from a Victron Combi....or can it be connected to a victron combi/BMV 712?

Not directly, since the Mastervolt and Victron protocols are both proprietary and different. Victron has a rather confusing array of different protocols depending on the specific device. see here: https://www.victronenergy.com/upload/documents/Whitepaper-Data-communication-with-Victron-Energy-products_EN.pdf

As far as I can see from a brief look, the 712 only has VE.Direct which is a TTL level RS232 connection. It doesn’t have a CANbus interface. I presume you have connected it to the Combi so presumably the Combi has a VE.direct port too? Of course “Victron Combi”is a bit vague since they have so many different models/generations. Some have VE.Bus which is RS485, some (later ones, probably) have VE.Can which is obviously CANBUS and compatible with NMEA2000 which is the latest generic comms protocol for boats. So a lot more confusing, varied and inconsistent than Mastervolt. On the plus side, they seem a bit more open about it than Mastervolt, who are very secretive (but I’ve hacked it!).

 

So in summary it could be made to work with your setup, but would require different software and some hardware modification to provide a TTL level RS232 connector. Anyway, it’s purpose is only to provide information to the alternator controller and to operate the emergency disconnect relay(s) and cell balancing FETs. I hope to start work on the alternator controller once the PCBs arrive in a couple of weeks.

[Dr Bob]

4 hours ago, nicknorman said:

 I hope to start work on the alternator controller once the PCBs arrive in a couple of weeks.

Ok, good stuff.

Awaiting progress with anticipation.

[nicknorman]

Got the alternator controller board back from Hong Kong, slightly delayed due to Coronavirus but not bad at around 3 weeks. Cost £8 for 5 bare boards. Now populated with the electronics and lots of connectors, RS232 on the near left, double CANBUS on the far left. There is an “emergency field disconnect” relay to the right of that. Main connector to alternator, ignition light etc out of sight at the back, and a 2.5mm jack plug for the alternator temperature sensor. Pads to connect to a SoC selector switch on the right (50%, 80% or 100%) and then the mini-usb for programming the microprocessor. Rather a large heatsink for the main regulator chip, I suspect it is far too big and can be much reduced but I didn’t really know how to calculate the heat dissipation so best to start “too big”. Board is about 70mm x 80mm. Push button to do “something” and a couple of pretty LEDs to show “something” (not yet fitted).

Now I “just” have to write the software to get it all working. If I can’t manage that, I am going to look a right plonker!

 

[jetzi]

OP here. Finally sat down and read this thread from start to end (or should I say cover to cover!). I didn't keep up past about page 6 and it was growing faster than I could read it. I've dished out some ✅'s and ?'s liberally. Treasure trove of information, absolute required reading for any LiFePO wannabe! There's a lot of debate on finer detail like charging regimens and cell balancing, though. While important, it can make lithium systems seem too complicated for a newbie.

 

Undervoltage is easy to understand, even for those with only LA experience. With a programmable shutoff that will cut the domestic supply at a voltage representing 20% or 30% SoC, and it seems hard to go wrong. Unless your lithium cells are horribly out of balance, which anecdotally seems to be quite rare if you start with balanced cells and maintain the 80/20 rule. Seems to me that the tricky bit with Lithiums is charging. I'd like to summarise some key observations from this thread, and ask that you please correct if I'm wrong!

 

Solar charging of LiFePOs is easy - our charge controllers already know how to switch to "float" when the battery reaches a configurable voltage. For LiFePOs, we set the float voltage trigger low enough such that charging is essentially "off" when batteries get to 80% state of charge. Job done.

 

Boaters use alternators to charge our batteries, which don't usually have a smart controller. We thusly use a battery voltage monitor (and/or a cell voltage monitor) to turn charging off when the Lithiums exceed a preset voltage representing 80% state of charge. Easy? Not quite:

• We can't abruptly turn off the alternator output under load, because the resulting surge can fry the alternator's rectifier diodes.
• The internal resistance of Lithium is so low, they can draw high currents even at low alternator RPM. Since alternators are cooled by an internal fan at the same RPM, they can overheat when charging Lithiums.

Solutions?

1. 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?)
2. 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.
3. 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.

I'm starting to buy the bits. But I'm being extra cautious and plan to keep my LAs and LiFePOs totally separate for now. I'll not touch the existing alternator -> split charge relay -> starter LA and 110Ah domestic LA. I'll just switch over my domestic loads to Lithium. I'm adding 1.5kW of solar to charge the LiFePOs. Since we're going into summer this should do for now, and drastically simplifies the system. I'll then have the equipment I need to experiment and skill up with the charging possibilities at my leisure, minimising risk of blowing up the batteries, the alternator, or both.

 

This is what I have in mind for Phase 1 which has no alternator charging. About 3 grand including delivery but excluding wiring and sundries. Yes I have gone with pricier Victron equipment and note that I have oversized some bits for future proofing. Please let me know if you can see anything glaringly wrong with this!

 

[MoominPapa]

Good price on the cells. To equalise the current through parallel cells, the link between the two blocks of cells need to be at the other end of the bus bar connecting the three cell terminals.

 

MP.

 

[Dr Bob]

1 hour ago, MoominPapa said:

Good price on the cells. To equalise the current through parallel cells, the link between the two blocks of cells need to be at the other end of the bus bar connecting the three cell terminals.

 

MP.

 

MP shouted at me for doing that as well!

1 hour ago, ivan&alice said:

OP here. ✅?

 

This is what I have in mind for Phase 1 which has no alternator charging. About 3 grand including delivery but excluding wiring and sundries. Yes I have gone with pricier Victron equipment and note that I have oversized some bits for future proofing. Please let me know if you can see anything glaringly wrong with this!

 

 

Looks good. Great price on the cells. For the lithium bit, you are around £1000 which is really good. Where are the cells from? If the pic is your cells with the red cable link in the wrong place -then I guess you got them from Jeremy?

 

On the layout, yes looks like you have got it right. The only thing I would say is that solar on a good summers day will be pushing in 50-90A which is more than my alternator does so you are likely to get up to 100% SoC easily. My solar is 20-30A so never really pushes me up to full. I assume the MPPT will switch to float when you need to but if you are charging at 90A, it may switch to float early as 90A may need a constant voltage phase in the charge cycle. MP should be able to advise as he does go that high in charging (at 30-40A charge, its easy to terminate at a chosen voltage). It will soon become apparent when the sun comes out! Cant see it being a problem though....just something to keep an eye on when commissioning.

I only use my auto disconnect as a last resort and the solar and alternator are set to cut back before the emergency cut out is needed.

 

 

[peterboat]

On 20/02/2020 at 16:40, Dre said:

Hi Dr Bob, 

I was gonna buy the Renogy 40a DC DC but it's out of stock at the moment (Simone from Renogy actually emailed me back to say they'll be in stock end of Feb). I would think the Sterling is a better option as it has a few more features. I don't think I would charge at more than 60a (and never more than half that bank rating) so I'm not worried about the balancing. I'd do an initial balance and monitor but don't expect to have problems with balancing at those charging or discharging C ratings. Moreover those batteries from that vendor seem to be very well matched. 

As to charging parameters I have seen reviews on the Sterling with Lifepo4s as being optimal in use. I'm also gonna monitor the charging while using the engine as I hope it won't be needed as much once I install a kW worth of solar. Also the bms is fully programmable and that will be a failsafe if the Sterling can't be optimally programmed. For the money I think either the Sterling or the Renogy are better and easier than anything else at charging lifepo4s via an alternator and they both have been used and reviewed extensively. 

As for splitting the battery bank I doubt it would make a difference as I'd have to balance 8 cells anyway as I'll be doing it manually and hopefully not for a long time after the first balance. I can parallel two battery banks with a bms each or parallel the two bms onto one battery but it's the same. Two bms will give me double the current regardless. 

I think there's a lot of technical and not so technical talk online about lifepo4 installations on the net but with the kit available nowadays and at those prices it all seem pretty straightforward and I don't expect any problems apart from maybe quality control issues from some of the cheap components. Lifepo4s are super safe and are successfully used in much more demanding applications, both electrically and mechanically speaking. I really don't think a few point volts difference in charging parameters will make much of a difference to the amount of cycles you get out of them and if I had to sacrifice a few hundred cycles of life or a few amps of capacity I am not too bothered as I can just add a few more cells at those prices.

I'll definitely keep you posted but I guess it might take a while still before I start building as I expect quite a delay for the batteries and bms from China. Poor souls have worse problems right now! 

 

https://battlebornbatteries.com/sterling-pro-batt-ultra-battery-battery-charger/

 

https://faroutride.com/b2b-review/

 

https://marinehowto.com/understanding-the-sterling-power-pro-batt-ultra-battery-to-battery-charger/

 

 

Just a couple of points inverter is 2000va which is way under 2000 watts and I would buy a midnite controller market leader and cheaper than victron. You should be safe with solar I have more than you and flat I have seen just under 80 amps. So enjoy your batteries 

[jetzi]

4 hours ago, MoominPapa said:

the link between the two blocks of cells need to be at the other end of the bus bar connecting the three cell terminals

Just checking - do you mean moving the red cable to here? This to me looks "better"...

 

 

Not to pass the buck but this is a photo that Jeremy sent me. He said that he'd make up the blocks with straps and end plates to compress the cells (which I didn't know was required!). I haven't taken delivery yet. he can't access his stock due to Covid unfortunately. Yes the price does seem good - he says the cells are a little older and not suitable for current draw > 200A such as for traction applications. I couldn't manage 200A if I tried anyway (maybe the starter motor?). Given the good price, do you think it would be outrageous to go for a 4p 4s configuration and get a whopping 640Ah? I want freedom to tinker and with 16 cells I could conceivably make a 24V or even 48V battery if I wanted to in the future...

 

2 hours ago, Dr Bob said:

on a good summers day will be pushing in 50-90A which is more than my alternator does so you are likely to get up to 100% SoC easily.

Theoretically, but I wouldn't do that though, I'd set the MPPT to cut out at 80%, or perhaps even 70% for the first few weeks or so, to be sure I reeeeeally know what I'm doing...

 

Here's a real newbie question: would it be sensible to get to 100% every few months (straight after balancing) to check on capacity and erase any "memory effect"? Of course, discharging the batteries promptly thereafter and not storing them at 100% even overnight.

 

2 hours ago, Dr Bob said:

alternator are set to cut back before the emergency cut out is needed.

Oh! Is this is a new addition to your setup? I recall that you advocated dumping the load from a switched off alternator into your hybrid LA bank. Is this now just a safety precaution?

 

1 hour ago, peterboat said:

inverter is 2000va which is way under 2000 watts

I don't understand VA! If I have a 2000 "watt" branded inverter and ran a 2000 "watt" AC load on it, what current at 12V would it be drawing?

 

1 hour ago, peterboat said:

midnite controller market leader and cheaper than victron.

Reasoning is that the victron kit all talks amongst themselves and there's lots of community support for interfacing with it on a DIY low level. Victron is also just extremely well known and has a great reputation for quality and customer service. I know I'm paying more than I need to but I'm OK to spend a couple hundred quid more on the whole system to get it blue branded.

[peterboat]

34 minutes ago, ivan&alice said:

Just checking - do you mean moving the red cable to here? This to me looks "better"...

 

 

Not to pass the buck but this is a photo that Jeremy sent me. He said that he'd make up the blocks with straps and end plates to compress the cells (which I didn't know was required!). I haven't taken delivery yet. he can't access his stock due to Covid unfortunately. Yes the price does seem good - he says the cells are a little older and not suitable for current draw > 200A such as for traction applications. I couldn't manage 200A if I tried anyway (maybe the starter motor?). Given the good price, do you think it would be outrageous to go for a 4p 4s configuration and get a whopping 640Ah? I want freedom to tinker and with 16 cells I could conceivably make a 24V or even 48V battery if I wanted to in the future...

 

Theoretically, but I wouldn't do that though, I'd set the MPPT to cut out at 80%, or perhaps even 70% for the first few weeks or so, to be sure I reeeeeally know what I'm doing...

 

Here's a real newbie question: would it be sensible to get to 100% every few months (straight after balancing) to check on capacity and erase any "memory effect"? Of course, discharging the batteries promptly thereafter and not storing them at 100% even overnight.

 

Oh! Is this is a new addition to your setup? I recall that you advocated dumping the load from a switched off alternator into your hybrid LA bank. Is this now just a safety precaution?

 

I don't understand VA! If I have a 2000 "watt" branded inverter and ran a 2000 "watt" AC load on it, what current at 12V would it be drawing?

 

Reasoning is that the victron kit all talks amongst themselves and there's lots of community support for interfacing with it on a DIY low level. Victron is also just extremely well known and has a great reputation for quality and customer service. I know I'm paying more than I need to but I'm OK to spend a couple hundred quid more on the whole system to get it blue branded.

Your inverter from victron is always branded as VA not watts I have a 3000 VA victron inverter its really 2400 watts,  a 2000 VA inverter wont run a washing machine,  read what it says on the inverter and spec same for their solar controllers check the maximum voltage in it needs to be 200 volts ideally 

[nicknorman]

Well for clarity a 2000va inverter can supply 2000w into a resistive load such as a 2kw electric kettle, 2 kw connector heater etc. The difference between va and w is to do with power factor, which is the relationship between the phase angle of voltage vs current. With a resistive load, the phase angle difference is zero and so va = w

 

But with a non-resistive load which includes anything with a motor, the current and voltage are out of phase and so the limiting factor is the va. The w will be somewhat less, depending on the phase angle (aka power factor).

 

Also things with motors tend to have a startup surge demand, so whilst a motor for eg a fridge might consume 100w steady state, at start up it might consume 1000w.

 

And just to answer your other question, at 12v and allowing for inverter losses and cable voltage drop, it is roughly 100A for every kw at mains voltage. So for 2kw it is 200A from the batteries.

Edited April 2, 2020 by nicknorman

[peterboat]

6 minutes ago, nicknorman said:

Well for clarity a 2000va inverter can supply 2000w into a resistive load such as a 2kw electric kettle, 2 kw connector heater etc. The difference between va and w is to do with power factor, which is the relationship between the phase angle of voltage vs current. With a resistive load, the phase angle difference is zero and so va = w

 

But with a non-resistive load which includes anything with a motor, the current and voltage are out of phase and so the limiting factor is the va. The w will be somewhat less, depending on the phase angle (aka power factor).

 

Also things with motors tend to have a startup surge demand, so whilst a motor for eg a fridge might consume 100w steady state, at start up it might consume 1000w

What it's supposed to do and what it does are two different things! My old 3000 watt inverter powered my 2800 watt kettle my victron 3000 VA inverter tripped out on overload and that is a resistive load. I now have a new kettle  Google says 3000 VA is 2400 watts which would be why my kettle overloaded the inverter 

Edited April 2, 2020 by peterboat

[jetzi]

1 minute ago, nicknorman said:

But with a non-resistive load which includes anything with a motor, the current and voltage are out of phase and so the limiting factor is the va.

With electricity I always think I get it, and then I get a curve ball like this!

 

There are only two things that the inverter needs to power, and never at the same time:

 

1) a 7kg washing machine with a true "cold wash" button that completely disables the heating element. This is probably around 500W and definitely not more than 1000W.

2) a 1200W hairdryer.

 

I am surprised that you guys use an electric kettle, I use a stovetop gas one.

 

31 minutes ago, peterboat said:

a 2000 VA inverter wont run a washing machine

I was under the impression that the water heating was the bit that consumes a lot of power, which is why I bought a washing machine that didn't do that. How can I calculate how large an inverter in VA I need to power this washing machine, then? I can't get a straight answer on peak W load, but it says it uses 1.05kW per "cycle", around an hour. So I guessed that the peak usage is around 1500W-2000W with the heating on.

 

I'd like to keep the option open of powering a dishwasher one day, but since this would have to heat water I feel like this is a pipe dream on a boat.

 

Perhaps I should be looking at a 3kVA inverter? Like https://www.bimblesolar.com/victron/victron3000w-12v
 

The larger inverter is actually not that much more expensive, I'm more worried about oversizing the inverter and unnecessarily draining the batteries when powering smaller loads.

[nicknorman]

25 minutes ago, peterboat said:

What it's supposed to do and what it does are two different things! My old 3000 watt inverter powered my 2800 watt kettle my victron 3000 VA inverter tripped out on overload and that is a resistive load. I now have a new kettle  Google says 3000 VA is 2400 watts which would be why my kettle overloaded the inverter 

Google doesn’t know what it is talking about. One problem with inverters is that they are constant power devices so if you want 3000 watts out (into a resistive load) you need 3000watts plus an allowance for efficiency in. Watts of course being current x voltage. If the voltage droops due to battery resistance, cable resistance, isolator resistance, connection resistance etc, the inverter compensates by taking more current. Which makes the voltage drop further, until the input voltage at the inverter is below the minimum and it trips.

 

if you plug your old kettle in and measure the voltage on the inverter input terminals, you’d find it drops below the minimum (say 10.5v) which trips the inverter. To drop 2v at 300A you only need 0.0066 ohms, which is not a lot.

8 minutes ago, ivan&alice said:

With electricity I always think I get it, and then I get a curve ball like this!

 

There are only two things that the inverter needs to power, and never at the same time:

 

1) a 7kg washing machine with a true "cold wash" button that completely disables the heating element. This is probably around 500W and definitely not more than 1000W.

2) a 1200W hairdryer.

 

I am surprised that you guys use an electric kettle, I use a stovetop gas one.

 

I was under the impression that the water heating was the bit that consumes a lot of power, which is why I bought a washing machine that didn't do that. How can I calculate how large an inverter in VA I need to power this washing machine, then? I can't get a straight answer on peak W load, but it says it uses 1.05kW per "cycle", around an hour. So I guessed that the peak usage is around 1500W-2000W with the heating on.

 

I'd like to keep the option open of powering a dishwasher one day, but since this would have to heat water I feel like this is a pipe dream on a boat.

 

Perhaps I should be looking at a 3kVA inverter? Like https://www.bimblesolar.com/victron/victron3000w-12v
 

The larger inverter is actually not that much more expensive, I'm more worried about oversizing the inverter and unnecessarily draining the batteries when powering smaller loads.

A washing machine without the heater on doesn’t use much, your inverter will easily power it. Ours runs fine with a 2.5kva inverter including when the heating element is on, although it is a small machine with a ~1kw heater.

 

Larger inverters don’t necessarily use more quiescent power, the determinant is the quality of the design aka the manufacturer.

Edited April 2, 2020 by nicknorman

[jetzi]

1 minute ago, nicknorman said:

inverters ... are constant power devices

So a 3000W inverter powering an LED and a 3000W inverter powering a kettle always use 3000W plus change for inefficiency? This is what I'm worried about - and why I want to size the inverter accurately to my load...