Cheap LiFePO4 BMS?

[jetzi]

Very sure. I balanced them at 3.6V last week and they were charged up to 13.9V on the whole battery this morning before switching to float. When not under load (when they are just under my small 12V loads) they sit at around 3.3V. Unless my BMV is inaccurate, I've only used 24Ah since this morning.

 

[TheBiscuits]

47 minutes ago, ivan&alice said:

Washing machines don't get specified in a nice "this many watts" label because they vary so much depending on what they are doing; they get specified in "this many kWh per year" which is an infuriating measurement.

 

They do have a maximum power rating in Watts on the label.

 

My old Candy is rated at 1300W but only ever uses more than about 200W when heating water for a hotter wash.

 

It's worrying if your machine wasn't loaded and wasn't heating water that it hammered the batteries that hard.

[nicknorman]

What I don’t quite understand is that either a hair dryer or a toaster will use a lot more power than a washing machine on a cold wash, and yet it seems these didn’t give you a problem.

 

Presumably you have checked that all the interconnects are tight / not getting warm under load?

 

There is a fundamental problem with using only low voltage to disconnect as you have discovered. There is nothing wrong with taking a Li cell to a fairly low voltage under heavy load. What is to be avoided is over-discharge, and voltage without regard for discharge current is not a good guide to over-discharge. Which is why my BMS will allow for discharge current x internal resistance before deciding to operate the isolation relay. We would want our Li to be able to run 2kw for a few minutes without breaking into a sweat even if not at a high SoC.

[peterboat]

1 hour ago, BEngo said:

A washing machine heater of, say 1.7 kW is going to want about 170 A into the inverter, so much more than you expect.

One of the LiFe downsides is they don't readily deliver high current like an LA.

N

It's strange but true my electric truck performance better on 4 x LifePo4s batteries than 12 AGMs! The capacity of the LifePo4s is 6 kWh and the Agm is 10 kWh.  Also on a quick Google says they hand out better CCA than LAs, I couldn't find any advantage of LAs over LifePo4s on my googling? Why do you think that LAs are better? Just curious 

[BEngo]

I don't.  They are the battery of tomorrow, or maybe late this afternoon. I was just pointing out that they do have some limitations, like LA, to go with a lot more upside than LA.

N

[nicknorman]

The Chinese LiFePO4 batteries i was looking at, around 100 - 150Ah, have a quoted internal resistance of 0.3 - 0.4 milliohms. So 100A would drop 0.03v to 0.04v which is not a lot!

[Col_T]

The data sheet for Winston (https://files.ev-power.eu/inc/_doc/attach/StoItem/2010/GWL-Winston-LFP300AHA.pdf) specify maximum discharge should be less than 3C; data sheet for Victron (https://www.victronenergy.com/upload/documents/Datasheet-12,8-&-25,6-Volt-lithium-iron-phosphate-batteries-Smart-EN.pdf) specifies 2C maximum discharge. Both manufacturers give a 'recommended' discharge current of half label capacity (C5?).

 

A couple of questions -

 

What is the result of exceeding maximum discharge rate for a lithium battery - immediately fatal? serious shortening of expected life? general reduction in life expectancy? no real world knowledge?

 

is there a limit to discharge rates on an LA battery, or is that what CCA is about? I think I understand that the greater the discharge current from an LA the greater the chance of damage e.g. plates shedding material or deforming, but haven't a clue about what kind of discharge levels these occur.

 

For clarity, this is about continuing my education rather than trying to make a point about something!

[MoominPapa]

9 hours ago, nicknorman said:

 

There is a fundamental problem with using only low voltage to disconnect as you have discovered. There is nothing wrong with taking a Li cell to a fairly low voltage under heavy load. What is to be avoided is over-discharge, and voltage without regard for discharge current is not a good guide to over-discharge. Which is why my BMS will allow for discharge current x internal resistance before deciding to operate the isolation relay. We would want our Li to be able to run 2kw for a few minutes without breaking into a sweat even if not at a high SoC.

Ivan's query made be go back and check the code in the MoominBMS. Low voltage shutoff is 12.52v, which is mainly to do with avoiding significant discharge of the parallel LA start battery. This is reduced by 0.78v per 100A discharge current for precisely the reason you give. The per-cell limit is 2.9v which is independent of discharge. In a perfectly balanced battery, the per cell limit will be triggered even though the whole-battery one isn't at somewhere north of 100A, so this is a bit of an uneasy compromise.

 

MP.

 

[peterboat]

1 hour ago, BEngo said:

I don't.  They are the battery of tomorrow, or maybe late this afternoon. I was just pointing out that they do have some limitations, like LA, to go with a lot more upside than LA.

N

I am currently building a lithium polymer pack for my little electric car. The old bank was Agm and weighed 359 kilo including the frame and is rated at 10 kWh. The replacement will weigh about a third and be half the size at 17.5 kWh, I would expect to over double my range, have much better acceleration and top speed plus will charge up much quicker! Lithium batteries are here now not this afternoon. 

[Richard10002]

11 hours ago, ivan&alice said:

Very sure. I balanced them at 3.6V last week and they were charged up to 13.9V on the whole battery this morning before switching to float. When not under load (when they are just under my small 12V loads) they sit at around 3.3V. Unless my BMV is inaccurate, I've only used 24Ah since this morning.

 

Have you specified your battery to Inverter cables correctly, to cope with voltage drop, (I cant believe you haven't, but a question worth asking, if only for elimination).

 

With a 2500W inverter about 10/12ft from the batteries, I used 95mm2 cable to cope with the possible 250A draw at the extremes.

[Dr Bob]

On 19/07/2020 at 22:26, ivan&alice said:

Wash was cold, a 15min test run. This washing machine has a "true cold wash" button that allows you to turn off the heating element entirely. Washing machines don't get specified in a nice "this many watts" label because they vary so much depending on what they are doing; they get specified in "this many kWh per year" which is an infuriating measurement.. But in various other threads we've discussed this washing machine and the general consensus was it uses around 500W. 1000W is less than 100A.

My domestic battery is completely isolated from the engine for the time being. I'm expecting this battery to be able to put out 200A. Is this unreasonable?

 

 

On 19/07/2020 at 22:29, MoominPapa said:

Just did a test for confirmation, hairdryer on max. 140A discharge, battery about 80% full, no cell below 3.17v, so your numbers look low.

 

MP.

 

 

There is something funny going on if your washing machine on a cold cycle took the voltage that low yet your toaster and hair dryer didn't. I agree with MP your voltages are low. Important to put the clamp meter on - and check all 3 to see their comparative draws.

Jeremy who sold you the batteries does check them by putting them under a load that is seen by EVs. Not sure if that is 1C, 2C or 3C, but likely to be at the high end. When MP and T&B got theirs, he said they were not quite there for EV usage so they got them cheaper. When I got mine he had run out of the 'weaker' ones so I had to pay a premium as mine were fit for EV duty. I am sure he will have told you which category yours fall into but even the weaker ones should be able to cope with any boat duty they are asked to do. My Tesla does 0-60mph in 4 secs. God knows how much power that draws!!!

As Richard says above, could it be a bad connection? Maybe one of the battery cell interconnects is not perfect? I think I remember you saying you had split the bank down a number of times? That would be my best guess but looking at the draw on the washing machine, the toaster and the hair dryer might help.

I know I have seen it but can you repost the diagram of how you have everything wired up.

Edited July 21, 2020 by Dr Bob

[peterboat]

3 hours ago, Dr Bob said:

 

 

There is something funny going on if your washing machine on a cold cycle took the voltage that low yet your toaster and hair dryer didn't. I agree with MP your voltages are low. Important to put the clamp meter on - and check all 3 to see their comparative draws.

Jeremy who sold you the batteries does check them by putting them under a load that is seen by EVs. Not sure if that is 1C, 2C or 3C, but likely to be at the high end. When MP and T&B got theirs, he said they were not quite there for EV usage so they got them cheaper. When I got mine he had run out of the 'weaker' ones so I had to pay a premium as mine were fit for EV duty. I am sure he will have told you which category yours fall into but even the weaker ones should be able to cope with any boat duty they are asked to do. My Tesla does 0-60mph in 4 secs. God knows how much power that draws!!!

As Richard says above, could it be a bad connection? Maybe one of the battery cell interconnects is not perfect? I think I remember you saying you had split the bank down a number of times? That would be my best guess but looking at the draw on the washing machine, the toaster and the hair dryer might help.

I know I have seen it but can you repost the diagram of how you have everything wired up.

My dishwasher is on now the batteries were showing 13.5 now on 13.3 on a heat and wash cycle. Something is wrong in Ivans setup either a connection or a duff battery.

So further on the washing machine has been on along with the kettle a couple of times and under load its 13.2 v but back up to 13.4 as soon as load gone

Edited July 21, 2020 by peterboat

[nicknorman]

On 20/07/2020 at 09:21, Col_T said:

The data sheet for Winston (https://files.ev-power.eu/inc/_doc/attach/StoItem/2010/GWL-Winston-LFP300AHA.pdf) specify maximum discharge should be less than 3C; data sheet for Victron (https://www.victronenergy.com/upload/documents/Datasheet-12,8-&-25,6-Volt-lithium-iron-phosphate-batteries-Smart-EN.pdf) specifies 2C maximum discharge. Both manufacturers give a 'recommended' discharge current of half label capacity (C5?).

 

A couple of questions -

 

What is the result of exceeding maximum discharge rate for a lithium battery - immediately fatal? serious shortening of expected life? general reduction in life expectancy? no real world knowledge?

 

is there a limit to discharge rates on an LA battery, or is that what CCA is about? I think I understand that the greater the discharge current from an LA the greater the chance of damage e.g. plates shedding material or deforming, but haven't a clue about what kind of discharge levels these occur.

 

For clarity, this is about continuing my education rather than trying to make a point about something!

Bearing in mind the absence of replies to your query I would propose “no real world knowledge”. But my best guess would be that exceeding max discharge current would cause local overheating which would likely shorten life expectancy. Whether you want to say “serious shortening” or “general reduction” would depend on the duration and frequency of the excessive discharge rate. Some of the LiFePO4 cells I’ve looked at quote a continuous discharge rate as well as 2 peak rates of medium and short duration.

 

LA batteries seem to dislike fast charging more than fast discharging and anyway the maximum discharge current is less than for Li of the same capacity. Probably once again the enemy is local overheating and with the higher internal resistance of LA that heating will be worse. But heat generation gives temperature rise in proportion to the duration, so as per Li, a short duration very high current discharge is inconsequential, whereas a prolonged discharge will create local hot spots which are detrimental.

[TheBiscuits]

On 19/07/2020 at 23:35, nicknorman said:

What I don’t quite understand is that either a hair dryer or a toaster will use a lot more power than a washing machine on a cold wash, and yet it seems these didn’t give you a problem.

 

 

You're not the only one who doesn't follow that logic ...

 

Measurement error? Toaster and hair dryer running at the same time as the washing machine?

 

It's probably worth checking again!

[Richard10002]

12 hours ago, TheBiscuits said:

 

You're not the only one who doesn't follow that logic ...

 

Measurement error? Toaster and hair dryer running at the same time as the washing machine?

 

It's probably worth checking again!

Yes..... The inverter doesnt have any idea what is drawing power from it, nor does it care - 140A is 140A. So, if the hairdryer and toaster draw between 100A and 200A, or whatever, without causing the shutdown, the problem cannot be due to the inability of the batteries to cope.

 

The conclusion being that there is something in the washing machine circuit that is causing a relatively huge voltage drop.

 

On current draw relative to bank size, Will Prowse often draws 1C with his hot air paint stripper thing, when he is testing the capacity of a bank, and ther is talk of 2C and 3C in a variety of places.

 

Ivan has a 640Ah bank, so 1C would be 640A. Never say never, and all that, but the chances of him drawing more than about C/2, are slim to nil. Even with degraded lithium batteries, with maybe 80% of nominal capacity, it remains unlikely.

[nicknorman]

On 17/07/2020 at 12:07, MoominPapa said:

Do you store parameter logs?

 

MP.

OK just for you, I logged some parameters from today. Charge rate set to slow ie maximum field current about half of maximum. Started at Audlem below lock 12 so a bit of faffing with rpm up and down as we went through the last 3 locks, hence charge current fluctuations because at that stage the field current will be pegged on its limit (1/2 normal value). Then we stopped briefly at Overwater (looking for new mooring lines, they didn't have). Then we stopped again at Hack Green due to lock queue (45 mins!). Battery voltage reported over LIN only has a resolution of 0.1v hence the rather castellated plot. Everything on LH axis except voltage which is on RH axis. Due to stops I didn't use a time scale on the bottom, just sample number, but each sample is about 20 secs apart, hence each of the major scale markers is 10 mins.

 

[MoominPapa]

49 minutes ago, nicknorman said:

OK just for you, I logged some parameters from today. Charge rate set to slow ie maximum field current about half of maximum. Started at Audlem below lock 12 so a bit of faffing with rpm up and down as we went through the last 3 locks, hence charge current fluctuations because at that stage the field current will be pegged on its limit (1/2 normal value). Then we stopped briefly at Overwater (looking for new mooring lines, they didn't have). Then we stopped again at Hack Green due to lock queue (45 mins!). Battery voltage reported over LIN only has a resolution of 0.1v hence the rather castellated plot. Everything on LH axis except voltage which is on RH axis. Due to stops I didn't use a time scale on the bottom, just sample number, but each sample is about 20 secs apart, hence each of the major scale markers is 10 mins.

 

Just look at that current decay! I sure don't miss LA. Thanks, those curves look as I'd expect. Your temperature curve looks fine, at least on slow charge.

 

MP.

[nicknorman]

14 minutes ago, MoominPapa said:

Just look at that current decay! I sure don't miss LA. Thanks, those curves look as I'd expect. Your temperature curve looks fine, at least on slow charge.

 

MP.

Yes, however I think the battery-sensed voltage and the “hard” regulation speeds the charge significantly compared to a conventional regulator. Still getting 80A (into 450Ah) at 90% SoC, which is pretty good for LA.

Edited July 23, 2020 by nicknorman

[Dr Bob]

Interesting info on the graphs Nick.

You havent got individual scales on the LH axis, so what was the max current and max temp. I was getting bit worried about the 14.7V until I realised this was for LA's!!!!

[nicknorman]

17 minutes ago, Dr Bob said:

Interesting info on the graphs Nick.

You havent got individual scales on the LH axis, so what was the max current and max temp. I was getting bit worried about the 14.7V until I realised this was for LA's!!!!

The scales are all the same for current, SoC and temperature. Just volts on the RH axis.

 

So max current around 95A (after the bottom lock and before Overwater, when the engine rpm was at a steady 1300). Max temperature was around 60C. Bearing in mind as previously mentioned, this is the temperature of the rear fan exit air (having been drawn past the diodes and the rear part of the stator). Obviously there is an element of mixing engine bay temperature with alternator workload, so at the start the engine bay is cold hence cool alternator even though it was working relatively hard, then engine bay temperature climbed as alternator workload decreased.

 

 

[WotEver]

2 hours ago, nicknorman said:

Still getting 80A (into 450Ah) at 90% SoC, which is pretty good for LA.

Yeah, I'd say that's 'pretty good'... it's nearly double what you'd expect with a conventional regulator!

[Craig Shelley]

On 21/07/2020 at 09:30, Dr Bob said:

 

 

There is something funny going on if your washing machine on a cold cycle took the voltage that low yet your toaster and hair dryer didn't. I agree with MP your voltages are low. Important to put the clamp meter on - and check all 3 to see their comparative draws.

Jeremy who sold you the batteries does check them by putting them under a load that is seen by EVs. Not sure if that is 1C, 2C or 3C, but likely to be at the high end. When MP and T&B got theirs, he said they were not quite there for EV usage so they got them cheaper. When I got mine he had run out of the 'weaker' ones so I had to pay a premium as mine were fit for EV duty. I am sure he will have told you which category yours fall into but even the weaker ones should be able to cope with any boat duty they are asked to do. My Tesla does 0-60mph in 4 secs. God knows how much power that draws!!!

As Richard says above, could it be a bad connection? Maybe one of the battery cell interconnects is not perfect? I think I remember you saying you had split the bank down a number of times? That would be my best guess but looking at the draw on the washing machine, the toaster and the hair dryer might help.

I know I have seen it but can you repost the diagram of how you have everything wired up.

Hi again, I've been keeping an eye on this thread for a while now, and was curious to know how many other people on here are using old cells from Jeremy at EVSupport?

 

On the subject of voltage variation, we've noticed the batteries appear to have a higher "internal resistance" when state of charge is lower. This has caused the low voltage protection to trip out when starting the Dyson. I don't think this is helped by the fact that we often have the inverter in power save mode, so not only is there the surge current from starting the motor, there is also the inrush current of the inverter charging its DC link capacitors. This hasn't been a problem in the past couple of years as we've improved the state of charge monitoring and now maintain a higher SOC during cold weather. 

 

I saw, a few weeks back, discussion of cell configuration series-parallel vs parallel-series. All of the points seemed quite valid, but there are a few extras I'd  like to throw into the mix. 

To simplify integration into the existing boat wiring, we went for 3 parallel packs, each consisting of 4 cells in series. The obvious downside is that much more monitoring is required, not that we've installed any yet. But this gives more flexibility for the cells to settle at their own preferred voltages. The danger we envisaged was that one cell becomes out of balance, and suddenly starts to dip in voltage. However, since the cells are all separated between the packs, a single lazy cell causes a reduced contribution to the overall current from its pack. We do notice this as we have independent ammeters for each pack. Sometimes one pack will be providing a lower current than the others. Later on the same day, the same pack might be providing a slightly higher current than the others. It all ends up balancing out, but hints at how imprecise and chaotic the cells really are inside.

[jetzi]

Sorry I haven't had a chance to reply - when I wrote originally it had just happened. I haven't retested properly yet but I will report back soon with some numbers and diagrams.

 

I did successfully run a cold wash, which suggests that everyone's hunch was right - I think that the heating element was on in the washing machine. Even so, it seems excessive that it it dropped the voltage quite so much that it tripped my BMS. Although it is not my planned mode of operation - I'd expect to be able to power my washing machine with heating element. Having that capability seems like a reasonable margin of safety.

 

On 20/07/2020 at 10:42, Richard10002 said:

Have you specified your battery to Inverter cables correctly, to cope with voltage drop, (I cant believe you haven't, but a question worth asking, if only for elimination).

 

With a 2500W inverter about 10/12ft from the batteries, I used 95mm2 cable to cope with the possible 250A draw at the extremes.

I used 70mm2 cable throughout. The inverter is less than two feet of cable away from the battery. I can't remember what the voltage drop of that was, but I based my calculations on a maximum 240A current, given the 3kVA size of my inverter.

 

On 22/07/2020 at 13:21, Richard10002 said:

Ivan has a 640Ah bank, so 1C would be 640A. Never say never, and all that, but the chances of him drawing more than about C/2, are slim to nil. Even with degraded lithium batteries, with maybe 80% of nominal capacity, it remains unlikely.

So I am right to expect my 640Ah battery to be able to power my inverter flat out (2kVA = that's < 0.4C = 240A right?)

 

On 20/07/2020 at 09:38, MoominPapa said:

Ivan's query made be go back and check the code in the MoominBMS. Low voltage shutoff is 12.52v, which is mainly to do with avoiding significant discharge of the parallel LA start battery. This is reduced by 0.78v per 100A discharge current for precisely the reason you give. The per-cell limit is 2.9v which is independent of discharge. In a perfectly balanced battery, the per cell limit will be triggered even though the whole-battery one isn't at somewhere north of 100A, so this is a bit of an uneasy compromise.

That's interesting, reducing the low voltage cutoff due to discharge current - doesn't that mean though that if you are busy running down your battery that you're at risk of overdischarge? Does that mean you're entirely relying on the per cell cutoff?

[Richard10002]

51 minutes ago, ivan&alice said:

I used 70mm2 cable throughout. The inverter is less than two feet of cable away from the battery. I can't remember what the voltage drop of that was, but I based my calculations on a maximum 240A current, given the 3kVA size of my inverter.

less than a tenth of a volt, less than 1%.... so should not be a problem at all.

51 minutes ago, ivan&alice said:

 

So I am right to expect my 640Ah battery to be able to power my inverter flat out (2kVA = that's < 0.4C = 240A right?)

I an no expert, but it's hard to see why they wouldn't. I used to power a 2kW hoover with my 2500W inverter charger from 450Ah of Trojan T105s, and 330Ah of Varta Hobbys.

 

Something is going on, and you/we don't know what it is........

[TheBiscuits]

2 hours ago, ivan&alice said:

I did successfully run a cold wash, which suggests that everyone's hunch was right - I think that the heating element was on in the washing machine. Even so, it seems excessive that it it dropped the voltage quite so much that it tripped my BMS. Although it is not my planned mode of operation - I'd expect to be able to power my washing machine with heating element. Having that capability seems like a reasonable margin of safety.

 

That's good!

 

Please either pull out your machine and read the power label or let us know the make/model so we can look it up.  If it's way more powerful than we think there might be tricks we can play - I know some people who have disconnected one of the two heating coils in some washing machines so it will work from an inverter, and I know a few people who have bought a different machine because they can't!