Jump to content

Balancing LiFePO4s


MtB

Featured Posts

2 hours ago, Mike the Boilerman said:

 

Aha! So unless we can figure out how it works/what it does exactly, we don't know if it will work with Li, or do we?

 

on the bumph supplied with it, it says it works with 3.3V cells so it should work with LiFePO4s.

It works on Tom's Li's.

Link to comment
Share on other sites

1 hour ago, Dr Bob said:

If I ate an extra sausage roll that would be 150 calories more so you woud have to apply the resistor thingy for an extra 30 mins.

No, I am afraid that the "calories" that you use for food measure are in actual fact Kilo-calories so you would have to run the thingy for an extra 30,000 minutes or 21 days. This goes to show that if you could invent the sausage roll powered narrowboat then you would be well in and Greg's share price would rocket.

  • Haha 1
Link to comment
Share on other sites

5 hours ago, Mike the Boilerman said:

6E68CCE4-75CA-4F20-9065-423C659B1865.jpeg.dfcd3b5dabad2b12c906a4477c3979df.jpeg

 

 

Here is the bank disconnected (i.e. with the isolator switch turned OFF) after an hour of charging. No current going in or out.

 

I suspect you may be setting too much store in the accuracy of the readings. It would be interesting to swap cells 1 and 4 and see if the lower voltage stays with the cell or with the measuring device.

  • Greenie 1
Link to comment
Share on other sites

4 hours ago, Dr Bob said:

One piece of detail to watch .....but the effect is small....is that the BG8S meter seems to take the power to run the meter from cell  1....ie the meter works on 3.3V. I see a drop of 4mV on cell 1 when I push the button to light up the display. This is not a big issue as the meter display is only on for 60secs or so. HOWEVER - I also have the older meter (the BC8S) which has the display on all the time.....I couldnt work out how to turn it off...and this also takes power from cell 1. Over a month or so that could take a bit of capacity out of cell 1. I have disconnected my BC8S.

 

Aha! 

 

Now I had my old BC8S connected to the battery bank for about six months prior to starting to use it. This *might* explain why cell 1 is lower than all the others. 

 

Just wondering if I can re-charge it somehow using this motorcycle LA battery I have lying around... Maybe bung it in series with one of them cup heater elements.... 

Link to comment
Share on other sites

4 hours ago, nicknorman said:

I suspect you may be setting too much store in the accuracy of the readings. It would be interesting to swap cells 1 and 4 and see if the lower voltage stays with the cell or with the measuring device.

No Nick. I very much doubt it. There is a chance you are right but not in my experience (or Tom's).

As you charge the cells you see differences in how they charge and you will find some of the cells accelerate in voltage as you get around the knee....ie some cells get there first and not necessarily the ones you think. Until you have a set of these, you cannot understand how they operate. If it was an accuracy offset you would soon see it. Also you can check with a good multimeter.

Nothing I have seen in Mike's data shouts a meter problem. It is exactly the same sort of readings I see. I have loads of cell voltage data but unfortunately it is on around 15 pages of handwritten recording so not able to post it. Yes the meter could have an offset from a true reading but mine are consistent and it is the comparison of readings that are important once you have mapped out your 'standard' data.

You could prove it by swapping leads but not sure how to do that. Cell 1 is the fist cell on the meter and the wiring starts from the -ve pole of cell 1 so it would be impossible to swap cell 1. Maybe you could swap cell 7 &8 but then cell 6 wouldnt work. Best to check against a multimeter. Also Mike has both the BG8S and BC8S so you would see any issue comparing the two.

Trust me.....I have spent a lot of time checking my system in this way as I did worry about accuracy. I also checked my multimeter accuracy against MP's system so was happy I was measuring 'good' values.

Link to comment
Share on other sites

3 hours ago, Mike the Boilerman said:

 

Just wondering if I can re-charge it somehow using this motorcycle LA battery I have lying around... Maybe bung it in series with one of them cup heater elements.... 

Look through the Nordkyn or 'marine how to' sites 'cause one of them talks about charging a bank of paralled 3.3V cells with a 12 V charger. As long as you watch the voltage on the cell and dont let it go over 3.6V then you should be able to charge it. I cant remember which site I read this on. I can rememeber them also saying to connect the 12 V charger to the batteries with thinner wire to reduce the voltage. I never tried this as I didnt want to knacker my boat charger ......but if you had an old 12V car charger that would be fine. :)

A motorcycle battery should be ok (are they 6V?) .......connect up with thinish cables, measure the voltage and current (with a clamp meter) and maybe try by putting say 5Ahrs into to the low cell then rest for 10 mins and check how much the voltage has gone up. Wouldnt bother with a resistor in line. Be careful though because if you try and get the voltage of cell 1 similar to the others at 80% SoC, you could find cell 1 runs away when you eventually charge to 100% with your boat charger. I would maybe close the voltage gap by a half then take up to 100% and then repeat if cell 1 is the laggard at the full charge. It takes a long time the first time you do it but if you record the voltages, it is very quick to do 2nd time round as you know how many minutes of charge or discharge is needed to raise or lower the voltage.

Keen to see how you get on. Breaking new ground!!!!!!:)

Link to comment
Share on other sites

33 minutes ago, Dr Bob said:

No Nick. I very much doubt it. There is a chance you are right but not in my experience (or Tom's).

As you charge the cells you see differences in how they charge and you will find some of the cells accelerate in voltage as you get around the knee....ie some cells get there first and not necessarily the ones you think. Until you have a set of these, you cannot understand how they operate. If it was an accuracy offset you would soon see it. Also you can check with a good multimeter.

Nothing I have seen in Mike's data shouts a meter problem. It is exactly the same sort of readings I see. I have loads of cell voltage data but unfortunately it is on around 15 pages of handwritten recording so not able to post it. Yes the meter could have an offset from a true reading but mine are consistent and it is the comparison of readings that are important once you have mapped out your 'standard' data.

You could prove it by swapping leads but not sure how to do that. Cell 1 is the fist cell on the meter and the wiring starts from the -ve pole of cell 1 so it would be impossible to swap cell 1. Maybe you could swap cell 7 &8 but then cell 6 wouldnt work. Best to check against a multimeter. Also Mike has both the BG8S and BC8S so you would see any issue comparing the two.

Trust me.....I have spent a lot of time checking my system in this way as I did worry about accuracy. I also checked my multimeter accuracy against MP's system so was happy I was measuring 'good' values.

You just have to look at this post to see the inaccuracy of one or almost certainly both devices. He did compare the two, and there were obvious errors/inaccuracies.

 

Obviously these devices have their uses - as the knee is reached the change in voltage is much greater and thus inaccuracies become less important. So they are fine for top balancing (or bottom balancing). But it would be a mistake to read too much into a mid-charge reading and worrying about seeming imbalance. As we know, the voltage stays very steady for much of the midrange and these meters aren’t sufficiently accurate to be able to determine that one cell is “flatter” or “less flat” than others and hence needs balancing. As I said, that can only be determined near full or empty.

Link to comment
Share on other sites

5 minutes ago, nicknorman said:

 But it would be a mistake to read too much into a mid-charge reading and worrying about seeming imbalance. As we know, the voltage stays very steady for much of the midrange and these meters aren’t sufficiently accurate to be able to determine that one cell is “flatter” or “less flat” than others and hence needs balancing. As I said, that can only be determined near full or empty.

I have continually said that - dont use the cell voltages in the midrange to do anything. IT IS only useful at the top and bottom.

It is not about meter accuracy at all.

Edited by Dr Bob
Link to comment
Share on other sites

19 minutes ago, nicknorman said:

You just have to look at this post to see the inaccuracy of one or almost certainly both devices. He did compare the two, and there were obvious errors/inaccuracies.

 

 

 

On 11/11/2019 at 10:11, Mike the Boilerman said:

 
Sooo... not very accurate at all!!! One or both...

 

Nick, if you are referring to Mikes post 28, then you cannot easily compare the voltages between the BG8S and BC8S as the system was on charge (or very soon after) when the BC8S pic was taken ie voltages around 3.35V but not when the BG8S was taken around 3.30V. The state of charge has a big effect on the cell voltage. The sort of differences seen are very typical of what I see on mine ie one cell will rise in voltage more than another with a charge on it.  Yes, there could be a difference in reading from one to the other meter but just look at one meter.

Once you get a set of Li's and monitor  them on charge and discharge then you will understand how the voltage readings vary.

Link to comment
Share on other sites

9 hours ago, Dr Bob said:

on the bumph supplied with it, it says it works with 3.3V cells so it should work with LiFePO4s.

It works on Tom's Li's.

 

What I meant was, what does it do, precisely?

 

Does it discharge the highest cells to achieve balance? Or charge the lowest? And presumably it can only work above the knee? Or below it? 

 

 

6 hours ago, nicknorman said:

I suspect you may be setting too much store in the accuracy of the readings. It would be interesting to swap cells 1 and 4 and see if the lower voltage stays with the cell or with the measuring device.

 

It would be, but impossible given the way the thing is wired up, with 9 wires. One wire to each interconnect plus one at each end of the 8 cells. 

 

 

Link to comment
Share on other sites

1 minute ago, Mike the Boilerman said:

 

What I meant was, what does it do, precisely?

 

Does it discharge the highest cells to achieve balance? Or charge the lowest? And presumably it can only work above the knee? Or below it? 

 

 

I assume it discharges from the highest cells to the lowest.

I would think it only works at the knees as that is when you see the imbalance. On my cells now at 60%, cell 1 is the lowest voltage but I know it will get to the high knee first so I cant see the unit balancing my system well unless I only use it at the knee. Tom says his works over the whole range.

Link to comment
Share on other sites

25 minutes ago, Dr Bob said:

I have continually said that - dont use the cell voltages in the midrange to do anything. IT IS only useful at the top and bottom.

It is not about meter accuracy at all.

It is about meter accuracy if you try to draw conclusions mid charge, as MTB has done!

 

Link to comment
Share on other sites

24 minutes ago, Dr Bob said:

 

 

Nick, if you are referring to Mikes post 28, then you cannot easily compare the voltages between the BG8S and BC8S as the system was on charge (or very soon after) when the BC8S pic was taken ie voltages around 3.35V but not when the BG8S was taken around 3.30V. The state of charge has a big effect on the cell voltage. The sort of differences seen are very typical of what I see on mine ie one cell will rise in voltage more than another with a charge on it.  Yes, there could be a difference in reading from one to the other meter but just look at one meter.

Once you get a set of Li's and monitor  them on charge and discharge then you will understand how the voltage readings vary.

Obviously the SoCs are different and thus so are the voltages. But the relevant point is the relative voltages - the first pic has the cells roughly in order of voltage, the second the voltages are all over the place. If the measurements were accurate the cells would surely have roughly the same relative voltages?

3 minutes ago, Richard10002 said:

Can I ask what the knee is..... I can hazard a guess, but no sense reinventing the wheel?

The knee is the point in the graph of voltage vs SoC during charging where, approaching fully charged, the voltage suddenly starts to rise rapidly. Ie an upward kink in the graph. Knee as in bent knee, not straight knee, obvs!

Edited by nicknorman
P
Link to comment
Share on other sites

22 minutes ago, nicknorman said:

Obviously the SoCs are different and thus so are the voltages. But the relevant point is the relative voltages - the first pic has the cells roughly in order of voltage, the second the voltages are all over the place. If the measurements were accurate the cells would surely have roughly the same relative voltages?

No, the first pic is the BG8S when the system was not under charge and you can see a range of voltages. The second shot was the BC8S where it was either under charge or reading taken soon after.  What you don't seem to understand ....as you 

have obviously from your comments never seen these LiFePO4s charging.....is that the order of the voltages will change. Cell 1 could be a laggard at no charge but become  a mid voltage cell under charge. In the second shot, they are not all over the place. They are just in a slightly different place due to the difference in charge. THAT is why it is not a good idea to use the mid range voltages as they are not representative in this range.....nothing to do with accuracy.

Your immediate reaction is look at the accuracy but in this case that is unlikely. Once you have these cells, you learn their subtle changes. One of my first posts after installation was to comment on how they vary on charge and discharge. Mike's are very similar. It is easy to check them with a good multi meter.

If you are not convinced, get yourself down here and we can spend a day looking at all my data and showing you the variation. I am living with this. You are not.

 

 

eta... Also as I said earlier, twice, ignore the SoC reading on the meter. It is not designed for these batteries....it is totally wrong.

 

Edited by Dr Bob
  • Greenie 1
Link to comment
Share on other sites

19 minutes ago, Dr Bob said:

No, the first pic is the BG8S when the system was not under charge and you can see a range of voltages. The second shot was the BC8S where it was either under charge or reading taken soon after.  What you don't seem to understand ....as you 

have obviously from your comments never seen these LiFePO4s charging.....is that the order of the voltages will change. Cell 1 could be a laggard at no charge but become  a mid voltage cell under charge. In the second shot, they are not all over the place. They are just in a slightly different place due to the difference in charge. THAT is why it is not a good idea to use the mid range voltages as they are not representative in this range.....nothing to do with accuracy.

Your immediate reaction is look at the accuracy but in this case that is unlikely. Once you have these cells, you learn their subtle changes. One of my first posts after installation was to comment on how they vary on charge and discharge. Mike's are very similar. It is easy to check them with a good multi meter.

If you are not convinced, get yourself down here and we can spend a day looking at all my data and showing you the variation. I am living with this. You are not.

 

 

eta... Also as I said earlier, twice, ignore the SoC reading on the meter. It is not designed for these batteries....it is totally wrong.

 

Seems a bit unlikely that the cell voltages would change a lot relative to each other. What is the mechanism for that? I’m sceptical because whilst I don’t doubt it is an effect you have observed, you are observing it via the el cheapo BMS and perhaps what you are observing is the inaccuracies of that BMS, not something fundamental going on with the cells. Can you prove me wrong?

Link to comment
Share on other sites

6 minutes ago, nicknorman said:

Seems a bit unlikely that the cell voltages would change a lot relative to each other. What is the mechanism for that?

 

Capacity. A bank of cells all differing slightly in capacity will swap around in terminal voltage depending on SoC and charge or discharge conditions, particularly as some arrive at or pass through the knee whilst others are still approaching.  And at the bottom end, different cells will tumble first. 

 

  

  • Greenie 1
Link to comment
Share on other sites

17 minutes ago, Mike the Boilerman said:

 

Capacity. A bank of cells all differing slightly in capacity will swap around in terminal voltage depending on SoC and charge or discharge conditions, particularly as some arrive at or pass through the knee whilst others are still approaching.  And at the bottom end, different cells will tumble first. 

 

  

Yes I understand that, but that effect would only be visible near the top and bottom, unless the cells had massively different capacities. Coming back to your post 28, as I understand it those two pics were taken shortly after one another? If so one can see a fair discrepancy between the two, not a fixed offset for all cells, but the cells are in a different order, eg cells 7 and 8. Displays to 3 decimal places are alluring, but when the accuracy is unlikely to be better than 1% tops, misleading.

 

Edit: cell 8 seems to have around 0.04v difference which is 1%

Edited by nicknorman
Link to comment
Share on other sites

8 hours ago, nicknorman said:

Seems a bit unlikely that the cell voltages would change a lot relative to each other. What is the mechanism for that? I’m sceptical because whilst I don’t doubt it is an effect you have observed, you are observing it via the el cheapo BMS and perhaps what you are observing is the inaccuracies of that BMS, not something fundamental going on with the cells. Can you prove me wrong?

The observations are not due to a faulty meter. I have checked the readings of the cell meter to a good multimeter. The readings aren't real. 

Your statement that cell voltages wouldn't change a lot shows you don't know what is going on. 

Cell voltages will vary for a number of reasons.

Mike has given te first one. SoC. My cells go from 3.2V at the bottom to 3.35V ish at rest low to high charge. I have not looked in detail at the graph below half way on cell only voltage (but have done on total bank voltage) and that plateau voltage is pretty much a straight line....so voltage is proportional to SoC. SoC can vary between cells. This must be happening when one cell reaches the high knee before the other. After my first 3 months, I found my highest cell was 80Ahrs ahead of the average so 20% SoC too high. That shows up as a different voltage.

That isn't the full story.

You have to consider capacity. I have 12 cells in a 3s4p configuration. Are all the capacities the same? I doubt it. If you charge to 100% then take 240 Ahrs out of my 480 Ahr bank then all cells should be at the same discharge Ahrs. It is highly likely many will then not be at 50% SoC if capacities vary hence voltage variation. Read the Marine how to site or the Nordsky site if you want to see other peeps experiences. 

The next variable is internal resistance. For a given SoC and capacity does the IR change the cell voltage? I don't have a clue but from these other sites ....and the massive 1000 page plus site on the cruiser forum, it suggests voltage will be depressed at higher IR. On my cells, I keep saying that cell 1 is lowest on the plateau but first to the knee. That is a fact an not inaccuracy on the meter. I would like to understand if that is due to IR.

Another 'emerging' theory (which I am following with interest but far from convinced) is the memory effect. This mentioned in the Nordsky site but have spawned a separate thread on the cruiser forum trying to collect data to prove it is real. In summary, as the bank reaches around 80% it 'gets stuck' if charging only goes that far and you can't charge above it. The peeps discussing it look at this as a total bank issue but could it affect individual cells more than others? Not a clue but the theory is being forwarded to explain voltage variation on charging.

 

Are there any other reasons? What about temperature linked to IR and capacity variation.

 

The bottom line is that there are many ways the voltage can and does vary cell to cell ( the series cells) and that is why they go out of balance. They do go out of balance. My BG8S is perfectly capable on monitoring voltage over the whole range As proven by checks against a good multimeter  so don't knock the meter for values you think my be wrong in Mikes data.

Edited by Dr Bob
  • Greenie 1
Link to comment
Share on other sites

Yes i know all that (except for the memory effect suggestion, which seems questionable) but it’s not relevant. If you are going to preface every response with “you don’t know what’s going on” I am going to think that having an article published has gone to your head.

 

Anyway all that is irrelevant to the key point which is that if you look at Mike’s photos of the two types of monitor which as I understand it were taken just moments apart and with no significant charge or discharge, you can see that there are significant differences in voltage, not just absolute voltage but also the relative voltages of the different cells. I can’t see any possible explanation for that except meter accuracy. Can you provide another explanation?

Link to comment
Share on other sites

33 minutes ago, nicknorman said:

Yes i know all that (except for the memory effect suggestion, which seems questionable) but it’s not relevant. If you are going to preface every response with “you don’t know what’s going on” I am going to think that having an article published has gone to your head.

 

You might be right there, but equally you could do your case some good if you accepted there is a possibility as a non-user, you really don't see the full picture. Even if you are right about meter inaccuracies on my meter in particular (I have form for getting sold poorly calibrated meters), it seems likely to me the effects seen by Dr Bob are still real given his meter is known to be accurate.

 

I'll do some more comparison photos in a day or so when I'm back on the boat.

 

Edited by Mike the Boilerman
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.