Charging: Are current Begode / Extremebull EUCs allowing battery balancing to happen?

[Cam]

36 minutes ago, Planemo said:

OK so what function are you saying the pack comms wire/cut on the distribution board is for? Simply an over-voltage on the input?

I have no idea, I'm just saying it (cutting the whole pack) can't do the same thing as cutting groups as they reach a threshold (which I believed to be classic as first part of balancing, I could be wrong..)

Regarding the thread linked by Bizra6ot it looks like there isn't a firm conclusion, except if I missed something.

edit: Here under the spoiler are the photos of the ecodrift teardown of the first Masters 33.6V pack (uses the same family distribution board). The article itself mentions bleed, but not cutoff (neither presence or absence).
 

Spoiler

 

 

 

Edited March 9 by Cam

[Planemo]

2 minutes ago, Cam said:

I have no idea, I'm just saying it can't do the same thing as cutting groups as they reach a threshold (which I believed to be classic as first part of balancing, I could be wrong..)

OK then I think I'm lost on what you mean. What I'm suggesting is that the BMS can balance just fine without a charge board if the cells are relatively well balanced to start with. If one has drifted off (eg higher than others) and charge current is still relatively high and the high cell hits 4.215v and tries to bleed off, can it bleed off enough of that input current or will it's voltage keep climbing? If not, 4.25v is when the charge cut off comes into play to save the day because without it I'm assuming the bleed resistor will go into meltdown and/or the cell voltage will keep climbing.

I don't actually know how much current the bleed resistors can cope with, hence theres still a lot of 'what if's' in the above.

2 minutes ago, Cam said:

Regarding the thread linked by Bizra6ot it looks like there isn't a firm conclusion, except if I missed something.

No, you didn't miss anything. I only mentioned it to say we have been here before with this question. Oh and because there was a pic of a Master BMS on there but it still doesn't make the situation conclusive. Without circuit diagrams or a physical test as per @RagingGrandpa comments I'm not sure we will ever get an answer either.

[Cam]

We might be a bit lost indeed, and I do not have enough solid knowledge to affirm anything on typical balancing.

From (the little) I know of bleed it happens pretty slowly, so yes I suppose it wouldn't cope with charging currents.
Speaking of which I worry that the distribution board lowers the voltage too fast even for bleed to do much ballancing ..

🤷🏼

I would really like to know if BG/EB have a plan here or is everyone is just surfing on their cells not getting out of balance..

Edited March 9 by Cam

[Planemo]

33 minutes ago, Cam said:

From (the little) I know of bleed it happens pretty slowly, so yes I suppose it wouldn't cope with charging currents.

Thats my understanding too.

33 minutes ago, Cam said:

 

or is everyone is just surfing on their cells not getting out of balance..

I think that could indeed be whats happening. Or more accurately, surfing on the cells not getting too out of balance, as a BMS without a charge stop will cope just fine if not. The sole purpose of a charge stop is to be there only for when something isn't right.

[Cam]

3 hours ago, Planemo said:

The sole purpose of a charge stop is to be there only for when something isn't right.

One could suppose so yes, but then it might not be well implemented..

edit: It seems to me like it just makes matters worse by not allowing proper balancing. Which could lead to more cell over voltage risk as the cells drift apart..

Edited March 9 by Cam

[Chriull]

15 hours ago, Planemo said:

No they haven't, which is the problem. It was requested that Roger try to establish this given he was the founder of the bypass mod but to my knowledge neither he nor anyone else has confirmed re the charge stop question.

That's a pity

15 hours ago, Planemo said:

What we do know is that these later wheels have a comms wire to the board which the older wheels didn't have.

That's the white wire shown in the master bms pics labeled b-co?

In @EcoDrift's  teardown linked by @Cam one sees the com wires beeing part of the connector beeing plugged in the power distribution board. And @EcoDrift mentions that with any battery disconnected the controller stops working - so the missing  "b-co" signal of any battery could shut down the wheel. Don't know if the ebcp works with just half a battery ((2s16)p1 instead of (2s16)p2) or if this just disables charging? Or it is just unaffected?

As the picture on aliexpress shows the white signal wire is not forwarded to the motherboard 

15 hours ago, Planemo said:

We also know that there are mosfets fitted to the distribution board which appear to be very similar to the charge cut fets that were mounted to Begode BMS's of old. Alternatively, these could be charge-on fets for the input.

Seems these are used to cut off prematurely charging ?at some fixed voltage? As without the power distribution board charging happens up to the charger voltage (within bms allowed range).

Could be that this was indented as secondary safety measure to prevent overall pack overvoltage in combination with draining down to this threshold. Unfortionatly this voltage threshold happened to be too low and now hinder proper charging. So it "perverted" from a safety measure to a hazard by preventing balancing...

And additionally the b-co signal of any bms is used to shut down for all packs. 

15 hours ago, Planemo said:

We simply don't know until someone, somewhere, pulls apart a fairly recent Begode BMS to see whether it looks like theres any charge cut fets on it.

The master bms pictures show 6 mosfets at P - after shunts. So this is firstly a very strong indicator that now begode bms have implemented overcurrent shutoff.

The two black wires seem to be temperature sensors. Used to shut down charging/discharging, signal via the b-co wire or just to start the ?buzzer? on the bms?

The balancing and monitoring of the cells and the temperature is handled by the two ICs. Two tracks on the pcb lead from these ICs two the "bulb of components" which perform the "logic" of b-co signal and the mosfet switching.

Without knowledge of these ICs and reverse engineering the "bulb of components" it cannot be said how the three input signals overcurrent, single cell (group) overvoltage and temperature are distributed to the mosfets, b-co and the ?buzzer?

14 hours ago, Cam said:

Cell group over voltage cut is part of CV phase balancing AFAIK, the cut of the distribution board cant do that function. I doubt (even Begode) would drop the former because of a second function that is not the same.. 

Cutting charge off all packs once on pack shuts off is an important measure for safe charging!

Imagine for your example one of the packs cuts charging in because of cell overvoltage. For the second pack in this series string charging is stoped too as no current can flow anymore in this string.

But the other string notices nothing of this and continues charging.

Over time by discharging the cell with too high voltage the overvoltage condition is cleared and the "charge cut off" reset. So these two, less charged packs are suddenly connected again to the other two fully charged packs!

With some 20 mOhm internal resistance per cell one has about 0.16Ohm in this 32s4p config. For some volts difference for a "slightly/medium" unbalanced pack the equalisation currents should be ok - for severe unbalanced packs with "dead" cells they could likely fry sone pcb tracks/components...

Another possible hazard could occur if one rides immediately after such a charge and quite fully discharges the one fully charged string. If the overvoltage condition is clearance in the second string took as long, one has about the maximum voltage difference of almost (4.2-3.3)V x 32 = 28.8V. Over the internal resistance of about 0.16 Ohm this leads to a (peak) current of up to 180 A! Nothing nice will result from this

 

13 hours ago, Cam said:

Regarding the thread linked by Bizra6ot it looks like there isn't a firm conclusion, except if I missed something.

Just read this, too. The concerns of @RagingGrandpa seem valid. There could be a good chance overvoltage cutoff is not implemented in the BMS.

13 hours ago, Cam said:

edit: Here under the spoiler are the photos of the ecodrift teardown of the first Masters 33.6V pack (uses the same family distribution board). The article itself mentions bleed, but not cutoff (neither presence or absence).
 

  Reveal hidden contents

 

 

 

Thanks for the link an the pictures!

Tldr:

After direct charging i'd strongly suggest to check the battery packs voltages to ensure non cut off charging. If the voltmeter shows strangely low voltage values this should be because the mosfets are still switched off. They perform a very high resistance voltage divider together with the internal resistance of the voltmeter. So half the supposed voltage values or less are easily possible.

To address the possible not existing cell overvoltage protection one could check if one of the packs voltages drops during "some" hours after charging more than the others as first "warning" sign every couple of charges.

Otherwise reconnecting the batteries should not be a problem.

Besides the power distribution board maybe draining again the packs?

Personally i'd disassemble and reverse engineer the power distribution board - does not seem to complicated:  

 

[Planemo]

2 hours ago, Chriull said:

That's the white wire shown in the master bms pics labeled b-co?

Yes.

2 hours ago, Chriull said:

As the picture on aliexpress shows the white signal wire is not forwarded to the motherboard 

Correct. So we know that the comms wire only goes as far as the distribution board. However we also know that the dist board is capable of cutting charge current. So it wouldn't be a stretch to imagine that the comms wire also triggers the charge stop on the distribution board for a case of string over voltage.

At this moment, my feelings are that the fets on the BMS are used solely for overcurrent cutoff, not string over-voltage cutoff, and that the string over-voltage cutoff is handled by the comms wire > dist board > charge input shutoff.

Ergo, there is no string over-voltage protection within the packs themselves.

But I would love to be proved wrong so I could be happier in bypassing the dist board, getting to full charge and likely better balancing too.

[Cam]

Thank you for the analysis and all the information @Chriull 

2 hours ago, Chriull said:

After direct charging i'd strongly suggest to check the battery packs voltages to ensure non cut off charging. If the voltmeter shows strangely low voltage values this should be because the mosfets are still switched off. They perform a very high resistance voltage divider together with the internal resistance of the voltmeter. So half the supposed voltage values or less are easily possible.

To address the possible not existing cell overvoltage protection one could check if one of the packs voltages drops during "some" hours after charging more than the others as first "warning" sign every couple of charges.

Yes I will check for that. Currently there is only 0.1V difference and it's not moving. I will see how / if this evolves after charging by bypassing the board. With some luck they should align after some cycles. I just need to wait some (long) weeks for the MR60 connectors.

 

2 hours ago, Chriull said:

Besides the power distribution board maybe draining again the packs?

Personally i'd disassemble and reverse engineer the power distribution board - does not seem to complicated:  

Yes the drainage seems to happen with everyone anyhow. This is why, if this doesn't work (charging via the other side of the board) I will try the packs individually with a 67.2V charger. This way the board won't discharge them before they get the time to balance. I hope I won't have to do this though.

If ever I were to replace the whole distribution board I will disassemble it and take good pictures of everything.

Thanks again.

[Planemo]

4 minutes ago, Cam said:

This is why, if this doesn't work (charging via the other side of the board) I will try the packs individually with a 67.2V charger.

Why not just cut out the middle man and go straight to charging each pack at 67.2v to start with?

I am very tempted to do this myself, it's only the potential lack of string over-voltage cutoff I'm nervous about but that still applies if you're charging all packs via bypass anyway.

[Cam]

3 hours ago, Planemo said:

Why not just cut out the middle man and go straight to charging each pack at 67.2v to start with?

I am very tempted to do this myself, it's only the potential lack of string over-voltage cutoff I'm nervous about but that still applies if you're charging all packs via bypass anyway.

Yes it would have the benefit of (probably) being more efficient. Ultimately I would prefer doing that. The reason I don’t plan to do that first is that it would require a lot of fiddling with the battery connectors (they are so stiff and high current), also because I’m not certain of the voltage of my 67.2V charger who outputs more than that due to being set for a KS16S.  Can do but not handy.

Edited March 10 by Cam

[Planemo]

1 hour ago, Cam said:

 it would require a lot of fiddling with the battery connectors (they are so stiff and high current),

Not sure I follow - the supply connector would be the same whichever way you do it and unplugging the packs from the board is easy.

1 hour ago, Cam said:

also because I’m not certain of the voltage of my 67.2V charger who outputs more than that due to being set for a KS16S.  Can do but not handy.

Fair enough, I thought you may have had a server charger or similar.

[Cam]

1 hour ago, Planemo said:

Not sure I follow - the supply connector would be the same whichever way you do it and unplugging the packs from the board is easy.

My battery connectors are so stiff I worry about breaking them off the board if I do it too many times. I’ll do it if I have to, but pulling my nails off (*1) on a badly shrink wrapped high DC connector (*2) while risking to tear off the whole socket of doesn’t inspire me..

*1: I can’t manage to grab and pull properly with nitril gloves
*2: The transparent wrap doesn’t go all the way down, exposing the solder. (Thinking of adding some Kafuter silicone)

I have some contact cleaner (PRF 6-68 and PRF 7-78) one of which is lubricant maybe that could help, I'll have to check the specs. They will probably have to be disconnected to re-check the voltage at some point anyway.

Edited March 10 by Cam

[Planemo]

On 3/9/2024 at 1:06 PM, Cam said:

 

Just happened to revisit this thread and noticed the diagram above is one way of doing it or you could simply switch the negative as I recently did. I didn't want to permanently wire the bypass (given we're still not sure re string over voltage cut off) nor have one charge port bypassed and one not bypassed because I prefer to use both ports together whenever over 5A.

I simply ran two lines of 12AWG silicone, one from the solder point on the main negative feed on the lower board that comes direct from the linked charge ports, and another solder point (again on the lower board) that the main motherboard plug (negative side) is fed from.

I haven't actually charged up using it yet but my OCD was struggling with only having one switch on one side of the wheel for my LED's so I wanted to even it up and a bypass switch seemed like a good option lol.

After thinking about it a while I do wonder how it all actually works because even with the bypass I assume something must still be going through the logic board because each pack port reads 67.2, not 134.4, which was what I was initially worried about when thinking about how this is supposed to work. It's prob pretty simple but my head kept saying 'you're gonna be feeding each pack port 134.4v....don't do it!'. If someone can clear up my confusion it would be appreciated!

Anyway's, I'm not convinced I'll be using it much except for test purposes, just to see if it does allow a full 134.4v charge but more importantly, if it allows a reasonable fast charge (10A) without shutting down at like 75% as has been reported without the bypass. If it does I may end up using it whilst out and about but will always revert to the stock method when charging at home. I'll report whatever findings I get.