ATTENTION - To all Li and Li+ Vectrix owners about equalization
Dear all,
Since day 1 I've been having a lot of problems on my Vectrix VX-1 Li+, the latest of which was/still is a severe loss of range.
After several months battling over with Vectrix Europe, and having some interesting feed-back from Vectrix USA, I finally got news today that I'll get a new battery soon.
You can see how my battery behaves on my own Vectrix journal, on these 2 pages, where you can see full discharge and full charge graphs:
http://prius-pt.com/cafe/forums/t/3503.aspx?PageIndex=27
http://prius-pt.com/cafe/forums/t/3503.aspx?PageIndex=28
Now why am I bothering you with this? Because you probably have the same problem as I do, or you are on your way there, and mainly because today I was taught something new by my importer:
1st) There was a large batch of bad batteries coming from GP Batteries that led to this situation. This is mainly a battery manufacturing problem.
2nd) Bear with me, this one is a killer: We need to do LOOOOOONG equalizations to our Li batteries at least once a month. Preferably once a week.
Let me detail this last point a bit better:
At least once a month we are required to put the bike to charge uninterruptedly for at lease 14h straight. Yes, that's right, over 12h without disconnecting from the grid!
There is an important catch however: You cannot plug the bike to the grid to perform the long equalization if the bike is already charged. That won't trigger the long equalization process.
In order to perform the long equalization you have to put the bike to charge coming from your regular commute, i.e. the battery is not full yet, and let it stay there charging for > 14h.
Now this information is not present on my user manual, nor is it in the appendix I was sent by email. My shop did not know about it, and not even my importer knew about it.
The technical manager from my Portuguese importer Masac was in Poland last week for training, and it was only then that he learned about this important fact.
I find it unbelievable that Vectrix knows that people are having all these problems and complaints about their Li bikes and they just sit on top of this information and do not release it in a preemptive fashion.
So take care of your battery, and do a full charge with full equalization once a week to be on the safe side.
Hope this helps.
-jprates
Thanks jprates for this important information!
So we do have be careful with version Li! Now I really understand why so much client want nimh...all know well problems nimh!
and doing this regularly solved your range problems?
Like I said in my post:
today I was taught something new by my importer
I don't know if it will solve my range problems, but I believe my battery pack is beyond recovery.
Did you look at the graphs I gave posted the links about? Did you notice the voltage drop on at least 4 cells is dramatic compared to others?
Vectrix recognized this, so I said again in the original post:
I finally got news today that I'll get a new battery soon
My post is not about solving a dead battery problem, it's about preventing the problem in the first place.
I know English is not my first language, but I thought I had left it clear... sorry about any confusion there.
I think the cells that reach full first (e.g. cells 5 and 7 here: http://prius-pt.com/cafe/cfs-filesystemfile.ashx/__key/CommunityServer.Discussions.Components.Files/127/6567.Full_5F00_charge_5F... ) are not the same as the ones that reach the low voltage cutoff first (as shown here: http://prius-pt.com/cafe/cfs-filesystemfile.ashx/__key/CommunityServer.Discussions.Components.Files/127/4135.batt_5F00_fully_5F0... )
It is however a bit difficult to work out from these diagrams because the colours of cells are similar.
It looks to me like an EQ charge might be all that is needed!
I've heard somewhere that you need to discharge below 90% to trigger the long EQ phase.
I've heard somewhere that you need to discharge below 90% to trigger the long EQ phase.
Just to avoid misunderstnadings:
Consider a battery full of energy-->100%
with a discharge of only 11% you'll reach a level "below 90%".
I think the cells that reach full first (e.g. cells 5 and 7 here: http://prius-pt.com/cafe/cfs-filesystemfile.ashx/__key/CommunityServer.Discussions.Components.Files/127/6567.Full_5F00_charge_5F... ) are not the same as the ones that reach the low voltage cutoff first (as shown here: http://prius-pt.com/cafe/cfs-filesystemfile.ashx/__key/CommunityServer.Discussions.Components.Files/127/4135.batt_5F00_fully_5F0... )
It is however a bit difficult to work out from these diagrams because the colours of cells are similar.
Mik you have graphs there separated by battery module, 20 cells each, I don't know why you insist on going to the 40 cells graph.
Of course having 40 totally different colours is difficult, that's why I did separated graphs on each module...
Anyway you are wrong. You can see on these 2 graphs for battery bank #1 that cells #17 and #19 are the ones that loose power first while discharging:
And you can see at this other graph where those very same cells are the ones that quite easily turn the table, i.e. recover faster and charge first while charging from the grid:
It looks to me like an EQ charge might be all that is needed!
I wish. I'm not an expert, but not that dumb either. I've done at least 3 complete discharges and full charges with full equalization with no visible effect.
It is not normal for a cell to go under 2,5V while others are still well over 3V in the same pack.
Anyway, moving forward, I've done one full charge yesterday, starting from about 25% SoC only, and let the bike be connected charging for over 12 hours.
The result was disapointing to say the least. Nothing extraordinary happened. The very same cycle toke place, just like the one detailed on this other post of mine (see first plot graph):
http://prius-pt.com/cafe/forums/p/3503/32107.aspx#32107
So this thing about letting the bike stay plugged over 12h to make sure a full equialization takes place is starting to seem like bulls**t from Vectrix Europe to me.
Either there is something else that needs to happen at the same time to trigger the full equalization, or this is just another of their lies.
I'm so sick of this bike...
Hi,
I own a Li+ myself, and it took me a couple of months to understand the way it works. Yes, I experience too the long charges (12 to 14 hours), but I have a pretty good range. I am not worried as I can get over 100 km.
LiFe batteries are pretty simple. Either you bottom balance them, or top balance them. Vectrix chose the latter as it is easier to engineer by the means of BMS, although bottom balancing is safer in the long run.
That said, you shouldn't have to balance them in a regular basis, as they SHOULD be of the same capacity. Or within say, 3% of the same Ah. Jack Rickard (St Jack) have tested Sinopoly LiFePo4 modules for several years, and they stay in the same voltage range naturally without BMS.
If we need to top balance our Vectrix every week to keep a good range, I am afraid that he inconvenient truth is, or appears to be, that GP batteries are of VERY different capacity and not well matched within a pack.
I would like to have Antiscab (an other V 'masters') input on this, but I am afraid it is so.
I am not complaining as my own range is pretty good. I avoid, on the contrary, long eq charges not to over-heat BMS. I usually do 3 hours normal charges (up to full bars). I do an EQ charge once a month. By doing long full charges each time, you're dissipating heat around already charged cells (at 365) for hours, waiting for low cells to charge slowly at 0.5 amp. I chose to balance my cells once a month but they never appear to be balanced... only of different capacity.
You are correct, jprates!
I did not look very carefully at the graphs. It's even visible in this one graph here: http://prius-pt.com/cafe/cfs-filesystemfile.ashx/__key/CommunityServer.Discussions.Components.Files/127/4135.batt_5F00_fully_5F0... that the same cells that have the lowest voltage under load reach the highest voltage during regen braking. I guess that means their IR is increased.
Hi,
I own a Li+ myself, and it took me a couple of months to understand the way it works.
I have mine since September, and I still don't understand how it works.
A good user manual would help, instead of a useless NiMH manual what tells you little about a different beast altogether.
Yes, I experience too the long charges (12 to 14 hours), but I have a pretty good range.
I don't remember seeing mine ever doing it. But once I plug it in in the garage I just walk away, so who knows...
I am not worried as I can get over 100 km.
I also had it in the beginning, but now I suffer to get 50 km out of it.
I'm almost at 15.000 km now, within 9 months of running.
Some days I do well over 200 km per day, so hence the idea of doing it once a week.
Vectrix says AT LEAST once a month, meaning the more time you spend without doing it, the more prolonged the eq process will be.
I usually do 3 hours normal charges (up to full bars)
If the battery is depleted it is impossible to charge up to full bars in only 3 hours, at least on my bike.
Anyway mine stays plugged in every night from 00:00 to 07:00, more than enough for a full charge.
Since I do it everyday, I don't think a top up and eq process is impossible on that time. My guess here.
Anyway, I was looking at the History Info in the bikes memory, and isolated the last charges where it went to full charge (gauge >= 254).
This is what I have on my last 3 full charges:
Time Stamp On Time Distance AmpHrs Fault Code Fuel Start Fuel End Charge AmpHrs High Bat Temp Hi Cell Volt Low cell volt Ride Time
11/21/00 07:33:53 0 0 0 93 254 26,52 46 3,67 3,19 00:00:00
11/18/00 04:04:10 0 0 0 134 254 19,76 48 3,67 3,22 00:00:01
11/15/00 05:22:07 0 0 0 89 254 27,18 48 3,67 3,19 00:00:00
What bugges me is not understanding why on earth the Low Cell voltage at the end of the charging sequence is only about 3,20V.
This does not seem like a full equalization. The top (Hi Cell Volt) is fine at the max, but the Low Cell is too damn low. WHY?
Another thing that I don't understand is the way the charging process takes place.
If you look at this post graphs for each battery module and its 10 cells being charged from empty, you'll see what I mean:
http://prius-pt.com/cafe/forums/p/3503/32981.aspx#32981
Why do the cells go up to some point almost in pair with one another, then suddenly they reach a peak, and each one starts coming down, and then slowly back again up, but at this later stage each cell goes up on its own pace?
Why did the charger not stop when they were all up in the first place?
Why does the voltage drop at some point so abruptly?
Something tells me that this last stage after the peak is the equalization phase.
But I can't seem to understand why each module behaves the way it does.
For example at module #0 one can see cell #5 almost keeping the voltage constant at the top 3,64V but on the same module you can see cell #10 coming down to almost 3,42 and then slowly going up again.
Does this drop of voltage occur when the charger changes from Constant Current to Constant Voltage?
What do you experts (compared to me which am an admitted ignorant) think?
jprates..
If you had the current on the plots it might make a little more sense. The charger outputs 10amps (CC) up to the point where the first cell reaches ~3.4V. This triggers the transition to constant voltage. The dip in voltage (current) insures that a fully charged pack does not overshoot the max voltage of 3.65V (side note...these voltage values will be different in the future). After the dip the current will slowly increase while keeping the max cell below 3.65V. When the first cell hits 3.65V the current will begin decreasing in order to maintain 3.65V. (thus constant voltage). Many of the other cells will droop under this condition because they are not as charged as the cell at 3.65V and can absorb more current than is available. Any cell that reaches 3.65V will have its discharge resistor ON, which can bypass 300mA around the cell. Eventually the current will drop down to 300mA as it tries to maintain the highest cell(s) at 3.65V. This is officially the EQ phase. Any cell at 3.65V bypasses the 300 mA and all the others charge at 300mA. They all equalize at different rates which is the spread you see at the end of the cycle. Eventually they all reach 3.65V which ends the charge cycle.
Indeed the DC current reading would help, but I've no way to get it with my tools.
I'm "limited" to the Vectrix Diagnostic software "service tab" which records individual cells voltage, and to a CurrentCost EnviR IAM unit that records AC grid current at the plug.
Do you know of a way somewhere in the Diagnostics software to record the cell currents? If you do, please let me know.
Your detailed answer is simply excellent.
Now I can understand what's going on and why the spreading of voltages when entering EQ stage.
Thanks a lot.
Indeed the DC current reading would help, but I've no way to get it with my tools.
I'm "limited" to the Vectrix Diagnostic software "service tab" which records individual cells voltage, and to a CurrentCost EnviR IAM unit that records AC grid current at the plug.
You can use the free Canhacker software to get the information.
I don't know if (and how much) the CanBus address list and general setup might have changed from the NiMH version.
If it has not changed, then you can use your PEAK Can reader and Canhacker to get all this information.
If the Canbus info has changed, maybe you can figure out what has changed? I guess the main CanBus ID's may have stayed the same.
You can find this out for free (because you already have the PEAK Can/USB adapter).
See this thread for how: http://visforvoltage.org/book/ev-collaborative-hand-books/6916#comment-40025 (Unfortunately I was just poking around in the dark and sometimes used the wrong technical terms to describe what I found. Someone with a computer programming background could work it out and describe it accurately. Such a programmer could also build a device to log (or display in real time) the voltage, battery temperatures, current, speed and a number of other Canbus messages).
The Canbus ID to watch is this one: ID00FEF105
BYTE 2 of ID00FEF105 shows SPEED during riding and (charge current x 10) during charging
.
Forward range:
09 =9km/h
0A=10km/h
10=16km/h
12=18km/h
18=24km/h
1D=29km/h
22=34km/h
27=39km/h
31=49km/h
3C=60km/h
46=70km/h
50=80km/h
5A=90km/h
64=100km/h
66=102km/h
68=104km/h
.
Reverse range: 00 to 04, meaning 0 to 4km/h.
ID00FEF105 during charging with the NiMH Vectrix:
6C = 108km/h (CP charging at 10.8A)
1C = 28km/h (CC charging at 2.8A)
0D = 13km/h (EC charging at 1.3A)
That would be interesting Mik.
Unfortunately the data cable is not mine, it was borrowed from Masac (local importer/distributor) so that I could do all the tests and recording without loosing my bike.
I need my bike for work and for personal transportation, I cannot be without it. And I was tired of having to travel to the shop to take one memory dump after the other.
I'm supposed to return the cable ASAP now that everything is determined.
Let me just finish this thread on my part by telling all Li/Li+ users that there is a BMS/Charger/Controler software upgrade pending that should be released soon.
I VERY STRONGLY advise all of you to update your bike to this new software version. Your battery will appreciate it. Sorry I can't go into further details.
To me this sounds like they again failed to properly develop and test their software before launch.
Both the software issues with NiMH and Li have and are costing them a lot of money.
Maybe Vectrix should do some "design for six sigma" training...
To me this sounds like they again failed to properly develop and test their software before launch.
Both the software issues with NiMH and Li have and are costing them a lot of money.
Maybe Vectrix should do some "design for six sigma" training...
Ridiculous, isn't it? They had all the time in the world to do a proper job of it. And this time, it's owned by the battery manufacturer, isn't it? You'd think they'd know how to charge and maintain their own batteries.
And this time, it's owned by the battery manufacturer, isn't it? You'd think they'd know how to charge and maintain their own batteries.
Or rather they tried to charge in a way that would make their cells fail massively just beyond normal warranty time in order to increase their profit?...
Or rather they tried to charge in a way that would make their cells fail massively just beyond normal warranty time in order to increase their profit?...
I'd like to think not, but who knows.
I won't go that far into the plot theory. If I were CEO of Vectrix, I would certainly prefer happy customers and reliable products that construct a good image brand in the long run. Happy customers are happy sales persons, and vastly help increase user base and sales. And doing so you make far more money than expecting failures just after warranty, why would give you only a few thousand euros in the very short run and bad publicity. I think we should all keep reason and stick with technical issues and how to understand and deal with them, rather than fighting an improbable conspiracy.
Sorry, I should have written :
which would give you only a few thousand euros in the very short run and bad publicity.
You are correct, Paco :-) It is more tongue-in-cheek what we are colluding here, because we also think happy customers will create far more business in the long run than short-sighted timed failure modes.
But it is hard to understand that New Vectrix seem to be going the "Chinese way" (*) of little development and only very basic testing at best and letting the customers sort out the problems, instead of doing their homework in the first place, before they launch the product.
.
.
.
(*) Of course not all Chinese companies show this sloppiness, but a lot of those involved in electric 2-wheelers certainly seem to work that way...
I'm getting sick of this crap.
Please see posts starting on this one:
http://prius-pt.com/cafe/forums/p/3503/33325.aspx#33325
The part that matters is in English.
What kind of company is this?
What kind of company is this?
All this is very sad: client of Vx Li are a rat of Lab!!!
Thanks all for your comments.
It is quite simple really, now that I have a full understanding of what was going on:
GP Batteries doesn't have a clue of how to charge its own cells. They just know how to build them. Period.
In the previous software version Vectrix relied on GP Batteries info to build the charge algorithm and motor controler for discharge, etc.
They just blew it. The too high currents on acceleration and the high charge algorithm would kill the batteries in no time.
This time they took the time to come up with Vectrix own algorithms based on field testing.
Will this field testing be enough? We're about to find out. The new version is out now.
The customers with problems will be the first ones served.
I'm told by my importer my new battery and software upgrade will arrive in Portugal by July 12th.
I hope Vectrix R&D team has come up with a good algorithm this time, but I feel totally disappointed mainly by 3 things:
1) Vectrix Europe is not 100% honest/transparent to its customers nor to its distributors;
2) GP Batteries as a main battery manufacturer should know a lot more about the usage of a product they build;
3) One has to battle too much for my linking in order to get things working even within the warranty period. What if my bike was out of warranty?
This makes you wonder what they were smoking when a short while ago they proposed to install fast chargers in the Li bikes...
Thanks all for your comments.
It is quite simple really, now that I have a full understanding of what was going on:
Are you sure? Do you really beleive their story, or is it just a way to clear the name for Vectrix.
I think it is only a cover for getting it wrong again.
My personal interpretation:
Their range claims were again too optimistic for only 30 and 42 ah and in order to get to the range claim they made the same error as with the original NiMH. There probably is not too much wrong with the charge algorithm from what I read from the X Vectrix posts. So the damage is done by the discharge in order to meet the range/accelleration claims.
My guess:
The new software will reduce max current during accelleration and raise the low voltage mode.
Sounds familiar?
I hope I don't offend anybody but I just wanted to share my thoughts which are totally not supported by any measurements just interpretation of posts on various forums.
Are you sure? Do you really beleive their story, or is it just a way to clear the name for Vectrix.
I think it is only a cover for getting it wrong again.
No I'm not 100% sure, because like you I also don't have proof. It's my personal opinion as well.
Anyhow I'm relying on someone with "internal" knowledge on Vectrix who up to this point has been absolutely impeccable helping out in several ways.
Please do understand that what I said is not Vectrix official position, it is my understanding of several sources of information.
So far in my official email exchange with Vectrix no acknowledge of error on their part what so ever has been seen.
On the other hand Vectrix is owned by GP, so saying that GP is to blame is not clearing Vectix name at all.
I think it is a disgrace that GP being one of the largest battery manufacturers does not know how to use their own batteries.
If they did know, I think they would be the first and foremost interested party on making sure Vectrix was depleting and charging the battery correctly.
On the other hand let me tell you that you are wrong about the range assumption.
The new software, with totally new cell voltage limits (lower and upper), new max current draw from the battery, and new charging algorithm results on MORE RANGE from the same battery.
We'll have even more range than before, not less range, with the same cells but totally different software.
As always, I don't totally agree. I understand that people with non-functioning bikes feel angry. But we shouldn't forget facts.
GP batteries have a chemistry (LiFePo4) and a spec sheet that gives information on charge, discharge voltages and rates.
LiFePo4 has proven to be a reliable all-around chemistry with hight number of cycles.
So GP batteries might (I say MIGHT) have problems regarding matching same capacity.
If you "understand everything" you should know GP spec sheets. I think it should be these :
http://www.evbtech.com/lifep04-cells/
We can see that 3,65v is pretty high for charge.
My Li+ has red warning under 3,00v which is pretty conservative (fact).
Li+ acceleration has be reduced compared to 2009 NiMh software which had already be reduced from 2008 - 2007 softwares (fact).
So what do you call exactly an "algorithm"?
They can reduce acceleration but why, it is far more conservative than all previous versions...
Why should they reduce low voltage ? 3,00v is ok (recommended 2,00v cutoff on spec sheet).
So what's wrong ?
How to brake a LiFePo4 cell ?
1. Overcharge (fact);
2. Overdischarge (fact);
3. OverHeating (fact).
Speed of charge curve MIGHT decrease overall life but not so as to see dead cells in a 6 months time...
So, if we want to learn, we should remain close to facts and away from anger.
I think they might reduce high voltage charge point. I bet on that. They will stop at 3.55v or something. To go from there to 3.65v, you ad up much time, heat, and pretty insignificant range.
I am afraid to say BMS might be part of the problem. Or be the problem. What if they fail due to heat? It is not safe to have those circuits giving heat on top of LiFePo4 chemistry for 10 hours to balance cells...
So I am pretty kind of optimistic, it is easy for me as my Li+ works like a charm, touching wood.
But the truth is :
1. Vectrix is the first and ONLY 125cc + scooter in the world.
2. Every company have faced major problems trying to put such a bike on the market (BRAMMO has delayed a lot, XO2 has delayed a lot, etc.).
3. 30 000 km / year is a lot to ANY bike, let alone raising technology. I know a couple of TMAx that have reached 100 000 km, but quite worn down (and in 8 years). 100 000 km is what you would do in 3 years. In France, I know Vectrix which have more than 35 000 km but since 2007 (NiMh).
I am afraid that we are lab rats. But it is not only Vectrix's fault, it is so. Nobody has tested LiFePo4 battery for 10 years. Ever. Vectrix hasn't. Nor GP.
I have tested lead-acid batteries on a mope. They lasted to 8 000 km with good range. And to 10 000 km with 80% or less range. They last one year at that rate. I can write a lot on that. I have tested that for 4 years. I have only be using my Li+ for 8 months and 8000 km...
But we shouldn't generalize on one bike. Your Li+ have a problem, but claiming that it is a known fact that touches every bike is the wrong answer to an uncomfortable position.
Assuming your post is somehow an answer to my posts, and not to others (a quote would help) I'll stick to the facts:
1) I said I knew what happened simply because I had feedback from both my importer and from one guy inside Vectrix who were honest to the point of hiding nothing from me;
I did not tell you that I knew the technical details behind the problem. The only thing I know for a fact, and I quote one Vectrix employee on this, is that "High charge cycles cause higher impedance in the batteries which will not be noticed until later and then it will be too late". This is a fact. I can see the effects on my recorded data. Period.
2) I know the tech sheet. The tech sheet does not tell you on what voltage and what current to charge, it does not tell you for how long, it does not tell you many things you should know about. Using your own explanation, it does not tell you if you should charge at 3.55 or 3.60 or 3.65. Got it? This is a fact.
3) I could not care less for what your bike does if it is fine. Enjoy it while it lasts. Why? Because having one bike working fine proves nothing, it does not prove the bike is well designed. Fact.
4) You can search the internet for the definition of algorithm. I used to be a senior software developer for over 20 years, so don't mind about me knowing what an algorithm is. Do you know it yourself?
5) I don't work for Vectrix, I don't know the details and "why's" of all the changes. I was only sharing with this community what I knew told by Vectrix employees and my importer.
6) I don't know about previous Vectrix bikes. I only have a Vectrix since September, and my life has been a nightmare with it. That's a fact.
7) "We shouldn't generalize on one bike"??? Are you kidding us? Have you read the previous posts or is there a problem with the English language? Here's some facts for you:
a) Several Li+ owners here in Portugal are experiencing the same exact problem, and we saw references here to the same in Barcelona/Spain = FACT ; source = myself and this forum;
b) One Vectrix employee has admitted the charging algorithm was causing a raise in internal resistance = FACT ; source = email from Vectrix employee;
c) My importer tech responsible was recently visiting Poland on training and confirmed the problem with them = FACT ; source = Masac importer on phone call;
d) Vectrix Europe has been stalling me and others who were complaining about the same problem, waiting for the new software to be released = FACT ; source = Masac importer and Vectrix employee;
e) Production in Poland was stopped for several weeks because they did not want to sell more bikes without replacing the software = FACT ; source = Masac importer;
f) Production has restarted now that the software was released, and they have a lot of catch up to do with delayed orders = FACT ; source = Vectrix employee;
Is this enough facts for you or what?
Can someone from admin or mod please delete this thread as it is duplicate?
My browser hang and apparently the refresh submitted the post once again.
Appologies.
jprates