Cell Cycling and Matching

Dave you can save time by doing the first two cycles charging a pair of modules in series (14.4 volts nominal), but keep the charging to modules spaced in the pack. The Hitec's are able to do this if the supply voltage is available.

John (Britprius)

 

As greasemonkey007 suggested unless people have a dedicated DC power supply like the OP did then the Hitec X4 AC plus versions are certainly easier because you dońt need more than 50w per channel to do the cycling and it´s alot easier to deal with when each has their own AC power supply built in.

I´d be very curious if this works well and how the charger settings need to be modified to make that work.

 

When a single module goes bad. Why do the other 25 modules need rebalancing?
Those remaining modules are allready ballanced. The hv ecu does module pair ballancing as well as module pair voltage monitoring.

Toyota puts alot of effort to keep batterys between 40 and 80% charged.
What sence is there for prius battery rebuilders to cycle 6.5ah modules to 7.0ah and above?
Cycling all ready ballanced batteries well beyond 100% capacity!

This is fine business for rc cars where 50 or 70 races per battery pack is acceptable?

Once the 40-80% charge capacity is exceeded the indivual cells in the module become unballanced.


I think the battery rebuilders would have better results with brand-new modules.

My tests with nimh 2400mah battery charged at 50mah for 48 hours generates no heat.
However the 49 and 50 hours of continued charging (over charging) the cell starts heating up
This heat is all from chemical reaction. The cell resistance decreases during this chemical reaction resulting in an increase in charge current and a thermal run away condition.

I think an alternate methode for ballancing a prius hybrid battery would be to simply put an approximate 232 volt 150 mah source acrossed all the series connected modules.
For an approiate amount of hours to bring the lowest voltage module equal to the others.
As the fully charges modules go into chemical heating (minimal) there cell resistance decreasing forces more voltage/current across uncharged cells.
This chemical heat reaction is far far less than the heat generated by battery rebuilders force cycling the batteries in excess of there rated capacity.

A spent personal computer power supply would be a economical source for 250 plus volts for charging prius batteries.
The pc supply ac is rectified and doubled acrossed two 200 volt large capacitors that are in series giving near 400 volts. The voltage could be dropped through some 100 watt incandessant light bulbes to reduce voltage and current. Some expermentation will be required to find the combination of required light bulbs to provide a 150mah hv charging source.
DONT TRY ANY OF THIS WITHOUT HAVING 10 YEARS ELECTRONICS TROUBLE SHOOTING SKILLS ! ! !

 

When a module goes "bad" it goes bad because generally one cell in the module has become out of balance with the rest. When the module is discharged that cell can be reverse charged rendering it useless.
To be able to say that after x tens of thousands of miles and countless charge and discharge cycles that all the other cells in in the remaining modules are still at the same state of charge (162 cells) would be miraculous.
Very small differences in the ability of individual cells to convert electrical energy into a chemical change, and back into electricity will unbalance them. Working on only 1 ma difference in efficiency "or 1/6500" of the capacity of each cell would accomplish the destruction of that cell in a short period of time.
Computer power supplies only rectify the incoming voltage to DC it then charges usually two 400 volt working capacitors to whatever the RMS value of the supply is.It does not double the input voltage. The capacitors are rated at 400 volts to allow for the RMS value with a safety margin of supplies such as in the UK at up to 250 volts AC.
The DC voltage is then turned into a high frequency variable pulse width supply into a transformer to lower the voltage and this turns the DC back to AC that then has to be rectified to DC and smoothed with capacitors again. This voltage is monitored and feedback to the high voltage side of the circuit keeps the output voltage at a constant level.
The mains supply can usually be anything from around 90 volts AC to 260 volts AC, "but in fact will work happily on DC" and the output will remain constant.
All this makes it easy and cheaper, and with less weight "copper" for the manufacturer to make a supply that will work on almost any mains AC supply of any frequency.

John (Britprius)

 

John your description is accurate for your countries 250 volt mains.
Rectifier charging two 400 volt capacitors . obviously capacitors in parallel.

Over here on 120 volt mains there are two 200 volt caps in series in our pc supplies.
The full wave bridge rectifer is configured like two half wave rectifers.
Each half wave rectifer portion charging a seperate 200 volt cap. Resulting in what I call a voltage doubling effect.

The 115/230 volt input selector switch configures (pc switching supplies) the supply like you've described (230 position) and in the 115 position like I've described.

Both input selector settings resulting in something like 325 vdc being presented to the dc-dc down-converter.

 

The power supply units we tend to get here follow the lines of the small brick units for lap tops. They are rated "without any switching" from 90 volts to usually 260 volts AC 50 or 60 cycles. I use a laptop version as an emergency 12 volt battery charger from the Prius HV battery. The only modification being an LM317T being used as a current regulator on the output to charge the battery in the event of something being left on discharging it. Pushing a button brings in a control relay to start the charging, and a ready signal cancels it.
I use the voltage doubling you describe on our 250 volts supply for my photographic studio home made flash system.

John (Britprius)

 

You are absolutely correct, and this has been done for years on the Honda cells. 350mA is an ideal power level as it allows for a 24 hour empty-to-full charge & balance while not generating excessive heat. HV battery fan is powered by external 12v PSU and used to clear any heat accumulation. The same technology will be released for the Prius in a few weeks, just need to finalize the HV cooling fan control unit. The FET is getting hot in high-heat thermal testing and needs a little cooling of its own.

 

Hello everyone, I am new here. For the last 2 weeks, I have been perusing the PC forum for DYI information about HV battery issues (I have a 2007 with the triangle ! and a P0A80 code). There is so much to learn in this space. So thank you all.

I have a few questions and would appreciate your suggestions very much. For the purpose of looking for a bad module, will performing the discharge/charge cycles in pair (two battery modules in series) still be a good idea? And why did you recommend doing all this for the first two cycles only? Would you mind explaining that? I'd imagine that it will work for all three cycles, won't it?

Suppose I want to do the setup in series as John (Britprius) suggested, does the following settings make sense?

Battery type: NiMH
Nominal voltage: 14.4V
Charge current: 2A (2A-5A, as I have seen on other threads)
Discharge voltage: 6V (can anyone tell me why 6V?)
Discharge current: 1A (or should I max this to 1.6A if the charger has discharge power rated at 10W? Is there any issue that comes with discharging at higher current?)
Charge capacity: 7250 mAh
DCHG>CHG: 3

Now, if I am planning on getting the HiTec ePowerBox 30 for DC power supply (Output 12-18V, 30A for a max of 540W), how should I set the volt and amp output on the ePowerBox to supply the HiTec X4 charger that will accommodate the settings above?

Again, thank you very much for your help.

 

Couple things......
1) I would suggest sticking to cycling individual modules because if one of your modules is likely bad and you need to find which one and you also want to see if a module is increasing capacity or getting worse over time (working in pairs also won't help you find what the individual capacities are for matching later when you reassemble the pack)
2) I edited your settings above to what I would be using for Prius module cycling
3) Make sure you also do load testing as outlined in the thread link below
4) Here's the thread you really want to be reading carefully (esp towards the end):

Gen II Prius Individual Battery Module Replacement | PriusChat

 

gil0806, as MTL states above charging modules singly does have advantages in some respects. If charging in pairs doing the last cycle individually will find bad modules, and show the AH capacity of the single modules for matching later. The discharge level with two modules in series will be 12 volts. One volt per cell being the minimum discharge level to help avoid the risk of reverse charging any cells within a module that may be at a low charge level compared with the rest.
I am sorry for the lack of backup to my posts over the last few weeks. This is due to serious illness within my close family requiring long round trip visits to hospital totaling 3500 miles in the last four weeks. I will however attempt to maintain some presents here at PC when I can.

John (Britprius)

 

Sorry to hear this John, our thoughts and prayers are certainly with you and your family during such a difficult time.

 

MTL, thanks for the link and the clarification of the settings. Very helpful! In regard to the load test, does the term just mean connecting the module to a 55 watt bulb (ryousideways' method) and measure the voltage at the module terminal as the bulb is lit? Should load test be done before or after the 3 discharge/charge cycles? Or both? The fourth cycle is to put the SOC within the 40-80% that Prius usually operate, is that correct? What happen if you stop after the 3rd cycle and leave the battery at 7000mAh?

John, thanks for the 12 volts discharge correction. Certainly makes sense. To echo MTL, definitely I'll send in some prayers for you and your family.

I cannot express enough how appreciative I am of the willingness people around PC have (you two and others) in trying help others! Kudos to PC and all those who help make this space such a cool platform for learning!

 

I've run packs in my test car straight off the charger and they do absolutely fine. The most I've seen for a nimh delta was around .6 volts and it quickly came down and self balanced just by driving it, with no difference compared to a parallel balanced pack. I've seen several people mention a nimh delta of 0.2v will throw a code and that is just not true. I think the confusion is coming from the "other" pid (the official one the car uses) that measures the delta v as a percentage and THAT one throws a code at a value of .2, but it's really reporting the percentage difference (20%) which is over 2v on a block pair.

 

I found it best to do the load testing after cycling the modules since I wanted to be sure how they would react when installed. If you have any low capacity modules you can weed those out first then load test to be sure volatge drop is consistent across all of them. I hooked up my modules to both the low and high beam (55 + 65 = 120w) and then measured the voltage drop after 60 seconds (bulb lit). Yes, the last cycle at 4500 mAh brings the modules into normal operating range for the Prius. If you don't do that, you can have error codes after the install.

 

MTL_hihy, I am wondering if using the discharge function of the hitec x4 will accomplish the goal of a load test. What I mean is to set the hitec to discharge (say at .7a) and record the the voltage drop of each module after 1 minute (or however long you want). Should the module with greatest drop in voltage than be weeded out? Does this make sense?

Does capacity matching mean combining one module with a higher capacity and one with a lower capacity? I am confused about this because it seems like some would simply move modules from the outside to the middle of the pack. Is this also a form of capacity matching?

Thanks!

 

The voltage drop with the Hitec charger on discharge is too small to be noticeable so to me using the headlight's high/low beams (around 6-10 amps) is a much better load test. You'll want to set a standard time for testing (I used 1 min for mine) then compare your results and look for any outliers (good modules will not vary much on voltage drop during the test).

Capacity matching is a bit of both maching modules to get a consistent average capacity per block (module pairs) but also moving the lower capacity modules from the center (they are usually lower due to heat exposure) to the ends will also help keep the average capacity more even over time.

 

Hello I this is my first post long time reader （1 year since I have a Prius)
I just received the triangle of death and borrowed a odb2 reader I retrieved codes P0a80 and P3016.
I have got some spare modules I will replace the faulty ones with but I want to cycle these first to test them ( and me).
I have just bought a Imax b6ac v2
If you could please confirm this is what I should do.

Step 1 clamp module ( do you recommend this?)
Set discharge voltage 6.0v
High voltage?? Or just 7500mah
Discharge 1amp or can I do max 2amp?
Charge max at 5amp or less?
Cycle 3 times?
Start with discharge then charge?

Thanks in advance

 

..and I think that's a money making machine for the dealerships