grid charger/balancer

[retepsnikrep]

Originally Posted by aisbell

I'm a bit concerned that without the constant current final power supply of your current charger design, my charger might be overheating my battery, especially during the initial charging. The input current is about 800 ma. which is considerably higher than what I have read might be an optimal output current. But I have no way to infer an output current from my measured input current.

Art

I'm not concerned, I tested it for several days and output current will be fine. However if you want to send it back and pay return postage, I will upgrade/swap it to the CC version free of charge.

[Hybrid-Battery-Repair]

Originally Posted by aisbell

I will be parking my Insight in an unheated carport in about 2 weeks for another 6 months. At least the battery temperature will be considerably lower than it was when my Insight was parked in Honolulu. I have read that self-discharge is slower at lower temperatures, so maybe the resulting imbalance will be less. But when I return, I'll be able to fully charge it using your grid charger before I drive my Insight anywhere, so hopefully that will help in the long run.

Art, after a couple of months, the imbalance will be there no matter what the temperature. You really need to charge it before you drive it, or you'll be taking the ferry to visit Peter next fall.

[Mesuge]

Obviously it's a different NiMH cell/pack but some results are worth to mention here as well. Many of you have probably seen it, the hybridinterfaces test with OEM packs for PHEV application. Namely, they charge upto 4x of them in parallel (<4kWh for driving) via hacked Zivan charger (formely Pb profile). Even with <0.15C charging they ran into potential thermal/issues at the end of the cycle (T2,T3); so not only proper cooling but some additional safety circuits are necessary (diodes, timer, .. as proposed in your thread also). So, yours 350mA CC looks much safer anyway, most importantely with regard the issue of possible build up of internal pressure in combination with elevated ambient temp and bad charging wiring/design could result in quite dangerous situations.

Thanks to Mike & Peter for documenting their charging setups based on MeanWell LED supplies, surely it could be adapted to Prius as well; hybridinterfaces level: 238.7VDC by 5x ~48V or rather

*Mike/Peter's approach: 4x ~47.6V CV + 1x CC 3V to 48V
suitable CC led supply ~500mA (still bellow 0.1C): PLC-30-48 Current&V adj.

There has been one "exploded" PHEV Prius using Zivan already and it's not pretty, basically the car is a write-off because of the violantly released toxic stuff everywhere. However, the install seemed as pretty unsafe to begin with, so can't blame the Zivan only, pumping in summer almost 0.5C (+ 2 OEM packs) without proper cooling and robust tight down of these packs looks rather foolish in retrospective..

OEM pack based PHEV charging

charger timer circuit/ideas

--
Btw. Mike charged subpacks 300mA upto 8.47V for full equalization (~14hrs), which translates into 237.16V for the Prius OEM pack
*http://www.99mpg.com/Projectcars/mimapackwhack/

[Mike Dabrowski 2000]

A few things to consider:

1. the 350ma meanwell supply can maintain CC between 4 and 48V
The larger versions increasingly have smaller CC ranges, (33-48 for the 500ma) therefore when we connect them in higher voltage series strings for charging higher voltage packs, the pack voltage swing between discharged and full may exceed the CC range.
If more current is desired, putting the 350ma CC supplies in parallel may be a better approach than using a larger single supply.
2. I used a 12V battery fan power supply to assure that thermal issues at the end of charge would not be a factor, granted, running the fan at 6V is probably more than sufficient to keep the pack cool.
Without the fan running, the pack may develop hot spots, and cell heating could damage cells in a way that would not be immediately obvious, so my advise is to run the fan, to be safe.
I know of several people that have built first gen and second gen prius chargers using my series supply technique with good results.
I have a lot of experience with the prius cells, and can assure you that heat and internal pressure is generated even at the low 350ma current. The prius cells flat sides make a great internal pressure indicator, and will begin to bulge outward just as the cells reach 100% SOC, pointing to a simple microswitch based charge terminator.
http://www.99mpg.com/blog/batterypac...subpacks.jpg,0

[Mesuge]

Thanks for the details, I'm not sure this equalization focused approach is worth scaling up to say the levels of that Zivan NG3 (.8kW output), which can charge 3.5-4.5kWh (4x parallel packs) in few hours time - as low as 7hrs (T1_5h-T2&T3_onlyfor2h), on the condition of all the safety precautions incl. passive and active cooling system. For Prius, I see the advantage of these MeanWell style cost effective grid charger balancers more in the role of service/maintanance the longevity of batt. pack than for regular PHEV charging application. Because ~100W output doesn't make it practical with these very long charge hours even at shallow %discharge driving cycle. Perhaps at least 4x scaled up version ~400W (roughly 2x slower than Zivan NG3) might be considered as barely practical solution - real charger. Alternatively, set it up as 4x standalone 350mA grid chargers for each OEM sub-pack, but that's 4x cost/work-wirelogy-complexity/space used/failure rate or perhaps not, as some features like the timer could be shared. Altough much safer, this individual type of charging will also to a degree defeat the major advantage of parallel charging function as a balancer.. So, perhaps the Zivan NG3 "rocks" afterall, lol..

Zivan C7 or SX (NiMH) curve issues explained:
http://priuschat.com/forums/gen-ii-p...stol-uk-7.html

[Mike Dabrowski 2000]

I have been experimenting with a two step charging system that would allow a full charge on an empty pack in 5-7 hours while still allowing a 350ma finishing charge.
Instead of using three 48V RS-25-48 supplies. I went with three RS -50-48 supplies. They have 1.1A output current VS the RS-25 .57A supplies.
I use two CC supplies in parallel,to supply a 1A charge the LPC-20-350 which supplies 350ma, and a second CC supply, the
LPC-35-700.
The second 700MA CC supply is powered via a latching relay on the AC input, that is latched on at the start of the charge.
Once the pack reaches an adjustable threshold voltage, the 700ma supply will be disconnected, allowing the 350ma supply to safely finish the charge.
It is looking like the threshold voltage should be in the 163V - 168V range.
The voltage while charging with a CC with the fan running, has a pretty reliable and linear relationship to the SOC, unlike the voltage while charging with the IMA, which has a widely variable voltage as we pump charge and pull assist from the pack.
I will build up a prototype when I get the supplies next week.
The same system will work on a Prius pack, with an additional fixed supply added to the series to bump up the base voltage.
The relay/threshold circuit and larger and additional power supplies should only add about $40 to the system.

[E27006]

I charge my pack installed in the Car using a constant current supply system,

I disconnect the PTC strip from BCM using a switch, and monitor the PTC strip resistance to indicate temperature.

Here are some figures from my last session:



Charge rate 600 mA.

Additional cooling (none) battery fan not used

Charging terminated at t=300 minutes(5 hours)

* = peak voltage of pack,

Ambient Temperature 4 to 5deg C, pack status cold (car not driven for 24 hours before charging)

Time (minutes) Pack Voltage PTC strip (Ohms)

0 153.5 17.6
60 161.8 17.6
120 165.6 17.6
180 169.2 17.8
210 171.3 17.9
225 172.2 18.0
240 173.1 18.2
255 173.3 18.3
265 173.5* 18.3
270 173.5* 18.4
285 173.2 18.6
290 173.1 18.8
295 172.9 18.9
300 172.6 19.0


Note that after 270 minutes of charging (600mA) the pack voltage has peaked and declines, this is due toa minor negative delta V effect as subpacks top out.

PTC strip figures remains almost constant until last 20 to 30 minutes, as cells top out they dissipate excess charge as heat which is seen in the rise in ohms.

Test carried out on a cold day.

19.0 ohms at termination is a safe figure, the sticks will still be cool to the touch.

I usually run at 600 to 700mA then reduce to 300 to 350mA as the pack nears completion, but for this test I ran 600mA.

The charging behaves exactly as Mike D and Peter P have predicted


Postscript: I laid out the table using tabs, they looked Ok but on posting the tabs are lost hence the funny layout

[RobertSmalls]

How hard would it be to build a delta-V detecting circuit that shuts off the charger? That seems like the best way for someone who wants to top off the pack every night to save fuel, without worrying about overcharging.

[Mike Dabrowski 2000]

Delta V detection will not be reliable on many packs, especially if they are badly unbalanced. Once the pack has been through several cycles and gets better balanced, delta V should be more reliable. On an unbalanced pack, some cells will be dipping as they get into the delta V range others will still be on the rising side of the delta V curve, and cancel the dip. The 350ma with the fan running will top off all of the cells, then level off and stay at a steady voltage as seen on this charge graph.
http://www.99mpg.com/Projectcars/mim...ns_graph.jpg,0

On the new circuit I am suggesting, the detection circuits could first detect when to drop from 1A to 350ma, based on voltage. The second detection circuit could detect when the voltage reaches the full charge point based on the voltage reaching a steady state for say 10 minutes. A simple timer could also stop the charger a fixed time after the drop to 350ma, since this time should be repeatable.
Using this every night to top off the pack, is what I had in mind.
Using the temperature rise on the ptc strip without running the fan, is another way to determine end of charge, but it depends on there being enough current to generate a clear temperature rise, which 350ma just marginally can do.
This also depends on the pack being at ambient temperature when you start. As you can see from the graph, this pack started off warm from the 50 mile drive to my place, and actually cooled off as we charged it.
To do a full charge in a night, waiting for the pack to cool before charging would take too long.

A similar charger to set to a higher voltage to charge a Prius booster pack would be another use for this system.

[jf_cole]

Had Retepsnikrep's charger fitted there yesterday. Its a neat little install and not as hard as I thought it would be. Didnt take long at all. Hoping to test it as soon as the temp climbs above 3 degrees!