Honda Insight Forum banner

grid charger/balancer

411K views 2K replies 174 participants last post by  carlford 
#1 · (Edited)
I did some research into HV CC power supplies, and think we can get off the shelf components to make one for less than $100.

The supply would be made with:

LPC-20-350---------$16.57
LPC-20-350, Mean Well LPC-20-350
Constant current supply

three RS-25-48-----$20.16-----$60.48
RS-25-48, Mean Well RS-25-48
Adjustable fixed voltage

one RS-15-12-------$18.14
RS-15-12, Mean Well RS-15-12
to run the battery fan


All supplies are universal AC inputs 85-260VAC, have isolated outputs.
The outputs of the three fixed 48V and one constant current supply would be connected in series to yield a 148 to 192VDC 350MA CC supply for the charging/balancing supply, and the 12V supply connected to the battery cooling fan.

The TRC people only sell to OEM, but I could order them through my business.
Of course if we did a group buy, the pricing could be better.
Output fuse and diode, and some leads and power cord, and we have a 350MA CC grid charger/balancer for just a bit over $100.

http://99mpg.com/Data/resources/downloads/gridchargerstuff/basicgridchargerv1.pdf
 
See less See more
#636 ·
Like a new battery!

Mike
Just wanted to let you know the Insight battery is acting like it's brand new!
I think you are going to sell out of the charger/balancers faster than you can build them.
Please hold a unit for me. You have my deposit. I'll give you the balance when I pick up the Civic. - Mark:)
 
#637 ·
Mike: Thanks for the detailed reply! I think you are going to have a runaway winner.

The modest size is especially attractive to me. The reason I asked about it is that after almost 2 years of unemployment as an IT project manager, it LOOKS like I will be starting a contract assignment in Buffalo, NY in about a week and a half. Since I live in Austin, TX, that means I will be buying one while in NYS, and need a unit small enough to take WITH me as I travel back home after the assignment is completed. The size and weight you talked about make that practical even in a small car like our Insights!

Uhtrinity & mbertsche: Thanks for the "real world" feedback.

Jim G
 
#638 ·
Charger is back out for another grid charge test.
The owner reports the following:

In Aug. I had P01447 and P01449. Took pack out and charged each sub. Put it back in and it lasted about a month and a half. Codes are back on now.

I don't expect the pack to fully recover, but it will be interesting if it can clear the codes.
 
#639 · (Edited)
Hi Mike, got question about the Prius ver.
In case someone would like to use it in the following parallel OEM batt. setup, how many temp channels can you spare for the upcomming multi-charger? This is of great importance for parallel charging setups, as you might recall that british girl blowed up (totaled) her unprotected diy phev Prius a year ago. What about at least 4x temp channels (i.e. OEM +3x addon batt. packs), ~around 8x would be ideal though. Is it feasible at all with your currently planned control boards? And if not would it be possible to offload such subfunction on some separately dedicated Prius pack accessory board..

You mentioned the design it's scalable, so apart from the 8Ah overnighter ver., more powerfull ~750-1000W (Prius PHEV) version is possible, that's basically the "mega charger" you prototyped earlier, only missing more 48V supplies for the Prius setup, right? http://99mpg.com/Projectcars/buildingahybridcar/

Thanks

see pictures midpage:
One Year of Operation in a Plug-in Prius with Norm’s BMS+ System
Releted tech. discussion on safecharging/temp probes/timer
 
#640 ·
This is of great importance for parallel charging setups, as you might recall that british girl blowed up (totaled) her unprotected diy phev Prius a year ago.
I didn't see any information about this anywhere, do you have a link? Were they in straight parallel or charged as separate units? I suppose it might not matter much each way as any imbalance causing cell overcharge would cause the Prius prismatic to bulge pretty easily from what I've seen.
 
#641 · (Edited)
Look for Nikki the former cohost of EVcast, I can't recall all the details off hand (there should be still some pictures and written accounts around the web), the setup involved zivan charger, it was not properly cooled (or temp sensed), and prior to the incident often driven as not properly secured to the vehicle, .. , it could have been anything, or combination of couple above mentioned factors.

It was definately not an issue of the overall feasibility paralleling these oem batt. packs in Prius, if done properly works wonders, see the other links. Because it's "cheap" and takes familiar oem chemistry, supports regen I see it as competetive against low eff. hymotion/enginer inverter addon diff. chemistry pack approach, these have got another advantages though (longer range, supersafe). But, as always your milage may vary, Prius PHEVs are always about low speeds-short errands, freq. opportunity charging scenarios, that's where the magic occurs, not for highway flyers.
 
#642 · (Edited)
I read about it, but I don't remember if she had used HV relays to connect the batteries or if she had a direct connection. In my implementation I use a 12 ohm resistor to separate the batteries when the car is off (and when charging). Max current flow between the batteries in this configuration is going to be Amps=Vdiff/12ohms where vdiff equals the voltage differential between the packs. When HV relay turns on then the batteries are connected directly. At this point pack temps can be monitored directly and the PTC system can turn off the relays if an overheat occurs.

Edit: to add something more specific to charging. I have used a charger (700 - 800mA peak) where the output was split through isolating diodes across that 12 ohm resistor (think a pyramid where two sides are diodes and the base is the resistor) for almost 2 years with no incidents.
The current charger (no pun intended) uses two parallel units @ 350mA each charging a pack on each side of the 12 ohm resistor. There of course can be a spillover current between the packs to keep them voltage matched. That spillover current is the same that I mentioned above.
 
#643 ·
Mesuge,

I am concerned with liability for even the 350ma charger, so making more big berthas is not something I want to get into.
I will likely require a signed waiver of liability as a requirement for purchasing one of the chargers.


The basic idea of using a 350MA CC supply for the final topping, is so the total wattage delivered is in the 50 watt range.

This wattage is distributed over the whole pack at the end of charge where all input is turned into heat.

50 watts with a strong airflow from my experience will not allow the batteries to get much more than a few degrees above ambient.

The 9 pin connector (I am kind of stuck with that as I bought 50 sets) is going to do 2 analog temp sensors (MIMA type), and the PTC strip.
On a prius, I could add two more analog temp channels on the PTC conductors. or I could switch to the serial temp sensors where up to 255 can be networked over 3 wires.
I feel this may be an overkill though.

In my opinion, charging a full prius pack, fan running, with inlet air temp, and outlet air temp sensors, we should be pretty safe.
If the delta Temp gets above the set max, the system shuts down.

As we gain experience using the charger, that delta T value where for normal safe operation becomes abnormal will become very clear.

The current cutback voltage will also be important, as this will determine when the charger cuts back from the max of 1 A. to the safer 350ma.

This value will be temperature dependent, so temperature compensation of that setpoint, and a careful look at the pack delta Temp will keep things safe.

I am getting pretty close to having the user interface and LCD display working as I want. The next phase will be connecting real hardware and working out the control and accuracy stuff.
This is the time for your suggestions, so keep them coming.
 
#645 · (Edited)
The 12 V supply can be bigger, I will consider that when laying out the final chassis and can put in another set of mounting holes.
Some test with a civic blower over the weekend show that the blower can draw as much as 3A when running at full speed, so the civic version will need to use an rs-50-12 (4.2A) for the 12V power supply
 
#646 · (Edited)
Mike, thanks for the detailed answer, so we are talking at least 2+2temp sensors possible in total, right? That should be enough for basic safety when charging upto 3x addons only, i.e. bypassing the oem/stock one, as per default when the entire car is shut off.

You see the single channel in/out delta temp is a nice approach for a single compact pack, but could be dangerous for this gargantuan parallel setup (hidden clue), where the primary concern (as demonstrated on past "explosions") is that only one module and friends inside one part of subpack could go very rapidly mad, that's why I linked the ChargeSafe unit with 8temp channels, which can tap directly inside portion of a subpack. Btw. yep, that I2S for 255x sensors won't be necessary now, but if you can design it as allowing for such future optional addon, that would be great. Seems like 3x3 is the magic number for 3x addon packs (max space limited), since it's better to allow for some regen headroom, why charge the stock pack as well, anyway if we also include this OEM pack, it's 12channels/probes: http://hybridinterfaces.ca/InstallThermistors.html
http://hybridinterfaces.ca/charging.html

My take on the ~750-1000W "big Bertha" charger unit for parallel addon phev Prius is more or less in the fashion of standalone "garage-rack" equipment, not to be hauled with the car, that in itself should increase the safety factor by a notch. My driving route/schedule luckily allows for it, obviously fixed onboard unit would be always preferable to get the best phev economy, and this unit could be of smaller power anyway. Alternatively as mentioned, spliting the overall parallel pack into subpacks for charging via add. HV relays/cabling is always the option, although it eats a lot of the precious real estate in that rear end area, which would be of even greater concern as HSD hopefully soon trickles down into smaller cars like Auris/Corola, Yaris etc. Should the Evert bros. ever offer their 70mph EV mode hack also in unbundled lite version with ~55mph EV mode, that would be great for these 3x addon conversions.. In terms of diy limits (not comercial phev addons), you can add/monitor upto 6x OEM packs now, for very real 20mi EV mode range, but that would most likely eliminate the rear seats, making Prius effectively 2seater big-cargo hatch. Or plug 5x 48V@20Ah lithium subpacks as one "virtual addon" ~236V pack, that could be fitted in the back, if these are highend prismatics like A123 pouches.

Mr Mik> very good point about the desire controling more 12V line cooling power if possible. Should be usefull for the Prius application as well, paralleled packs stuffed in small space need a lot of cooling (as backup if all the precautionary charging/timer electronics goes gaga)..
 
#647 ·
Got the civic fan driver working nicely today, it will have a fan speed pot that can be used to set the best speed for the temperature, and since it is a microcontroller, I may just tap the pack inlet and outlet temperature signals and use that information to change the fan speed automatically. Since this blower is a pretty serious air mover, that will probably not be necessary.
The controller will do a safe switch over of the fan circuit from charger control to stock control.
When the charger is plugged into the car, the fan board will operate a switchover relay.


On the grid chargers,
The microcontroller can save information in non volatile memory, what can we do with that function besides storing the setup and auto shutdown parameter data.

We could store things like :

What the voltage at final settling of the pack was?

What the rate of charge/voltage rise /minute/ was at a particular part of the charge slope?

With a constant current charge, we may be able to do some battery health predictions based on changes in charge rate and final voltage?
Any ideas?
 
#648 ·
I'd be interested in an Ah count going into the pack. ...or maybe just some duration of the charge and I could do the math from there. Max temperature during charging perhaps. These things are not critical and I'm not sure what you are monitoring already and what is feasible, probably wouldn't be worth it to add those two things if it made a cost difference to implement, just brainstorming ideas.
 
#649 · (Edited)
here is the plan:
MIMA Honda Insight Modified Integrated Motor Assist - Building a hybrid car grid charger

The following data will be available for display.
Batt Volts (Real time)
Charge mA (Real time)
Inlet temp F (Real time)
Outlet temp F (Real time)
Temp gain (Real time)
PTC ohm (Real time)
Chg mAmpHr (how many mAh have been put out by the charger after the start button was pressed)
Charge Time (total time in charge)
Top Time (time in topping area, where volts are no longer rising)

the following are setpoints for the shutdown safetys which will be stored in flash memory

Max charge mAh* (shutdown after Chg mAmpHr reaches this point)
Max PTC Res* (shutdown after PTC ohm reaches this point)
MaxTmpGain* (shutdown after Temp gain reaches this point)
Cuttbak V* (Reduce current when Batt Volts reaches this point ----Overnight charger only)
TopTimeMN* (shutdown after Charge Time reaches this point)
MaxtimeMN* (shutdown after Top Time reaches this point)

The two lines of the display can be easily scrolled through the list of variables, the bottom row as well, but the bottom row will allow entry of setpoint data which is stored in non volatile eeprom.

So the question is what other information would be usefull to store in the eeprom?

It cost nothing except programming time to make the software better,and that cost is a one time development labor cost.

I can now do floating point math, which was difficult in assembly, and can display as many datapoints as required.

Plenty of eeprom and programming space left, so now it the time to get all the ideas implemented.
 
#650 ·
Could you measure battery IR by switching between the 1A and 350mA charge current and measuring the difference in voltage? Maybe that could be a rough battery health indicator. The voltage difference between 1A and 350mA charge current is likely different at low and high SOC, so that could be held in memory and then be compared at the end of the charge. No idea if it really means something, though! :D

When the PTC max cutoff is reached, will the cooling fan continue to run after shutdown? That would be very important, because a lot of heating occurs after the completion of charge.
 
#651 ·
Ad grid chargers/microcontroller, what about adding simple serial data output via rs232 for remote monitoring, is that feasible with your layout and hardware used? You can then easily log, plot, manipulate, .. the data on standalone computer/any such device or even over network, all that with already existing applications be it basic spreadsheets like Excel (and its ouput functions to RSS channel). I think it has been recently discussed at Perkin's UK forum.
 
#652 ·
My experience with the cell level tester project makes me think that the difference between 350ma and 1A is too small to yield any reliable IR information.
It is worth a try though.

I was going to run the fan all of the time the charger is connected.

Serial in and out is also part of the plan, but I have not started that part of the code yet.
I don't have enough memory in the processor to do much internal datalogging, but could add a flash memory chip for a couple of dollars that could record the voltage each minute for the whole charge, or I can simply broadcast the data on a serial line, and one could log it or not?
Since I don't expect to be powering the BCM, I don't think that Peters BCM data will be available, but I will include a bi directional serial com if we decide that we want to try that.

With a constant current, I expect that the charge data will be pretty boring, with the voltage rising slowly and finally settling out to some max. The curves get more meaningful as we get closer to the cell level where the differences between subpacks or cells can be readily seen.

The topping area of the curve may show if the pack is badly out of balance or well balanced, as a well balanced pack should reach the peak more quickly than a pack with a bunch of cells that are way behind the leaders?


Lots of possibilities to explore.
I hope to have the prototype charger with display and controller running next week for some real world test.
 
#653 ·
I don't think I would bother with any on board storage.

Just squirt the data (your parameters) at a TTL serial port every second/minute whatever.

Then with a simple serial to usb lead like i use with the BCM gauge and an excel speadsheet or terminal program people can capture the data to a laptop and analyse it to their hearts content. IMHO People are unlikely to want to capture data for every charge, and most people have a laptop or netbook or can access one for the rare occasions they want to look at the data.

The free PLX-DAQ software/spreadsheets work well, I used it today to gather data from my system. If we use one core system then we can share spreadsheets and data via the forum for comment analysis. I'll be posting some tomorrow as part of the more current project. They are easily modified.

I suggest data ouptut should be 9600, 8,N,1 standard for compatibility. I appreciate you like that labview software but it you can make data output compatible with both that's great.
 
#655 ·
The 5V TTL serial to usb lead I use has a simple 3.5mm jack plug on the end. So I use a cheapo 3.5mm stereo jack socket. Cheap as chips. I don't see any need for a dB 9 etc. Most people will need to buy a lead anyway.

USB TTL Serial
 
#657 ·
Any cheapo stereo 3.5mm socket will do but i suggest use one that will be insulated from your chassis. I don't have a pic of mine except in the gauge thread itself somewhere hanging on a bit of wire!! I'll post the connections pinout shortly. Is your pcb board having the 5 ICSP pins as well please?
 
#659 ·
Quite a few people have bought the cheap usb programmer if you can add the 5 sil pins as well as your mima type socket to the pcb then they can upgrade without having to make an adapter cable for the programmer. Or are you inteding that units are returned for updating? Do you have a schematic we can look at? If the pcb is already done then they will just have to make a little adapter cable.
 
#660 · (Edited)
I am still working off the easy pic 6 development system, so the hardware is still completely open right now, so the SIL pins will not be a problem, but I will need the pin asignments when the time comes.

No schematic yet, as I am still getting the software to function as desired
and the ports can be shifted around later on the actual hardware, after I get it all working.

Things are starting to go easier with the C code, with the user interface pretty close to where I want it, and the shutdown logic is all working.

Need to look at the analog signal conditioning, to figure out the best way to do that.

Need to design the in car harness for the Insight and Civic,

Need to try out the optoisolated SS relays for turning the 48V supplies on and off under full load.


Yard is 4 " thick in leaves.
Not enough wood stacked up and ready for winter
My insight EGR valve needs looking at
Just received my Microchip 3 phase motor development board.

Seems like there is never enough time.
 
#661 ·
Yard is 4 " thick in leaves.
Not enough wood stacked up and ready for winter
My insight EGR valve needs looking at
Just received my Microchip 3 phase motor development board.

Seems like there is never enough time.
I can empathise with that :plusone:
 
#662 ·
Hi Mike!
I have a 2001 have not had any battery issues. But I have noticed that my big battery charge either goes way up or way down. I commute 150 miles a day.
Are you still selling those deep battery charge machines? Please let me know! Thanks! Rich
 
#663 ·
Mike, Thanks for linking me to this thread. I just received my first Honda Insight about 12 hours ago. I have been reading threads here and also ecomodder for the past couple of weeks. All this info is a little much, but I am trying to digest as much as possible.

My first drive to work in my 2003 Insight, I managed to get 52.6 mpg. To my understanding the IMA was replaced less than 1000 miles ago. I am very interested in building one of these grid chargers. I should be able to, in my free time, I install full car stereo systems.

I also wonder how a car audio system would have an effect on the Insight's mpg, if I was to add an amp and a cap.

Thanks for everyone's input as well!

Ryan_C - A Proud new owner of an Insight!
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top