DIY V1 Grid Charger/Discharger
For less than $70!

Latest update:  11/25/2019  Mean Well power supply found for V2 grid charger
.
This version of the grid charger really can't be built anymore because the LED drivers it used aren't sold now.  Use this link V2 Grid charger article to go to the latest version of the charger.


Return to Insight home page                          



WARNING!  (SEX)
(Now that I have your attention, please read the next sentence carefully.)

There are dangerous high voltages inside the charger and IPU.  You can be KILLED if you are careless and don't follow all the safety instructions when working within the battery box.

At the same time, if you follow the instructions it is quite safe to work on these high voltage systems.  If you are not comfortable working with high voltage, try to enlist the help of someone that knows what to do.




Select articles on this page from the list below.
Background

The first generation Insight high voltage IMA battery is composed of 120 special ~"D" sized Ni-MH individual cells arranged in 20 "sticks" of 6 cells each. The nominal voltage of the battery is 144 volts.  A new pack will typically run 156 to 165 volts while in use.  Ni-MH cells have a rather high self discharge rate of 1% per day and over time the capacity of the individual cells will vary.  At the same time the internal resistance of the cells will tend to rise and not be equal between all the cells. This causes the pack to become out of balance and the full capacity of the battery can't be utilized while driving. More information about all this can be found at the  insightcentral.net/forums/.  Do a search for "grid charger".

The IMA battery pack is normally charged (if needed) by the kinetic energy of the car when slowing down or coasting to a stop while in gear. This is called "regeneration".  The battery can also be charged while the engine is running in what is called a "background charge".  The background charge does not always show up on the dashboard instrumentation.  Normally there is no way to charge the IMA battery with an external charger.
The battery control module (BCM) has many features to safely control the charging of the battery so that it operates between 20% and 80% of it's capacity. If the capacity of the battery goes below ~10% the BCM will not charge the battery or allow it to assist the IMA system.  This is done to protect the battery.  The digital dashboard will also display a "Check engine" warning and an error code will be stored in the OBDII system.

Enthusiasts on insightcentral.net/forums have devised a way to rebalance the battery cells by slowly charging the pack to 100%.  Many of the  2000 to 2006 Insight IMA batteries can be brought back to life by giving them this low current charge for an extended period of time (24 to 35 hours). Typically 0.20 to 0.35 amp is used. The this low level charge allows all the cells of the battery to become fully charged without overheating the better cells of the pack which will reach full charge first.  The battery will charge to ~174 volts.

Even new batteries ought to be grid charged once every 3 or 4 months after they are 6 to 9 months old.  This keeps them balanced and will help them last longer.  Think of it as maintence task to extend the life of your battery.

The charger that does this is called a "grid charger" because it derives it's input power from the normal home power grid (the wall sockets).  The charger is basically what auto mechanics call a "trickle charger" because it charges at a very low rate.  A normal 12 volt car battery charger can NOT be used on the IMA battery because the voltage and current is not correct for our purpose.

The grid charger also powers the normal Insight 12 volt battery cooling fan to insure that the IMA battery pack doesn't overheat due to the extended low level charge or discharge. There is also a fan inside the grid charger to cool its components.



Information about LED driver change

 Read this, it is important;   (4/4/2019)

You MUST use current limited power supplies for a grid charger at this time. Not constant current supplies. The difference is subtle but most constant current supplies when wired in series will not track properly. Unfortunately eBay Chinese sellers of LED power supplies have switched over to constant current supplies and the current limited supplies aren't listed on eBay anymore. If I find a current seller of a 45-90 volt supply I will post it here.

I have tested several different constant current pairs and typically the first supply that reaches it's maximum current (considering the load) will prevent the other supply from delivering 1/2 of the voltage to charge the battery due to a race problem when the supplies start up. This will cause one supply to operate above it's wattage rating. I will describe this further near the end of this section.

Note, even though the current limited LED drivers are not available anymore, the rest of the information in this article will help you understand what grid chargers are used for and with luck how you can build one.

The following are the results of two typical identical Chinese 60-90 volt, constant current 300 ma supplies that I recently tested.
Test 1: Test both supplies individually for consistency.
Supply #1: (Tested by itself)
No load voltage: 126 volts
Loaded with two 40 watt light bulbs in series: 89.9 volts @280 ma (after 10 minute warm up)

Supply #2: (Tested by itself)
No load voltage: 131 volts
Loaded with two 40 watt light bulbs in series: 88.5 volts @275 ma (after 10 minute warm up)

Test 2: With both supply's outputs wired in series
No load voltage: 261 volts (NOT 180 volts as you would expect from two 60-90 volt supplies in series.)
This is actually normal for a constant current supplies. The no load output voltage has to be higher than the expected loaded voltage to deliver the maximum specified constant current to the load.

A limited current supply has a maximum voltage it can deliver (for instance 60-90 volts) and will if necessary, also limit the current so as to not exceed it's rated maximum output current. That is the subtle difference between the two types of supplies. The lower rating (60 in this case) is the lower limit for normal operation.

Test #3: Check individual load voltage of the two supplies connected in series.
Load with two 40 watt light bulbs connected in series to simulate charging current: 120 volts @ 239 (that sounds reasonable)
Output voltage of each series connected supply: (Wait till you see the next test readings @ 239 ma!!)
Supply #1: 4 volts (Yes, FOUR volts!)
Supply #2: 116 volts

Obviously these two typical non-precision constant current supplies are just not going to work for our grid charger since each supply is rated as a 25 watt supply. One supply is overloaded and doing most of the work and the other one basically took the day off.

Here's what caused one supply to only output 4 volts. When the AC power switch is set to ON there is a voltage/current race problem between the two supplies wired in series until one of them reaches it's constant current condition.

Since the supplies are wired in series, power supply #2 with the lower constant current capability with that load will have also reached it's maximum capable voltage at that maximum current. Supply #1 then fills in the remaining voltage (only 4 volts in this test) because of the current supply #2 is already forcing through both supplies. The higher current capable supply #1 can't force more current through the series connected supplies to balance the voltages because supply #2 won't allow any more current to flow.

Current limited supplies don't have the race problem because they each put out the their maximum voltage listed (90 volts) and if necessary limit the current to their maximum specified current. Yes the lower max current supply rules but only after it reaches it's specified voltage. Since the lower current sets the current the other supply is able to also output 90 volts for a total of 180V.

Return to top of this page 



Purpose of this article

This article is an idea provoking one to present how my DIY grid charger looks and to present the schematic with a parts list so you can build one for yourself.   I am not going to give a detailed mechanical dimensions of where to drill holes etc because the parts you buy, the case you use, the individual parts, power supplies etc may vary in size from the parts I used. And truthfully I don't want to encourage people to copy my design to sell  [6/11/13].

Ready to use grid chargers from simple to very elaborate computer controlled models can be bought through  insightcentral.net/forums/  and eBay. This article describes a simple grid charger (no fancy computer control) that can be built from components found on eBay etc.  If you are familiar with reading schematics, soldering and using simple hand tools you should have no problem building your own charger.  You also have to open the battery box (IPU) behind the Insight seats to install a charging harness.  You can find information how to install the harness (some with videos) at the Insight Central forums.

When I built this charger in February 2013 the main electrical components cost me less than $30 not including the used computer power supply case with it's fan, the switches and other mechanical parts.  Naturally your cost will vary depending upon how many parts you already have.  I had most of the parts.  The parts price now is typically less than $70.  You can usually find a local computer repair store to buy a junked ATX power supply for the case, fan and  the AC connector etc. With luck you may also find many other parts on the printed circuit board(s) to use in your DIY charger/discharger.

Return to top of this page
   


Features of this charger

The case I used for the supply is an old ATX computer case with the fan and AC power input connector retained.  Try to use a case with a top mounted (larger) fan.  

The original power supply printed circuit board(s) and wiring were removed and discarded.  The original printed circuit board threaded mounting spacers (PEMs) in the case are used on my first grid charger.  Luckily I found a nice case in my used parts collection that had ventilation slots on the top and one side.  This allows me to position the charger on various sides and still have enough air flow.  You might be able to use some of the components from the power supply PCB.

The unusual features of this simple charger are that two LED meters are used and it has a connector to plug your discharge load into.  The discharge connector allows you to use the built in metering to monitor the discharge voltage or current.  I won't use the 12 volt meter for any other grid chargers I might build. (The voltage always stays the same.)

One of the meters displays the 12 volt supply output voltage and the other meter is a dual function meter that has a front mounted push button that allows the meter to now read either 0-199.9 volts or 0-500 ma.  The multimeter micro processor is set up to use an external shunt that originally allowed the meter to display 0 to 50.0 amp.  I use a different value shunt which allows the meter to display a reading of 0 to 500 ma.  Unfortunately the decimal point is preset and will be in the wrong position when reading current.  The meter will be displaying 0 to 50.0 which is really 0 to 500 ma (normally we charge at 350 or less which will display as 35.0).   Otherwise the meter works fine.

The multimeter is very handy to monitor the IMA battery system while charging or discharging it.  The multimeter requires a separate small power supply with an isolated output voltage between 6 to 15 volts at less than 50 ma to power it.  I used a small 6.25 volt cell phone charger supply but that is not a common voltage so I listed a 12 volt wall wart instead at the end of this page.  I have built another charger and I was able to modify a 5 volt cell phone charger to output 7 volts with it still regulated.  The supply must have the output isolated from the AC input.  This is necessary because the DC negative lead of the small power supply is connected to the negative lead of the high voltage being measured by the meter. Switch mode power supplies have the output leads isolated from the AC input wiring.  The common 5 volt cell phone chargers are not rated to power the multimeter without modification.  

The IMA battery charging voltage is created by two "LED driver" current limited supplies wired in series. Each supply I used is rated to deliver 45 to 90 volts and are rated for 300 ma output current.  The grid charger will output 178 to 181 volts no load and ~255 ma when charging the battery.  These supplies are normally used to power strings of LEDs for lighting or special affects. The supplies I used are protected for short circuit, over temperature and limit the maximum output to 300 ma. The cases of the supplies are aluminum and will dissipate heat much better than plastic cases.  By mounting them to the bottom of the metal case of the power supply more heat can be dissipated.

[4/11/2019 Update] Since the current limited LED drivers aren't listed on eBay anymore I have been investigating how to adapt the constant current LED power supplies for a grid charger. It is slow work because it takes a month or more to receive power supplies from China to test. I hopefully have configured a safe method of modifying the constant current supplies now listed on eBay to act as a limited current pair for the grid charger.

I am waiting for another set of isolated power supplies to hopefully confirm what will become an updated V2 of  my DIY grid charger/discharger. The good news is that a dual 4 digit V/A meter and LED supplies are much cheaper than the equivalent components listed in the V1 bill of materials.
---

A good feature of the LED supplies I used is that they have an internal diode connected across the output leads so that when two supplies are used in series as a grid charger they will be protected if one of them quits working (as an example, from overheating).  The charger won't output any charge current but at least it won't let the magic smoke get out of either of the supplies.  In use I have found the charger supplies run at whatever the ambient temperature is.  The supplies do not feel abnormally warm to the touch after charging for many hours.

I also use a series output diode within the grid charger to keep the IMA voltage from being present on the supplies when they are not operating.  This allows the charger multimeter to display the IMA battery voltage when the supply is connected to the car but not charging.  This also allows you to read the voltage and discharge current while discharging the battery.

The 12 volt, 2 amp rated switching supply I bought seems to be a typical Chinese supply in that it really doesn't put out 12 volts AND 2 amps at the same time.  12 volts is output with a 1 amp load.  At two amps load the voltage drops to 11.1 volts.  On the other hand, even 11.1 volts at 2 amps should drive the grid charger case fan, the Insight 12 volt Insight battery cooling fan and the BCM if you want to run it while charging the battery.  I set the no load voltage to be 13 volts with the built in calibration potentiometer.

NOTES:
  1. If a diode(s) is used within the IPU to prevent battery voltage from appearing on the charging cable, the multimeter will not be able to read the IMA battery voltage when the battery is not being charged.  Unfortunately you will also not be able to do a discharge cycle of the battery to increase it's life with a diode inside the case.   For that reason I would recommend that you not install any diodes in the charging harness.
  2. While you have the IPU open to install the charging cable, check the gap around the battery cooling fan. Typically there will be a ragged gap in the plastic housing with a space up to 1/4" along the fan sides.  Use some RTV or flexible caulking to fill in the gap to increase the cooling affect of the fan.
  3. As a safety feature I would install a fuse on either high voltage lead when you install the charging harness.  A fuse holder can be mounted outside the IPU case or a panel mount fuse holder can be mounted on the IPU case so you can change the fuse without opening the IPU case.  I prefer the panel mount fuse because you are trying to protect the harness wiring inside the case if there is an external short.
  4.  I used a 2 amp slow-blow DC fuse in line with the positive battery lead charging harness.  Use a fuse rated for at least 200 volts DC.  A DC fuse is designed to quench any arcing if the fuse blows to prevent it from starting a fire inside your car!
Return to top of this page



Construction and description of how to build a grid charger

grid charger side fan 2nd supply top fan

This is a preview of what two chargers I've built look like with a no load voltage of 178 volts.  The charger on the left is described in detail below and is is my proof of concept charger.  The right side picture is a charger I built for a local Insight owner and is how I would build more chargers (minus the 12 volt meter).  Lately there are dual display volt/current meters listed on eBay that would require modification to read 500 ma.  They display the volts and current at the same time.  Please notice that most of the cheaper ones do NOT display tenths of volts though.  It is better to have tenths shown so you can get a more accurate voltage reading to have an idea how the charge or discharge is going along. Do yourself a favor and buy a 4 digit high voltage meter.

The switch on the upper left is the master switch that turns on the 12 volt supply, the 6.25 volt supply for the HV LED meter, the grid charger fan and the IMA battery ventilation fan.  The right hand switch turns on the high voltage power supplies.  If you want to run the BCM while charging the IMA battery you have to apply the 12 volts to the BCM before applying the charging voltage.  This allows time for the BCM computer to boot up and do its chores.

The voltage or current selection push button for the multi-meter is located at the lower right hand corner of the meter bezel. If you look closely through the ventilation slots you can see that the NE-2 neon bulb is glowing to indicate that the high voltage power supplies are energized.  The neon bulb also acts as a low current bleeder to discharge the two power supplies output capacitors when switched off. When the neon bulb goes out the HV is a little below 80 VDC so you have to wait ~20 seconds or more for the secondary bleeder resistor to drain the voltage below 10 volts or so.



To see an enlarged view of these pictures, left click on a picture or right click and select "View Image".
top view

This is an overall top view of the grid charger.  Notice the 3/4" w x 1/8" thick aluminum strip of metal that holds the two LED power supplies in place. The strip also is a mount for the 6.25 volts cell phone charger supply and the 12 volt, 2 amp supply.  All the supplies are high efficiency switch mode type power supplies.

The two LED driver supplies are mounted at an angle because that is the only way they would fit in the case.  Notice the 1/4" spacing between them so cool air can circulate around the cases.

cell charge

This is a close up of the 6.25 volt power supply which is used to power the larger LED front panel meter.  It is mounted on the hold down strip with a #4-40 nut/bolt and hot melt glue.  I cracked the case open so I could solder the AC input wires to AC terminals on the left side of the PC board.

The hold down strip also is used to mount the 12 volt, 2 amp power supply located in the lower left of this picture.  One of the mounting screws for the supply is partially hidden and is located just above the small yellow voltage adjustment potentiometer.
LED spacing

There is a spacer located under each end of the LED supply hold down strip. The spacer length is chosen so the lower side of the hold down strip is ~1/16" below the top of the LED supplies.  This clamps the supplies in place and also helps to dissipate some of the heat they produce.

The hold down screw for the 12 volt supply is located on the right bottom of the supply below the green LED. The green and white heat shrink is used to protect the LED high voltage leads when the cover is placed on the case.

LED supply spacing

This is the other end of the hold down strip showing the spacers and the screw that also mounts this end of the 12 volt, 2 amp supply.  The spacers are just hollow tubes that the mounting screw passes through.  The PEM studs attached to the case are threaded for metric screws.

The 12 volt supply has many holes in the cover for ventilation so it is OK to mount it against the case with its ventilation holes.  If your case doesn't have ventilation holes in that area space the 12 volt supply away from the case.
AC wiring

The AC input wiring is on the left of this picture.  Also shown is an end view of the 7 lug terminal strip I used.  All the output voltage wiring connects to the terminal strip or the green stand off which is the high voltage positive output terminal.

terminal

This is a close up of the terminal strip with the high voltage (HV) wiring on the 4 left end lugs and the two 12 volt wires on the 2 right end lugs. The 5th lug is the mounting (grounded) lug. It is located under the green stand off terminal in this picture.
fan

The fuse holder is mounted where the 120/240 switch was originally located.  A normal computer power cord is used.

After these pictures were taken I installed a female 4 pin Molex connector on the "front" side of the case for the charging harness to plug into the charger.  A discharge 2 pin connector is also mounted on the front panel.
top panel meters

The case is made of steel so you will have to find a way to cut the two rectangular holes for the LED meters. It took quite a bit of careful measuring to make sure the larger meter clears the fan.  The fan fits into the clear area above the meter and to the left of the switches in this picture. The two calibration pots can be seen on the multi-meter. The two white connectors and wires came with the meters.

Inside view 2nd charger

These two pictures are of the top fan grid charger I built for a friend.  The case is a standard ATX power supply.  The main advantage of  the top mounted fan is that it allowed mounting the meters on the front side of the case rather than on the removable lid. That eliminated all the wiring from the cover except for the two wires for the fan.  I used the normal fan 2 pin connector so the top cover can be completely removed and placed out of the way.

The LED supplies are bolted to the bottom of the case which eliminated the hold down strip the first supply used.  The two amp ventilated cover supply is mounted to the rear panel with #4 hardware.  I cut away a good bit of the two amp power supply's ventilated cover where it touches the rear panel to increase the air flow through the supply. The supply cover is mounted to the rear panel first and the supply is then slid onto and under the cover from the right side of this picture.  Note that the terminal strip is also mounted on the supply. All of these details take a lot of measurements and trial fittings before you drill any holes.

The cell phone charger with the brown PCB is mounted with two long #4 screws using several #4 nuts to space the supply above the LED supplies.

Return to top of this page
2nd supply outside

A short charging cable with a 4 pin male Molex connector comes out the back of the case near the 120 Vac plug.  The charging cable coming from the IPU case uses the normal ATX power supply female Molex connector.

The two advantages to using the top mounted fan case are the fan is very quiet and there are no permanent wires or switches attached to the cover. Since the fan and cover is removable the charger was easier to build and if necessary, will be easier to repair.

This charger has a neater layout and mounting set up compared to the first supply which I built for myself. The first charger was actually a proof of concept test charger.

I have found that the LED supplies run at the ambient air temperature while charging the battery and there is more than enough air flow with the top or side mounted fan.  The top mounted fan makes a much nicer looking power supply internally and externally. The fan discharges the air upward.

Just to the left of the 12 volt meter you can see the covered hole where the original computer DC power leads came out of the power supply.



  Schematic (Click on the schematic to see a clearer view.)
grid charger sch


Bill Of Materials
Part designators
D1
F1
Fan
LP1
J1
J2
M1
M2
P1
PS1, 2 (Note 2)
PS3     (Note 1)
PS4     (Note 1)
R1
R2
R3
SW1, 2

             
Description                                                                                   Vendor
1N4007, silicon diode 1amp, 1000  Vr  ......................................... Various suppliers
2 amp fuse, 250 volt and holder ......................................................Various suppliers, fuse holder  bgmicro.com/FUS1008.aspx
12 volt fan ....................................................................................... Part of power supply case
NE-2, neon pilot lamp  ................................................................... Various suppliers  (Not absolutely required but nice to have)
Male line cord plug ......................................................................... Part of power supply case
Molex male 4 pin ............................................................................ Mate to computer Molex hard drive power connector
Dual red LED meter, 199.9 volt, 50 amp  ....................................... eBay item   (Note 2 below)
3-30 volt red LED meter  ................................................................ eBay item  (Not absolutely required but nice to have)
Female 2 pin connector ................................................................... Any 2 pin connector rated for 200 Vdc for discharger cable
45-90 volt, 300 ma LED current limited power supply  ..................  eBay item   (Do NOT buy constant current supplies!)
12 volt, 800 ma cell phone charger  ................................................ eBay item   (To power the dual LED meter)
12 volt, 2 amp regulated switching power supply  ........................... eBay item    (To run the cooling fans)
1 megohm, 1/4 watt film resistor  ................................................... Various suppliers  (Any tolerance value will work.)
220 k ohm, 1/4 watt film resistor  ................................................... Various suppliers  (Any tolerance value will work.)
0.15 ohm precision resistor  ............................................................ Various suppliers (Must be +/-2% or better for meter stability)
SPST mini toggle switch  ................................................................ Various suppliers



[4/11/2019 Update] Since the current limited LED drivers aren't listed on eBay anymore I have been investigating how to adapt the constant current LED power supplies for a grid charger. It is slow work because it takes a month or more to receive power supplies from China to test. I hopefully have configured a safe method of modifying the constant current supplies now listed on eBay to act as a limited current pair for the grid charger.

I am waiting for another set of isolated power supplies to hopefully confirm what will become an updated V2 of  my DIY grid charger/discharger. The good news is that a dual 4 digit V/A meter and LED supplies are much cheaper than the equivalent components listed in the [above] V1 bill of materials.
---

Read this, it is important;   (4/4/2019)

You MUST use current limited power supplies for a grid charger at this time. Not constant current supplies. The difference is subtle but most constant current supplies wired in series will not track properly. Unfortunately eBay Chinese sellers of LED power supplies have switched over to constant current supplies and the current limited supplies aren't listed on eBay anymore. If I find a current seller of a 45-90 volt supply I will post it here.

I have tested several different constant current pairs and typically the first supply that reaches it's maximum current (considering the load) will prevent the other supply from delivering 1/2 of the voltage to charge the battery due to a race problem when the supplies start up. This will cause one supply to operate above it's wattage rating. I will describe this further near the end of this section.

The following are the results of two typical identical Chinese 60-90 volt, constant current 300 ma supplies that I recently tested.
Test 1: Test both supplies individually for consistency.
Supply #1: (Tested by itself)
No load voltage: 126 volts
Loaded with two 40 watt light bulbs in series: 89.9 volts @280 ma (after 10 minute warm up)

Supply #2: (Tested by itself)
No load voltage: 131 volts
Loaded with two 40 watt light bulbs in series: 88.5 volts @275 ma (after 10 minute warm up)

Test 2: With both supply's outputs wired in series
No load voltage: 261 volts (NOT 180 volts as you would expect from two 60-90 volt supplies in series.)
This is actually normal for a constant current supplies. The no load output voltage has to be higher than the expected loaded voltage to deliver the maximum specified constant current to the load.

A limited current supply has a maximum voltage it can deliver (for instance 60-90 volts) and will if necessary, also limit the current so as to not exceed it's rated maximum output current. That is the subtle difference between the two types of supplies. The lower rating (60 in this case) is the lower limit for normal operation.

Test #3: Check individual load voltage of the two supplies connected in series.
Load with two 40 watt light bulbs connected in series to simulate charging current: 120 volts @ 239 (that sounds reasonable)
Output voltage of each series connected supply: (Wait till you see the next test readings @ 239 ma!!)
Supply #1: 4 volts (Yes, FOUR volts!)
Supply #2: 116 volts

Obviously these two typical non-precision constant current supplies are just not going to work for our grid charger since each supply is rated as a 25 watt supply. One supply is overloaded and doing most of the work and the other one basically took the day off.

Here's what caused one supply to only output 4 volts. When the AC power switch is set to ON there is a voltage/current race problem between the two supplies wired in series until one of them reaches it's constant current condition.

Since the supplies are wired in series, power supply #2 with the lower constant current capability with that load will have also reached it's maximum capable voltage at that maximum current. Supply #1 then fills in the remaining voltage (only 4 volts in this test) because of the current supply #2 is already forcing through both supplies. The higher current capable supply #1 can't force more current through the series connected supplies to balance the voltages because supply #2 won't allow any more current to flow.

Current limited supplies don't have the race problem because they each put out the their maximum voltage listed (90 volts) and if necessary limit the current to their maximum specified current. Yes the lower max current supply rules but only after it reaches it's specified voltage. Since the lower current sets the current the other supply is able to also output 90 volts for a total of 180V.

Return to top of this page


Notes:
  1. If you don't want to power the BCM you can buy two of item PS3, don't buy PS4 and use one of the PS3s as PS4.  Use whatever combination is cheaper at the time you buy the parts.  How to mount them is left up to the student.  Sorry 'bout that.
  2. "eBay item" are links to the current supplier of those parts. Use the pictures of the parts to find current vendors. The dual meter I used is also hard to now find on eBay. There are several dual voltage/current meters listed on eBay now but please by only the 4 digit meters to be able to read 1/10s of a volt. The vendor that I used previously for the two LED drivers isn't on eBay anymore. I have linked to another eBay vendor that seems to have the same 45-90 volt drivers I used. Do not buy similar looking units that do not output 45-90 volts DC. There seems to be many more drivers now that have either too high or too low a voltage range. You need a 90 volt DC output at ~300 ma current limited supply. The total no load output voltage of the series connected supplies should be 175 volts to 180 volts..  
  3. I've had good luck buying parts from B&G Micro for many years and you will find most of the small parts on their website. With luck you can probably find the resistors and other small parts in the computer supply that you can use for the case.
  4. I used a  male 4 pin Molex (white) computer power supply connector cut from a cable adapter on the end of the grid charger cable.
  5. I used a normal Molex 4 pin female connector from a computer power supply cable on the charging harness from the IPU.
  6. You will also need a AWG 22 as the minimum size four wire wire cable for the charging harness inside the IPU case.  You can make a cable with 4  AWG 22 insulated wires and place them in insulated tubing instead of a dedicated cable.  Use stranded wire, not solid wire.  My charging cable comes out of a gap at the bottom of the rear right corner of the IPU case next to the floor of the car body.
  7. The multimeter reads current when the push button is released (OUT) and voltage when it is pressed and holds IN.  Ignore the decimal point when reading the current.  It is preset by the meter uP and from what I can determine it can't be changed.
  8. You will need a 8 lug terminal strip with 1 ground mounting lug for wiring the various DC wire ends from the meters etc.
  9. You will also need a 4 lug terminal strip with 1 ground mounting lug for wiring the AC power wires.
Return to top of this page




First check of grid charger

The multimeter will read voltage when the small button on the lower right side of the meter bezel is pushed IN and current when the button is pushed so it is OUT.  Do not connect the grid charger to the car's charging harness at this time.
  1. Push the multimeter selector button so it stays IN.
  2. Set both switches to OFF and connect the AC cord to the charger.
  3. Plug the AC cord in to a wall 120 Vac socket.
  4. Set SW2, the power switch ON.  The power supply fan should run and both meters should light with the smaller meter showing ~12 volts DC.  If desired you may be set the 12 volt, 2 amp power supply to 13 volts or more with the pot on the 12 volt power supply.
  5. If the fan doesn't run or the meters don't light etc then trouble shoot and correct the problem. (Check the wiring and then individual components.)
  6. Set SW1, the high voltage [HV] switch ON.  The multimeter should show ~178 volts.  There's no power supply adjustment for this voltage.
  7. If you installed LP1, the 1/4 watt neon lamp it should be lit whenever the HV switch is ON and the voltage is above ~80 volts.
  8. Set both switches to OFF and read the notes below.
Notes:
  1. If the charger is not being used to charge the battery is is OK to disconnect the charger cooling fan while working on the charger. If you use a two pin fan connector don't forget the fan connector is polarized positive and negative so you must connect it properly to the 12 Vdc supply connecting it..
  2. If the cover is off the grid charger, be aware that there are open 120 Vac open connections in addition to the 178 Vdc connections just waiting to shock the didily out of you.  The same voltage that the IMA battery has!  So be careful.
  3. To power down the charger I always turn the HV switch OFF first and wait until the HV meter shows less than 20 Vdc.  Then I turn the power switch OFF and unplug the AC cord before sticking my hand(s) into the grid charger.  The neon bulb LP1 will remain lit until the HV falls below ~80 volts.  But 80 Vdc can still shock you so have patience and wait for the HV meter to go down to 20 Vdc.
  4. The ONLY time I will work on the charger with it plugged into a wall socket  is when calibrating the multimeter or trouble shooting it. 
Return to top of this page



Calibrating the multimeter

Once the initial power supply checks have been done you need to calibrate the multimeter.  The voltage calibration pot is closest to the multimeter power connector at the lower edge of the meter PCB.  The current calibration pot is located at the top of the PCB.  You will need a small very short flat blade screw driver to set the 10 turn calibration pots.  You can also use a small needle nose pliers to set the pots.

The current calibration is done first.  An accurate DVM is used as the calibration standard to set the multimeter. To simplify calibrating the current meter I use an adjustable low voltage power supply to light a small 12 volt pilot type bulb.  The bulb should be rated to draw at least 1/4 amp at 12 volts.

In affect the low voltage power supply takes the place of the LED drivers and the 12 volt light bulb simulates the IMA battery being charged. The multimeter doesn't care what voltage is being used when reading current.  The low voltage supply is fine for calibrating the meter and avoids having 180 Vdc floating around in the charger while sticking your hands into it.  Just be careful and don't touch the AC power terminals etc.


How to connect the external low voltage supply to the charger.
  1. Do not plug the AC power cable into the charger at this time.  Set both power switches on the grid charger to OFF.  Set the external power supply to OFF.
  2. Connect the positive voltage lead of the external variable power supply to J2-1.  (The "-1" or "-2" is the pin number of the connector).
  3. Connect the negative voltage lead from the external supply to J2-2.  (J2 is the grid charger output connector to the charging harness.) 
You can also connect the external supply leads to the appropriate points inside the charger case.


How to connect the DVM and 12 volt light bulb (dummy load) to the charger.
  1. Set the DVM to read at least 500 ma (current).
  2. Connect one lead of the bulb to P1-1, the positive terminal.of the discharge connector.
  3. Connect the other lead of the bulb to the DVM positive lead.
  4. Connect the negative lead of the DVM to P1-2, the negative terminal.of the discharge connector.
Note:  You can use jumper wires to connect to the appropriate points inside the charger.


How to do the current calibration.   (Use caution since there are exposed 120 Vac terminals inside the charger.)
  1. Set both power switches on the grid charger to OFF and then plug the AC power cord into the charger.
  2. Set the charger power switch SW2 to ON, leave the HV power switch SW1 OFF.  The grid charger fan should run. The 12 volt meter should show ~12 volts or whatever you set the 12 volt power supply to.  You can disconnect the fan for current and voltage calibration if you want to.
  3. Set the multimeter to read current by pushing the bezel button so it stays OUT.
  4. Set the external supply to the lowest output voltage and turn the supply ON.  Slowly increase the voltage until the external DVM current meter reads 250 ma (1/4 A).  The light bulb may only glow at this low current and that is OK since we are only interested in the current.
  5. If necessary adjust the multimeter current calibration pot to read the same current as displayed on the DVM.  Ignore the decimal point as it is pre-wired in the meter and can't be turned off when reading current.  (250 ma will read 25.0)
  6. Set the charger power switch (SW2) and the external power supply switch to OFF.  
  7. Remove the external power supply, the DVM and the light bulb.

How to do the voltage calibration  
(The HV output of the charger can't be changed but we do want to calibrate the multimeter.)
  1. Set both power switches on the grid charger to OFF.  Do not plug the AC cord in the charger.
  2. Set the DVM to read at least 200 volts DC.
  3. Connect the positive  lead of the DVM to J2-1, the HV positive terminal.of the HV output of the charger. 
  4. Connect the negative lead of the DVM to J2-2, the HV negative terminal.of the HV output of the charger.
  5. Push the multimeter button so it is IN to read the HV.
  6. Connect the AC power cord to the charger.
  7. Set the charger power switch SW2 to ON. The meters should be lit and the 12 voltmeter will read voltage.
  8. Set the HV power switch SW1 to ON.  The DVM should read ~178 volts.  As long as the voltage is above 175 volts the charger will correctly charge the IMA battery.
  9. Set the multimeter voltage calibration pot to read the same voltage as displayed on the DVM.
  10. Turn both power switches to OFF and disconnect the DVM.
  11. Your meter is now calibrated.    -- Enjoy! --
Return to top of this page



Make a discharge load

First off you might ask, "Why even discharge the IMA battery?"  Good question.

In use the individual cells of the IMA battery will tend to deteriorate over time and the cells will have different levels of watt hour capacity. This is called an "unbalanced battery".  When driving the car, the weaker cells limit the time and amount of assist the battery pack can deliver.  The batteries can usually be brought back to good capacity by properly grid charging them.  This is called "balancing the battery" or "a balancing charge".  New batteries ought to be grid charged three or four times a year after the first 6 months to a year of use. This helps to keep the battery pack balanced over time.  A weak battery should be first charged to balance them as much as possible, then discharge the battery, and then charge again.  Really weak batteries respond better to doing the charge/discharge sequence several times in a row.

The purpose of discharging the battery is to discharge all the cells down to what is considered the discharged condition.  Individual Ni-MH battery cells are considered discharged when they read 0.9 volt per cell.  Of course we can't read each cell of the pack while the cells are installed in the car so we have to fudge the discharge a little here.

Unfortunately when discharging an unbalanced battery pack the weak cells will discharge to 0.9 volts much quicker than the good cells.  In fact really weak cells can not only discharge to 0 volts but they can be slightly "reversed" charged by the discharge current while the good cells in the pack continue being discharged.  At high current levels, reverse charging cells is not good!  Using too high a discharge current with cells reversed charged can result in those cells overheating with venting which will cause all sorts of problems over time.

Through experimentation, the general consensus on the Insight Central  forum is that if the cells are reversed charged at a low current, that minimum or no damage will occur.  What that low current  is, is still up for some debate.  But 200 ma or less range at the end of the discharge should be OK.  And if your battery is on it's last legs anyway it may be worth to do this procedure on the chance you will get more life out of it. Several people have discharged weak IMA battery packs to zero volts and had very good results even after the first grid charge to 100%.  Two charge, discharge, charge cycles are better.

As the say in the TV advertisements, "But first .........."  
If you have installed one or two diodes in the high voltage charging harness you will NOT be able to discharge the battery. So if you want or later need to discharge the battery remove any diodes in the HV lead(s) of the charging harness.



OK, here's how you actually make the discharge load.

The actual discharger I presently  use consists of two ceramic light sockets mounted on a short piece of 1" x 6" board.  One socket has a pull chain ON/OFF switch on it. The actual discharge load can be made by using 120 volt light bulbs of various wattage combinations wired in series. Nothing exotic about that, just wire the two sockets in series like Christmas tree lights with a single length of wire and attach the 2 wire discharge cable to the single free terminal on each socket.  I used a long 120 volt power cord from a vacuum cleaner as the discharge cable so I can have the light bulbs inside my garage to keep an eye on the discharge while it is running.

The current drain when using light bulbs will vary depending upon the battery voltage.  With two series connected 60 watt bulbs and the battery at 162 volts the discharge will start at ~265 ma. When the battery is down to 115 volts the discharge will be ~240 ma.  If you want a lower discharge rate vary the wattage of the two bulbs as the current changes depending up on what final current you want to discharge at.  Lower wattage bulbs give lower discharge currents.

I would recommend the discharge to be below 200 ma near the end of the discharge if possible.  But again these numbers are still being discussed and tests on the Insight Central forum are still being done as you read this.

I did some bench tests on various wattage incandescent light bulbs to help understand the current vs voltage. I also checked the current flow using a 500 ohm, 100 watt resistor for comparison. The data is presented on the chart below. Use this chart as a guide to what wattage bulbs to use depending upon what current you want to discharge the battery at.  

While using a resistor as a discharge load will work; you do have to take into consideration that the resistor will be dissipating up to 60 watts when starting the discharge with a fully charged pack.  Even mounting my aluminum cased 100 watt resistor on a 3" x 5" finned heat sink did not keep it cool enough so it wouldn't cook your finger if you touched it.  A computer fan did not help much.  As a test I finally used a microprocessor dual heat pipe heat sink with a large shrouded fan to be able to keep my fingers on the resistor for a short while.  

Again, I recommend using two old fashioned light bulbs.  We know they are going to be hot and generally don't put our fingers on them.  And they do give some indication of how the discharge is going along.

Click on the chart to see it full size.

I vs V for light bulbs * 100 Ω resistor

One thing to notice is that the current vs voltage for an actual IMA discharge does not exactly correspond to the bench measured currents using the same bulbs by ~3%.  I rechecked the calibration of my grid charger at room temperature and the voltage and current readings are correct using the same DVMs used for the bench measured data and the grid charger calibration.

Other people have also noticed that the current limited LED driver supplies do not always run in current limiting when charging the IMA battery. When driving a resistive load my supplies current limit at ~300 ma when the voltage is above 145 volts DC. In use the grid charger maximum output is ~255 ma. 

 But they do the job so we use them.

I would take the chart currents to be a close approximation of what the discharge currents will be as a discharge load.  With the current meter built into the grid charger you can check the actual discharge current and perhaps use different size bulbs in various combination to get the current you want at the beginning or end of the discharge.

Old fashioned incandescent light bulbs are getting hard to find and CFL bulbs will NOT work.  Halogen flood lights work fine and have the same curved discharge current vs voltage as the normal old fashioned bulbs.

Before I could wire the ceramic bulb sockets I had to route out an area on the board underneath where the sockets would mount with a ~1/2" wide trench between the sockets for the wires to go between the sockets.  There is another short trench from one socket for the two wire discharge cable to leave the board.  I used a matching 2 pin male connector on the end of the cable to plug into a 2 pin female connector on the grid charger.

The picture on the left shows the routed out area underneath where the sockets mount.  The right side picture shows the sockets mounted on the board.

Discharger openDischarger sockets mounted

I used a long two wire cord from a vacuum cleaner (with the plug changed) so I could have the discharger sitting in my closed garage to have an idea how the discharge is going.  It would be possible to also have provision on the board to mount a voltmeter to keep track of the discharge.  I didn't do that because I use my gird charger multimeter to read the current and voltage and run the cooling fans.  

Some people use dual bulb "Y" sockets designed for wall mounting.  They have two wires on each socket that can be wired in series.  You don't have to route the board with a "Y" socket. The problem with the "Y" socket is that there is no switch on the ones I've seen.  I like the remote cut off feature of having a switch on one of the sockets.

While the battery is being charged or discharged the batteries get warm so you MUST run the battery pack cooling fan at the same time.

Return to top of this page



Information about rejuvenating a battery


Normally when you buy most small rechargeable batteries the manufacturer recommends that you charge them at low current for 16 hours or so. Especially so if the batteries have been sitting uncharged for awhile.  This used to be called "the forming charge" when the batteries are made. In affect, rejuvenating a battery is re-forming the battery.

To rejuvenate a battery you should do at least 2 or 3 charge/discharge cycles with a final charge.  One cycle would be, a low current charge to fully charge the battery followed by a low current discharge.  Different discharge cutoff voltages have been put forward by members on insightcentral.net/forums/  but I normally use the manufacturers recommended voltage of 0.9 volt per cell on batteries for the first discharge. To make it easy to read on the meter I use 100 volts as the cut off point for the first discharge.

For batteries that are really on their last leg it has been suggested to do at least 3 discharge and full grid charge cycles with the cutoff voltages decreasing from 100 volts to 75 and 50 volts for each of the three discharge cycles.  The problem is that the lower voltages will certainly result in some still bad cells being reversed charged during the discharge cycle if the pack is still unbalanced.  This may cause problems later on.  Probably not ... but ... maybe.

Recently tests have been done by completely discharging the battery pack to five or even zero volts.  There is a technical paper here that indicates that there is some merit in this approach.  The results are encouraging on batteries that had to be grid charged very frequently.  But at this point the experimenters aren't sure if the fix is a long term fix or not.  I recently talked to the owner of an aircraft battery maintence company.  If necessary he discharges Ni-Mh batteries to zero volts and then puts a dead short on them overnight before charging them back up to a full charge.

It has also been reported that batteries that have sat unused for several years and have self discharged to very low voltages were brought back to life by doing several rejuvenation charge/discharge cycles.  Individual sticks of 6 cells that were in very weak condition were also brought back to life after sitting for long periods of time by cycling them.

I recently found an Insight that sat for 10+ years outside of a home.  The IMA battery had self discharged to a total pack voltage of 0.816 volts. After  talking to the owner he allowed me to remove the battery and do rejuvenation at my house.  After two cycles the average stick voltage was 16.42 V +/- 0.02 Volt.  The car is in very bad shape due to water leakage because the radio antenna base being sheared off in a hurricane, front bumper half ripped off and just general deteriation from sitting in the weather for many years.  For that reason we couldn't actually drive the car to see how the battery would perform.

"eq1" an  insightcentral.net/forums/  member has done a lot of battery work and has come up with some very good results when completely discharging battery packs. I would highly recommend reading this post of his and this one on how he sets his discharge current.

When the battery pack voltage gets to 30 volts or lower during the discharge the current also be down in value.  You may want to change the light bulbs to ones with a higher rated wattage (when used at 120 volts) to raise the discharge current back up again.

While eq1 discharges at quite a bit higher current than I use,  I would shoot to keep the discharge at 100 ma or so when going below 50 volts.  My idea is to give the chemical reactions within the batteries time to complete.

Be careful when changing the bulbs. The bulbs may be dim at the lower voltages but they are still very hot if you change them without wearing work gloves.  I would turn the discharge off with the socket switch, wait awhile for the bulbs to cool off, change them to higher wattage bulbs and then start the discharge again to the final volt level for that cycle.  You will find the pack voltage rises quite a bit while waiting for the bulbs to cool off but the voltage will also go back down as soon as you resume the discharge.

Storing a stick or a complete battery for a long time doesn't seem to be a problem.  There is no danger of a cell becoming  reverse charged by allowing a battery pack to self discharge for a long time with no load on it.  Ni-MH batteries self discharge roughly about 1% per DAY.


 Damage can occur by improperly charging a dead battery when starting to use it again.

If you find a car that has sat for many months or several years without running it is recommended that a slow grid charge be done before using the IMA battery again. In other words, don't just start the car and take off on a long trip home with it.  The cars normal charging current has much too high a charge current level to do a proper balancing or re-forming of the battery.
 
If you have a short trip you can switch the IMA battery master switch to OFF, start the car on the 12 volt battery that has been charged and drive the car home. The switch is located under a small cover at the center of the battery box (IPU) at the back of the car under the rear carpet. You will need a 10 mm wrench to remove the small cover.  There is a red safety clip over the handle of the switch that needs to be lifted off the handle so you can switch it OFF.  Replace the clip so you don't loose it.

CAUTION:
With the IMA battery switched OFF the 12 volt battery will NOT be charged so the distance you can drive is limited by how long the 12 volt battery can supply enough voltage to run the gas engine.  It is possible to open the IPU battery case and disconnect one of the connectors on the BCM computer box and safely run the car as a gas only car for as long as you want to. The 12 volt battery will be charged and some people have run their Insight for years that way. Instructions how to disconnect the IMA battery can be found at  insightcentral.net/forums/    Google "disconnect battery" on the forum.

Please note that the car will have a much slower acceleration rate at low engine speeds with the IMA battery not being used.  The gas engine is tuned so it has to be revved up to develop its peak power (torque) at ~4500 RPM.  Normally the IMA system (electric motor) fills in the low RPM power band.

Return to top of this page



Do a rejuvenation

IMPORTANT: When discharging the battery you must connect the grid charger to the charging harness so the IMA battery cooling fan can run while doing the rejuvenation since discharge cycles usually follow charging  cycles.  The battery will continue to rise in temperature for many hours after a charge so the cooling fan should be running even while doing a discharge.

DO NOT TURN THE GRID CHARGER HIGH VOLTAGE POWER SWITCH ON WHILE DISCHARGING THE BATTERY!

The method I use to have an idea when it's time to stop the discharge is to record the battery voltage at the start and at each hour thereafter as the voltage drops to the voltage when you want to stop the discharge (or discharge) at.  You will find the volts per hour change will start at 4 to 6 volts/hr and will taper off to below 1 volt per hour.  A weak battery may drop from 115  to 100 volts very quickly.  So keep an eye on the voltage so you don't reverse charge many weak cells if you are discharging at high currents.  With a low current discharge the cells probably won't be damaged anyway with the fan running but why take a chance?

By recording the volts per hour change as the battery goes below 115 volts you can guesstimate how weak the battery is.  A quick voltage drop of 1 volt in 5 or 10 minutes is a sign of weak a cell(s) becoming completely discharged.  Pay close attention to the battery voltage at that point if you are not discharging to zero volts.

After you terminate a discharge the battery voltage will rise up to some higher value rather quickly with the load disconnected.  A weak battery discharged to 100 volts will typically rise back to 120 volts or more.  A really good battery will not rise as much. It is permissible to start the discharge again to pull the battery back down to the goal voltage but keep an eye on the voltage because it will drop to your desired cutoff voltage in a shorter time after the initial discharge.

I usually let the battery rest an hour with the IMA fan running to dissipate some of the battery heat before starting the next phase of a charge or discharge cycle.  I also try to start the charge in the early evening so it runs when the ambient temperature is below 86 F.  Here in South Florida I usually don't have to worry about the temperature being too cool.

When charging the battery, the high cut off voltage should top out at ~174 volts for a battery in good condition.  The exact voltage will vary depending upon the battery temperature and charge current. A weak battery may show higher than 175 volts due to higher than normal internal resistance (IR) of the cells.  The higher the charge current, the higher IR voltage drop which will cause the total battery voltage to appear higher while charging.  As you can see there are a lot of variables when working with a 120 cells in series battery.

By watching the charge voltage change per hour you will see that the voltage increase rate per hour will start at 4 to 6 volts per hour and then taper off to ~0.2 volt per hour when the voltage is nearing 170 volts.  At that point the cells of the battery that are are 100% charged will be producing a lot of heat. So it is important to not turn the cooling fan off while charging or discharging.  If you use a 3 digit voltmeter you will not be able to see the slow 0.2 volt per hour change.

The typical charge time for a good battery may be 24 hours to 35 hours. The discharge time varies greatly and is determined by how much capacity the battery has etc. A complete rejuvenation is going to take days to complete.  But if you can get a year or two more life out of a weak battery it may be worth it.

The money you save on gas can pay for a new (better than OEM) battery from one of the forum vendors.  One Insight owner says he not only paid for the new battery by grid charging it but for the car too!


Return to top of this page

Return to top of the Insight home page