OlRowdy01 V2 Grid Charger Build (Illustrated)

1: PARTS LIST AND SOURCES

Ol’ Rowdy is an electronics hobbyist and has a lot of “organized junk” lying around from which he was able to identify and scavenge parts. I ordered most of my parts new. Circa January 2020, these were my sources and costs. Part numbers in parentheses (eg, PS1, R1, etc) refer to labels on Ol' Rowdy's circuit diagram…


GRID CHARGER
$31.11 ... Mean Well MLG 60H-C350A 100-200 volt, 350 ma constant current power supply (PS1)
Link: HLG-60H-C350A - Mean Well Authorized Distributor
This is the part that does all the work of charging the battery, and what makes the V2 design work. It’s a nice heavy-duty power supply available from multiple sources. As a constant-current supply, it tries hard to hold the current steady somewhere around 350 milliamps even as the voltage changes (say, as the battery charges).

$0.76 ... 1N4007 silicon diode, 1amp, 1000 V (D1)
Link: 1N4007 Si Diode, 1000V 1A - NTE125
$0.29 ... 100K ohm, 1/2 watt resistor (R1)
Link: 100K Ohm 1/2 Watt Resistor
$2.49 ... 2 amp fuse, 250 volt and holder (F1)
Link: 5x20 (GMA) Fuse Holder - Panel Mount
$2.99 ... SPST Round Illuminated Rocker Switch On-Off 16A 125VAC neon red (SW1)
Link: SPST Round Illuminated Rocker Switch On-Off 16A 125VAC Neon Red
$2.99 ... SPST Round Rocker Switch On-Off 10A (SW2)
Link: Philmore 30-16060 - SPST Round Rocker Switch On-Off 10A
These are the various bits needed to complete the circuit. I decided to make SW1 an illuminated switch so I could visually verify charge/discharge mode.

$3.85 ... 200V 10A dual digital meter 4-digit precision (M1)
Link: DC 0-200V 10A 4 Bit 5 Wires Voltmeter Ammeter Red+Red LED Amp Dual Digital Volt | eBay
These meters seem to come and go on eBay all the time. They ship from China and are still less than $5. Crazy world. Be sure to get one that has two displays of four digits each, one for voltage and one for amps. Ol’ Rowdy recommends red instead of blue so the numbers don’t wash out in bright sunlight.

$6.99 ... 15A/125V orange 3-wire M3 female connection for discharge harness (J2; Menards #3638617)
This is a bright orange female plug designed to be installed on the end of an extension cord. It can be adapted to make a sturdy discharge socket.
$1.09 ... 1/4" hole grommet 6-pack (Menards #3648469)

$FREE ... Computer power supply case
$FREE ... 12v DC cooling fan (F1)
$FREE ... Molex female 4-pin 2x2 (J1)
$FREE ... Male 3-wire AC power cord (P1)
I pulled the standard-sized power supply from an old Dell PC for the metal case. Ol’ Rowdy recommends using an extra-large power supply, but I was able to make everything fit in a standard one. The 12-volt cooling fan mounted in the power supply worked fine. It turned out that the square 4-pin power connecter from the power supply to the motherboard matched the connector on the RB Batteries harness (see below), so I was able to use this part as well.

$FREE ... 5 volt switching power supply => powers meter (PS2)
$FREE ... 12 volt, 1 amp min regulated switching power supply => powers cooling fans (PS3)
These are standard plastic wall-wart style chargers that I happened to have laying around. The 5-volt supply was for an old cell phone, and the 12-volt supply was for an old cable modem. The 12-volt supply needs to be a minimum of 1 amp to drive both the fan in the grid charger case and the large fan on the IMA battery in the car.

I also was able to use heat-shrink tubing, electrical tape, wire, crimp-on electrical connectors, wire nuts, rubber feet, and zip ties I had around.

TOTAL COST OF GRID CHARGER $52.56 not including shipping.


HARNESS
$30.00 ... Charging Harness
Link: Add-ons
I opted to order RB Batteries’ harness for the Insight because it looked nicely constructed and was designed for easy installation. Plus, if my build failed, I could just buy RBB’s grid charger. You could save quite a bit of money making your own harness for the car.


DISCHARGE LOAD
$2.39 ... Porcelain lamp holder w/ switch (Menards #3634101)
$1.29 ... Porcelain lamp holder (Menards #3634091)
$4.99 ... Orange extension cord 10’ (Menards #3700433)
$1.99 ... 25W incandescent bulbs, 2-pack (Menards #3533160)
$1.99 ... 40W incandescent bulbs, 2-pack (Menards #3533161)
$4.99 ... 60W incandescent bug lights, 2-pack (Ace Hardware)
Old-fashioned incandescent light bulbs, not LED bulbs, are required for the discharge load. The new standard incandescent light bulbs are often “energy efficient,” so that a bulb rated at 60W actually only uses 47W – check the fine print. Look at the specialty bulbs for simple incandescents. I finally had to find an old Ace Hardware store with some dusty 60W incandescent bug lights on the shelf!

TOTAL COST OF DISCHARGE LOAD $17.64

2: POWER SUPPLIES AND CASE CONSTRUCTION

One of the challenges Ol’ Rowdy encountered building his new grid charger design is that the new Mean Well power supply is large. He had to find a super-sized computer power supply to have a big enough case.

I started with a standard-sized power supply from an old Dell computer. After stripping out the original internal components, I experimented with the three power supplies – the Mean Well, and the two plastic wall-wart transformers. I discovered that by setting the Mean Well on its side, bending the mounting tabs at one end, and removing the mounting tabs at the other end, I could get it to fit.



The bent tabs were angled so I could attach one end of the Mean Well to the case with a small machine screw. The tabs at the other end were removed by carefully bending them back-and-forth with a pair of pliers until they came off. In hindsight, I would recommend testing the Mean Well power supply in the circuit BEFORE bending or cutting the tabs!

The wall-wart transformers were labelled to indicate their voltage and amperage output. I cut the plugs off the end of the wire, separated the wires, and tested them with my multitester to see which was positive and which was negative. Label the positive (+) wire, as this is important later! The transformers are designed so the positive wire is the same no matter which way the wart is connected to 120V.



I marked the location of the power supplies on the bottom of the case, then used a Dremel to cut slots in the case on each side of the three supplies. Later in construction, I used the slots to run zip ties around the three transformers and strap them in place. Don't mount the Mean Well permanently until the very end - it is bulky to work around, and you may have to make adjustments later that can't be done with the power supply mounted in the case.





By test-fitting other components, I decided to place the fuse holder and discharge socket on the back of the case, and the meter, two switches, and charging harness connector on the front of the case. I used a combination of Dremel and drill to cut the necessary holes. The round switches I used have a tab on one side so they don’t spin in the hole; Dremel to the rescue.



The plug I chose for my discharge socket is a heavy-duty three-prong female plug designed to be installed on the end of an extension cord. By cutting off the wings I could install it neatly in the case.



I chose this so that I could put a standard 3-prong cord on my discharge load, in case I ever want to use it as a light in the workshop. The discharge socket should be clearly labelled on the case “200V DC DISCHARGE” so no one mistakenly plugs in a standard 120V AC tool!

3. AC WIRING

After fitting components and cutting holes, I began by wiring the 120V AC side of the circuits. I always find it best to start at one end of a circuit and methodically work it through. I began at the AC plug on the back of the case. After testing with my multimeter, I confirmed that the stubs of Dell’s wiring on the back conformed to standard color codes: brown is the “hot” side of the AC, blue is neutral, and the green striped wire is the ground.

I happened to have some wire left over from a trailer project that matched the color coding, and so I tried to be consistent throughout the circuits to make troubleshooting easier. If you need different colors of wire, small hardware stores like Ace often sell it by the foot.

The green striped wire (ground) from the plug on the case is grounded to the case itself. Even though the ground pin on the discharge plug won't be used, I went ahead and wired it to the case ground as well.

The brown wire (hot) from the plug connects one terminal of the fuse holder. The other terminal of the fuse holder connects to pin 2 on the main power switch (SW2 on Ol' Rowdy's circuit diagram and the parts list).

The blue wire (neutral) from the plug needs to connect to the blue wire from the Mean Well, to one spade on each of the wall warts, and to pin 3 on the illuminated switch (SW1; see below).

In the photo below, I'm using wire nuts to make temporary connections while I am testing.



In my box of electrical parts, I found some crimp-on female spade plugs large enough to slide on the spades of the wall-wart transformers. I cut the spades shorter, then used the crimp-on female connectors to attach wires. Throughout all the wiring, I made extensive use of heat shrink tubing to cover and insulate all the electrical connections. There’s a lot of wires and a lot of volts in this little box, and you don’t want anything getting mixed up!



The main power switch (SW2) is not shown in the above photo. Pin 2 should connect via brown wire to the fuse holder. Pin 1 should connect via brown wire to the second spades on the wall wart transformers, and to pin 1 on SW1, the charge/discharge switch.

The illuminated rocker switch (SW1) did not come with a description of pin function. By googling the part, I found a diagram showing that pins 1 and 2 are the single-throw/single-pole (SPST) switch, and pin 3 is the ground for the light. To wire the switch so it lights in charge mode but is off in discharge mode, connect pin 1 of the charge/discharge switch (SW1) to pin 1 of the main power (SW2). Pin 2 of SW1 connects via brown wire to the Mean Well's brown (AC hot) wire. Pin 3 of SW1 connects via blue wire to the main neutral wire coming off the wall plug on the case.

With the 120VAC side of the circuits wired, you should be able to temporarily connect the DC outputs from the 5V wall wart to the digital multimeter, and the DC outputs from the 12V wall wart to the fan. Make sure you connect the positive (+) wire you marked earlier to the red wire on each.

Now you can power up the main switch. The fan should run and the meter should illuminate. When you flip the charge switch on, the switch should illuminate. (If you've actually wired the Mean Well into the circuit, you now have 200V DC on the output side. Careful!)



If the switches work, you can snap or glue them into place. This is also a good time to zip tie the wall warts in place. Leave the meter loose for now to make wiring and adjusting it easier later. And again, don't permanently mount the Mean Well until final assembly!

4. THE PIGTAIL FOR THE HARNESS

As mentioned in the parts list, I chose to order a $30 charging harness for the car from RB Batteries because it looked like it was well made and included mounting hardware to make installation easy. Plus, it matches at least some of the commercial chargers out there in the wide world.

The construction is simple but nice, with hardy wire-loom protecting the wires from the plug back. The wires at the installation end are labelled, and the battery connections have the right connectors to make installation easier.



Fortunately, the 2x2 plastic connector on the end of the harness exactly matched a 2x2 connector on the Dell computer power supply that fed power to the motherboard. I clipped this off of the old power supply to use as the pigtail connection on the grid charger.

To identify which wires did what, I connected it to the RBB harness, and then used my multitester to check continuity from the labelled ends of the harness to the ends of the pigtail wires. I used color-coded heat shrink tubing and a fine point sharpie to label the pigtail wires. I also used larger heat shrink tubing to wrap all the pigtail wires into a neat bundle.



To install the pigtail in the case, I drilled a hole and inserted a 1/4" inner diameter rubber grommet. I slid the pigtail through from the outside further into the case than the final position. I then wrapped electrical tape around the pigtail inside the case to taper it wider at the point where I wanted it to stop, and tugged the pigtail back out until the tape wedged into the grommet.

5. DC WIRING AND ELECTRONICS

With power to all the transformers, it's time to start hooking up the DC side of things. Begin with the simple circuits.

The 5V DC from one wall wart is used only to power the meter. You probably hooked it up and tested it in step 3 already. Make sure that the positive (+) wire from the transformer is connected to the thin red wire from the meter. The other wire from the transformer connects to the thin black wire from the meter. At this point you can solder these connections and insulate them with heat shrink tubing or electrical tape.

The 12V DC from the other wall wart powers the fan in the grid charger, and the fan on the IMA battery through the harness. Make sure the positive (+) wire from the transformer is connected to the red wire from the fan in the grid charger case, and to the fan (+) wire on the pigtail. The other wire from the transformer connects to the black wire from the internal fan, and to the fan (-) wire on the pigtail. Solder and insulate these connections.

Stop for a moment and power up the grid charger to make sure the meter lights and the fan runs. Yay.

Ol' Rowdy included two small electronic parts into the circuit for the high voltage grid charger. The diode keeps the battery from back-feeding power to the Mean Well when the grid charger is turned off. The resistor helps bleed off high voltage when the charge mode is shut off.

I started with the thick red and black pigtail for the meter. Connect the thick black wire from the meter to the negative (-) wire from the DC output side of the Mean Well and to one end of the resistor. If you are using heat shrink tubing, think about how it's going to work and stage the tubing in advance... (experience speaking)

In the photo below, you can see the end of a small vice grip clamped onto the lead for the resistor. The vice grip makes a great heat sink when soldering, so that you don't accidentally cook the resistor and let out the "magic smoke."



Next comes the diode, which is directional. Typically the striped end is the cathode (-) side, and the other end is the anode (+). Make sure you get this right, because the charger won't work if it's backwards.



The other end of the resistor, the anode (+) side of the diode, and the positive (+) wire from the DC output side of the Mean Well get soldered together. Use heat sink(s) to protect the diode and resistor! In the photo below, you can see how I've used heat shrink tubing to insulate and protect the diode and resistor.



Finally, the cathode (-) side of the diode connects to the thin yellow wire from the meter, to the IMA (+) wire on the pigtail, and to the "hot" pin on the discharge socket. Use a heat sink to protect the diode and keep that "magic smoke" inside. (grin)

To finish the DC wiring, the thick red wire from the meter connects to the IMA (-) wire on the pigtail, and to the "neutral" pin on the discharge socket. I wrapped the end of the discharge socket with electrician's tape to protect the exposed screws from shorting on the Mean Well or other metal inside the case.

At this point, the wiring is done. It looks like spaghetti, but hopefully somewhat neat spaghetti that you can trace, and that can be tucked inside the cover of the case.



Before you hit the power switch, trace all the wiring one more time. Since I wired forward from the 120V AC plug, to double-check I traced the wiring backwards from the IMA and discharge sockets.

Don't strap down the Mean Well yet, and don't permanently mount the meter. You may need to make some adjustments on both.

If all looks good, try to power it up. When you flick the main power switch, the fan should run and the meter should light up with zeros. When you flip the second switch to charge, the switch should light up and the top, voltage readout on the meter should jump up to 180-210 volts (ballpark). When you flip the second switch to discharge, the meter should start to drop as the resistor bleeds off voltage.



Does it look good so far? Yay, again!

6. BUILDING THE DISCHARGE LOAD

The discharge load is simply two incandescent light bulbs wired in series. Buy two cheap porcelain light fixtures at the hardware store and an extension cord. I matched an orange cord to the orange discharge socket. You'll also need a chunk of 2x6 or 2x8, and some screws to mount the fixtures to the board.

Cut about a foot off the female end of the extension cord. You'll use this short end later during testing.

Working with the long end, strip the insulation off the cord far enough for the wires to reach across the bottoms of both sockets. Cut the green ground wire off, or wrap it around the cord and bind it with electrical tape.

Run the long black wire to a screw on the far socket.

Cut the white wire shorter and run it to a screw on the near socket.

Use a spare piece of wire to connect the unused screws on the two sockets.



Cut a groove in the board just big enough for the wires, then flip over the sockets and screw them down. Voila, a simple discharge load!



With the sockets wired in series, the current flows through one bulb and then the other before completing the circuit. The one switched socket controls both bulbs. Wired in series, the resistance of the two light bulbs is added together.

By using a standard electrical cord, the discharge load can also be used as an extra light in the garage or shop.

Go ahead and strip the white and black wires on the short end of the cord that you cut off earlier. The green wire can be cut off or folded over and taped off. You'll use this little cord to help with testing in step 7.

7. TESTING AND ADJUSTMENT

Now we're ready to test and adjust the grid charger. For this step you will need some 60W and 40W incandescent lightbulbs and a multitester.

There are three adjustments to make:
(1) Setting the output current on the Mean Well. (You didn't strap it in yet, did you?)
(2) Adjusting the amperage readout on the digital multimeter.
(3) Adjusting the voltage readout on the digital multimeter.

I didn't have much luck using the calibration and adjustment method suggested by Ol' Rowdy - my grid charger could not light two 60W light bulbs in series. At first I thought my charger build was at fault. I have since tested two other Mean Well power supplies and found that they couldn't power two 60W bulbs either. So the following instructions are my own adaption...

First, a primer on how to use a multitester to read voltage and amperage. In the picture below, a bare Mean Well has been set up to drive the test load.

To read amperage, the multitester on the left is actually wired into the circuit, so all the current flows through the tester on the way to (or from) the load. The brown (+) wire from the Mean Well is connected to the black lead of the meter, and the red lead of the meter is connected to one wire from the test load, so that current flows through the meter. The other wire from the test load is connected to the blue (-) of the Mean Well. (Note that the meter is reading -0.34 amps because I have it wired backwards, but the numeric value is accurate).

To read voltage, the multitester on the right just has to touch the leads from the tester to two different points on the same circuit, upstream and downstream of the load. The red lead from the tester is connected to the (+) side of the circuit, and the black lead from the test is connected to the (-) side.



Don't worry if you don't have two multi testers. We only need to measure one thing at a time. Be very careful doing this. Remember, this is 200V we're talking about - you don't want to touch it!

Begin by plugging the short end of the extension cord into the long end coming from the discharge load. Screw in a 60W and a 40W bulb. Wrap one of the wires from the short extension cord around the IMA (-) end of the charging harness.

----
To adjust (1) and (2) on the list above, we need to measure current.

Take the black lead from your multimeter and connect it to the one wire on the short extension cord, and tape it. Take the red lead from your multimeter and connect it to the IMA (+) end of the charging harness.

Set your ammeter on "amps" - probably 1 amp or maybe 200-300 milliamps if you have different settings.

Connect the charging harness to the pigtail on the grid charger.

Turn on main power to the grid charger, then set it to charge mode.

Use the switch on the discharge load to turn it on. If the bulbs don't light, or only flash, turn the load off and on again until the bulbs light. (Still won't work? Make sure both bulbs are good, and are screwed in. Or try a 40W/40W bulb combination. Also make sure your multitester is configured to read amps and is turned on.)

----
Step (1) - With the bulbs lit, your multitester (NOT the meter in the grid charger) should read around .350 amps (350 milliamps). If it doesn't, we need to adjust the Mean Well's output. Set the unit to discharge and wait until the voltage drops to near zero. Then shut the grid charger off and unplug the power cord.

Carefully slide the Mean Well sideways until you can access the little rubber plug that covers the adjustment screw. Pop the plug out. Align the Mean Well so you can insert a philips screwdriver and turn the adjustment screw.



Now, plug everything in, turn it on, and fire up the charger. BE CAREFUL! There are a lot of places you can shock yourself right now!

With the bulbs lit, carefully turn the screw until the multitester (NOT the meter in the grid charger) reads approximately 0.350 amps (350 milliamps). This is not a very precise adjustment - just get it as close as you can.

Once you have the current set on the multitester, shut everything down, let the voltage bleed off, and unplug everything. Put the rubber plug back into the Mean Well and slide it back into position. NOW you can strap the Mean Well down.

----
Step (2) - Turn everything back on and light up the bulbs. If the lower number on the grid charger's meter doesn't match the amps reading on your multitester, the meter in the charger needs to be adjusted. There are two tiny screws on the back of the meter which you can turn with a tiny flat-tip screwdriver, preferably one with an insulated plastic handle. They are probably labeled VR (volts) and IR (amps).



(Note that in the picture above, the meter is not connected to the grid charger. When you are making these adjustments, the meter needs to be fully wired in but loose from the case, so that you can carefully adjust the screws on the back while reading the numbers on the front.)

With everything powered up, very carefully adjust the IR screw on the back of the meter until it matches the reading on the multitester. This is a finicky adjustment and you probably can't get it perfect. Just aim for close. And watch out for all those places where you can get zapped!

----
Step (3) - To adjust the voltage reading on the meter in the grid charger, you need to reconfigure your multitester to read voltage. Power everything down and unplug it. Leave the (-) wiring in place. Disconnect your multitester leads, and connect the IMA (+) wire from the harness directly to the wire from the discharge load.

Now power everything up and turn on the bulbs. Set your multimeter to DC volts (probably the range is 200VDC). Touch the red lead to the IMA (+) connection on the wiring harness, and the black lead to the IMA (-) connection on the wiring harness. The multitester will read the voltage in the circuit.

If the voltage on the meter in the charger doesn't match the multitester, use your teensy screwdriver to adjust the VR screw on the back of the meter until it matches your multitester as closely as possible.

Now power everything off and unplug it. At this point you can mount the meter in the grid charger case.

Ta-da! Your charger is working and adjusted on the bench.

8. INSTALLING THE HARNESS IN THE CAR

This step takes a while, but was relatively easy because of the RB Batteries harness I was using. They have a great YouTube video showing the process...
Pull out the carpet and the storage well to expose the metal shield covering the battery. Ignore the radioactive warning sticker installed by the current owner as a joke.



Remove the switch cover and turn off the battery. The switch may have a red plastic lock on it that you need to pull off, then reinstall after flipping the switch.

Remove the 25-ish (!!) 10mm bolts holding down the cover. Some are hidden under the lip of the carpet covering the Tom Mix bar. You also need to remove the little tray on the passenger side, which involves pulling loose the liner in front of the taillight and removing a couple more screws. There are three plastic rivets connecting the two metal trays which are more likely to break than release. You can get replacements at a decent auto parts store.



With the screws out and the cover slid off, it's a good time to use your multitester and make sure the battery is really turned off!



Pull the foam pieces off. Disconnect the two big orange battery cables. Feed the harness wires under the edge of the metal shield. I chose to do this on the passenger side where the gap is just big enough to get the ends of the wires through.



The IMA (+) connection goes to a spade terminal awkwardly located between the battery and the computer. Wiggle the existing spade off, plug in the harness wire, and then connect the original wire to the extra spade on the RBB harness connection. In the photo below, I had unscrewed the fuse holder to temporarily get the rest of the harness out of the way.



The IMA (-) wire bolts onto the top of the battery where the negative (black end) cable connects.



The RBB kit includes nice crimp-on connectors for the fan (+) and fan (-) wires. Be sure to connect fan (+) to the positive fan wire, probably blue.



Use the included zip ties to safely tuck wires out of the way of the fan and the cover. I wrapped black electrical tape around the harness where it went underneath the shield as an extra precaution.



Before you button everything back up, plug in your grid charger and power it up. The fan on the IMA battery should spin when you turn on the power, and the meter voltage will show the battery voltage and 0.000 amps. When you turn on the charging switch, the meter should show approximately 0.350 amps (350 milliamps) and the voltage will start to climb very slowly.



If it looks good, you can put all the covers on. It's also a good idea to test the discharge load off battery power. Remember that the grid charger will not read amps during the discharge.



Almost done...

9. FINAL TESTS AND OPERATION

Once I knew everything worked, I carefully installed the lid on the grid charger case.







The first day, I was only able to do an overnight grid charge. This is in a garage hovering somewhere between 40-55 degrees F. The voltage readings climbed as I expected based on IC posts about grid charging...

3:30p 157.7v 0.346a 52F
4:31p 163.9v 0.346a
5:33p 166.4v 0.346a
6:33p 167.6v 0.346a
7:33p 168.3v 0.346a
8:33p 168.8v 0.346a
9:29p 169.2v 0.346a
10:27p 169.5v 0.346a
11:33p 169.8v 0.346a
5:40a 172.0v 0.346a
6:50a 172.4v 0.346a

Just this overnight charge a week ago made a noticeable difference. I'm looking forward to trying the full three-cycle charge / discharge / charge process. Maybe this coronavirus shutdown will finally give me time. (grin)



In summary, thanks again to Ol' Rowdy for a very satisfying and useful project. He was kind enough to consult with me by email and help me through some of the bumps along the way. This design works well and is within the capability of a pretty modest tinkerer. Go for it!

- Park

I've been taking advantage of extended downtime right now to do a multi-cycle charge & deep discharge on my battery using the new grid charger and discharger. I'm live-logging the data to a Google spreadsheet which can be viewed here...


Here's a photo from early in the discharge cycle.



As the image above shows, I've rigged my multimeter into the discharger circuit to track discharge amps. I unhooked the wire connecting a terminal on each light bulb socket and connected the leads from my multitester so that the current can flow through the tester and be measured. The light switch is wired parallel to the tester so I can bypass the tester and turn it off between measurements.

The above photo is with two 60w bulbs, early in a discharge process. The voltmeter on the grid charger shows 166.1 and the current flow measured by the yellow multitester is 0.41 amps (410 milliamps). The multitester only has two-digit precision at the setting needed to handle the higher discharge current.

I'm using the three-cycle charge and discharge method described most recently by S Keith in this thread (combining a few posts)...

Grid charging, deep discharging

If your power supply is 350mA, you charge for 10,400/350 = 29.7 hr at the start of each phase.

Each discharge phase length will change on each cycle. Try to check it at least once per hour. If it's bed time, and it's over 120V, install the 25W and let it run overnight. Take what you get.

A) Grid charge for 10,400mAh input

• Discharge to 144V w/2X 150W wattage bulbs IN SERIES (goes pretty quick)
• Discharge to 120V w/75W bulbs IN SERIES (takes awhile)
• Discharge to 96V w/25W bulbs IN SERIES (takes a long time)

B) Grid charge for 10,400mAh input

• Discharge to 144V w/2X 150W wattage bulbs IN SERIES (takes longer)
• Discharge to 120V w/75W bulbs IN SERIES (notably quicker than 1st time)
• Discharge to 60V w/25W bulbs IN SERIES (notably quicker than 1st time)
• Overall takes less time than cycle A

C) Grid charge for 10,400mAh input

• Discharge to 144V w/2X 150W wattage bulbs IN SERIES
• Discharge to 120V w/75W bulbs IN SERIES
• Discharge to 60V w/25W bulbs IN SERIES
• Like cycle B but less pronounced a jump than what was experienced between cycles A and B. Overall will take a little less time than B.

D) Grid charge for 10,400mAh input

If you can monitor the discharge current and record time, voltage, current periodically, you can calculate the capacity of your pack. If you do this, make sure you record 144V, 132V and 120V readings.

Discharge time changes between cycles because you're "moving" the capacity you discharged at lower currents/voltages to higher currents/voltages.

Click to expand...

My variation from the above method is using lower wattage bulbs. This slows down the discharge rate. I can't check the process as often as I'd like, since I'm still working intermittently and also like to sleep at night. (grin)

I'm curious to know how current and voltage readings at 144V, 132V, and 120V can be used to calculate the pack capacity. I'll have to do some research on this. In the Google sheet, these approximate voltages are marked in bold.

Here's another photo late in a discharge cycle with the 25w bulbs...



Here the voltage reading, somewhat obscured by the refresh rate on the meter, is 117.4v and the current is 0.13a.

My Google spreadsheet (linked above) includes graphs of voltage and current over time for the charge and discharge cycles. Amps are multiplied by 100 in order to show meaningful data on the graph. Each charge or discharge graph includes the rest cycle at the end.

Some observations so far...

Ol' Rowdy's V2 grid charger does a great job on the charge cycles. The current remains very close to 0.345 amps (345 milliamps) while the voltage steadily increases. According to S Keith's 10,400mA calculation, my pack really should charge for 10,400/345=30.1 hours. I've been cutting it off somewhat shy of the mark.

I let the pack rest for an hour between cycles. After a charge cycle, the voltage drops a bit. After a discharge, the voltage rises a bit. Mainly the rest time is to allow the pack to cool off and stabilize.

On discharge cycles, I've been amazed how steady the current stays during each step. If you look at the amps on the discharge graph, you can see clear stair steps where I switch to lower watt bulbs, and otherwise the discharge current is almost steady. I expected it to taper more as the voltage bleeds down. Not so.

So far, things are working like a charm. Thanks again to Ol' Rowdy, S Keith, and the many other smart people on Insight Central who've pioneered and shared all this stuff.

- Park

Whoops. Discharge #2 turned into a deep discharge on accident.


I left the discharge process in the care of my brilliant grad student daughter. She dutifully logged data all day, and I showed her how to use the pull-chain switch on the discharge load to turn off the lights and change to 25W bulbs when the voltage hit 120V.

The battery lingered for a loonng time over or close to 140V, then tipped over a cliff. In one 85 minute interval, it dropped from 136.5 to 128.2V. In the next 76 minute interval it fell from 128.2 to 30.22V.

At that point my daughter called me for advice. I told her to turn off the light bulbs. Unfortunately, she turned off the grid charger instead of the pull-chain on the discharge load. So over the next 69 minutes the battery dropped from 30.22 to 8.4V, and did so without a cooling fan. Ouch. At least the discharge current was fairly low, and my garage is pretty cool.

So far the battery is recovering well on charge #3.

The good news is that it looks like my battery can hold a nice long steady discharge at a pretty consistent 140V+.

Onward...

- Park

olrowdy01 said:

Man that charger looked like a bowl of spaghetti before you buttoned it up.

Click to expand...

It looks worse in the final photos before I buttoned it up, for instance while being tested in the car, because I cut back into the wiring and rigged up a second Mean Well (tucked behind the charger) while I was diagnosing the circuit. So there are two Mean Well's worth of wires there, with one disconnected, and more wire nuts.

After the pictures, when I figured out the original Mean Well was okay, I cleaned things up quite a bit before closing the box. Still, a spaghetti bowl. I prefer to have a little extra wire instead of not enough... ?

olrowdy01 said:

What voltage did the battery recover to after the discharge to 8.4 volts?

Click to expand...

You're referring to the end of discharge #2. I'm not sure what the recovery voltage was, because my daughter was implementing my (apparently not very clear) directions by text message, and the only data I have is what she wrote down. She couldn't explain it to me very well, but here's what I think happened...


What she described doesn't make a lot of sense to me.

According to your description of the Mean Well on your website, it shouldn't start charging unless the battery is at a minimum of 100V. Maybe it automatically kicked in once the battery recovered to 100V+, sometime between 17:04 and 18:11?

It was definitely charging at 18:11 last night. As of 9:11 this morning it had pushed the battery up to 164.6V. The charge curve looks the same as the previous two charge cycles so far.

We'll see. I definitely appreciate the more detailed write-up of the charge and discharge cycles you added to your website.

- Park

Finished discharge #3, and was able to monitor the discharge myself and more closely. Actually, I did two discharges on this cycle.


The first, long discharge (#3a) took the battery from 169.1V to 60V over about 22 hours. As before, there was a very long slow discharge until the battery dropped below about 138 volts, and then the curve rounded into a steep descent. At this point I began checking every half hour, then every fifteen minutes. I let it drop to 60V.

With the load disconnected, the battery quickly rebounded to to over 120V in about fifteen minutes. After resting overnight the voltage was 137.3V.

I wanted to see how long it would take the battery to reach 20V, so I started another discharge (#3b) this afternoon using the 25w/25w load, monitoring every five minutes. It only took 1.5 hours to bring the voltage down from 137.3 to 20V. With an hour's rest, the voltage bounced back to 131.3V.

olrowdy01 said:

The power supply will not charge the battery if it is below 100 volts.

Click to expand...

While the battery was recovering from the deep discharge and still well below 100V, I tried turning on the charging circuit. As advertised, it did not try to charge the battery.

Now I am on to the last charge cycle. 30 hours to go, and the car will be back on the road. We'll see how she runs then.

Learning points so far...

(A) The 3-cycle charge and discharge battery conditioning takes a LONG time.

(B) The battery discharge requires careful monitoring once the voltage drops below 140.

(C) The battery voltage bounces back quickly from a deep discharge, but the battery will quickly drop again under load.

None of this is news to old-timers, but it's interesting for me to experience for myself.

- Park

Final update...

I finished my full three-cycle charge and discharge battery conditioning. It took about eight days from start to finish. You can see the graph of volts and amps during each cycle in links elsewhere in this thread.

Here's a photo from near the bottom of the final discharge cycle...



The discharge is down to 41V and 0.08 amps (80 milliamps). The dual 25W bulbs are barely glowing. At 20V the lamps were not visible illuminated at all.

After the third charge and discharge cycle, I did one final grid charge for 29 hours. The voltage peaked at 173.4V, and after resting overnight the battery read 165.4V.



--------
The point of this of course is to rejuvenate the battery. So how does it drive? WAY better.

Before doing this reconditioning cycle I could drive around town without generally tripping the IMA light, if I was careful. I was still seeing a lot of recalibration and positive charging on the dash. If I got on the interstate or climbed any kind of hill, the IMA would trip out.

After the reconditioning, in city driving the car is flawless. Good boost, good recharge, no recalibration. In fact, the battery meter barely moves.

To really test the battery, I headed east out of La Crosse on State Road 16, which climbs steeply for 1.3 miles from the Mississippi River up to the top of the bluffs. I was able to mostly maintain 55mph in fourth gear with the boost almost pegged. I shifted down to third gear as the system suggested, and with boost was easily able to accelerate up the remainder of the hill. All this fun only depleted the battery gauge about 20%, and descending the hill with regenerative braking topped off the battery again.



I haven't had a reason to get out on the interstate yet, but I expect it will be fine on the big road as well.

Overall, I'm very satisfied with Ol' Rowdy's V2 grid charger and discharger. For less than $100, I've been able to refresh my battery to a very usable state. The car really is much nicer to drive with a solid IMA!

Based on previous experience, I hope to get about six months of service before I need to do the reconditioning cycle again.

I hope this thread encourages you to build your own grid charger/discharger. It is a fun and satisfying project. Feel free to ask any questions. And, one more shout out of thanks! to Ol' Rowdy for the design and assistance.

Blessings,

Park

One last tweak to the discharger...

Ol' Rowdy's design doesn't read the amps on the discharge cycles - this is a quirk of the digital multimeter used in the design. I rigged my digital multimeter into the circuit to see the amperage data when discharging. Now I've found an old fashioned mechanical ammeter to give me that info.

MparkH said:

DISCHARGE LOAD
$2.39 ... Porcelain lamp holder w/ switch (Menards #3634101)
$1.29 ... Porcelain lamp holder (Menards #3634091)
$4.99 ... Orange extension cord 10’ (Menards #3700433)
$1.99 ... 25W incandescent bulbs, 2-pack (Menards #3533160)
$1.99 ... 40W incandescent bulbs, 2-pack (Menards #3533161)
$4.99 ... 60W incandescent bug lights, 2-pack (Ace Hardware)

Click to expand...

Add to that list...

$1.97 ... 85C1 Ammeter DC Analog Mechanical Pointer Type 500mA (AliExpress, ~$5 with shipping)
LINK: US $1.83 17% OFF|85C1 Ammeter DC Analog Current Meter Panel Mechanical Pointer Type 1/2/3/30/50/100A 50/100/200/500mA|Current Meters| | - AliExpress

TOTAL COST OF DISCHARGE LOAD $19.61

Be sure to choose the 500mA meter out of the several listed. In my testing, I never saw the discharge load go much over 400mA.

---
The meter is plastic but seems to be sturdy. Before installing the ammeter, I wired it into my discharge load to check that it would read on the positive side, and marked the terminals on the back with the wire colors.



The trickiest part was drilling the holes and cutting the notches in my discharge load's board to insert the meter.



I wired the ammeter, checking the color codes I'd marked earlier. The ammeter is in series with the light bulbs, so one wire attaches to a screw terminal on one socket, and the other attaches to a screw terminal on the second socket.



Now when the discharge load is connected to the charger, I can read the volts off the digital meter on the charger, and the amps off the analog meter on the load itself.



With two 60 watt light bulbs and a fully charged battery, the ammeter reads around 400 milliamps.



I didn't check the meter against my digital multimeter, but 400 milliamps is about right based on my full charge/discharge cycles I performed a few weeks ago.

The car is still running great, by the way. The IMA works like a charm. This is a great project for anyone who wants to do some home maintenance on their Insight's battery.

- Park