Using the Universal grid charger

[Calpod]

Mike, I noticed the voltage tops out an 168 volts for the Gen 1 Insight, I thought it would be higher?

My pack took only 2.5 hours to charge until the charger stopped, it had 0 bars when I removed the battery after sitting there for 3 weeks.

[aisbell]

Originally Posted by Mike Dabrowski 2000

The software is Version 1.2. we of course tried to think of everything and anticipate everything, but the real world has a way of coming up with the things we did not think of no mater how hard we try.

Originally Posted by Mike Dabrowski 2000

We are already working on Version 1.3, which will have easier more natural navigation of the setup and other menus, and a couple of very minor bug fixes.

For those of us with PIC programmers, is it possible for us to update the code as new versions are released?

[Eli]

I also believe that the max charge voltage is a little too conservative out of the box, especially with the cold temps most of the country is experiencing right now, but like Mike says that's easy to change. I just discharged my spare pack through a 300W lightbulb and am now recharging it. I didn't have a way to measure the total amp hours out, but it was around 4+Ah (2A+ for ~2 hours).

Note that it isn't recommended to use this method of discharge on a pack of questionable or unknown status, as it risks cell reversal. This is my 2 year old, healthy, balanced stock pack.

The charger only inputted 4.1Ah back into the pack. The pack was deeply discharged, so this doesn't represent full. I'm going to muck with the settings and try again.

Grid Charge 1 Battery Voltage

Grid Charge 1 Input mAh

Grid Charge 1 Temps

An interesting note is that with this 2 year old stock pack, even under 1A of charge, the voltage taps were all within 0.08V of each other at any point that I measured them. This is in contrast with the weak old pack that I've been testing in car #3, which shows a 0.26V spread under a mere 0.35A of charge.

[Eli]

Doing discharge #2 so I can test the new grid charger settings. Used two 400W work lights drawing about 4.5A to speed up the process while manually monitoring the voltage taps about once a minute or so.

Once the pack hit ~132V, voltage started falling faster and I began monitoring which I knew was the weakest tap. To my horror, it fell from 12.4V to 9.0V in a matter of about 10 seconds, while total pack voltage was still ~128V! I quickly flipped the switch to drop the load. I now discourage this method of discharging your pack even more, unless you go with a "safe" voltage of something like ~144V total or are monitoring the voltage taps.

After dropping the load, the voltage on the weak tap climbed quickly back up to ~13.6V. However, all the rest of the taps climbed back up to 13.9-14.1V, perhaps showing some sight imbalance in them as well.

I then switched to a small 0.35A load. Continuing to monitor the strongest and weakest tap, I let this load go for another 30 minutes, until the weakest tap started dropping faster than 0.01V/second, which happened at about 12.0V. This happened while the strongest tap was at 14.07V, a difference of a whopping 2.07V. Perhaps this pack isn't as balanced as I once thought.

Total discharge approximately 3.8Ah. This makes sense, due to our ~4.1Ah input on the previous charge.

I have modified the settings to 7200mAh max input, 174V Cut-backVPx, 45 minute Sample Time and enabled an Extended Soak of 60 minutes. We'll see how much is inputted. It should be somewhere around 7,000mAh to ensure that the cells in this weakest, fully discharged tap are fully charged, and to account for charging inefficiencies. When I tested my first spare pack, most sticks took ~7000mAh of input @ 5.5A to output ~6400mAh @ 5.5A. I am not sure if upping the Sample Time was necessary or warranted, but it will be interesting.

WARNING: Only change your grid charger settings if you understand what you're doing.

After inputting only ~450mAh, the tap that was the weakest now shows the highest voltage. Hmmm... I smell a weak cell in my 2 year old pack.

[Mike Dabrowski 2000]

Excerpt from the operating instructions:
http://99mpg.com/Data/resources/down...ctionsv1.2.pdf

There are many ways or Reasons for the charger to stop charging, in order to automate and optimize the
process of charging, and protect the battery. The reason the charger stops charging is displayed briefly
as it happens on the LCD, sent out on the serial port with the regular status reports, and kept and
displayed as the "R" indication on the "Not Charging" flash message. This message is shown one
second out of 6 seconds. (Reason indicated as R 1, or R 2, .... etc.) There are fourteen of these reason
codes at present, which advise the reason why the Grid Charger last stopped charging:
R1: Manual Stop, STOP key was pressed.
R2: The maximum allowed Voltage has been reached. (ChgStopVP1*, ChgStopVP2*, ChgStopVP3*,
depending on the CarProfile selected)
R3: The maximum allowed total charging time has occurred. (Max CGTime*)
R4: The timed end of topping mode occurred. (TopTimeMN*)
R5: Battery now too hot to continue. (MaxPTCVal*)
R6: The TEST button was pressed.
R7: The maximum allowed milliAmp-Hours charge was supplied to the battery (Max_chg_AH*)
R8: End of Topping charge due to Voltage plateau detected.
R9: Temperature Differential (Outlet - Inlet Temperature) target reached, meaning the charging energy
received by the battery is now going into heat rather than raising the state of charge.
R10: Higher current than expected through power supplies
R11: Too HOT Outlet Temperature
R12: Too COLD Inlet Temperature
R13: Too HOT Inlet Temperature
R14: Too Cold Outlet Temperature

Depending on the reason for the charge termination, we can make some adjustments.
R8 is the standard plateau was detected reason, which is adjusted by the sample time. The default is set at 30 minutes, but it can be raised to 45 or even higher. This is the number of consecutive minutes that the voltage remains within one raw count of the first sample.

If the charge stopped due to temperature difference in to out, we can adjust the raw temperature counts allowed between inlet to outlet with the TDifMaxRaw variable. the default is set up for ~ 7 degrees f. 35 raw counts /(5counts / degreeF)

I advise using the soak time on setting (setup menu) for at least one charge, which will allow the charge to continue for 9 hours or the TopTimeMN variable set at 540 minutes.
This gives any cells that have fell way behind to fully top off. The soak time flag will reset to off the next time the charger is powered up.

The chargers can be programmed in the field with a programmer. I use the microchip IDC3, but expect that most pic programmers will do the job. Jason has sent me an ICP01-V1 programmer from Complete Microchip PIC Solution - PICcircuit.com, which I will try out when I get a breather.

I would also advise that you guys data log a charge so you can see what is happening. The voltage graph and ability to see the sample timer count up will give a very clear picture of what is happening at the critical end of charge detection zone.

I have seen packs top off at as low as 165, and as high as 174, so a lot depends on the condition of the pack, and the ambient temperature.

[Eli]

It's interesting in the logging file from my first discharge,

The Voltage Plateau was reached after the 3rd Sample Time, even though the voltage rose from 165.5V to 165.8V 16 minutes into the sample?

BattV  ChgCur  SysSt   StopR  ChgMode   Time    mAHrs	InT	OutT  TScntr

165.5	353	193	0	9	273	3926	64.1	64.9	1
165.5	353	193	0	9	274	3932	64.1	64.7	2
165.5	353	193	0	9	275	3938	63.9	64.7	3
165.5	353	193	0	9	276	3943	64.1	64.7	4
165.5	353	193	0	9	277	3949	64.1	64.7	5
165.5	353	193	0	9	278	3955	63.9	64.9	6
165.5	354	193	0	9	279	3961	64.1	64.9	7
165.5	354	193	0	9	280	3967	64.1	64.9	8
165.5	354	193	0	9	281	3973	64.1	64.9	9
165.5	353	193	0	9	282	3979	64.1	64.9	10
165.5	353	193	0	9	283	3985	64.1	64.9	11
165.5	353	193	0	9	284	3991	64.1	64.9	12
165.5	354	193	0	9	285	3996	64.1	64.9	13
165.5	353	193	0	9	286	4002	64.1	64.9	14
165.5	353	193	0	9	287	4008	64.1	64.9	15
165.8	353	193	0	9	288	4014	64.3	64.9	16
165.8	353	193	0	9	289	4020	64.3	64.9	17
165.8	353	193	0	9	290	4026	64.1	64.9	18
165.8	353	193	0	9	291	4032	64.1	64.9	19
165.8	353	193	0	9	292	4038	64.1	64.9	20
165.8	354	193	0	9	293	4044	64.3	65	21
165.8	353	193	0	9	294	4049	64.1	64.9	22
165.8	354	193	0	9	295	4055	64.1	64.9	23
165.8	353	193	0	9	296	4061	64.1	64.9	24
165.8	353	193	0	9	297	4067	64.1	64.9	25
165.8	353	193	0	9	298	4073	64.1	64.9	26
165.8	353	193	0	9	299	4079	64.1	64.9	27
165.8	353	193	0	9	300	4085	64.3	64.9	28
165.8	354	193	0	9	301	4091	64.3	65	29
165.8	353	193	0	9	302	4096	64.3	65	30
165.8	353	1	8	0	302	4096	64.3	65	0

[Calpod]

Eli, were you able to get 4096 mAH into the battery? I just did another charge after taking the car out and depleting the battery and after 4 hours charging I was only able to get 2249 mAH before it finished charging, I extended the soak time to 45 minutes. final voltage was 169.4

[Eli]

Yes, but the battery was drained on the bench, not in the car, so it isn't directly comparable. Remember the car has a 60% charge/discharge window(80% to 20%).

Interesting. Looks like I need to study the settings more. The reason the charge stopped this time was R9 - Temperature Differential reached? That's a little odd. The temp difference was 7 degrees. I guess that's quite a lot. I think the pack is much closer to actually topped off this time though.

Also, changing the Sample Time didn't seem to do anything, but maybe that's because I had Extra Soak enabled? I expected it to count to 45 minutes for each sample in the log, but it counted all the way to 111 minutes. Still need more fiddling to fully understand things I guess.

Bench Charge 2 Battery Voltage

Bench Charge 2 mAh Input

Bench Charge 2 Temps

Bench Charge 2 Spreadsheet

[Mike Dabrowski 2000]

The soak disables a shutdown due to plateau being reached,so the sample time means nothing, and the charge should run for the full (TopTimeMN* )540 minutes after the topping zone has been entered (Topping zone is entered when the voltage is > 166V the cutback V)

The plateau determination will consider a +1 increase in the smoothed raw data as no change,and since the system voltage resolution is about .2V per count in the raw data from the A/D, 165.5 and 165.8 are within the window.
The topping code that is executed when the sample timer times out:
TopDif = (Bat_VoltRawAve - BatSnapRaw) ; //TopDif = average value with a snapshot of the present value
if (TopDif < -1 || TopDif > 1 ) break;// Is the difference greater than 1 in either direction?
If it is within 1 in either direction, we consider it unchanged and stop the charge.

Calpod, What was the reason for the charge stopping,R? and did you modify the TopTimeMN variable to reduce the soak time?

Another important thing to remember, is that the SOC gauge may or may not register that the charge happened, so if the SOC did not do a positive SOC, the gauge hitting bottom may not have fully discharged the pack, therefore the smaller number of AH going into Calpods pack. it seems more likely to do a positive recalibration if the SOC was low, than if it was midscale or higher.
The SOC reset device I am testing should help to get us to the 19 bar SOC after a charge so the uncertainty is reduced.

R9 - Temperature Differential reached
The TDifMaxRaw variable is set to 35 raw temp units which translates to~ 7 degrees. That may be too close, and was based solely on the test charging of my pack with the TDifMaxRaw set to 100 (20 degrees) so it would never stop for that, and the soak was turned on, and the value set at 1440 minutes or 24 hours.
It charged for 20 hours, and at the end it was still running when I looked at the TDif and saw 7 degrees.
If the charge terminates for that, one can bump it up a degree or so.
Each degree is 5 raw units, so a 40 would get us up to 8 degrees, 45 to 9, and 50 to 10 degrees.
The fact that you saw 7 degrees, does mean that the pack had pretty much went into the zone where most of the energy was making heat instead of charging, so in reality one could say it was balanced.

[Varak]

What additional hardware is required to do the data logging that Eli is doing? I'm assuming it is from the RS232 port. Is that only available on the overnight charger, though? (appears that way from the pictures).

Also, might want to make this thread a sticky.