Battery Replacement - Checking CAC

Did they pass on the ESS will take over a month properly balance and accordingly adjust the CAC? If not please let them know to do so in the future so others don't get worried. Just give it some time. These packs are stored for a long duration around 30-50% SOC, so that throws them out of balance as well as that they are typically mixed with good sheets pulled from returned packs.

Doing the CAC calibration up front won't get you much. It needs to first balance out, if the pack is not balanced it won't calibrate anything! My refurbished pack took over 2 months to fully balance and continued to climb for almost a year.

The refurbished pack has a 1 year warranty, so that should give you some confidence. After 2 months have passed with regular driving the CAC / range should be as good or better than the pack it replaced when it was working correctly.

I agree. My refurbed battery took about a month before it got back to where the previous one had been before the sheet failure. Hang in there.

Turn the parking lights on overnight. The warning chime will keep the car awake. The 1.5 only balances while awake. The 2.0 and 2.5 seem to balance at a much lower rate when not awake and the coolant is flowing. SOC needs to be at least 79% for the highest brick. The roadster balances the battery by discharging the bricks with the higher voltages, until the difference between high and low bricks is < .03 volts. So if you have one low brick, the other 98 have to be discharged until they all are within .03V. Not the best way to do things, but that's the way it is. It takes awhile. You can read the min and max voltages in the service menu (ESS/V&T).

MLAUTO said:

Turn the parking lights on overnight. The warning chime will keep the car awake. The 1.5 only balances while awake. The 2.0 and 2.5 seem to balance at a much lower rate when not awake and the coolant is flowing. SOC needs to be at least 79% for the highest brick. The roadster balances the battery by discharging the bricks with the higher voltages, until the difference between high and low bricks is < .03 volts. So if you have one low brick, the other 98 have to be discharged until they all are within .03V. Not the best way to do things, but that's the way it is. It takes awhile. You can read the min and max voltages in the service menu (ESS/V&T).

Click to expand...

NO!!! The 1.5 (nor any other Tesla Roadster version) does not balance when *only* awake, where did you come up with that non-sense? If that was true, nobody would EVER have a balanced pack.

Also as soon as you turn the key to the on position, balancing stops. It balances when its 82% or above SOC which is basically a little under a completed std. mode charge without a key turn, once that key turns all bets are off and balancing ceases until the next charge complets. Just keep it plugged in in Std. Mode and it'll balance out over time. If you want a faster balance, do a full range mode charge and leave it there for a few days, then repeat. But I don't like my ESS sitting at a high SOC for long periods of time.

Here's a good reading discussing passive (as well as active) balancing which the Roadster and Model S use. The BMB (battery management board) is responsible for this task:

Battery Management Considerations for Large Cell Count Systems | Battery Power Magazine

wiztecy said:

NO!!! The 1.5 (nor any other Tesla Roadster version) does not balance when *only* awake, where did you come up with that non-sense? If that was true, nobody would EVER have a balanced pack.

Also as soon as you turn the key to the on position, balancing stops. It balances when its 82% or above SOC which is basically a little under a completed std. mode charge without a key turn, once that key turns all bets are off and balancing ceases until the next charge complets. Just keep it plugged in in Std. Mode and it'll balance out over time. If you want a faster balance, do a full range mode charge and leave it there for a few days, then repeat. But I don't like my ESS sitting at a high SOC for long periods of time.

Here's a good reading discussing passive (as well as active) balancing which the Roadster and Model S use. The BMB (battery management board) is responsible for this task:

Battery Management Considerations for Large Cell Count Systems | Battery Power Magazine

Click to expand...

My 1.5 sat for 2 weeks with an out of balance pack (.06v between min and max) and the voltage difference never decreased. Every day I would have to wake it up to get back into the service menu and it would always show the balancing as active. After two weeks I lowered the ess and checked the balancing resistor inside and it was room temp. After I left the parking lights on for three days the pack was within .03v and the resistor got warm during that process.
When unplugged, the balancing stopped when the max brick SOC reached 78%.
The pack will only get way out of balance if it is left unplugged for any length of time, as the daily wake up call is enough to keep it balanced. (unless of course you have a weak cell or do some sheet swaps).
These are my observations, but since I can only get data when the car is awake, it is hard to tell exactly what goes on when asleep.

Like the others, I had the same thing happen when I had my battery replaced.

My service tech told me to range charge the car at let it sit at full charge for a long time until it balanced. I used the service display to see when it was balancing, and it took about a week sitting at full range charge. After that, it's been great.

And I know people will object to leaving the battery full for that long, but it was only once, it was nearly five years ago, and my battery is one of the outliers (on the good side) of the mileage/range curve. I've got 80K miles and it charges to 170-172 miles standard range. So it probably didn't do too much damage.

MLAUTO said:

My 1.5 sat for 2 weeks with an out of balance pack (.06v between min and max) and the voltage difference never decreased. Every day I would have to wake it up to get back into the service menu and it would always show the balancing as active. After two weeks I lowered the ess and checked the balancing resistor inside and it was room temp. After I left the parking lights on for three days the pack was within .03v and the resistor got warm during that process.
When unplugged, the balancing stopped when the max brick SOC reached 78%.
The pack will only get way out of balance if it is left unplugged for any length of time, as the daily wake up call is enough to keep it balanced. (unless of course you have a weak cell or do some sheet swaps).
These are my observations, but since I can only get data when the car is awake, it is hard to tell exactly what goes on when asleep.

Click to expand...

This was either a coincidence or something else stopped the normal balancing process. 2 weeks sounds too quick to see if the pack has fully been balanced honestly. Right when I thought my pack was close to being balanced (after 2 - 3 weeks) where balancing grid turned for the most part back all to 0's with about four to five ones, balancing started all over again where the majority of the grid turned back to all 1's. Also the balancing is based upon the SOC LIM and MIN (which are really the same) not MAX. How did you connect the ESS to the Roadster if it was pulled out?

I haven't used the min/max/avg voltages to gauge the balancing, only the grid. If those are live voltages, which they should be since you can see the bricks rotate out, it would be a way to see how out of balance the sheets are like you said.

I can see what your saying about the Roadster needing to be awake to balance, but you're always waking it when you view the grid / voltages, so its a little like the guy who turns on the light inside your refrigerator. I've had 3 ESS's (orig, 1st refurb & 2nd refurb that replaced that), did the range mode balancing on the 1st refurb, and std. mode balancing on the 2nd refurb carefully watching the balancing behavior. Each time you can tally up the status, and just having the Roadster sit while sleeping I've always seen progress. Sometimes slow, sometimes fast. I found its all dependent on how many sheets and how far out of balance they are.

Possibly you found some hack to speed up the balancing? But I don't see how or why it would using that technique nor have I seen Tesla ever mention it internally or externally.

As for the balancing SOC trigger, I've never seen the balancing grid initiate for a balance at 78% SOC LIM/MIN. I wanted to find where this cut-off was so I could unplug while in the middle of a std. mode charge, but still have the benefit of the pack being balanced. What I found was the threshold to be 82% for SOC LIM/MIN that created the active grid of 0's populated with 1's. Others have mentioned 80% or above, and I never seen it kick on with 80% either and think that was just slang.

wiztecy said:

This was either a coincidence or something else stopped the normal balancing process. 2 weeks sounds too quick to see if the pack has fully been balanced honestly. Right when I thought my pack was close to being balanced (after 2 - 3 weeks) where balancing grid turned for the most part back all to 0's with about four to five ones, balancing started all over again where the majority of the grid turned back to all 1's. Also the balancing is based upon the SOC LIM and MIN (which are really the same) not MAX. How did you connect the ESS to the Roadster if it was pulled out?

I haven't used the min/max/avg voltages to gauge the balancing, only the grid. If those are live voltages, which they should be since you can see the bricks rotate out, it would be a way to see how out of balance the sheets are like you said.

I can see what your saying about the Roadster needing to be awake to balance, but you're always waking it when you view the grid / voltages, so its a little like the guy who turns on the light inside your refrigerator. I've had 3 ESS's (orig, 1st refurb & 2nd refurb that replaced that), did the range mode balancing on the 1st refurb, and std. mode balancing on the 2nd refurb carefully watching the balancing behavior. Each time you can tally up the status, and just having the Roadster sit while sleeping I've always seen progress. Sometimes slow, sometimes fast. I found its all dependent on how many sheets and how far out of balance they are.

Possibly you found some hack to speed up the balancing? But I don't see how or why it would using that technique nor have I seen Tesla ever mention it internally or externally.

As for the balancing SOC trigger, I've never seen the balancing grid initiate for a balance at 78% SOC LIM/MIN. I wanted to find where this cut-off was so I could unplug while in the middle of a std. mode charge, but still have the benefit of the pack being balanced. What I found was the threshold to be 82% for SOC LIM/MIN that created the active grid of 0's populated with 1's. Others have mentioned 80% or above, and I never seen it kick on with 80% either and think that was just slang.

Click to expand...

I have found that the easiest way to determine the balance is to look at the min and max brick voltages. I have seen the back balance with .03v difference, but never with .02V. The process seems fairly simple-just bleed off excess charge on the upper voltage bricks until the required difference is reached. It doesn't matter if it is plugged in or not, balancing will continue as needed until the max voltage brick SOC reaches 78%. The higher voltage bricks are the ones being reduced, so that appears to be what determines when to stop. The lower voltage bricks stay at the same number (unless you get some initial drop after charging). Charging in range mode gives more voltage to the lower bricks, and if they stay above 78% SOC the balancing should continue until complete. I'm not sure if there is a higher SOC required to start balancing than 79%, I just know it shuts off at that point.

Judging by how much heat the bleed resistor creates (especially if you need to bleed down 98 bricks due to one low one) I find it unlikely Tesla would allow that to happen without coolant flowing though the almost sealed ESS. When I had one low brick I did see 98 1's and one zero, so the current from each brick has to be very small, The wires coming from each brick to the BMB any tiny (20 ga maybe?). The bricks are numbered on the balancing screen from top to bottom and then left to right (1-9 first vertical column, 10-18 for the 2nd, etc.), so if you look at the lowest voltage brick, it will always have a zero. I had one brick in one ESS that would self discharge faster than the balancing could keep up, I finally had to drop the pack when the min SOC dropped below 30%, while the max SOC was 83%. It was still trying to balance.

My test vehicle has no body panels at all, even the rear bulkhead is out , so the battery is accessible while in the car (with the front cover off). It sits on a cart just plugged in and not bolted to the car. Takes about 10 minutes to disconnect and raise the car up when I need to access the sheets.

Here is a pic of the bleed resistor. Its sits about an inch away from bricks #3 and #4. And yes, that is a smoke detector (only on the 1.5).

supersnoop said:

... Am I to assume the pack still is not really balanced?

Click to expand...

I would just use thumb drive and grab logs to a USB Flash drive i.e. --> Kevin Sharpe's decreased Roadster range - Page 36

Then open the log file in excel, see thread here --> ahr.log file - seems useful for keeping an eye on your brick/sheets health

Look at Voltage column for all 99 bricks, divide it by 8192, then compare it's voltages with each other to see if they are in balance or not.

At .03V it will balance, at .02v it will be done. Your first balancing screen shows brick #2 being discharged, which it shouldn't. It seems to jump around a lot right after charging, maybe some kind of self test. In your case brick #10 was reduced to 4.06 and balancing stopped. Because there are 99 bricks, the min and max brick numbers (not voltage) will change if the voltage of different bricks are the same. What you want to keep an eye on are the max and min brick numbers before and after charging. I weak brick will tend to always be the min one when the battery is low, and if it is the max one when charged, then you have a problem.

Here's the CAC history of our car, Roadster VIN 217. The top graph is the CAC, the bottom graph is miles driven each day.



We had three sheets fail the bleed test in September, 2010. Our battery pack was replaced on November 15, 2010, at the bottom of the big drop to 146.5 Ah. We were told to do a full range charge and then let it sit until all of the cells dropped to 4.0V. I did a range mode charge when we got it back and let it sit. On December 3, I did another range charge and took it on a long drive (156 miles over a 4,000-ft pass, in 23-degree weather, with a charge from 29% to 88% in the middle), then did another range mode charge and let it set for another 24 days. As you can see from the chart, the CAC continued to rise until July when it peaked at 158.1 Ah.

Verified today that the ESS initiates balancing at the following threashold:

SOC:
LIM=82% MIN=81% MAX=83%

Ideal miles in std. mode: 171
Note that ideal miles will vary some per Roadster

I stopped charging LIM at 78%, 80%, 81% and finally at 82% LIM. Balancing only triggered at 82% LIM.