Honda Insight Forum banner

Grid Charging in Q3 2016?

21K views 69 replies 17 participants last post by  S Keith 
#1 · (Edited)
I am seeing a lot of new commentary on Grid Charging that seems to counter methods used over the years. This makes for a very confusing mix.

It seems that the initial process stemmed from the R/C crowd's experiments and apparatus for small battery packs in radio controlled cars and planes.

Can we use this thread to work out what the most optimal current procedure should be for grid charging and discharging of Gen 1 Insights at this point in time?


Some suggested parameters & topics:

- Consideration for battery pack's age & condition
- Grid Charge Current
- Grid Charge Start Voltage
- Grid Charge End Voltage
- Grid Charge Duration
- Grid Charge Frequency (schedule)

- In-Car discharging pack (by driving) before grid charging
- Preparing pack for load discharging
- Discharge Load : Incandescent Bulbs vs. Resistors
- Discharge Current
- Discharge Start Voltage
- Discharge End Voltage
- Discharge Rate
- Discharge Frequency (schedule)

- Natural Discharge (unattended packs for weeks/months)

- Stepped Discharge/Recharge process

- Pack Cooling

- Compare & Contrast Stick Level charge/discharge vs. Pack charge/discharge

- Commentary on climate, ambient temperature and pack life

 

Attachments

See less See more
2
#2 ·
Just my 0.02$...

Grid charge frequency: once a year ore more often when your battery is weak and gives IMA light or recals frequently.

Grid charge end voltage: does not matter. What matters is: when voltage has not changed for 1-2 hours while still charging or when it even drops a bit, keep it going for 2-4 hours or so.

About discharging pack (by driving) before grid charging: it is even better to let the car do auto stop with headlights on and fan on (A/C off). The only problem: when the HV battery is empty from the car's point of view, the 12 V battery feeds it all. I have not checked that exactly, but it was no longer than 1,5 hours of auto stop on my last (and first) discharge session.

This was followed by single bulb discharging. Stage 1: 100 W/230 V down to 135 V or so. Stage 2: 18 W/230 V down to 70 V.

YMMV, but I thing that this helped in keeping my battery healthy. Besides, my pack is in good condition. I have never seen IMA codes.
 
#6 · (Edited by Moderator)
I think the big issue is there is no wiki or faq, so instead one must sift through 35 pages of opinions and discussion. It's hard to distill actual answers. Even if there were 3-4 differing opinions, it'd be nice to have them in one spot. I'm not knowledgable enough about them to compile it.
 
#7 ·
I agree. Unfortunately there are multiple projects with huge amounts of discussion, i.e. OBDIIC&C with 247 pages or IMAC&C aka IMA Control with 47 pages. I don't know what to do about such threads because to distill them down to pertinent info and instructions takes a great deal of time. I'm going to IF to find out the answers.
 
#8 · (Edited)
I know my answer was a little dickish... at least I felt that way when I typed it due to frustration at the idea that anyone can give definitive information on the vast majority of it. There are too many variables.

TL;DR
Buy a Hybrid Automotive grid charger and follow Jeff's directions

https://hybridautomotive.com/pages/recon

Long version:

Here's my asshole, erm... opinion (YMMV):

Charging:


  • The WHY of charging: Grid charging is the single most important aspect and essentially the ONLY way to actually "balance" a pack. By "balance", I'm referring to each and every single CELL in the pack being at the SAME state of charge (SoC). When you've fully balance charged your pack, all cells will be at or so very near 100% SoC.
  • Preventative maintenance on a healthy pack should only be done every 3 to 12 months depending on circumstances. If the car sits a lot or is highly cycled due to hilly terrain - 3 months. If the car is in a mild climate and driven normally without exceptional hardships - 12 months. Extreme environments (Phoenix): once in April to prepare for summer, once in October to recover from the summer.
  • Grid charging as a preventative maintenance item on a pack that has thrown codes or is regularly recalibrating should only be done frequently enough to restore acceptable performance, e.g., the first recal witnessed 30 days after the last grid charge, two recalls within a week, one recal per day, etc.
  • Force charging within the car's normal operational range is an acceptable option to reduce grid charge time.
  • On a pack in an operational state (presumed 60-80% SoC), charge to peak voltage for 2-4 hours, 10 hr max @ 500mA max.
  • On a healthy pack from a discharged state, 10,400mAh input limit at 500mA max (based on 0.1C charge for 16 hours)
  • On a deteriorated pack from a discharged state, 8,450 mAh input limit at 500mA max (based on 30% charge inefficiency on a deteriorated pack).
  • Periodically recording total voltage as well as tap voltages is a very useful tool in determining when the pack is full. Constant voltage for 2-4 hours has you close enough to 100% SoC.
  • NEVER EVER EVER NEVER EVER NEVER grid charge without cooling. With a 350mA charge and no cooling, I've seen pack temps at 130°F when ambient was 37°F. Note that if you can monitor temps, cooling is not necessary provided no cell exceeds 110°F.
  • Precaution: extended charging of NiMH at low currents induces voltage depression with subsequent capacity loss. The practice of grid charging eventually necessitates the need for a deep discharge.

Discharging:


  • The WHY of discharging: Is done to eliminate "memory" associated with voltage depression. Voltage depression: Every time a NiMH cell is charged without being FULLY discharged, the phase of the terminal is altered. This altered phase produces current at 0.8V (depressed) instead of the nominal 1.2V. Deep discharging "consumes" the capacity stored at 0.8V. When the cell is recharged following a deep discharge, the proper phase is restored at the terminal and that capacity is again available at 1.2V. It's not uncommon for a deep discharge to restore > 25% capacity.
  • Should be done only when grid charging does not produce the desired results.
  • Current should be monitored ($5 harbor freight multimeter in 10A current mode in series with the load).
  • Time, Voltage and Current should be periodically recorded during a discharge (every 5 minutes for the first 30 and every 30 minutes thereafter).
  • The above record of time and current can be used to compute actual capacity. This can be compared between discharges to assess pack deterioration.
  • Tap voltage snapshots at 144V and 120V while UNDER 25W load can give insight into stick health.
  • High current (>25W) discharging should be limited to >1.2V/cell.
    Below 1.2V/cell, discharging should be limited to about 200mA (25W).
  • Capacity extracted below 144V should be limited to 1000mAh (about 5 hours on a 25W bulb).
  • Initial discharge target voltage should be 0.8V/cell, but the 1000mAh limit is more important. When capacity is computed, one can actually see the improved capacity above 144V on subsequent discharges. If there is a 20-25% improvement after the first cycle, subsequent cycles are likely not needed. If a second cycle only produces <10% improvement, a 3rd is likely unnecessary. If the computed capacity of a discharge to 144V @ > 25W is > 80% of rated, subsequent discharges are unnecessary.
  • Bulbs or resistors may be used. It doesn't matter which provided the voltage and current limits are respected. Bulbs behave a little more like constant current loads and resistors taper current linearly with voltage drop. Resistors will take longer.
  • Discharging in-car within the car's normal operating range is an acceptable option for reducing discharge time; however without current monitoring, the ability to compute actual capacity is lost.
"natural" discharging:
This encompasses MANY months and even a year or more of sitting. NOTE: A truly healthy pack WILL NOT DISCHARGE much BELOW 144V OVER MANY MONTHS. Yes, for deteriorated cells, extended self-discharge is very effective at eliminating voltage depression and restoring the capacity of cells/sticks/packs to their full potential, but a pack that self-discharges below 1.2V/cell is an unhealthy pack. My personal experience with extended self-discharge was an HCH2 pack (in NH, 110K, IMA light) that sat for 2.5 years ending at 0.45V/cell. The sticks all tested at 90% rated capacity, and the pack performed well in-car for 10 months. Note there is evidence that extended self-discharge improves IR measurably.

See my sig for stick level reconditioning.

Tidbits:

  • ANSI standard for NiMH capacity testing: 0.1C charge for 16 hours, discharge at 0.2C.
  • Per Energizer, NiMH cells are polarity reversal tolerant up to 50% extracted capacity, i.e., if you have a battery pack and one cell reverses polarity after 2000mAh extracted, significant cell damage should not occur until an additional 1000mAh is extracted. This must be done at low current. High current can force electrolysis in the reversed cell and destroy it prior to the 50% limit being reached. Momentary reversals of cells for short durations do not cause measurable damage. I've done it at 20A for about 5 minutes, and there is no measurable decay over the next few test cycles. I would not deliberately repeat the reversal for any reason but data gathering on a stick that I would never put back in a car.
  • Tap variations at rest and under moderate load (13A or less) should be less than 0.2V from Max to Min. A truly healthy pack will be < 0.1V min/max variation at rest.
  • Tap variations under max load (>80A) are typically 0.4-0.6V.
  • Cell variations within a fully charged stick should be less than 0.03V, >10 hours after charging.
Steve
 
#34 ·
My personal experience with extended self-discharge was an HCH2 pack (in NH, 110K, IMA light) that sat for 2.5 years ending at 0.45V/cell. The sticks all tested at 90% rated capacity, and the pack performed well in-car for 10 months. Note there is evidence that extended self-discharge improves IR measurably.

Steve
Just to confirm, this pack got IMA light after 100K, then sat for 2.5 years self discharge, after which it recovered 90% capacity with nothing other than a balancing charge of 36 hours or less?

Also, did it operate for 10 months without any additional grid charging before throwing IMA light again, or with periodic grid charge? Did it keep going strong after 10 months, or had deteriorated by then again and needed additional work?

TIA for clarifications.
 
#10 · (Edited)
#68 · (Edited)
grid

Source:

overview: https://hybridautomotive.com/pages/recon

charge: https://hybridautomotive.com/pages/guide

discharge: https://hybridautomotive.com/pages/sd#termination


00-06 Insight

Charge
Number of cells in battery pack 120

Normal Operating Range
(Filling Phase) 144-168V

Peak Voltage Range
(Approx. voltage of balancing phase) 168-172V



Discharge
First Discharge 96V
Second Discharge 60V
Third Discharge 12V
My first impression was - what kind of equipment is being used.
Because although there are generally adhered to grid charging methods
that seem to work well
The different equipment that is used by IC members can make it hard to follow
specific procedures if you are using different equipment than the one being described in any specific thread.
If you have enough patience, reading comprehension, tenacity and time you can usually figure out what works best in your specific situation.:birthday: :chainsaw:
 
#12 · (Edited)
I'll bite:

Background:
The various design choices and methods are all the result of a horribly implemented IMA system that guesses as much as we do on the overall SoC. The result is an unhealthy pack by design:
-No per-cell measurement circuitry, so can't analytically determine good/bad cell status.
-No bypass circuit in parallel with each cell, so can't nondestructively charge less-full cells once first cell hits '100%'.
-Some newer packs (even from Honda) lack series thermistors on each cell, so the only real method the computer previously had to tell when at least one cell was full (due to overheating <horribly, shudder>) is no longer present.
-Relying on NiMH's "self-recovering" overcharge condition - recombining oxygen and hyrdrogen back into water inside the cell - to overcome the above design decisions, yet not effectively trickle charging the pack when full during normal use.

In short, Honda really dropped the ball on the battery management system.

Most people agree that constantly trickle charging the IMA pack at ~350 mA is the safest method to balance the pack. I don't think you'll find anyone that disagrees with the method. Some people charge at higher current; others perform multistage charging. The recipes vary, but the theory is solid. You can easily build your own grid charger and install it with all the connectors for under $100... others prefer spending many times that buying some other system and even more getting said system installed, and I can't knock that because while I know electronics very well, others do not. I wouldn't know the first thing about performing heart surgery (how difficult most people think installing a grid charger is), but I could probably draw someone's blood with a little instruction (the actual difficulty).

Most people are also in the 'deep discharging' camp. I've been skeptical about the effectiveness, so I recently gathered some data on my own failing pack, since I couldn't possibly care less about its longevity*. In general, 'Deep Dischargers' make two claims:
1) "Deep discharging pulls all cells to 0 volts." It doesn't. Instead, it horribly unbalances the cells, particularly the weakest ones. After a weaker cell reaches 0 volts (i.e. runs out of juice), it starts 'charging' in the opposite direction due to the current still flowing through it from better cells. At some point, the overall voltage across all cells sums to 0 volts, even though the strongest cells are still very much charged, whereas the empty ones might be several hundred mV NEGATIVE. Eventually, the sum of all cell voltages is zero, so no current flows through the pack.

Test:
0: Discharge pack with a programmable load set to 1 amp. A light bulb and/or resistor bank aren't going to cut it for this test. Get an electronic load with programmable current sinking, remote sense, and very low dropout (less than 100 mV @ 1 A).
1: After a few hours, the pack still hovers around 50 Volts. Then suddenly the pack can no longer deliver 1 A. The programmable load does what it can.
2: A few minutes later the load can no longer sink more than 10 mA because the battery no longer generates a cumulative voltage. When this occurs, place a DMM across the pack (or use the one built into the load if you're using remote sense). You'll measure 0.00 volts.
3: Unscrew the semi-transparent "YOU WILL DIE" cover from the side of the pack. Take your DMM and probe stick pairs 14S at a time. Note that even though you might expect to read 0V across all the sticks, you'll actually see that there's a voltage soup ranging from 0 volts all the way up to plinus 12 volts!

Conclusion:
-Deep discharging certainly doesn't bring all cells to 0 volts.
-In fact, is does the opposite: Cells that have more charge to begin with (the good ones) stay positive, whereas cells that have less charge to begin with (the bad ones) end up at a negative voltage.
-If there's a chemical benefit to pulling cells negative, the 'worst' cells get the best medicine.


2) "Deep discharging changes crystal/chemical structure." Not in my test data. Methodology:
0: I charged the entire pack at 1 A for 3 hours, fans blazing high and monitoring PTCs.
1: I charged the entire pack at 100 mA for 48 hours, fans blazing high and monitoring PTCs. Total accumulated Ampacity: 7.8 Ah.
2: I disassembled the pack to 20 sticks.
3: I discharged each stick @ 30 A until the voltage 'necked', indicating the weakest cell was empty. I recorded the Ah on the stick in black.
4: To perform A/B testing, I paired sticks that performed similarly into twin pairs.
5: To maintain a control group, I divided the twins into two groups:
GroupA ("charge only") and GroupB ("discharge & charge").
6A: I did nothing. Being a control is easy.
6B: I continued discharging at 1 A until each stick output 0 V @ 10 mA load.
7: I charged all sticks as per 1 & 2 (above).
8: I discharged all sticks as per 3, and recorded the Ah on the stick in red.

Conclusion:
-There was no distinguishable difference between GroupA & GroupB cells.
-Deep discharging doesn't appear to have any benefit based on my test data.
-The only benefit I see to discharging the pack is that it allows you to charge the pack more; I believe any potential chemical/crystalline benefit occurs while charging. To fully test this theory, each control 'twin' would need to be charged the same number of additional Ah that the test twin discharged to reach 0 volts, then discharged the same amount. This neglects charge/discharge inefficiencies, which are difficult to calculate and are also part of the value we're trying to measure, so there'd be room for foul data.


Overall conclusion:
Deep discharging with an external setup takes a ton more time and effort than simply going for a drive and accelerating hard until the pack is again empty, then charging more. It's possible that the added benefit people attribute to deep discharging is actually the result of additional low-level charging.

*Full disclosure: I'm working on a competing project right now - Linsight - that will replace the entire OEM battery system and solve these problems.
 
#13 ·
Overall conclusion:
Deep discharging with an external setup takes a ton more time and effort than simply going for a drive and accelerating hard until the pack is again empty, then charging more. It's possible that the added benefit people attribute to deep discharging is actually the result of additional low-level charging.

*Full disclosure: I'm working on a competing project right now - Linsight - that will replace the entire OEM battery system and solve these problems.
You are a ridiculously smart guy, and you know your ****. Your position on this baffles me every time you express it.

The pack is NEVER empty in the car. Never. Every reference indicates capacity is lost when NiMH is charged before it's fully discharged. How do you explain the capacity improvements obtained from cycling that have been documented here likely hundreds of times? Deliberately running the battery as low as the car can get it doesn't do anything to increase available capacity. It simply increases the depth of discharge, which increases "wear."

Your termination criteria is a bit confusing. Why did you not just terminate the discharge at the prescribed 1V/cell? How consistent was your termination voltage? Did you also compute extracted energy in Wh to see if there was a difference?

I can't tell you how many times I've seen it happen either from cycling or deep discharge. My first deep-discharge of an HCH2 pack increased the usable capacity by 25%. I grid charged for 24 hours, discharged to 132V (1V/cell) in 48 minutes at an average current of 4.2A (500W Halogen). I continued the discharge to 106V w/60W bulbs. I then recharged it for 24 hours and repeated the 4.2A discharge to 132V - which took right at an hour. The extracted energy in Wh increased even more than the Ah as the average voltage was a little higher.

I currently have a Prius pack on my bench. I have conducted a 10400mAh input charge, discharged to 1.05V/cell, deep discharged at 0.2A. I still need to repeat the input charge. I will then repeat the discharge to 1.05V/cell. I will post the results of each of the 28 modules before and after capacities.
 
#14 · (Edited)
I'm still relatively new here. Both mudder and S Keith are ahead of me in many ways.

mudder's approach of bringing the pack to 0 volts is not what I've seen recommended. Maybe its recommended somewhere but I haven't seen it.

Yesterday, I did a detailed read of Hybrid Automotive's (HA) user notes for their bulb-based dischargers. That's because I'll build my discharger this weekend; my charger build last spring went well.

HA's user process runs three discharge cycles of the pack. Each discharge cycle is preceded by grid charging fully, to cell saturation. Each discharge cycle uses a series of three bulbs in decreasing wattage, ending up with a 25W bulb as the load.

Bottom line is that they don't recommend going lower than 0.1V per cell, which they only do in the last of the three discharge cycles. That's 12V for the full pack. HA's first and second pack discharge cycles only go down to 0.8V and 0.5V per cell, respectively. After reaching those limits, their process is to end the discharge cycle and recharge the pack.

Personally, I don't want to take my packs down to 0.1V per cell, or 12V/pack. Since I'm doing the whole pack as a unit, I don't want to risk pulling individual cells down below zero, which could happen if the whole pack is down at 12V. I'll probably end my final discharge at about 40-60V (pack), which is 0.3-0.5V per cell, assuming incorrectly that all cells will be at equal voltages.

Note, that 12V pack target from HA is with a 25 watt bulb for a load. At about 100V (pack voltage) when first connecting the 25W bulb, that's only a 0.25A load. Much less than the 1 amp load mudder discussed using.

As mudder touched on, bubs (which I'll use) have oddities when used as load resistors - their resistance changes with current and temperature so it's pretty much impossible to say what the ohms and loads really are (though you can clearly see the voltage). That said, all the bulb based discharge schemes I've read, use lower wattage bulbs late in the process, to reduce amps load. That way they have slower voltage drop, so the user is able to stop the process long before voltage goes to zero.

Hybrid Automotive bulb discharger user guide
 
#15 · (Edited)
I'm still relatively new here. Both mudder and S Keith are ahead of me in many ways.

mudder's approach of bringing the pack to 0 volts is not what I've seen recommended. Maybe its recommended somewhere but I haven't seen it.

Yesterday, I did a detailed read of Hybrid Automotive's (HA) user notes for their bulb-based dischargers. That's because I'll build my discharger this weekend; my charger build last spring went well.

HA's user process runs three discharge cycles of the pack. Each discharge cycle is preceded by grid charging fully, to cell saturation. Each discharge cycle uses a series of three bulbs in decreasing wattage, ending up with a 25W bulb as the load.

Bottom line is that they don't recommend going lower than 0.1V per cell, which they only do in the last of the three discharge cycles. That's 12V for the full pack. HA's first and second pack discharge cycles only go down to 0.8V and 0.5V per cell, respectively. After reaching those limits, their process is to end the discharge cycle and recharge the pack.

Personally, I don't want to take my packs down to 0.1V per cell, or 12V/pack. Since I'm doing the whole pack as a unit, I don't want to risk pulling individual cells down below zero, which could happen if the whole pack is down at 12V. I'll probably end my final discharge at about 40-60V (pack), which is 0.3-0.5V per cell, assuming incorrectly that all cells will be at equal voltages.
First, you need to understand that you will be pulling DOZENS of cells below zero before you get to 0.8V/cell. Reversals start happening between about 135V and 144V (nominal). The Genesis One grid charger has reversal detection as a termination criteria, and it's pretty common for it to trip in that range.

A good rule of thumb is (1 - volts/cell) *100% = the % of reversed cells, e.g., at 0.8V/cell, you've reversed 20% of your cells. This breaks down at about 0.5V/cell. When you start to get really deep, total voltage is reduced as much by reversed cells showing more negative voltage thus offsetting the non-reversed cells.

Again, you need to adjust your perspective. You will begin reversing cells very quickly between 1.1 and 1.2V/cell. You need to get past it. :)

Per my list above, the key is to keep the current low and limit the total extracted capacity by measuring current and computing capacity.

Note, that 12V pack target from HA is with a 25 watt bulb for a load. At about 100V (pack voltage) when first connecting the 25W bulb, that's only a 0.25A load. Much less than the 1 amp load mudder discussed using.

As mudder touched on, bubs (which I'll use) have oddities when used as load resistors - their resistance changes with current and temperature, so it's pretty much impossible to say what the ohms and loads really are (though you can clearly see the voltage). That said, all the bulb based discharge schemes I've read, use lower wattage bulbs late in the process, to reduce amps load. That way they have slower voltage drop, so the user is able to stop the process long before voltage goes to zero.

Hybrid Automotive bulb discharger user guide
Per my list above, they behave MORE like a constant current load. For a halving of the voltage, you'll see about a 30% drop in current. The $5 Harbor Freight multimeter in 10A ammeter mode provides invaluable current data that takes a LOT of the guesswork out of it... it enables you to CALCULATE your actual extracted capacity. It is surprisingly accurate.

Given that you're going DIY, there's no excuse for you to omit that $5 piece of hardware.

Your conclusions are incorrect. First, you'll almost never get a pack to zero volts. I've inadvertently been as low as 1.2V on a Prius pack (168 cells). It was with the equivalent of a single 12.5W bulb (2X 25W in series), and the current flow was around 50mA.

Second, the use of lower wattage bulbs has nothing to do with any desire to allow the user enough time to terminate before getting to zero. It has everything to do with keeping the current flow low to minimize damage potential to the dozens cells that are in polarity reversal.
 
#17 ·
Thanks for that link. At some point, I checked my charts and just remembered the 30% current drop for a 50% voltage drop relationship. I see the same on three of the wattages I checked on your charts. Confirmation is good.

The resistor takes all the guesswork out of it. The 500Ω resistor puts the current at 240mA @ 1V/cell and 290mA at 1.2V/cell. Those are pretty good numbers. The only downside is the extremely long discharge time. IIRC, Peter uses a 300Ω resistor, which would help expedite the trip to 144V.
 
#18 ·
More thanks, to S Keith and olrowdy01. The more I thought I knew, the more I discovered I had yet to learn. I'd planned my procedure on Hybrid Automotive's writeup but I like Steve's attention to detail. So here's my plan, per SK as best I can absorb what you've written.

0) Start with a fully grid-charged pack (SOC display is meaningless here).
1) I can discharge with a 150W bulb (or for me, 2-75W bulbs in parallel), but only to a point. 144V, or 1.2V/cell, should be the cutoff for the 150W phase. Note the time to reach that point, for comparison later.
2) I think I'll modify Hybrid Automotive's intermediate stage of discharging with a 75W bulb. I'll use 2-25W bulbs instead, total 50W. I expect some cells will likely start going into reversal in this phase, so I want lower current to avoid causing damage while in reversal. Good idea? Terminate that phase at 110V, or maybe at 100V.
3) After the 50W phase, I will use 1-25W bulb. I expect initial voltage to be higher than the 100-110V termination of the 50W phase due to the reduced load. Following your recommendation, I'll terminate the 25W phase after 5 hours (1,000mA) or when it reaches 0.8V/cell or 96V/pack. Whichever comes first.
4) Recharge fully after discharge.
5) Discharge again with 150W bulb or equivalent. End at 144V and note the time to discharge. If time is +20% or more from the time in step (1), consider it all complete.
6) If time to full is increased by less than 20%, continue into another deep discharge phase using a 25W bulb. This time, continue discharge down to 60V/pack, but again, limit time to 5 hours/1,000 mA.

Hmmm. If I get into a step (6), it could take longer to reach 60V than it did to reach the 96V limit of the first cycle. I'll just have to see how that plays out.
 
#19 ·
Modifications in bold based OMHO:

More thanks, to S Keith and olrowdy01. The more I thought I knew, the more I discovered I had yet to learn. I'd planned my procedure on Hybrid Automotive's writeup but I like Steve's attention to detail. So here's my plan, per SK as best I can absorb what you've written.

0) Start with a fully grid-charged pack (SOC display is meaningless here).
1) I can discharge with a 150W bulb (or for me, 2-75W bulbs in parallel), but only to a point. 144V, or 1.2V/cell, should be the cutoff for the 150W phase. Note the time to reach that point, for comparison later. Wire in an ammeter in series. Record time, voltage and amps at 1, 5, 10, 15, 20, 25 and 30 minute intervals and every 30 minutes thereafter.
2) Deleted
3) 1-25W bulb. Initial voltage WILL be higher than the 144V termination of the 150W phase due to the reduced load. Following your recommendation, I'll terminate the 25W phase the sooner of 5 hours (1,000mA) AFTER 144V is achieved under 25W load or when it reaches 0.8V/cell or 96V/pack. Whichever comes first. Record time, voltage and amps at 1, 5, 10, 15, 20, 25 and 30 minute intervals and every 30 minutes thereafter. Record tap voltages at 144V and end voltage under 25W load.
4) Recharge fully after discharge.
5) Discharge again with 150W bulb or equivalent. End at 144V and note the time to discharge. If time is +20% or more from the time in step (1), consider it all complete.
6) If time to full is increased by less than 20%, repeat step 3 except terminate at the sooner of 60V/pack or 5 hours/1,000 mA

Hmmm. If I get into a step (6), it could take longer to reach 60V than it did to reach the 96V limit of the first cycle. I'll just have to see how that plays out.
We're talking $5 and increased monitoring/data recording to give you actual numbers you can compare to in the future AND data to share with the group. Sure, time alone is valuable, but it's a special kind of chubby knowing where your pack stands (might just be me, though). :D

Steve
 
#21 ·
Steve K,
Yes I hav a multi tester that will read a few amps in-line. I haven't used it that way but I'll hook it up and get the readings.

Sorry I won't get stick pair tap voltage reads. Opening up the hatch takes me like 46-60 minutes every time. I'm interested in a pair of connectors like some have discussed but I won't be able to do that this weekend, if that's even sourceable.


Sent from my iPhone using Tapatalk
 
#24 ·
Mudder,

NOW I UNDERSTAND. Your perception is a bit off, but the scenario you described has been done; however, I doubt anybody maintained a 1A load to anywhere near zero as even bulbs taper their current with voltage drop.

Thank you for the clarification, and I agree it's ineffective as you presented it.

As indicated, I have completed a before/after deep discharge test on 28 Prius modules:

Start condition: all modules discharged @20A to 6.3V - initial "as-removed" capacity was 1108-2626mAh.

Total voltage spread in the as-removed condition was 7.53-7.57.

Column1: 20A discharge to 6.3V following 8450mAh input grid charge.
Column2: 20A discharge to 6.3V following 0.2A discharge to 4.0V and 8450mAh input grid charge:

3673, 5263
3589, 5318
3705, 5163
3703, 5211
3658, 4876
3572, 4912
3999, 4853
3965, 4822
4043, 4960
4068, 4945
3789, 4902
3928, 4794
3655, 4811
3795, 4953
3859, 4747
3642, 4700
3750, 4943
3733, 4924
3655, 5030
3713, 5183
3843, 4968
3612, 5250
3780, 4915
4035, 5022
3780, 4915
3728, 5216
3890, 4990
3997, 5365

I only noted the capacity extracted, and they were between about 900 and 1450mAh. The capacity consumed between 6.3V and 6.0V was about 700mAh, so the additional capacity extracted after the initial 20A discharge during the deep discharge efforts was around 1600-2150mAh.

While I don't have hard data handy to support it, the above results are also typical for healthy IMA sticks.

This 2004 has lived in AZ for its entire 165K mile life. The battery failed because module 11 sprung a leak and created a HV path to ground. Oddly enough, even the module that spring a leak is right at the pack average after leaving a nasty trail of KOH.
 
#29 ·
Hi guys this seems to be the expert level group. Can you point me to the best thread that has charging basics, or a favorite youtube video, etc? I follow most of what you are discussing here to a point, but, being that I don't own my G1 insight yet (still hunting), I'd like to get educated about:
-types of grid chargers (home made vs commercial)
-attachment points/harnesses
-do's and don'ts for reviving and for maintaining a pack
-terminology and acronyms

Thanks
Probably a tall order and too broad but maybe you're familiar with a great thread or just have a favorite bookmarked (Maybe one of you wrote it!)
 
#30 ·
I'm no expert, but here's my approach.

-types of grid chargers (home made vs commercial)
homemade, using a slightly modified version of the Dabrowski schematic
http://99mpg.com/Resources/downloads/
-attachment points/harnesses
I use these:
https://www.amazon.com/gp/product/B00YVC46QM
https://www.amazon.com/GLS-Audio-Speaker-compatible-Neutrik/dp/B00LW8VLRU

-do's and don'ts for reviving and for maintaining a pack
I do a complete discharge, followed by a 30+ hours grid-charge. Results have been nothing short of amazing on 3 of 4 packs.

-terminology and acronyms
YMMV
 
#38 ·
There is a "apples to oranges" issue here. The original HCH2 cells are different than the original HCH1 cells. They have lower capacity and the claims are they have lower IR. Results don't necessarily translate from one to the other always. As with everything in this area, YMMV.

Given that your pack is a replacement, it's impossible to say if they're the same cells. if you test sticks, and they're >5500, then there's a good chance they're the same as the HCH2, but I really doubt it.

IMA in AZ get eaten. HCH2 AZ failures are MASSIVE SD that rarely can be recovered with cycling.

The anomalous behavior you describe is not anomalous as it's expected since you've demonstrated your pack already has SD problems.

And your pack should NEVER drop anywhere close to 140V. Period. 140V is a horribly dead pack. If I haven't driven it home yet, if a NiMH battery open circuit voltage is EVER below nominal, it is a "Danger Will Robinson, Danger!" alarm. Something is seriously wrong. If you're seeing 140V, you need to pull the pack and start cycling it. NOTHING to be gained from the proposed experiment(s) or by leaving the pack in the car for another single second unless the goal is to collect noise.

Sitting may or may not help. I've had sticks that sat for a year and showed no improvement. The pack in that description you quoted was born and lived in NH before it found its way to AZ.

BTW... you have yet to pick a single piece of fruit. You've targeted it, you've even kinda tapped at it... given it a sniff, but you have yet to pluck it. Fly turd arguments aside, you still haven't done the years-long, widely accepted charge/discharge reconditioning cycling as described on HA's website.

A single grid charge showed significant improvement over its prior state. Given that, it follows whole-pack reconditioning may further improve it.

Pick the damn fruit, please. :)

Your SD comments - that's no hypothesis there. It's just how they work. A healthycell at 100% SoC has a higher SD rate than when it's at 80% SoC. The trend continues. SD never goes to zero, but it gets pretty close. I'd love Balto to chime in and tell me what that stick I sent him measures if he hasn't used it. A few months after I sent it to him, it was still near 8.0V.

I have data showing that cells can readily SD 10% of their capacity in the first 24 hours, yet take another 20 days to SD another 10%, and the majority of the extra 10% takes place in the first week.

You already have some data confirming the first portion of the above - your mAh/hr SD rate was decreasing before everything went wonky. I suspect the wonkiness was from non-uniform results from your single grid charge. Some improved, others didn't.

Charge again for 30 hours (10,400mah input), then follow the HA reconditioning cycles. That's the lowest hanging fruit you're going to find.

That other piece of fruit you see over there? You know... the one where you just discharge it to near zero and hold it there overnight? That's just rot with a pretty fruit skin on it. You've already demonstrated the battery in its current condition is imbalanced. The best way to go is up, not down.

You're free to continue the "people have been doing this successfully for years, but I think I'll try something different" plan, but you said "low hanging fruit." Fruit is up in the tree. Digging for taters and carrots ain't the same. :)
 
#39 ·
The 140V is under load giving assist. I believe that's an acceptable voltage reading under load.

Car is originally from CA. No idea where replacement pack was born or initially installed in 2011.

Are you proposing that doing the HA stepped cycling regiment, 95V/60V/12V can actually reduce self discharge? I haven't seen anyone experience or demonstrate this, so that would be new to me.

Also, why do you say deep discharging is just rot with a pretty fruit skin on it? It seems many people have experience great performance improvement on deep discharge including yourself.
 
#40 ·
Yes, 140V under moderate to heavy load is fine. I thought it was resting. Heavy assist can pull voltage easily into the 130s, and your voltmeter smoothing may be too heavy to see the variation.

I'm proposing that cycling CAN reduce SD. It just depends. Most don't demonstrate it because they're not paying attention to the numbers. Most just do the process and it works. They have no idea what happened besides an IMA light/CEL was there, but now it is gone, or it doesn't work. The data that cycling can improve SD is on this site, and you collected some yourself. You semi-cycled your pack (very near 0% and certainly 100%), and SD appears to be better than it was.

My rot comment was specific to your "low hanging fruit." Deep discharging your pack from its current state is NOT the low hanging fruit. Starting over with the HA process is the low hanging fruit.
 
#41 ·
Right, from what I've read, one deep discharge is all you need for say, a year or two, maybe even 5-6 years, with insufficient data to recommend any particular frequency of deep discharge other than "when your pack is garbage, and you have nothing to lose, definitely worth the risk." Additional deep discharge soon after the first one seems to bring no observable performance improvement. I have no intention of doing another deep discharge until I see significant deteriorated performance from current levels. Consequently I should point out that "deep discharge" used to be defined as ~70V or so a few years back. These days I think most people consider it deep discharge with at most 12V on the pack, and usually under 1V, with some positing 0.4V.

The "my pack is fixed and I have no idea what fixed it" is exactly why I'm taking a more methodical approach than I've been recommended by you and others. Hopefully I can collect a little more data, anecdotal or otherwise, that may give more insight as to what treatment affects what performance improvement parameter..

So far, the one cycle deep discharge and balancing grid charge appears to have done quite a lot of good. The pack is still not where I'd like it to be. But then it's a bit tired and potentially abused from harsh environmental conditions. At the least, this one cycle has transformed the pack from barely useful to quite utilitarian.

Last night I effectively cycled the pack at high amperage ~3 times in the middle range of SOC from my hill tests. After overnight self discharge from 169V to 159V, I ran another test this morning driving up that same hill. This time I minimized assist until I reached the hill. I also took a different freeway entrance that may be 20% up the hill already. While climbing the hill, I limited assist to maybe 6 bars at most by using 3rd or 4th gear. I also watched the voltage and it did not drop below 150V during assist. The battery did not recal and dash gauge was still ~55% full at the top. There are a lot of variables at play here, so it's hard to say whether the lack of recal is due to the cycles I put it through last night, the more gentle assist usage while going up the hill, the voltage not dropping too low to trigger recal, or some combination of these and other factors I haven't thought of. As I get more familiar with how the car drives, I'll be able to tell with much more certainty what the battery pack is now capable of. At that point, I can perform additional non-driving pack maintenance and try to correlate specific actions with observed performance. Hopefully I can figure out what additional observed driving performance improvements are afforded by the default recommended stepped discharge regiment without mucking up perceived improvements with other treatments. One hypothesis I'm trying to test is whether high amperage cycling after single cycle deep discharge and balancing charge via driving does just as much good as stepped discharge regiment. If confirmed, this could provide an alternative approach for first time revival of a known deteriorated pack, albeit it helps to have a good hill to perform.

Later today I'll go up the hill again from the bottom with fuller assist, and try to trigger a recal. I'll try to make sure the pack is newly full before doing that. This is to suss out whether the pack can hold newly received charge short term and have another point of validation to confirm self discharge. I'll also end the day at lower voltage and see if that reduces magnitude of voltage drop overnight.

I observed voltage as high as 182V during heavy regen, and very often in the high 170s. I also observed limited regen of 3 bars when SOC is on the high side. This is undoubtedly "normal," just new to me as this is really the first time I've been able to drive the car enough to start learning its behavior. There are multiple combinations of factors that will trigger limited regen. My guess for some of these factors are too high pack voltage, too high temperature on PTC, and potentially too much current has flowed through current counter. I'm somewhat dubious too high temperature is a factor as I didn't hear the fan turn on during these limited regen periods.

I'll collect a bit more data before updating this thread again with additional learnings and confirmations. So far the take away is, yes a single deep discharge at low currents (57W bulb) to under 0.4V (0.2V in my case) followed by a single balancing grid charge (also low current, ~350mA) can do a whole lot of good to make a pack useful again.
 
#43 ·
^Curious how I've never really seen in the threads you've contributed to over the past year or so the genesis of your 180 degree turn. At some point, I think when you got into the 'Prius' stuff, you started to change your tune... And now we're at that 180 degree about-face from deep discharge back to 'cycling'...

....Last night I effectively cycled the pack at high amperage ~3 times in the middle range of SOC from my hill tests. After overnight self discharge from 169V to 159V, I ran another test this morning driving up that same hill.

....I observed voltage as high as 182V during heavy regen, and very often in the high 170s....
A couple things: That fall from 169V to 159V tends to suggest you've got some high self discharge cells. Kind of depends on how soon after you charged the pack that you were reading that 169V value. But, for example, it'd make me nervous if I saw my pack drop from 169V to 159V; it'd make me nervous if I saw it drop from 169V to 162V, overnight... 163-164V, Bvo reading on the OBDIIC&C, is about par for the course... If my pack slips lower than 163V, from a full car charge definitely, but generally up around 70%+ nominal, it's a subtle sign something's a bit off. It may be that I haven't let the pack do its hanging pos recal for a while. Or it may be that it's been a long time since I last grid charged and there's some imbalance setting in...

On your 182V observed voltage during heavy regen - I think that's a strong indicator of ... imbalance of whatever kind. If your pack is balanced I'm pretty confident that you should be seeing around 188-192V (again, Bvo on the OBDIIC&C). If the cells in one tap are more charged than the others, or perhaps the cells are deteriorated and have high internal resistance, that tap will get pegged to the high voltage threshold - 19.2V, 1.6V per cell average. Yet, the other sticks/taps have a lower voltage. That's the only way you can see a low total voltage... In other words, your regen charging gets throttled when one of the taps hits the upper threshold, while the other taps aren't at that threshold; thus, the pack voltage is lower than the sum of the max tap voltages... Seeing 190V+ during charge (high current*) is an indicator for me that the pack is pretty balanced. When I see lower than that it's an indicator of a slipping pack - imbalance is creeping in...

*"high current" here really means under certain circumstances. If the state of charge is low then I wouldn't expect to see such high voltage; if state of charge is high and I'm doing heavy braking regen I can expect to see around 190V; if I'm doing a 3500 RPM rev charge (about 22 amps), I should see 190-192V when the pack positively recals to a nominal 75% state of charge, that is, if the pack is well balanced...
 
#42 · (Edited)
I'd like to clarify "deep discharge"...

I believe the infamous thread of that title refers to deep discharging to ridiculously low voltages.

IMHO, "deep" discharge is anything below an average of 1.0V/cell where cell polarity reversals are guaranteed if there is any imbalance at all.

"deep discharge" is nothing more than a shortcut. It's saying "I'm skipping the progressively lower discharges prescribed by the developer of this whole concept (with supporting test data), and I'm just going to shove it down to near zero."

That's it. It basically turns 3 cycles into one that's probably about as effective as 2-2.5 progressive cycles. It works exactly the same way - elimination of voltage depressed capacity and restoring 1.2V capacity. Period. There is also typically some improvement in SD and IR from full range cycling, but this is an extrapolation of stick-level results.

I can tell you that "deep discharging" a stick has no more benefit than 3-4 cycles to 5.4V (0.9V/cell); however, that first "deep discharge" cycle will have more benefit in terms of capacity recovery than the first 0.9V cycle by a significant amount.

The only difference between a 6 cell battery and a 120 cell battery in terms of "deep" discharging is the number of cells in reversal. With 120 cells near 0V, it's almost a certainty that you will have 60+ reversed cells. At pack nominal, you will likely start reversing cells around 135V under low current. At stick nominal, it's highly unlikely you reverse any cells at all.

EDIT: Premature post...

I'm not aware of any benefit of cycling in car. It doesn't follow that you should seen any significant improvement, it's the operation in-car that causes deterioration. Bigger cycles and higher currents continue to cause deterioration.

"High current" is used in the context of grid charging, typically intended to be around 1C.

it's the cycling through the full range of SoC of as many cells as possible that improves capacity and performance.
 
#49 ·
"deep discharge" is nothing more than a shortcut. It's saying "I'm skipping the progressively lower discharges prescribed by the developer of this whole concept (with supporting test data), and I'm just going to shove it down to near zero."
What IS this?? Who is this "developer" of the concept with "supporting test data"? Also, going to zero rather than doing cycles may or may not be a shortcut. It depends. But they're both shortcuts to deeply discharging sticks or cells...

I can't help but say that you've been a perpetuator of a lot of misinformation lately. I'm not sure what's gotten into you... And it really chaps my hide when people wave the magic wand of some mythical authority - just to convey an impression of having superior knowledge. You do it here with this 'developer with test data' missive and you do it above with the tangentially related research article. Can't you just explain your ideas and let them stand or fall on their own?
 
#51 ·
Mike is a very long time member and tester/developer here at ICN. PERIOD.

Willie
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top