Glad to help. Keep thinking about the deep cycle discharging too, it will provide added benefit.

 

ditto on the discharge. Having done stick-level work, full pack grid charging only, and a number of full pack deep discharges to different levels followed by full grid charges, I have found that only the deepest of discharges (to near zero volts) followed by a full grid charge fixes things. You'll get some improvement with the partial measures - and it's easy to mistake that improvement for a real fix. But in my experience the improvement pales in comparison to what a super deep discharge does, plus what improvement you get doesn't last very long...

Based on what I've read and come to understand about what's likely the main problem, what I've experienced with these different treatments makes a lot of sense: over time the cells become voltage depressed and imbalanced, and the way to get rid of the voltage depression and imbalance is to bring each and every cell down to a very low voltage (something like 0.45V at a very low current rate). But with a pack of 120 cells in series, it's impossible to bring each and every cell down to that level without taking the whole thing down to near zero volts.

Grid charges alone achieve some level of gross balancing - but only within a voltage depressed, limited range. You extend the capacity of the cells to some extent at the top end, but achieve nothing at the bottom end. The cells still remain imbalanced and the pack will soon revert to its grossly out of balance, dysfunctional state.

Deep discharges to not-so-low levels, like down to 120V, 100V, even 50V and 25V, followed by full grid charges, extend the top end more than grid charges alone; extend the bottom end for those cells that dropped low enough in the process; and achieve more balancing than grid charge only. But due to the nature of the NiMH chemistry and how they discharge - or are supposed to discharge, with the discharge voltage pretty much independent of the state of charge, a very flat discharge curve - these not-so-low deep discharges on the whole pack still leave many cells above 1V, like about as many cells as the voltage - 50 cells at 50V, 100 cells at 100V, etc. You're fixing some cells, but not all of them. You'll see some improvement, more than with just a grid charge, but again, it will be short-lived, the pack will soon revert to its grossly out of balance, dysfunctional state...

I'm absolutely convinced that one needs to do both a deep discharge and a full grid charge to recondition packs. And I'm fairly well convinced that the deep discharge portion needs to go down to near zero volts on a full pack. Anything else leaves too many voltage depressed cells that will continue to mess up the performance of the pack...

I'm still trying to figure out ways to quantify and display the impact of different treatments. At some point I might have some hard numbers, or at least some soft numbers...

 

Worked for me

I'm absolutely convinced that one needs to do both a deep discharge and a full grid charge to recondition packs. And I'm fairly well convinced that the deep discharge portion needs to go down to near zero volts on a full pack. Anything else leaves too many voltage depressed cells that will continue to mess up the performance of the pack...

A month ago, my battery had been deteriorating and the IMA was lighting up. I performed the descending discharge (120V/100V/50V) and recharge cycle and thought things were fixed; only to find the fix was temporary (lasting perhaps 3-5 days).

With the helpful advice of eq1 I discharged down to .20V and grid charged for 8000mAh. The Insight has been running strong for the last month. I plan to repeat the process when needed.

I have one of the original low voltage (350mA) 99MPG chargers and needed to modify the default settings to achieve the 8000mAh charge. The default time had been 900 minutes or 15 hours, and the cut-back voltage had a default of 167V which I increased to 179V.

eq1 had informed me that the 8000mAh was a very important level and the fact that my charger would shutdown after 15 hours would not let the charge reach the desired level. I really appreciate eq1's help and time assisting me rectify my battery issue.

If you are having similar issues I would suggest you attempt the deep discharge and full grid charge method.

 

Glad we could help!

eq1 is right, deep discharging 2-3 times a year and charging every 8-12 weeks is going to help keep the battery strong for as long as possible. I have been draining several Insight packs to 10V or lower (in two or three full cycles) and having nothing but very good results.

 

Glad to hear things worked out. I wouldn't say '8000mAh' is important in and of itself, but rather, that you had been only charging for 15 hours and that the cutoff was 167V are the critical shortcomings... Actually, if you were only charging to this level after your earlier successive deep discharges, then I'd say it's more likely* that's why your treatments were short-lived, ineffective, rather than how deep you were discharging... But in any event, I still suggest going low...

*edit - I take this back a bit. I guess I can't say it's "more likely," but maybe 'equally likely', possible...

 

Still running great after 700+ miles...

My average mpg has only been 41, but it has been cold and I let my car warm up for 5-10 minutes before the first drive of the day.
Thanks,
Joanna

 

My average mpg has only been 41, but it has been cold and I let my car warm up for 5-10 minutes before the first drive of the day.
Thanks,
Joanna

That's good to hear.

Suggestions (above) for the discharge part of the process are worth keeping in mind.

I just took my mt through the charge/discharge/charge sequence, huge difference in performance and available capacity (as opposed to grid charge only). Especially notable on the mountain climb to work. Time will tell the rest.

 

Great to Hear, the fix is long lived!

I know just discharging mine really helped it, so far. When I get my GC back, I plan to re do the whole process except GCing it after the deep discharge (I just used the car to multiple ties reset the IMA to go a bit higher with the car charger than normal. SO far two days and no IMA yet...

 

While most results are favorable, this fix is of completely unpredictable duration. With 120 cells in the pack, results from pack to pack are highly variable. Given that you didn't finish with a topping charge to 100% SoC (the multiple 12V reset method doesn't push it past 80% by any significant amount), you should expect to need the charge/discharge routine more frequently due to accelerated imbalance. For any cells with higher internal resistance, they will be less than 80%, and for cells with the lowest, they will be greater than 80%.

I don't think it will be a major factor, but if your pack could go 8 weeks between charge/discharges, your technique might knock that down to 6-7 weeks. OFC, I might be completely off my rocker.

FWIW, If you couple your 12V reset charge method with about 8 hours of grid charging, I think you would accomplish the same thing as a 24hr grid charge.

Steve

 

While most results are favorable, this fix is of completely unpredictable duration. With 120 cells in the pack, results from pack to pack are highly variable....FWIW, If you couple your 12V reset charge method with about 8 hours of grid charging, I think you would accomplish the same thing as a 24hr grid charge.

I wouldn't say "completely" unpredictable, or "highly" variable; I might say "quite unpredictable" and "somewhat variable." On the latter, the real wild card probably has most to do with the environment and circumstances under which a given pack was used - like hot climate and higher mileage=lower chances of success, or the longer one drives with a pack on the fritz, probably the greater chance single cells have been damaged beyond repair. Yet if the strategy seems to work at the get-go, the first time you do it, I'd say the results are more likely to be similar from pack to pack than dissimilar - if you make some kind of adjustment for pack age, mileage, etc... I guess this all applies to both phrases - "completely unpredictable duration" and "highly variable results." If the strategy seems to work once you do it, then how long it lasts depends more on the specific pack's 'life profile' than on the treatment per se, I think... Basically, I'm saying we can make some ballpark estimates given some info about the packs being treated, and I think they're more likely to be right than more likely to be wrong...

I don't know, maybe you're right...

Oh, on the extra grid charge thing, my understanding, or maybe just 'sense', is that, in general, you shouldn't expect a fast, variable charge to do as good a job as a slow, steady charge after a deep discharge. It's kind of like the way you shouldn't expect as full a charge from a fast rate at high SoC versus a slow rate at high SoC...

 

I guess with sufficient input, variability can be minimized. My recent work with two HCH1 packs showed that while deep discharge restored a significant amount of lost capacity for low-drain discharges (the typical ~25% improvement), the packs were unable to perform in the car w/o IMA lights. These things were way gone. Using the best 20 of 40 sticks to build a "best" pack resulted in inconsistent behavior (occasional IMA/CEL, but improved performance with occasional regen).

Whatever the results, IMHO, it's always worth trying!

While I would expect improved fast charge characteristics as you suggest, IR will still influence the results, an 8 hr grid charge on top of the car charging would be the best of both worlds.

 

I am glad to read this. I was wondering on (Store - Hybrid Automotive - Honda & Toyota Grid Chargers) is it better suggested for the Hybrid Battery Discharger Single or Dual Bulb for best results and what bulbs do you guys suggest? standard incandescent 60-200w?

I was considering http://amzn.com/B005EF15Q4 but not sure if that is really the better method even if it does seem a bit cheaper.

What are your thoughts eq1/mountain/others?

 

The charger you linked is not at all useful for grid charging. It is only partly usable for stick work. If you insist on doing stick work, I violently recommend the Turnigy Reaktor 300W from hobbyking.com.

 

I'd vote (what I used)

Single. I used a 200w bulb then went down for the remainder (see thread) for less damage/chances with reversals.(from HA).

 

Single. I used a 200w bulb then went down for the remainder (see thread) for less damage/chances with reversals.(from HA).

Thanks Thrillskier I did not see that part.

 

IMax charges at 50W. Reaktor charges at 300W. Imax discharges at 5W. Reaktor discharges at 20W. In other words, a Reaktor has 6X the charge capacity of an Imax and 4X the discharge capacity for 3X the price. It is also a might higher quality product than the "clone" you linked that costs $22. In fact, I bought several from that Amazon supplier.

I am speaking from experience. I have 14 Imax B6. I wish I had never bought them.

With my 2 Reaktor chargers, I can do 3X cycles of the 22 HCH2 sticks in 10.5 hours at 15A charge and 20A charge rates. You'd need 20 Imax B6 to do that in the same time, and the capacity results would have no real-world significance.

When you consider the regenerative discharge capability of the Reaktor at 300W, the choice is clear; however, this takes some effort on your part to utilize. An automotive 12V battery would be a good regenerative sink and allow you to discharge your sticks at 20A instead of the paltry 800mA the Imax can permit. You can also optimize your electron flow by strapping two empty sticks in series to act as the regenerative sink, so as you discharge your test stick, you are recharging those two. Due to the inefficiency in charging, you will only lose about 30% of the juice.

The ONLY reason to get to the stick level is to identify and replace bad sticks. The Imax has almost no usefulness in this regard except to DEEP discharge sticks to near zero; however, the Reaktor does this much better as it will taper the current once the target voltage is reached, i.e., if you are discharging at 20A to 6V, once it hits 6V, it holds 6V and starts tapering the current to as low as 1A (it's a setting, 5-100% of the start current).

The Reaktor also logs a bunch of data and allows you to download it to logview for analysis.

In order to properly assess sticks, you need to be able to subject them to a 90A load for 10 seconds to confirm they can support that load at >5.4V. Neither the Imax or the Reaktor can do this. Furthermore, you need to check capacity at a 10A load. Imax can't do this either, Reaktor can. You need to repeat that 10A test a week later to check self-discharge rate. Imax can't, reaktor can.

I hope you're seeing that this isn't an easy or particularly cheap thing to accomplish. The most productive option is the pack charge/deep discharge (minimum effort for maximum results). I have had great success with discharging to 120V with a 500W halogen (4-5A load) followed by a deep discharge from 120V down to near 0V with a 40W bulb. I don't personally recommend the 500W halogen because the operation at >120V takes its toll on the bulb, and a bulb will only survive 1-2 tests.

With all the ability that I have to do stick-level work very efficiently, I would choose to do a grid charge/deep discharge at the pack level before ever tearing into one.

To address your specifics, PACK discharges always follow "soak" grid charges of the pack. That way you ensure all cells are at their max capacity. I do a ~24hr grid charge followed by a 500W discharge to 120V and then 40W discharge to ~0V (1.8V is the lowest I've gone, stopped due to impatience). A repeat grid charge followed by another 500W discharge to 1V/cell never fails to provide a 25-33% improvement in capacity AT THAT DISCHARGE RATE (4-5A).

I don't bother multi-cycling either packs or sticks (I only gave a time above to give you a reference). I start any single cycle with a "soak" charge (350mA for a pack or C/10 for 16 hours for a stick). I take sticks to 1V/cell at 20A, which tapers to 1A, then I take those sticks to 0.2V total at 1A which tapers to 50mA. 15A charge with repeat 20A discharge to 1V/cell show a 20-25% capacity improvement. On HCH2 sticks, this has been true across 5 packs. On 2 packs of HCH1 sticks, the improvement is similar, but only at 10A discharge rates and they have significant capacity still available at a 5A load (800mAh or so). I attribute this to a higher internal resistance inherent in the design as well as age/abuse.

Steve

 

The most productive option is the pack charge/deep discharge (minimum effort for maximum results). I have had great success with discharging to 120V with a 500W halogen (4-5A load) followed by a deep discharge from 120V down to near 0V with a 40W bulb. I don't personally recommend the 500W halogen because the operation at >120V takes its toll on the bulb, and a bulb will only survive 1-2 tests.

With all the ability that I have to do stick-level work very efficiently, I would choose to do a grid charge/deep discharge at the pack level before ever tearing into one.

To address your specifics, PACK discharges always follow "soak" grid charges of the pack. That way you ensure all cells are at their max capacity. I do a ~24hr grid charge followed by a 500W discharge to 120V and then 40W discharge to ~0V (1.8V is the lowest I've gone, stopped due to impatience). A repeat grid charge followed by another 500W discharge to 1V/cell never fails to provide a 25-33% improvement in capacity AT THAT DISCHARGE RATE (4-5A).

I don't bother multi-cycling either packs or sticks (I only gave a time above to give you a reference). I start any single cycle with a "soak" charge (350mA for a pack or C/10 for 16 hours for a stick). I take sticks to 1V/cell at 20A, which tapers to 1A, then I take those sticks to 0.2V total at 1A which tapers to 50mA. 15A charge with repeat 20A discharge to 1V/cell show a 20-25% capacity improvement. On HCH2 sticks, this has been true across 5 packs. On 2 packs of HCH1 sticks, the improvement is similar, but only at 10A discharge rates and they have significant capacity still available at a 5A load (800mAh or so). I attribute this to a higher internal resistance inherent in the design as well as age/abuse. Steve

First off S Keith Thank you very much for that very in-depth explanation.

I was considering something else for the "In order to properly assess sticks, you need to be able to subject them to a 90A load for 10 seconds to confirm they can support that load at >5.4V." However after reading what you have stated I agree if someone with experience in this especially as much as you have still suggests grid charging instead of doing each stick I will follow that wisdom.

Next questions for you:

#1 What is a "soak" charge? This is the first time I have read this term in any forum post. If I am reading this correctly that is the first initial charge before the first discharge?

#2 Also you are suggesting in the Hybrid Battery Discharger Dual Bulb to use 500w halogen bulbs? Or just in the single? I figured the dual bulb unit was only $5 more and could still drop down to only using the single 40w bulb as you suggested from 120v to ~0v or in your case 1.8v.

#3 Or would 2x 500w be too much discharging speed and cause battery issues even if I watch it and stop it at 120v?

#4 I am curious about is how you know when I am assuming the whole grid pack here is at the deep discharge rate? Where do you see these measurements? My guess is that meter connected to the battery discharger?

#5 The last paragraph was very confusing for me as a novice to electrical measurements. I was able to follow you are saying you do not multi-cycle that it was just a reference. (Why do you not multi-cycle?) After that point I was lost. Can you explain that in plain english in terms of what you use like when you said the 500w bulb or something?

Sure be nice if their was a youtube video of someone doing a discharge with one of these things explaining it while doing it. I guess it is not that hard and the instructions are provided with the product or somewhere on the site and I have not found it yet. Thank you again for the answers fellow insighters!

 

you will have cell reversals by the time the total pack voltage drops below 100-120V DC.

you can use the ammeter in your DVM and measure the current while discharging into a light bulb to keep track of the current.

if you use two light bulbs in series then the voltage across each bulb is 1/2 and the bulb will not get as hot.

once the stick voltage drops to 5-6V or so i try to keep the current below 200mA to reduce any damages from the cell reversals.

i recommend removing the pack, using the balance charger to fully charge the pack up to full voltage and the doing a discharge on the bench so you can measure the stick voltages as it discharges.

if you discharge the battery into a dummy load like these light bulbs then you can use a wattmeter to measure the total Ah of the discharge to see which stick reaches the 5.4V level first and then you can record the capacity of that stick.

at this point where one stick is totally discharged and the others are still full of charge up around 7+V then i will use a little balancing RC charger called the imaX B6 to pump charge back into the depleted stick so that it is charged up close to the others in voltage.

then i continue the discharge to find the next lowest capacity stick, record the capacity in Ah, recharge it with the imaX and then continue the discharge to find the 3rd lowest capacity stick, and so on until you have found the 4-6 worst sticks in the pack.

i then take all of the sticks out of the pack and tie them in parallel and discharge them in parallel down to .1V so that all the sticks are exposed to identical discharging, to the first order allowed by this procedure.

i then cycle the sticks 2 more times. i use the imaX to charge all of the sticks up to full charge around 8.77-8.80V which usually takes almost 30% more charge than the sticks are rated to hold. the imaX will tell you how much it has pushed into the 20 sticks in parallel. it takes a long time because it is a small charger.

when the sticks have charged up then i discharge them again, down to .1V for a 5S4P pack i make using my wire jumpers. by using 5S of the sticks then the discharge current into several light bulbs can be rapid and since there are 4 sticks in parallel they can be discharged at higher than the 200mA, up to 1A or so initially as the sticks drop below 5.4V for each stick. again, you will find that one group of sticks reaches the 5.4V level before the others even though it has 4 sticks in parallel. this is not an issue now because you are not measuring individual stick capacity.

after the second discharge to .1V is finished, i do one more full charge, then discharge down to .1V of the 5S4P pack again. then i recharge the pack back up to about 30% SOC and put it in the car with the balancing charger leads connected to the pack.

i use a 36V lifepo4 battery charger to charge up my 5S4P pack to 43.8V too so i don't have to use the slow imaX B6 to recharge each of the sticks. the charger pushes 2.1A so it only takes a few hours to recharge the 5S4P to full charge.

you can then use the imaX B^ and connect all the sticks in parallel again to charge it up to full charge to make it balance after charging with the lifepo4 charger too.

pics of my 4P wiring harnesses made from some HF jumper wires i bot for $3.20/pack with a 30% off coupon. the little balancing charger made from the little led drivers with the orange endcaps from dragonsmart, installed next to the battery in my insight, then a second little led driver balancing charger from the all white cased led drivers, and mounted back to back with foam insulating separator and the diode on top visible sticking out to the right.

 

dnmun I am sorry but I am concerned with the information you're providing. Many members have done just what you're saying is going to cause cell reversals by the time the total pack voltage drops below 100-120V, while going as deep as 2v with grid charging. One of which is eq1 who has done stick, grid, and every other kind of charge/discharge cycling I have heard/read on. He has suggested in threads cycles as deep as .25v and S Keith just stated he dropped down to 1.8v so even if I was to only go down to 5v - 2v double cycle I believe it would be safe from cell reversals even if everyones pack is different.

 

You can have cell reversals, well, virtually anywhere in the pack voltage range, depending on how unbalanced the cells are, but more realistically in the 140-150V range. Basically, any full pack discharge-ers should be switching to their low load configuration by about 140V. Not that I think reversals at even the 'high' loads we're using would be that critical, but why tempt fate?. Switch to about 100w or so at 140V for the rest of the discharge...

People have to think a bit about what all this load switching and what-not is all about and then you'll understand. NiMH cells give up the ghost at about 1V, i.e. above about 1V, voltage more or less just hangs while the cell's discharging. But once a cell hits about 1V, voltage plummets fast. What you want to avoid as much as is feasible is having a cell 'drop out' at say 140V with a high current load and then driving current through that dropped-out cell in reverse, at that 'high' current rate. The problem is you most likely have an imbalanced pack - that's why you're working with it. So some cells will be at 1V and about to 'drop out', whereas others will still be humming along at 1.2V or so. That's why you can't just say 'discharge the pack to 120V' (i.e. 1V per cell average) then make the switch to a low load; you have to provide some wiggle room. Panasonic actually has a 'formula' for such a thing:

"Number of Batteries Arranged Serially
1 to 6 (Number of batteries x 1.0) V
7 to 12 (Number of batteries - 1) x 1.2) V"

Well, they have it for up to a 12 series. Pretend you're discharging ten 12 series sticks. 12 minus 1 = 11, 11 X 1.2=13.2V, 13.2V X 10=132V.

It's not perfect, but it's the same idea. If I were really concerned, I'd probably use a formula like (120 X 1.2V) X 'imbalance factor', where 'imbalance factor' has a value ranging from 0.83, little to no imbalance, to 1, bad imbalance... I don't know, I'm just making this up as I go along...

Oh, one more thing: doing full pack deep discharges, you're gonna have cell reversals; it's just the way it has to be. The whole low load thing and when to switch is simply about limiting the amount of current that will be driving a given cell into reversed polarity. I wouldn't worry about it. It just sounds bad, and internet people have made it sound like the kiss of death. The cells just pop back to normal polarity, or at least trickle on back...

 

Cell reversal is a non-issue. Period. Concerns with cell reversals are rooted in NiCD, which can't tolerate reversals. It also applies to CONSUMER NiMH cells that are far less robust that these high end cells used in these packs and can be damaged by reversals.

Mike D has REVERSE CHARGED cells at 15A to -1.8V with no measurable deterioration when compared to cycle/capacity testing results of an identical cell that was not reduced.

Based on my testing efforts including individual cell monitoring when discharging sticks, cell reversals were noted in almost 100% of cases when discharges were taken below 0.9V/cell

I have personally discharged hundreds of sticks to 0.2V total stick voltage at 1A with nothing but positive results equating to 20-25% improvement in capacity at a given discharge rate (typically tested at 20A).

Don't worry about cell reversals. It's a non-issue. You'll find multiple examples of "I forgot to check and the pack drained overnight" with subsequent positive results afterwards. You won't find a single account of "I drained the pack to zero" where a subsequent deterioration was noted.



Steve

 

eq1 and Steve thank you so very much you have help me feel a lot more confident about this. You guys are the best!

 

I performed my first grid charge over the weekend thanks to George V. for letting me borrow his charger.

I used a free iPhone app "Lapse It" to capture a photo of the digital read out every 15 minutes for charting the voltage vs time, attached...