9C-12C NiMh Battery NEW!!!

[xyster]

I find your argument rather academic, Safe, because better lithium cells are no more expensive.

The Li-ion are cheap, but then you have to do the whole BMS thing. Also, the Li-ions only make sense if you are willing to buy $500 or more. (the 1C rating requires a big initial investment) For someone who just wants a pack for $200 that can pull 36V 30 amps for 600 cycles the high discharge 10C cells are more economical. LiFePO is still expensive.

Try to build a Li-ion pack for 36V 30 amps and you need a lot of Li-ion cells.

Xyster... try running your "cheapest price scenerio" for a Li-ion 36V 30 amp pack... include all the chargers, BMS, etc...

The whole point is being cheap, cheap, cheap...

The 36v 30amp li-ion pack can be built with 100 cells (10s10p), cheap single cell chargers, and no BMS for about $500. NiMH, as you know because you were involved in the threads too, requires cell-matching as they roll off the assembly line with a high degree of variability, and each cell has a much higher failure rate. Ebikes.ca is exiting the NiMH pack business for this reason, yPedal and others here have had significant problems with NiMH cells dying and requiring replacement. Lithium does not require a BMS. NiMH does require separation of the strings for parallel charging. As widely noted at rcgroups.com, a single string of NiMH sub-C's pushed to 30 amps will live a very short life -- more like 60 cycles, not 600 cycles. So amortized over its service life, your $200 30 amp NiMH will be far more expensive to maintain than a $500 lithium pack.

 

[Ypedal]

Go with Justins 5C nicads !!!! 30 amps 36V 8Ah NiCad, 5C Max Rate
6.04 kg (13.32 lbs) = 260 $ 2 of these for 16ah would be better.

 

[safe]

Ebikes.ca is exiting the NiMH pack business for this reason....

From ebikes.ca:

http://www.ebikes.ca/store/batteries.php

"We are clearing out old stock in preparation for new NiMH, high-rate NiCad, and lithium packs which will be arriving in coming weeks."

You really didn't read my posting before...

The point is that trying to run the cells at the 10C rate is a waste because the moment the capacity drops just slightly you all of a sudden can't pull the amps anymore. If you cut the rate to 5C then that extends the life to at least 300 cycles. At 10C you can see that your batteries will not last for more than about 60 cycles.

The bottom line is that the Lithium solutions require a bigger initial investment. Once you invest then it becomes economical, but it does have at least a $500 startup cost.

The NiMh solution can be done for $200 and should last for at least 300 cycles... which makes it about the same as my SLA's that weight 86 lbs.

 

[safe]

Go with Justins 5C nicads !!!! 30 amps 36V 8Ah NiCad, 5C Max Rate
6.04 kg (13.32 lbs) = 260 $ 2 of these for 16ah would be better.

I got extremely close to buying their old stock... I still might...

 

[safe]

The Lost Book of Nostradamus.... I mean Safe...

They say a picture is worth a thousand words.

These two charts show the discharge charactoristics of my existing 38 Ah SLA batteries verses these 3.5 Ah NiMh cells.

If you look for 40 amps on the 38 Ah chart (1C) and compare it to 5C on the other chart you get the same visual result. Sometimes numbers give the wrong idea... it's a pattern that you are trying to understand, that every battery has extremes of performance and you want to be in the "sweet spot" for that particular battery.

5C for the NiMh equals the 1C for the SLA...

 

[xyster]

Ebikes.ca is exiting the NiMH pack business for this reason....

From ebikes.ca:

http://www.ebikes.ca/store/batteries.php

"We are clearing out old stock in preparation for new NiMH, high-rate NiCad, and lithium packs which will be arriving in coming weeks."

Great! Last I heard Justin had given up on NiMH because of the high failure rate. Apparently he believes these new cells will be more reliable.

You really didn't read my posting before...

Got me.

I noticed he's also selling raw, 5C NiCd cells:
bottom of the page
http://www.ebikes.ca/store/

Why not try those instead of the already known to be sketchy cheap NiMH subC's?

 

[vanilla ice]

I could buy an old, beater Chevy for $2000 and ask how I might fix it up so that it runs reliably. Or I could buy an old, beater Honda for $2000 that already reliably performs adequately. Unless I just like fixing things as a hobby, why should I bother exploring the former?

Which looks better with green flake paint, 13x8 daytons and a gold plated grill? Which has the front bench that seats four? Which sounds better with dual flowmaster exhausts? Which rear deck can fit the most 6x9 speakers? Easy decision.

 

[safe]

Why not try those instead of the already known to be sketchy cheap NiMH subC's?

The one person that posted a negative comment never replied about which color label they used or anything about the battery history. I just don't think that the major labels can deviate more than a few percentage points on average and get away with it. If they publish their specs you should be able to run thought experiments on them and get fairly reliable results. My SLA's, for example, perform pretty much exactly to the spec I downloaded. Download the NiMh pdf and take a look... they appear to be very precise about their testing methodology and conform to the international regulations.

http://www.all-battery.com/datasheet/sc3600mah.pdf

My guess is that people have tried to take 10C at "face value" and not done their homework to know that for economy you need to run the 10C cells at only 5C. Just because the cells CAN do 10C doesn't mean it's economical to do so. So the problems might have more to do with misuse than problems with manufacturing... they don't claim that at 600 cycles that the capacity would be any more than about 75% which seems right. At 10C discharge rates you would drop below your intended current draw almost immediately and then blame the batteries for your own ignorance.

A conservative estimate would be that at 500 cycles your pack would be at roughly 50% of it's starting capacity. So if you want to make it to 500 cycles you need to set up your machine so that when it pulls at a "relative" full discharge rate it can still pull after the capacity decline. So drop the demand in half and then the weakened by half battery still works up to 500 cycles.

Make sense?

In the beginning the pack draws at a rate that is not very hard for it to do, then as the pack gets older it gets closer and closer to it's natural internal limitation based on wear. When the pack is down to 50% the batteries are no longer usable, but you got the full possible life out of them so you got your money's worth.

 

[xyster]

Why not try those instead of the already known to be sketchy cheap NiMH subC's?

The one person that posted a negative comment never replied about which color label they used or anything about the battery history. I just don't think that the major labels can deviate more than a few percentage points on average and get away with it.

I'm talking about the plethora of reviews of NiMH subC's at rcgroups.com describing cycle life on an order of 10% the spec when drained near the cell's drainrate spec.

 

[safe]

I'm talking about the plethora of reviews of NiMH subC's at rcgroups.com describing cycle life on an order of 10% the spec when drained near the cell's drainrate spec.

And again you're really not getting my point... :?

The "point" was that you DON'T want to get anywhere near the peak drain rate because if your system is set up like that the first slight decline in the capacity due to wear will mean that you have to junk the batteries. It's very likely that after 100 cycles that the cells are down to 90% capacity... that would fit the cycle life chart pretty well. Factor in some extra wear and tear because of the extremely rapid discharge (above the cycle chart decline rate) and it certainly matches expectations.

My "point" (I'm still getting there) is that you watn to set up your machine so that your peak discharge rate FROM THE BIKE is as far away as possible from the peak discharge rate of the battery.

Let's make a little thought experiment...

Assume your BIKE drain rate is 10 amps.

Now assume that you are using two sets of subC's that have a total of 60 cells.

You now have a pack (like the one described previously) that is 36 volts and 7 Ah. The peak discharge rate is 10C so that means it could pull a total of 70 amps... but we don't... we only pull 10 amps so we are running at about 14% of the capacity of the pack.

In theory, you could now go through the entire cycle life of the pack and weaken the batteries all the way down to 14% capacity and still get the same performance that you did on the first cycle.

Is any of this getting through?

I know you have preached the gospel of oversized battery packs before, so this should make perfect sense to you. The only difference is that we are talking about a higher "C" rating... rather than discussing a 1.0C rate for Li-ion we are discussing 10C for NiMh. But we know that for Li-ion it's better to stay down in the 0.5C range if you want to actually get to 1000 cycles... if you use the full discharge all the time is wears out the batteries.

Does it all make sense now? Are we on the same page?

Drain rates are always RELATIVE. Whether you use SLA's, NiMh, or Li-ion it's all the same... stay within conservative drain rates...

 

[xyster]

I'm talking about the plethora of reviews of NiMH subC's at rcgroups.com describing cycle life on an order of 10% the spec when drained near the cell's drainrate spec.

And again you're really not getting my point... :?

The "point" was that you DON'T want to get anywhere near the peak drain rate because if your system is set up like that the first slight decline in the capacity due to wear will mean that you have to junk the batteries.

Then in order to use ~4ah NiMH cells, you have to run separate strings in parallel. Your overarching argument is that a small NiMH pack can be built cheaper than a pack of 18650s because fewer cells are required. Three strings of 30 cells (36v) is 90 cells. 90 NiMH cells at ~$3 each is no less expensive than 90 lithium cells at ~$3 each. Either of which could make a decent ebike pack, but the NiMH cells have that pesky parallel-charging problem, in addition to about twice the weight and the need to buy capacity-matched cells.

Let's make a little thought experiment...

OK, I like those!

Assume your BIKE drain rate is 10 amps.

Now assume that you are using two sets of subC's that have a total of 60 cells....

If the drain rate was only 10 amps, you could likewise easily get by with 60 lithium 18650s in a 10s6p configuration -- each pulling about 0.35C each.

I know you have preached the gospel of oversized battery packs before, so this should make perfect sense to you. ...

Funny. Is a gospel of adaptable reason still a gospel?

Does it all make sense now? Are we on the same page?

In this case, what you say about drainrate and cycle life makes sense, and fits the evidence -- if considered in a vacuum. But your larger argument, espoused on the previous page, is that a pack of small-cell NiMH can be constructed for less than half the cost of small-cell lithium. I don't see it because it takes about the same number of similarly priced cells to obtain the spec'd 300-500 cycle life of either.

If you're talking about a pack to put in parallel with SLA, you've still got the same dynamic: fewer lithium would be required just as fewer NiMH would be required.

 

[safe]

Three strings of 30 cells (36v) is 90 cells. 90 NiMH cells at ~$3 each is no less expensive than 90 lithium cells at ~$3 each. Either of which could make a decent ebike pack.

But wait...

In theory the NiMh can do 10C so the NAIVE builder will try to get away with only one string. It's the "one string error" that ruins their pack. If you go to two strings you are still going to be safely within the performance envelope that you want in order to get good pack life. So if it takes three strings for the Li-ion to survive it only takes two strings of the 10C's to survive. It only takes $161.40 to buy two strings of 30 cells each @ $2.69 with the NiMh.

The 10 amps number was hypothetical, it was to demonstrate the fact that the more lopsided the relationship of peak discharge of the bike to peak discharge of the pack the better. In the real world I'm planning to pull 30 amps at 36 volts and you can do that safely with two strings of NiMh because they will be running at about 5C that way. You would need many strings of Li-ion to do the same.

There is a point of intersection that occurs at $400 - $500 where the Li-ion comes in as a good deal, but at the low introductory level you still have the advantage with the NiMh.

If you want to try again the fact of the matter is I'm building my #003 bike to need a 36v 30Amp "peak drain rate" pack... so far I see that I can build an entry level pack that will have respectable cycle life if I spend $161.40 for the cells plus shipping, plus the chargers, plus... on and on... the usual stuff.

Just use the original 36V 30Amp number... not Ah, but "survivable amps" through the controller.

The range will be terrible... something like 8 miles, but it's enough to get the bike running and I can always add more range later. I'm likely going to have a "long range pack" to put into my battery box location eventually.

 

[fechter]

10C is an unrealistic number for most of the Nimh batteries on the market. If you get some batteries out of a Toyota Prius or a Honda hybrid, then you could peak at 10C. Even 5C is pushing it with most of them. Don't believe their published specs. Find actual test data.

There are some good cells out there, but they're not likely to be the cheapest ones you find on eBay.

A single string would be fine if it had enough capacity, both in terms of amp-hours and in terms of maximum discharge rate. Even the Toyota batteries are small in the amp-hour department. There must be a reason they don't make them in larger sizes.

Using the Toyota approach, you could just double the voltage and cut the current in half to get the same power. 60 cells in series or something like that. Of course this would take a different motor and controller, but in therory it would work.

 

[Johnbear]

The last 3C 5ah (10AH Total) in parallell (30AMP Max) NiCads I bought from ebikes.ca are running on a 20 amp controller. They have about 20 cycles on them and they run about 6.5 ah on the drain brain and they are toast. Not impressive. Maybe the 5c ones are better. Any one try them?

 

[safe]

The last 3C 5ah (10AH Total) in parallell (30AMP Max) NiCads I bought from ebikes.ca are running on a 20 amp controller. They have about 20 cycles on them and they run about 6.5 ah on the drain brain and they are toast. Not impressive. Maybe the 5c ones are better. Any one try them?

NiCads have the "memory effect"... are you sure you aren't dealing with that as a problem?

You also might have a single cell that went bad... another reason to build packs with tubes is so you can remove the cells and test them. All it takes is ONE bad cell and the whole pack goes bad. The whole idea of just buying a pack and blindly using it (and expecting it to work at the full 3C) seems wrong... you need to design with a 50% rule of keeping the current draw at 50% of peak. So in your case your peak draw should be 15 amps. However, you are having a problem that doesn't sound like wearing out due to cycle life... you have some sort of serious problem than MIGHT be repairable if you took the pack apart.


How to restore and prolong nickel-based batteries
Crystalline formation is most pronounced if a nickel-based battery is left in the charger for days, or if repeatedly recharged without a periodic full discharge. Since most applications do not use all energy before recharge, a periodic discharge to 1 volt per cell (known as exercise) is essential to prevent memory.

Nickel-cadmium in regular use and on standby mode (sitting in a charger for operational readiness) should be exercised once per month. Between these monthly exercise cycles, no further service is needed. No scientific research is available on the optimal exercise requirements of nickel-metal-hydride. Based on the reduced crystalline buildup, applying a full discharge once every three months appears right. Because of the shorter cycle life compared to nickel-cadmium, over-exercising is not recommended.

Exercise and Recondition - Research has shown that the crystals ingrain themselves if no exercise is applied to nickel-cadmium for three months or more. A full restoration with exercise becomes more difficult the longer service is withheld. In advanced cases 'recondition' is required.

Recondition is a slow, secondary discharge applied below the 1 volt/cell threshold. During this process, the current must be kept low to minimize cell reversal. Nickel-cadmium can tolerate a small amount of cell reversal but caution must be applied to stay within the allowable current limit.

Tests performed by the US Army have shown that a nickel-cadmium cell needs to be discharged to at least 0.6V to effectively break up the more resistant crystalline formation. Figure 2 illustrates the battery voltage during a discharge to 1V/cell, followed by the secondary discharge to 0.4V/cell.

http://www.batteryuniversity.com/parttwo-33.htm

 

[magudaman]

Safe it doesn't seem that you are really reading what people are posting here. It seems that you are really stuck on the nickel chemistry idea. You would think that with your advanced bike designs you would be trying the new advanced lithium chemistries to be on par. These will have higher discharge rates, lighter weights and more cycles than any other chemistry on the market (lifepo4 comparison). In the end the little money you save from going with dying nickel or lead could have been spent on a pack that in the end would save you money in the long run. And if you over design a lifepo4 pack it will last even longer than when run at ratted specs and 100% DOD. Less depth and less C rate discharge and charge you could see well over 2000 cycles.

You also worry about BMS with lithium and charging. Well I find charging Nimh much less certain than lithium. Lithium has a nice CV curve to know exactly when it is balanced and full but nickel has a crazy peak detection method that has a lot more variables. And balancing requires cooking the pack at a low rate endlessly so that the full cells over charge and the out of balance finally come up to charge. I just prefer the certainty of knowing the battery is charged and balanced every time I ride, not worrying of false peaks and imbalanced cells.

 

[safe]

Well it all comes down to how much you want to spend.

I've done a lot of cost conscious calculations and when you consider the start up costs the winner is the Sub C NiMh cells. Trust me, the moment the costs of the newer battery options come down to equal or surpass the lower cost solutions I'll switch to them. Even SLA is a cost effective solution at the low end. (you have to do like I did and carry a lot of it though... can you imagine 3000 miles for $138?)

When startup price is not factored into the equation (meaning the idea of spending $700 or more for batteries doesn't make you wince) then the other chemistries are better.

I'm planning to spend about $300. (Sub C's can give you 475 Wh for only $264)

Money changes everything...

 

[TylerDurden]

Well it all comes down to how much you want to spend.

Plus what you want in a bike.


If you want a cheap electric motorcycle that you don't pedal, lead may be a fair choice.

If you want an electric bicycle that does what all bicycles do and has electric power too, lead sucks the big one.

NiMh ain't worth the $$ IMO, NiCd would be better over time.

 

[bobmcree]

i was perusing the various threads, as it is raining and i have read all the books i have. it seems some people have overlooked the simple solution to parallel NiMH packs that I have been using and recommending for years - diodes. I got about 200 cycles out of the last batteryspace D cells i was using on my bike, only because my homebuilt charger fried them did they finally die. (i now put a fail-safe one time thermal fuse in all my packs to prevent this).

these less expensive NiMH cells tend to have higher impedance, so they exhibit a lot of voltage sag at ebike currents. putting 2 packs in parallel really reduces the problem, and with good schottky diodes combining the strings i can charge n strings with n charger channels without unplugging anything. If a cell fails in one string, the other takes the load until it sags to the voltage of the weak string, and they share.

one of the things people tend to ignore in batteries is that power loss in the battery goes up as the square of current. i have two strings of 40 10 Ah Powerizer D cells on my Trek. each cell has an impedance of about 7 milliohms for .280 ohms total, and at 30A that is 900 * .280 for over 250W of power dissipated as heat inside the batteries. if i parallel two of those strings the impedance is cut in half, but the current in each is also cut in half, so each string loses 15A squared * .280 or 63 watts in each string. I have thus cut my power loss in half and am operating the batteries where they will have much better cycle life.

i typically use 40CPQ060 dual schottky diodes to parallel packs less than 60v and they have a forward drop of only .55V so in a 40 cell string you are throwing away about 1% to get the safety and convenience gained.

i used to race RC trucks and bought matched 3700 maH 6 cell subC packs for up to $100 to get a couple of seconds edge. i discharged these cells at up to 35A or close to 10C, and they started to be useless for racing after 50 cycles or so. these high rates kill batteries quickly, but it is fun...

as for NiMh being difficult to charge, that is not my experience. there is a very pronounced increase in the rate of cell heating that is commonly used to terminate charging. I have been running my 20 Ah Nexcell DD pack for a couple of hundred cycles, often discharging as low as 90%, and have never needed to balance them at all. i will probably do it this week since i discharged them yesterday to 20.5 Ah, but the lowest voltage i saw at 35A was 39v and there is no sign of cell reversal.

I know Justin had a big problem with the Nexcell D cells delivering low capacity, and he is closing those out at heavy discounts now. We have both had more problems with bad welds and other assembly problems than with the batteries themselves. The main reason he is getting out of the NiMH business is that the price has gone up so much and other emerging technologies are becoming competitive.

 

[Ypedal]

My biggest beef with NI was the self discharge, and the heat during charge-discharge.

A lithium pack has 10 cells instead of 30 for NI, lower probability of problems, and they can be charged right away after a ride.. and used right away after charging... without any cool-down period.

 

[Johnbear]

Batteries I bought in July 2007 30x NiCad 5ah capacity

 

[Johnbear]

I also pedal alot when I ride. I rarely hit above 15 amps on my 20 amp controller. I can ride around 35 km on 5-7 amp/hr.

 

[Johnbear]

Thanks for the advice, I will discharge the batteries to 1v per cell and charge them. We'll see how they fare then.