My NiMH -> LiPo battery rebuild for i167. Success!
 

It's been a while since I posted, and indeed, since I rode my Segway. Hi y'all!

I've purchased six $400 NiMH packs over the years, and continue to disparage the battery design as the weak link in Segway -- certainly from an operating cost standpoint. (If anyone of the 60 NiMH cells falls out of balance and goes bad, the whole pack is junk..) With the batteries discontinued and nothing to lose, I thought I'd try building my own pack. This is the result:

http://lh5.ggpht.com/_qdk3lsNat6E/TR...25718-2477.JPG

For now, I'm using six $9 3S LiPo batteries, each 2200mAh and nominally 11.1v. It's a fit! The pack is much lighter, and I remove the packs to charge them on a proper balancing charger, but my Segway is road-worthy once again!

http://lh4.ggpht.com/_qdk3lsNat6E/TR...32312-2471.JPG

Of course, I can't use either of my Segway battery chargers to top off, and a short circuit would burn down the whole show, but at $100 all-up, I'm in business! Most importantly, when one cell fails, I can diagnose & service just that pack without discarding the rest.

With a little more Dremel work and tighter wiring, I think I can fit 5000mAh cells or larger into the same space.

Well done! This is a really interesting piece of work! Do post how it performs and what sort of range you are seeing. I cannot remember off the top of my head what the capacity of the standard NiMH and lithium packs is. How big a charger do you need for such a pack? Ideally you need to have an external charging socket on each pack so you can charge without dismantling. While clearly this isn't a perfect solution, given that you cannot fly with the lithium packs anyway, why not go this route, even though it is a bit 'quick and dirty'.

Driving around on my second set of lithiums, now fading into senescence after three years, I definitely agree with you that battery costs are the elephant in the room when considering the running costs of a seg. This may not be the perfect solution but it definitely will move the level of knowledge forward.

Nice hack!

How are you limiting the voltage so it doesn't go below less than 3 volts per cell?

How does the segway deal with it? I thought there were some cross checking circuits that monitor the battery. Is it satisfied? Does the battery indicator work on the display? What kind of range do you get?

This is a very interesting mod. I realize that you are taking out some of the safety levels, but also are dropping the price hugely. I would not do this for others, but as you are doing it for self, good luck.

Keep us posted.

All good points. This is not a fool-proof solution, and has a lot of "manual safety" for now.

The Segway seems satisfied with the battery. (I still have the original 12.0 no-lipo-support software.) I can't see how it would even know. A 11.1v 3S pack has replaced each of the 10S NiMH packs that was there, so the operating voltage is within acceptable range. I kept all of the original NiMH circuitry and temperature sensors. I'll keep an eye on it, but my suspicion is that it will actually be a little bit pessimistic. If it allows NiMH to get down to 1.0 volts per cell (10v a string), then the LiPo is safely within limits, and limiting might kick in a bit early. If it allows 0.8volts per cell (safe for NiMH), then manual intervention is required.

For now, I have a self-contained LiPo monitor (pictured) installed on one or two packs. It will sound an alarm if any cell goes below 3.1v.

Charging is a bit of a hassle, since all 9 cells on each side need to be balanced. I have a Junsi iCharger 3010B with 1000W output, and can charge half the pack (9 cells) in 30 minutes or so (2C), twice per battery pack, four times total.

I have not yet measured range, but I don't expect it to be impressive. The packs are 2200mAh, and I will be limiting discharge to 70% or so. (The NiMH cells were SAFT 4.5Ah) That's still much further than the dead packs got me!

It's all experimental at this point. It would not be expensive ($20 or so) to add pack-monitor/alarms to each battery, so that's likely the next approach, just to make sure no one cell drops below limits.

External taps for charging would be nice, so I don't have to keep taking the screws out.

I also have to tape the 110v AC charging port shut, so nobody carelessly plugs in a power cord and blows up the cells with overvoltage.

Wet weather use is right out, since it's fully open on top.


R/C LiPo batteries have plummeted in recent years. I paid $8.00 each for those 3S packs at HobbyKing.com. 5000mAh packs are available for $22.00 or so, complete with hardcase. If you're willing to hog out most of the battery shell, there is a LOT of room in there. Range might also improve just because the batteries are so much lighter than NiMH. The whole Segway feels lighter. I'll have to measure it.

If anyone knows the pinout/protocol of the battery pins, I sure would like to understand this link better.

Here's the link to the $8 LiPo pack:
http://www.hobbyking.com/hobbyking/s...idProduct=6306

The 45A Anderson PowerPole couplers cost half as much as the battery. :)

Hobbyking has connectors too you know? ;)

Yeah, I know.. I do like the XT60 connectors, but it seems like HK switches to a new connector every year.

I just ordered another 12 Zippy 3S packs, and XT60 pigtails, so the next battery overhaul should go faster.

$67 in shipping.. ouch.

That's a bit steep. I've paid quite a bit, but never that much through them!

Maybe you could put a bead of silicon around the outside top of the battery housing to act as a seal?

Have you seen these?

http://www.maxamps.com/Lipo-MV5250-Pack.htm

You could fit a crap load of them in there!

I have a couple, but only because I won them!

Can you post a close up of the main connector and the buss which all the lipo's are connected too?

Cool
 

I like the DEKA reference on the inside case of the batteries in the opening photos.

The DEKA batteries have an odd story. I found two i167 NiMH batteries on eBay back in 2006.
They arrived with none of the standard serial numbers or markings, just "DEKA" stickers. They were also mostly-shot, of course. :( I never got more than a mile out of them. The seller went mute, of course.

I don't think they were commercial Segway product, but some kind of DEKA pre-production battery. Who knows.

The standard batteries have ID stickers on each pack of 10 cells. These had none.


Does anyone know WTF that GLUE is that Segway uses in the battery packs?? It's insane.. I call it Demon Snot, but I'd love to know what it sells as. I haven't found any solvent that releases it easily. I'm just using a heat gun to soften it up in the extreme, so I can apply crazy amounts of force to lift the old NiMH packs out. With enough swearing, it works.

I have no intention of sealing these LiPo packs, since I don't ride the in the rain, and I have to take each of the packs out for charging and balancing anyway.

I wonder if you could incorporate a mechanical A-B switch kind of device, so that you could have one lead that comes out of a future iteration of this battery system. OF course, it would have to be an A-B-C-D-E switch, but you could plug in your charger, then just flip to another letter every couple hours.

It may not be too hard to incorporate a slide or rotary switch into the pack, and have just the knob (or a slot for a screwdriver) thru the case of the battery, which will allow you to put the covers back on.

This is not so much for gliding in the rain, but because this is a lot of power to have uncovered, since the road is full of all sorts of hazards, including puddles, and other potential hazards.

It's true, Karl. I'd be better off with a sealed pack (or mostly sealed) even in dry weather.

I think team222badbrad was getting at a similar idea -- discharging in series, and then switching over to a parallel arrangement for one-tap charging.
If the four series "lobes" of the Segway power are discharged to the same level, I will look into doing that..

Charging as a 9S4P configuration, or even a 3S12P would be much easier to wire externally.

I checked the in-battery circuit board, and there are taps between each 12v module, and a whole lot of digital circuitry in there, so I don't think I can pull one over on the Segway by wiring more directly.

Still, ANY miles is better than zero miles and a BEEP BEEP BEEP "get off" going uphill.

Another update -- I had to build TWO conversion packs in order to ride. (Alas, I have not a single viable NiMH pack to ride with.)

So far, so good! These cheap 3S battery packs are not a perfect fit to Segway riding -- there are complications.

First, the overall loaded voltage of 9S LiPo is slightly lower than what the Segway wants for NiMH. So right off the bat, with charged batteries, the Segway goes to 3 bars (half) charge indicator. After a couple miles, it's down to 2 bars.

The good news is that this seems to bother it not at all -- it stays on 2 bars for a long time. Without more detailed voltage monitoring, I'm not going to find the limit, but it's good enough to call the Segway serviceable once again.

The 2200mAh 20C packs are probably borderline for hillclimbing or demanding rides. I rode full-speed up some moderate hills, and the packs got just a little warm to the touch -- still below body temperature. I wouldn't try sustained extreme climbs in hot weather, since I effectively have no temperature monitoring on these, and no airflow.

Charging turns out to be easier than I thought.. I just bus all six of the packs in PARALLEL for charging, since they're all discharged the same amount. I bus all the balance taps as well. No problem. The charger treats it as a 3S 13.2 Amp-hour pack, charges at 25 Amps and has them fully done in 20 minutes, balanced.

I removed one of the circuit boards from a junk battery pack, and it's far more complicated than I thought. I was hoping to adjust its notion of "charged" to match the min/max of LiPo, but forget it.

Lipo battery hack
 

Greetings fellow San Diegan,

You are to be complimented on your DIY flair, just watch those hobby cells so that they don't add flare to flair ;-)

Have you thought to order up a single 2200mAh cell to add to your 66.6V string? I think that this voltage will still be compatible with the 12.0 firmware, but I'd like to here from you on this.

How much range are you getting?

Keep us posted, please.

Jim

Since the batteries attach to the BMS in 12V clusters he would have to add cells in multiples of six. Unbalanced battery clusters would result in reduced performance if a single cell were added. We have done the same thing with small SLA and LiFePO4 batteries. It can be done but the like always the OEM charger is not compatible.

No BMS here
 

Hi Jason,

Good to hear from you.

These are 3.7V cells, so I am confused by your suggestion that they would go together in groups of 6 to form 12V clusters. Please clarify for me, if you would.

I believe he is not using the stock BMS board at all, but rather charging in parallel and balancing with a hobby charger.

Thanks,

Jim

Correct. I dare not use the Magnetek external or the Segway internal chargers with these batteries -- I'm certainly it would blow them up with a NiMH charge profile.

I'm removing the battery cases from the Segway, and then removing the 3S Lipo's from the batteries to charge on a bench... that's 12 3S batteries, for anyone keeping track. :)


As you know, the Segway effectively has four strings of NiMH batteries, two in each battery pack. Each string is made up of three 10-cell strings.
So for one voltage rail the Segway has 10+10+10, four times.
It *does* monitor the voltage of each of those 10-cell strings, so I can't spoof it by replacing all three with one large battery pack.

Instead, I have 3S + 3S + 3S in LiPo, each one replacing a string of 10 NiMH.

I would very much like to try "3S + 4S + 3S" to see if the Segway complains about inequality, or abuses the 4S pack, or just corrects the battery gauge.

Given the discharge curve of LiPo, I'm not all that concerned with the low battery bars, since it will just park at "two bars" for most of the useable range.

First, this is what the PCB inside the battery looks like:
http://lh5.ggpht.com/_qdk3lsNat6E/TS...11740-3896.JPG

It's quite a bit more than I was expecting, and dashed any hopes I had of a full roll-my-own battery pack. Wow.

Notice the steel battery tabs at the top, which are bent over the top of the PCB. For my second battery pack conversion, I thought I'd try to be cleaner and replace the battery tabs entirely, by soldering the 12ga wire to the PCB.

That wasn't an effort I'd like to repeat. It required drilling holes in the back of the battery just to get to the solder tabs:
http://lh5.ggpht.com/_qdk3lsNat6E/TS...10100-3891.jpg

The resulting battery looked fine, and had fewer sharp edges, but was a more strained fit with the large wires coming over the top:
http://lh5.ggpht.com/_qdk3lsNat6E/TS...10049-3888.JPG

And the Zippy/HobbyKing XT60 connectors are actually pretty awesome, especially since they come on the batteries already. HK sells the female connectors, with leads already soldered, so this is pure Win. I like them.

I think I'm going to make a third assembly, using XT60 connectors (again) and soldering to the cut-off battery tabs. Then I'll have all 12 batteries on the same connector, which would be nice for massively-parallel charging.

If I was doing this over again, and not so bent on using cheap $8 batteries, I think 4000mAh or even 5000mAh of the right dimensions would be the way to go.

I'd like to find a way to bring the required leads out of the Segway for charging without so much screwdriver time, too.

Roll your own battery pack?
 

My confusion, here. Ground loop was talking about NiMH, and I had convinced myself that Safions were LiPo, when it seems they are LiFePO4. The hobby packs are made from stacks of prismatic cells. Adding another identical cell in series with those 1P packs should not unbalance anything. I'd try carefully to open up and remove a cell from one of the $8 packs (you did buy extras?)


I have only taken apart the Lithium pack. Thanks for the details and the photos. You might spoof it effectively if you were to feed it N fake voltage signals, or copies of one rescaled voltage value for the whole pack. If the goal is to keep your lithiums from undervolting during a glide, the stock NIMH circuit may not be calibrated to save an 11.1v cluster if it expects a 12V cluster as I think Jason wrote.

If the Seg's undervolt threshold is 0.9v per cell, then it will stop you (presumable shake stick and shut down) at 7.2v per cluster. Lithium Polymer cells, I've read, have a cut off at 2.5v per cell or 7.5v per hobby pack. If the Seg's cutoff is 1v, then you are fine. May I suggest, if you haven't seen them, inexpensive audible alarm mini-boards for your packs. Some can be seen in the accessories section at Hobby shack.

Do you or anyone else know what info and in what format the Segway control board "must"get from the battery control board? Aside from battery power, I only saw four pins leaving the battery housing; anyone know what they have on them? If we have clusters of 12v, then these pins can't be sense pins for the 6 clusters. Please don't tell me there is a CAN channel here. I'd just as soon be rid of the batt control board if it isn't performing a charging function, and the main board can be convinced to do without it.


I think the right number of 3.7v cells in series for the 12.0 firmware is likely 19, not 18 and not 20. Prime numbers can be inconvenient ;-)

If you build a third assembly, let me offer you use of my Fein cutting tool and help you part the top at the original weld line. You should be able to get at the old charge board this way. I have been searching for hobby-priced cells that will fit in a larger Lithium Segway battery shell I recently opened. 10Ah per shell may be possible - 1.5kWh per Glide!

Either way, I have an electric bike project and want to buy some of the $8 packs, so, If you want to combine shipping, let's talk. I also have a battery test unit and some data logging capabilities in my shop in El Cajon, in case you are interested.

How do you intend to integrate 3.7V nominal cells into the BMS? The BMS is setup for a 23S configuration on a 3.2V nominal cell. If you change to a 19S config with 3.7V nominal cell there are many hurdles to overcome just to get the Segway to communicate with the battery and turn on. PM me if you want a second opinion on concepts. I'll help out if I can.

Thanks for the kind offers, L3Research and Jason.

I think you might be giving me too much credit here -- I'm not messing with the BMS, charger, or anything complex.

I'm just substituting 9S LiPo (33.3v nominal) for the 30S NiMH (36v). All the circuitry, and as much of the case as I can retain, stay exactly the same.

The 9S LiPo will get down to 28v safely, 27 is probably okay if they're balanced.
So the question is, how does the Segway allow the NiMH to get under 1v per cell? 0.9v?

The other thing to watch out for is starting with fully charged LiPo and then going on a long descent.. the Seg would probably overcharge them with regen power. Who knows; I'm not going to try it.

I'm aware of what you did and appreciate the simple yet effective approach. Certainly it worked and there are limitations on how much you get out of it due to the lower nominal voltage. I was actually referring to L3Research when he said about a 19 cell configuration for the lithium ion packs. It's a totally different animal than the NiMH and cannot be fooled so easily.

I made another one, and I like this one best:
https://lh4.googleusercontent.com/_q...50516-2511.JPG
I left the middle strip 'mohawk' of plastic intact this time, to retain the battery connector and add some structural rigidity.

Now I'm all matched batteries (twelve 3S batteries), XT60 connectors, and cleaner wiring.

I also got a chance to run the Segway under more conditions with just LiPo power. First, the full performance envelope is available. With my resuscitated NiMH batteries, they would sag under load and get the straight-face slowdown.

I also added some battery monitoring of my own to track performance in motion.

I rode the Segway with a young passenger to bring it to 200lbs load, up and down hills at full speed and around the block. I measured a peak draw from each battery of 129 Watts. That works out to 11.4 Amps at peak load.

I made a second ride with 170lbs load and worked it as hard as I could. Rapid starts, rapid stops, full-speed turns, inclines, etc. The highest I could hit was 192 Watts, 17.5 Amps.

Overall, this puts the whole battery array at 2300 Watts peak.

As most of you predicted, battery cutoff became an issue. The Segway went to blinking-red when the packs reached 11.0 Volts. At this point, I had used only 0.58 Ah out of 2.2 Ah of capacity, so that's way too early.

None of the 3S batteries dipped below 10.97 volts at any time. (They're rated for 45A or so, so this is a cakewalk). Temperatures stayed slightly above ambient.

So what does all this data show? I believe the batteries are well-matched to this application, operating well within both average and peak load demands.
However, 11v per array is clearly the "zero bars" cutoff for the NiMH Battery monitor, so I'm going to have to experiment with added cells.

Next I will try 3S+4S+3S and see if the segway is monitoring each 10-NiMH string individually, or the one-of-four assembly as a whole. Based on the complex monitoring circuitry, I suspect it will not fall for this, but who knows.

If I had it to do over again, I would go with 4S (14.8v) packs all around. At 45v peak, that's still within the 'float' charge you might find on 30S NiMH, so I doubt it would fry anything. Anyone experiment in this direction? Does anyone know the resting voltage of the Segway/Valence Lithium packs?

Have you considered using prismatic LiFePO4 cells instead of Polymer so you end up with a 12.8V nominal and a safer battery chemistry?

Nominal voltage on Valence pack is 73.8V but thats irrelevant since your using a NiMH BMS and not a Lithium BMS.

Thanks MTO.
I just asked about LiIon voltage since I was thinking of feeding my segway 45v on each rail instead of the nominal 36. Seems like it will be fine.

The prismatic LiFePO4 or even A123 cells look sweet, but this is a low-budget operation I'm running over here. :)

Does anyone know if the Segway will run if it gets Power on the correct pins, but no connection on the other pins? It's a longshot. If I could build a battery spoofer that lies about voltage, that might also work. I'd like it to run down to 30v on each rail, to keep using the 9S LiPo arrangement.

http://www.hobbyking.com/hobbyking/s...dProduct=16575

:D

4S will give you more than 45V on each side when fully charged. Fully charged that pack should be reading ~16.8V per cluster which would be ~50.4V per half or 100.8V per battery. I would be concerned about voltage ratings on PCB components (capacitors, voltage regulators, etc.) inside the Segway as they are likely not rated for that type of voltage.

The Segway will not run on just power. It must have communication with the BMS in order to operate.

That's why I was asking about the Valance pack voltages.
If my i167 can be software-upgraded to work with the Valance batteries, then surely the components are cool with the ~50v range, no?

I had no idea HK carried A123 cells. Hm! And at $8/ea, cheaper than EPBuddy.com, where I've been buying them.

I'd like to just bring the power leads out and go to a tethered battery of some sort. Maybe a pack that mounts to the shaft or something. Tired of pulling all these screws out. :)

~84V is about the max the Valence packs achieve.

Hi , just signed up so I a newbie. What a great forum ! Back to topic, I want to ask how the Segway knows it is a NiMH I have a Lion pack and want to modify it to Lipo (I am very familliar with Lipos) without all the feedback the segway wants . Does anyone have a pinout for the NiMH packs.
Thanks

replacement NiMH batteries
 

Hi, as I am in need of replacing my 167 original NiMH batteries and am not enthusiastic about spending a fortune., I found this post looking for info about rebuilding my packs.

I know this post is old, but hoped people still check it...

For simplicity sake, I was thinking of just replacing the current NiMH cells with new NIMH cells. I was considering using these pre built 10 cell packs...

(I can't post links as I'm a newbie)
all-battery.com/rectangularnimh12v4500mahhighdrainbatterypackwithb areleadscustomize.aspx

can anyone offer an opinion as to whether they would be a problem?

thank you,

Jason

would recommend MTO battery's rebuild pack, it's design for Segway, with proper insulation and vibration damper, also it is using much thicker nickle plates to handle high current under hill climb and steep decent.

your link with the 10 pack Ni-Mh might work but really not recommended.

also with the plastic package, the battery temperature sensor will not work properly, and will send false signal to the BMS, causing the battery to under charge and might enter emergency shut down mode during hill climb and decent.

that cable is not heavy enough to carry the current either, it will get very hot on heavy load causing the insulation to melt.

I'm not that far away and can certainly help you. Please review our website at www.mtobattery.com and you may find what you need. If you still have questions please feel free to email us at sales@mtobattery.com or call me directly at 717-751-2705 and I'll do my best to help you make your decision on how to proceed.

Thank you for the reply... those are the kinds of things that I worry about - obviously... I don't want battery melt-downs or Segway shut-downs when I least expect it. I don't use the Segway nearly as much as I "should" so to spend a lot on batteries is nearly unjustifiable, but a dead Segway is no use either.

That site also sells individual Cells but I figured the pre-built packs would be better than what I could attempt to "Remake." In the interest of safety, I will look at MTO.

Thank you,

Jason

Hello everybody, I have just got a Segway from my friend, but unlucky for me is the Segway's battery does not work anymore. I try replace my battery with other about 80V and I got new problem: when I plug key I saw the red led flashing . Anyone can help me. New battery too expensive for me. Thanks .

Your problem is not new and the problem is explained in your last sentence.
May God bless your Main Board.


sent from my iPhone using Tapatalk APP

I am no battery expert, but there is a lot of communication between the segway and the battery, with most of the charging circuit in the battery itself, and mostly just the power supply in the segway.

Because of this, I do not believe you can get a LiIon battery to act like a NiMH battery. It is much more complicated than that.

Again, because of all the voltage and current sensing that goes on between the battery and the segway main boards, I do not believe you can do what you said you planned to do in your post.

If you deny the segway all the feedback it wants, as you put it, the segway will just stop operating, or may not even turn on.

Good luck. Keep reading. Talk to the people who really know, instead of the rest of us who make believe we know. There are a few names of good folks here who will come up when you read, that offer good advise, and data.

When you finally get to it, check in with segway or MTO. In the long run, or the short run, eventually you will need the support of either of those two when it comes to replacement of you batteries.

Hi Jason,

Just a thought. Talk to some Segway dealers about used batteries. I personally know of some trade-ins that still took a charge but didn't get the distance needed. Better than buying new or blowing up your Segway! :)

Good luck!