r/r + Shorai battery= epic fail, SHORAI POSTS 3-24
 

Took the bike out for a run today. I actually measured voltage before I took off as the battery is new and I wanted to make sure everything was in order. I noted a charge rate of about 14.2 volts @4K rpm. Anyway, I was blasting down the road and everything was fine, till the bike cut out and died. I coasted to the edge of the road to find this:

Attachment 26359

I managed to get the battery out with very little problem, but in the course of all the look-over determined that the R/R was fried as well. I am still unsure as to which caused which (that investigation is still underway).



I am tending toward a failed r/r, though, as there was some overvoltage damage to the headlights as well (the only incandescent bulbs on the bike). Most everything made out ok, though with only minimal fire damage.
Anyway, I have all the new parts ordered, but though you all would appreciate some good bike carnage!

Ouch!!

I've been thinking about getting a Shorai. Do you intend to put a new one in or go to something else?

I hate being the one to say "Told you so", but I have said this before... To all considering trying an expensive battery of the LiFeP4 type, whatever the maker, or DIY... The stock R/R just won't cut it... Doesn't matter if you measure it and it works just fine... It simply will not cut it... Swap it out for a known good Mosfet type R/R or stick to SLA/GEL type batteries that does have a larger margin for error... (Not something I usually consider a good thing, but it keeps things like this from happening)

This type of battery can soak up a lot more, and a marginal R/R will fail once you stick such a battery on, and if you are unlucky it fails in "overvoltage mode" like yours... With decidedly expensive results...

True enough! I ordered the R1 R/R last night, but am unsure on my next battery selection. I am still weighing options there!

I guess I count as "unlucky"!:D

Probably since the odds of it failing in "undervoltage mode" is actually higher (depends on how the diode grid is set up), and that would have avoided letting the magic smoke out... Then you would just have needed a charger and a new R/R to be up and running again...

Sorry about the damage, but glad you didn't incur any similar meltdown/damage to your body! That battery pic is hard to look at!

+1 !

Well one thing this tells us is that it doesn't use normal LiFePo cells.. Some kind of plates.

Sorry you had the meltdown, but perhaps we can all learn from it.

Well, so does all the LiFePo's... In the 26650 (cell size) configuration they are just rolled up... What's inside the Shorai is most likely primatic cells... Ie the same "sheet", but just not rolled up...

See already learning stuff..... but sorry to see this, just glad it didn't cause more damage.

That's actually what I was trying to do... open it up so we could see what was inside!:D

I am waiting to hear back from Shorai now. Not sure if much will come of that, though. Got the bike mostly sorted, just waiting on parts, now...

On a lighter note, I was contacted by Shorai today. They want to know how much it will cost for the entire repair of my bike, my labor time included, and are going to reimburse me for that amount. The president of Shorai is supposed to be in contact with me by the end of the week, to help figure what went wrong. Evidently, in the event of a failure, these batteries are supposed to be protected to prevent this sort of thing happening. So if this all happens, I may just be part of the compensated R&D, and may come out the better yet! I am trying to be positive, and so far, my experience with Shorai has been positive. Stay tuned for the outcome!!

So are you selling that battery in the picture at deep discounts? If they pay for the repair you realize how truly rare it is that that happens. Very fortunate (if you dont count the burnt bike part). Plus you weren't hurt which is always good.

I am interested to see what happens on this front: perhaps if it works for their R&D you could fill them in on the nuances of an older Honda R/R. It may help them narrow the problem down...

I am impressed that he took the time to snap a pic of the smoke pouring from the bike.

They are in fact prismatic cells.

I disassembled my R/R tonight and found solder melted all around within the bedding material. I am still unsure as to why this failure occurred with this battery change, though. I tend to thing that it has to do with the different charge characteristics of the Lipo4, but am not sure exactly what. Tweety, perhaps you would be able to shed some light on that, as was unable to determine the exact R/R wiring due to its severely melted state. I was hoping to draw a schematic of the R/R and thereby determine if the battery could have caused this, or if it was coincidentally the R/R's time to go. I seems that the deep cycle nature of Lipo4 may figure into this as well.

I will have my R1 R/R later this week, so I should be able to tell more then. I tend to think I may try the Shorai again, though.

Well... It's a diode based Rectifier and a Thyristor based Regulator stuck together, so it's simple as mud...

Couldn't find a decent image so I made a schematic myself... Lemme know if it's incomprehensible...

Attachment 11146

Oh, BTW... The Odd looking diode with the extra leg on it is the Thryistor...

Drawing 3 phase AC to DC was a PITA though, so here I managed to find an image...

http://upload.wikimedia.org/wikipedi...fication_2.png

The top is what comes in from the stator, the bottom is what the rectifier (the diode grid I drew in the previous post) puts out to the Regulator...

Basically it takes AC that has a peak at around 70 V (-35 to +35) and makes it DC... So then the Regulator gets rougly 35V or normally around 30V with some losses... Note that this is peak, but at no point does the voltage drop below the 16-17V volt reguired to go through the Regulator... That is unless things go wrong...

It then converts all above the required voltage into heat using the heatsink... On the stock R/R that's the bike's frame... On the "upgraded" Honda R/R and the other type that's the fins...

What happens when things go wrong is that either the diode breaks like a fuse, ie no voltage coming through either way... This is the preffered way since the result will now be a small dip in that ripple wave in the lower image up top...

Or, the not funny way... The diode decides it want's to have fun, fun, fun with your 'lectronics and start pumping voltage either way, just to piss you off... End result, part time it lowers the voltage by negating the wave that should have gone through, ie a dip like above... And part time you get the fun effect of two waves added on top of each other, giving you double the fun...

Now, the Regulator that is supposed to keep fun to a minimum, but is a very, very stupid circuit slams the door shut when there is a too high voltage peak... The thing is every time Mr Big Musle does that he gets warm and sweaty... And after a while he gets heatstroke... So then the door is open... End result... Smoke...

BTW This part, the Rectifier is the same in the Diode/MOSFET and Diode/Thyristor based versions... Or in principal at least...

It's the Regulator part that's different... The MOSFET's are much more efficient than Tyristors for regulation, so they create less heat, have less basic variation and those two effects combine since the thyristor varies more the hotter it gets...

The thing is, the Regulator is what kills the Rectifier, and the Rectifier is what kills the Regulator... And both are done in by heat...

The Regulator gets warm from disposing of excess voltage, and that heat makes the component in the complete package including the black plastic mass it's sealed in age... The diodes age, the mass gets old and cracked... If you are really unlucky moisture gets in, and all hell breaks loose...

If not, it just makes the diode's age and build up resistance... Resistance means heat, and heat means more ageing... And well... You get it huh?

Tweety = Scotty :D

http://www.jimis-cyberstore.com/stor..._star_trek.jpg

Hum... Nah... Not Scotty... He's to organized... I think...

On a more serious note... I'm not too sure I even understand my own explanation... Dammit... I have to make this more undertsandable and organized...

Ok, lets try this a more organised way... ;)

This is a better image...

Attachment 11147

The bottom part, the Rectifier, makes AC into DC... The diodes block current going the wrong way, and let it through the way we want it, ie from - to +...

Attachment 11148

Every time a diode blocks current that creates a small amount of heat, and every time it lets through current it creates a miniscule amount of heat from the resistance the diode adds (very, very small)... This amount of total heat is small, very small compared to the Regulator... But when the diodes age and the connectors corrode, the resistance builds up, and the rectifier creates more and more heat the older end more "worn" it get's...

The top part, the Regulator, takes a look at the output, and shorts one phase to ground by activating the "switches" that are either thyristors, or MOSFET's to ground to keep the peak voltage in check... Ie the Regulator does on purpose what a defective diode will do later when all things are about to goo poof... It makes a short, and burns of energy as heat...

The heat comes both from the actual current flow, and with thyristors also from the abrupt on-off nature and the fact that they will always be a bottle neck creating resistance and thereby heat... The MOSFET's however are "smoother", and open more fully, ie less heat from the switching, less heat from the bottleneck and a more precise voltage...

Both are however arcaic in their principle, ie we have to much power, shunt it to ground... So both create heat...

The difference is that MOSFET's make a lot less heat, and they are stable ie even if you heat them up they still produce the same voltage levels... Thyristors change a lot with temperature, and that makes voltage swing wildly... And this makes them have too work even harder to keep voltage in check, creating more switching, creating more heat... And the loop goes on...

As a resultof this, a typical thyristor based R/R will produce 13.5-14.5V if it's healthy... And in semi healthy condition usually 13.2-14.8V... But depending on the temperature in the R/R, RPM and how fast the RPM changes it will swing wildly between these values at random...

A MOSFET based R/R in new condition usually provides 13.5-14.5V, and in semi healthy condition (which takes about 20-30 times longer than the thyristor's to age into) the same 13.2-14.8V...

But... And that's a pretty important but... The MOSFET R/R tends to go towards middle voltage at just off idle RPM, and then towards lower voltage at high RPM, with only small peaks towards higher voltage usually when the RPM's change... Also the swings are slower, more controlled...

A SLA want's 13.8-14V or soo to charge, a GEL usually want's 14.1-14.2V to charge... Most SLA's "boil" and get reduced lifetime if the spend to much time above roughly 14.2V, same goes for GEL's at 14.5V... Drop to low and they simple stop charging... Just below optimal charging and you get "maintainance mode" as in most chargers...

This means that as long as the battery is in good condition it has no problems coping with a semi reliable R/R of either type... But a thyristor based R/R will age it sooner, and ages itself sooner... And then you get problems...

A MOSFET based R/R keeps the battery lasting longer, keeps the voltage more constant, which is good for the ECU/CDI, the electronics in the gauges, the fuses and also keeps the lightbulbs in your headlight happy since it likes just above 14V to make peak light output (provided you have decent wiring too it)...

Both types will make fireworks and smoke when they battery boils over if a diode in the Rectifier decides to go wide open, and both will stop charging the battery if it fuses... But a MOSFET Regulator takes a very, very long time to go "bad" enough to create the heat needed for damaging the Rectifier diodes... Corroded connectors are obviously something that affects both equally in terms of resistance/heat...

Now for the LiFePo's... They like to be charged at 13.6-14.4V, and very, very optimally at around 13.8V... They highly dislike going above 14.4V since that charges them very rapidly with no real way for them to dissapate the heat, and charging to much at lower than 13.4V will build up internal resistance which reduces lifetime...

So a thyristor based R/R in peak condition will work decently... But only in peak condition... A MOSFET one will work even in semi decent condition since it rarely peaks and if it does it's a short time... It might reduce lifetime, but it's unlikely to blow stuff up...

The LiFePo's are no more volatile than SLA or GEL batteries when they go poof, infact they tend to make less damage since the chemicals don't eat through aluminium, and also since they contain much smaller volumes of chemicals simply becuase they are much smaller...But the margin for error is a bit less...

Thanks Tweety, that is a lot more than I had expected. I have a couple questions now, though.
In the heavy industrial systems that I deal with, rectification is accomplished in the same way. However, regulation is handled by varying field winding current, rather than shorting a leg. So if I understand your description, during regulation on the SH, one of the three yellow wires on the R/R is shorted to ground as required to regulate voltage. The resulting heat is supposed to be handled through the frame of the bike or the cooling fins? This makes sense as to the condition of my R/R, then.

In my reasearch on Lipo4, I have pretty much come to the conclusion that the whole issue would likely have been much worse had I had a lead acid battery installed, at least from a safety standpoint. I am tending to think that the R/R failure was coincidental in the whole issue, and not directly related to the charging characteristics of the battery.

Thanks again! As soon as I get some parts in, I will do some more checking, since I would like to verify all this. I would really like to use the Lipo4 battery, too!

Holy smokes, Tweety!

Ok, it's gonna take a few reads to fully take this in, and I know this is probably another area of electronics, buy why do three phases of AC come in from the stator? Is this just what is convenient (and I suppose appropriate to reduce the dips in between peaks), or is there some rule as to how current is received?

sounds like t-ford had the worst combination of all of 'em.

Because stators (alternators really) generate AC current. The switch from DC generators was done for weight and cost reasons. A DC generator requires large permanent magnets to get the same output that a small alternator can make.

Yeah, most likely just a coincidence... Perhaps aided by the fact that the battery doesn't keep swallowing every thing the R/R dishes out, but kind of gets "full" and forces the R/R to start bleeding off more... And yeah, a SLA is the least fun battery to clean up after if it blows... The GEL type aint no picknic either...

Yep, in most R/R's the overproduction is handled through just simple shorting it out to ground and burning it off as heat... Not the most elegant solution... More like stone-age tech... ;)

The regulator is a tiny bit more involved than the image... There is standard shunt type regulator in there, so a few components more (since you probably know what that is, pretty unimportant for those of you that don't know...)

It's more than possible to make a R/R that uses series type regulation instead... It's probably just an issue of complexity and cost cutting, and/or some lazyness... I'm tempted to build myself one, but I'm happy using a MOSFET shunt type regulator from an R1...;)

Lazn handled that one beutifully... I'll just add that a DC generator would be about twice the size of the current stator to do the same taskbecause of higher losses at these "low voltages", and would likely have less lifetime... Plus added weight...