TOPIC:

I bought a Shorai lithium battery because I'm sick of the lead acid battery giving me issues, but I didn't do a ton of research beforehand.

I figured there would be no problem by just hooking the wires to the battery terminals, but I found some threads on here that *recommend* that you get a mosfet rectifier/regulator from Oregonmotorcycles, Rick's moto, or Spark Moto.

However, the Shorai battery that I got (LFX model with 210 cold cranking Amps) said: the battery "requires a charging system output of 13.1 Volts or higher at idle, and must not exceed 15.2 volts at maximum output."

From what I've read, the stock rectifier maintains a voltage of 14.5V. Why would I need a rectifier/regulator that outputs a reduced voltage if the stock part is right in the middle of the range the battery likes to be at? I think I'm missing something here.

I've been using Shorai batteries for about 5 years in my Z1A, Z1B, and my 1976 LTD and I have only had one problem and I don't think it was a battery/electrical miss-match. I had a voltage regulator go bad and fry the battery once on the Z1A while I was riding but I replaced it and haven't had any problems since. I think the regulator was bad when I put the battery in(I didn't check it until the battery fried).

I have read that some folks buy the battery, then the "special" charger, then convert to the SH-775 regulator / rectifier and really sink some money into this setup. Personally, I don't think the stock system should hit more than 14.5 volts as long as it is functioning properly.. Check the voltage on the brown "sense" wire at the regulator if your bike has such an animal. I found my sense voltage to be operating ~2 volts low after i replaced my stator. After I replaced my stator, I installed a ice-cube relay to feed B+ voltage directly to my stock regulator, to stop the overcharging. I thought my battery was bad, but it turned out to be a shorted stator, and low sense voltage causing it to overcharge..



I still have the same gel-cell battery in it 2 years later.

Man, I wish my electronics looked as neat as that....

But I purchased a LiFePo (Lithium IRON) battery, not a Lithium-ION one. I've read some places that you don't need to update the regulator/rectifier for these, but I don't know why.

Is this true?

It's a good question. Regardless of the type of regulator, they are all *voltage* regulators and simply try to maintain the system voltage and/or battery voltage at a specific average voltage. So as long as it is controlled by an accurate circuit, it shouldn't matter if it is mosfet, SCR, shunt or series-pass. The end goal is the same... a constant voltage.

From my understanding of lithium batteries, which is very limited, they don't like just getting a steady voltage applied. After they are fully charged they need the current to be lowered (which would be done by lowering the applied voltage). I don't think there is any voltage regulator for a bike that will do that.

A lead-acid battery is a brute and can handle a lot of abuse, so a simple voltage regulator is ok. The stator output is in pulses, so the regulation on commonly available regulators are imprecise. And the regulators don't really need to be very precise since most of that precision gets lost in the harness etc.

On old bikes, like most (but not all) KZ's, the most common cause of high voltage to the battery is bad wiring and contacts. When the switches and connectors and fuse holders get dirty, they drop voltage. This voltage drop causes the regulator to incorrectly assume the battery voltage is low, so the regulator increases the output to the battery. Thus the battery ends up up with too much voltage applied to it. This can happen with a perfectly good regulator. I would think this should be the main concern when using lithium batteries... making sure the regulator is getting an accurate battery voltage reading. This happens through the brown wire on KZ regulators that have a dedicated sense line. If I was to install a lithium battery, I personally would devise a system to ensure the regulator was getting direct battery voltage to regulate from.

Perhaps the caution is to get a regulator which uses its own output wire to the battery as the voltage sensing line. That way, the rest of the bike's wiring becomes irrelevant to the regulator. As mentioned, on KZ's the voltage sensing line on the regulator is the brown wire. If you get a 5 wire regulator (3-yellow, 1-red/white, 1-black) the sensing is done directly from the battery through the red/white wire.

Interestingly, Kawasaki, for a short time, had regulators that used the output wire as the sense line, then after a couple years they went back to having the dedicated sense line (brown wire). I should clarify... the brown wire on the main harness provides power to many things, but on the regulator it is used to sense the system voltage and assumes the battery voltage is the same or near the system voltage.

I think Kawasaki went back to the sense line (brown wire) regulators because back in the days of refillable batteries, they could handle a bit of over-voltage and the user would just have to top off the water periodically. It would also explain why the old mechanical regulators specify a pretty wide range of acceptable voltage including 15v. And by using the sense line, the regulator could compensate for those voltage drops and make sure the lights were bright and the ignition fired reliably, even if it meant the battery would need extra maintenance.... which may be hazardous on a lithium battery.

So - basically - the LiFePo or Li-ion batteries are not really designed for the way the stator works on old bikes?

A guy I work with has installed a Li-ion battery in his Lotus Exige, which he's had for at least a decade, but hasn't had any problems with the battery. He is no expert in batteries or circuits, but he mentioned that someone once told him that it wasn't just the overcharging that can make the batteries fail, but also - like you mentioned - the pulses.

He didn't know, and I sure as hell don't know, what about the pulsing is the contributing factor to a failing Lithium battery. I wonder if it is the frequency, the voltage variance between discrete pulses, a combination of both, or something else entirely.

dpivas7 wrote: So - basically - the LiFePo or Li-ion batteries are not really designed for the way the stator works on old bikes?

Lithium-ion... probably not. They don't even allow you to put those into the checked baggage on an airplane anymore. You have to carry them on so if they catch fire they can be put out. So maybe they are just hazardous period, not just when being charged.

I think LiFe-po claims to be fine for vehicles. Is your buddy sure he has a Li-ion battery and not a LiFe-po? The aftermath photos of Lithium battery fires in vehicles are pretty gruesome. That's the main reason I haven't tried them.

dpivas7 wrote: So - basically - the LiFePo or Li-ion batteries are not really designed for the way the stator works on old bikes?

A guy I work with has installed a Li-ion battery in his Lotus Exige, which he's had for at least a decade, but hasn't had any problems with the battery. He is no expert in batteries or circuits, but he mentioned that someone once told him that it wasn't just the overcharging that can make the batteries fail, but also - like you mentioned - the pulses.

He didn't know, and I sure as hell don't know, what about the pulsing is the contributing factor to a failing Lithium battery. I wonder if it is the frequency, the voltage variance between discrete pulses, a combination of both, or something else entirely.

The pulses are no different than from an automotive alternator. Automotive alternators are also 3 phase and there are 6 rectifier diodes in an automotive alternator for the 3 phases. A single phase rectifier is called a full wave bridge. The same as in a motorcycle rectifier and almost every power supply known to man. TV's computers, etc. They also use filter capacitors to smooth out the ripple and make it suitable for solid state electronics. The battery should act as a filter capacitor that will smooth out the ripple in the AC voltage produced.




The main limiting factor would be the voltage exceeding the proper levels. The battery itself should be smoothing out any ripple. Like my friend stated, make sure the brown sense wire is at B+ voltage to prevent overcharging. (I hope you got that out of his post) This is the reason I pulled fused B+ from the start solenoid B+ post, and used the brown wire to trigger it. I don't have a schematic on the circuit, but those relays are used on a lot of mods, some use it to supply higher voltage to the ignition coils, some use them for lighting, I used it to keep from frying my stator again / stop overcharging.

dpivas7 wrote: ... He didn't know, and I sure as hell don't know, what about the pulsing is the contributing factor to a failing Lithium battery. I wonder if it is the frequency, the voltage variance between discrete pulses, a combination of both, or something else entirely.

The system voltage, because it is pulsing DC, is actually a combination of a DC component and an AC component. The regulator is trying to control the average voltage, which is only the DC component, and the DC voltmeter is only reading the DC component The AC component is being neglected.

The AC component accounts for the ripple on the voltage signal, which can be significant.

If you have two voltages that read exactly the same on a DC voltmeter, but one is pure DC and the other has an AC component, and they are both connected to identical loads, the voltage with the AC component will deliver more power .

In our case, without doing the math, (which i'm willing to do later), I would assume the AC component only contributes the equivalent of 1 volt or less (maybe much less). But if it is 1 volt, then I guess it would be in the danger zone for lithium batteries.

But my speculation on lithium fires on vehicles is still with regulator failures (rather than design) or with wiring failures which cause the regulator to compensate with higher voltage.

Also, a bad battery won't do much filtering, since it places no load on the ripple.

You can tell a totally bad battery by how many amps it places on a charger . If the load is small, the battery is either fully charged or totally gone.

You can also see how much ripple is on a circuit by switching your meter to AC while checking voltage. Lots of AC ripple = no ripple filtering. You could add a large filter cap, but is not necessary with a good battery.

That said, your buddy might have had a bad alternator diode's , loose connection, low sense voltage, etc.. the list is endless.. A lot of people are clueless to this info, including MOST dealer and small shop tech's. It does behoove you to make sure things are functioning properly no matter what battery you use.

I replaced this NAPA battery after only 1 year, when in hindsight, it was probably the stator and low sense voltage / erratic charging making the battery seem bad. I was overlooking the stator tests, and just ass-u-me the charging system was working properly since I had 13.5 volts @ ~ 2-3K rpm.

Quality lithium batteries should have a battery management system circuit that would have some protection on the input.
Consider modern MOSFET reg/recs like the Shindegnen, these have fast switching and tend to put spikes into the loom.
A lead acid battery irons out much of this noise, but a lithium equivalent would probably require some sort of filtering, like a ferrite or a tranzorb.
Old style shunt reg/rec designs would probably be somewhat less noisy, especially transistor types.

I came across the same problem when putting a lithium in my KZ. In normal operation the battery voltage was 14-14.5V - a little high for my liking, but OK. But when you turned the headlights on it would reach up to 15.5V at higher revs - just too much. Even when I cleaned the contacts it was still too high with the headlights on.

After cleaning all contacts and adding relays to the headlights, it's pretty stable now.