I pretty certain that what I'm doing is safe, but I'm hoping for some feedback from you guys with more experience just in case there's a potential issue I'm unaware of.

I'm going to be distributing capacitors for guys to connect to the starting batteries on their motorcycles and I want to be sure there's nothing more I can do electrically (protection circuitry) to make them reasonably safe, from a fire perspective, before giving them out.

These are for off-road motorcycles, and if the battery is completely dead or damaged the bike will not start or run. Adding a capacitor across the battery leads resolves the issue, so I'm distributing these to a bunch of guys so they don't get stranded in the middle of nowhere by a battery failure (it has happened many times). They will be mounted in the airbox under the seat, away from heat, and connected in parallel with the starting battery. It's simply a capacitor with wires on the leads, dunked in plasti-dip for durability. Electrically speaking, is there any real danger to this?

Tech details: 12v electrical systems, peak voltage is around 15v. I'm using this capacitor (data sheet here), it's 10,000uF, 25V, 85C. Here's what it looks like; the leads are insulated with heat shrink then bent around and zip-tied to the can and the whole thing dunked in Plasti Dip for durability:

 

This was a very popular thing for Triumph and BSA motorcycles made into "choppers" back in the 1970's. Here's a company still selling them - http://www.oregonmotorcycleparts.com/BEC.html They use a higher capacitor value than your suggesting though, 22,000uF. Think I'd go with a 50V cap too.

 

That cap charged to 12V supplies the same energy as a hundred watt lightbulb.

Well, if you switch it on and off after a fraction of a second.

Plus if placed parallel to the battery they are just as discharged as the thing you wish to backstop.

Save your money and buy some jumper cables and a cell phone.

 

@shortbus - Yep that's exactly what I'm after! Thanks for the link. Why would you go 50V? Do you think 25V isn't enough margin?

@ErnieM - The capacitor definitely works, that's not the question. With a dead or missing battery these bikes won't run, but with a capacitor they kick start and run just fine (they have both kick and electric start). These bikes are fuel injected, and with a completely dead or damaged battery the power from the alternator is horrendous. A capacitor cleans it right up. Here are some shots from my scope testing out the cap:

No battery, 1 kick, will not start. The voltage is bouncing between 0v and 20v every few mili-seconds:


1 kick with capacitor, started right up and voltage settles just under 15V:

 

Automotive (thus motorcycle too) electrical systems have many voltage spikes in them. In a non automotive system the 25V would probably be OK, but going to 50V caps won't hurt anything, just makes it less immune to any spikes.

 

50V for sure... or more..
Its not uncommon to see 125V transients in automotive applications.

 

OK thanks guys, I'll take your advice. And I'm glad I asked about this! I was unaware of spikes being common in automotive systems. Does anyone need a bag of 40 10,000uF capacitors for cheap?

Question that I don't see answered in the data sheet; assume my capacitor is rated 10,000uF and 15V max. I assume this would be 10,000uF AT 15V? i.e. at 10V the total stored charge would be less than 10,000uF? If this is the case, then if I change to a capacitor that is rated 50V or 100V, I would also need to increase the rated capacity if I wanted to get equivalent stored charge @15V. Have I got this correct?

 

That's the safe working voltage of the capacitor. You can use this to calculate energy storage for a given capacity and voltage.
http://www.calctool.org/CALC/eng/electronics/capacitor_energy

 

No. The voltage rating on a capacitor is the "maximum" voltage the capacitor can safely be exposed to..

 

Automotive (thus motorcycle too) electrical systems have many voltage spikes in them. In a non automotive system the 25V would probably be OK, but going to 50V caps won't hurt anything, just makes it less immune to any spikes.

The old Lucas systems that were sometimes modified for off road, with a capacitor instead of the battery, had a 100W 15V zener as the sole form of regulation.

With such a zener fitted - a 16V cap would most likely be adequate, anything over 25V an extravagance.

 

...................
Question that I don't see answered in the data sheet; assume my capacitor is rated 10,000uF and 15V max. I assume this would be 10,000uF AT 15V? i.e. at 10V the total stored charge would be less than 10,000uF? If this is the case, then if I change to a capacitor that is rated 50V or 100V, I would also need to increase the rated capacity if I wanted to get equivalent stored charge @15V. Have I got this correct?

Not quite.
The capacitance value is not the stored charge (measured in coulombs) , it specifies how much charge is stored for a given voltage.
The stored charge is proportional to the applied (not rated) voltage and the capacitance (Q = C × V) where C is the capacitance, V is the applied voltage, and Q is the charge stored.
The rated voltage is just the maximum voltage you can apply. It has no effect on the stored energy below the rating, thus two capacitors of different voltage ratings but the same capacitance will store the same amount of charge at the same applied voltage.

 

Not quite.
The capacitance value is not the stored charge (measured in coulombs) , it specifies how much charge is stored for a given voltage.
The stored charge is proportional to the applied (not rated) voltage and the capacitance (Q = C × V) where C is the capacitance, V is the applied voltage, and Q is the charge stored.
The rated voltage is just the maximum voltage you can apply. It has no effect on the stored energy below the rating, thus two capacitors of different voltage ratings but the same capacitance will store the same amount of charge at the same applied voltage.

If you raise the applied voltage very carefully - the leakage current grows the oxide layer thicker on the anode foil, that increases the voltage the capacitor will stand but reduces the capacitance.

Conversely; if the capacitor is consistently operated below its rated voltage, the caustic electrolyte will slowly etch the oxide layer thinner - capacitance will increase, but it will no longer stand its rated voltage.

 

If you raise the applied voltage very carefully - the leakage current grows the oxide layer thicker on the anode foil, that increases the voltage the capacitor will stand but reduces the capacitance.

Conversely; if the capacitor is consistently operated below its rated voltage, the caustic electrolyte will slowly etch the oxide layer thinner - capacitance will increase, but it will no longer stand its rated voltage.

I was referring to more or less ideal capacitors, of course.
The "wet" (non-solid) type of electrolytic capacitors may behave as you stated but the solid electrolyte types generally don't.

 

Thanks guys! I did some reading, and I think the equation "C = q / V" is exactly what you're saying, where C is capacitance, q is charge in coluombs and V is voltage. So exactly what you said, all 10,000uF capacitors at 10V will have the same charge stored. But the 10,000uF capacitor that is capable of 100V can actually store much more charge (coluombs) when charged to the higher potential. I learned something, thanks!

PS-> The zener is a cool idea, I will have to consider that since I already have a bag of 25V capacitors. Is there any significant difference, safety wise, between using a zener with the 25V caps, or changing to 100V caps?

 

Thanks guys! I did some reading, and I think the equation "C = q / V" is exactly what you're saying, where C is capacitance, q is charge in coluombs and V is voltage. So exactly what you said, all 10,000uF capacitors at 10V will have the same charge stored. But the 10,000uF capacitor that is capable of 100V can actually store much more charge (coluombs) when charged to the higher potential. I learned something, thanks!

PS-> The zener is a cool idea, I will have to consider that since I already have a bag of 25V capacitors. Is there any significant difference, safety wise, between using a zener with the 25V caps, or changing to 100V caps?

The transient suppressers I've seen advertised mostly clamp at about 68V - voltages above that will probably damage things that are needed to keep the engine going.

A 100W 15V zener will have a certain amount of dynamic resistance, so in extremes the voltage could maybe reach about 17V. The zener failure mode is most likely short circuit - nothing gets damaged, but you're going nowhere fast.

The general subject of motorcycle alternators has been discussed recently on another thread, so it shouldn't be necessary to repeat everything here.

 

Thanks for the info, knowing that the failure mode is a short is enough. Some guys may leave these attached to their batteries and if for some odd reason a zener fails shorted and isn't fused that could definitely cause a fire. So higher voltage caps it is. Thanks everyone!

 

Thanks for the info, knowing that the failure mode is a short is enough. Some guys may leave these attached to their batteries and if for some odd reason a zener fails shorted and isn't fused that could definitely cause a fire. So higher voltage caps it is. Thanks everyone!

British motorcycles used zener regulation for many years. I've never heard of any fires due to zener failure, but there may have been the odd one or two in all that time.

They most likely had a fuse in there somewhere.

An off load alternator can get up to several hundred volts - so a 100V capacitor won't survive if there's nothing to control the voltage.

 

British motorcycles used zener regulation for many years. I've never heard of any fires due to zener failure, but there may have been the odd one or two in all that time.

They most likely had a fuse in there somewhere.

An off load alternator can get up to several hundred volts - so a 100V capacitor won't survive if there's nothing to control the voltage.

Commonly there's a large fuse or fuse link between the battery and all the circuitry and generator (except for the starter of course).

Generally even a dead battery would have a low enough impedance to prevent the alternator voltage from going that high.

 

The bikes are fuel injected, so there will always be an ECU, various sensors, a fuel pump and a fuel injector running, so I don't think the alternator will ever be fully unloaded. But that said, I guess surprises can happen. Maybe I need to integrate in a fuse and a zener... Then again, these guys are selling just a capacitor, I wonder what their failure rate is?

Also as @crutschow indicated, assuming they leave the battery connected it should provide a reasonable amount of buffering. Some may choose to run sans-battery entirely though if the battery is damaged somehow.

 

Commonly there's a large fuse or fuse link between the battery and all the circuitry and generator (except for the starter of course).

Generally even a dead battery would have a low enough impedance to prevent the alternator voltage from going that high.

My errr............not permitted here battery desulphator, had developed up to 210V across a sulphated battery.

More likely on a functioning motorcycle; an uncontrolled alternator can gas out all the distilled water - if the electrolyte gets lower than the bottom tips of the plates, its an open circuit for all intents and purposes.