Riding on a cap instead of battery?

[leo]

Nov 11, 2012 10:44:55 GMT -5 qwertydude said:

The CDI is on its own coil and is AC operated so you can totally ignore its effect on the operation of the electrical system.

all CDI's are not AC fired, mine is DC.
the only AC on my scoot is the stator which connects directly to the regulator.
the autochoke on my ride is also AC and connects directly to one of the stator phases.

So with a capacitor there won't be a voltage ripple and the regulator will want to maintain too high a voltage.

this depends on the design of the regulator.
if the regulator has SCR's that are phase fired then it's IMPOSSIBLE for the output to rise above what the regulator is set at.

. . . but in this case it has no effect because a circuit only uses as much current as the applied voltage divided by the resistance, this is ohms law. And applies to DC circuits only.

wrong.
the power law applies equally to BOTH AC and DC with the following 2 differences, in AC circuits capacitors and inductors are replaced by the corresponding reactances and the RMS value of the AC is used as the source voltage.

[qwertydude]

There is of course reactance from inductors and capacitors and ohms law does apply if you use it properly and understand how the phases of current and voltage lead and lag each other. But what I'm getting at is people who take a DC view of AC circuits never take into account those reactances and wonder how you can have more power running in wires than you have power being used by the circuit. Or how, with pulsed currents, having the same average current can lead to more heat in a circuit if the peaks are shorter and higher vs longer and lower.

The second quote you're misrepresenting how the regulator is set. The voltage may never rise above the peak voltage the regulator is set for but that peak is set higher so that when it's connected to a battery it averages 14.4. In the example regulator build I linked to it was set at 14.6. In my practice connecting capacitors to these systems it rose to 15-16 volts. So no, the voltage won't rise above what the sensing circuit is set to, but the sensing circuit is set to a higher voltage than 14.4.

And DC CDI's are pretty rare on scooters. It makes it pretty much impossible to kick start if your battery craps out. So a majority use AC cdi's.

[prodigit]

but why do you keep on saying about stator voltage?
The Cap won't be placed on the stator, what good would that do?

It's placed between the ground and the output of the voltage regulator.

[qwertydude]

Even though the cap won't be placed on the stator it still experiences quite a significant inrush current due to the voltage ripple the regulator outputs. With such a low internal resistance of the capacitor compared to the stator and regulator, power transfer between the stator/regulator and the capacitor reaches nearly 100%.

This means the heat generated by the stator and regulator will be significantly more. For the times on the ripple where the voltage is rising you essentially get maximum power transfer (not to be confused with maximum transfer efficiency) that the voltage difference between the capacitor and stator/regulator will output. This get progressively worse if you start adding loads to the DC electrical system such as lights.

That's how a capacitor on the DC side still acts like an AC system because the system is still only half wave rectified so DC pulses still act like an AC system. You can think of it as sort of an offset AC signal with all the same characteristics of inductance and capacitance as a standard AC signal.

And here we are back to where I said it would be, back to assuming this behaves exactly like a DC system.

[volvonerd]

Do you think it would be a good idea to add a capacitor to reduce ripple voltage even though the frequency varies? Would you size it for the frequency at idle? I know the equation I just wanted you opinion.

[volvonerd]

The more I think about it, the ripple voltage on a 3 phase full wave rectifier shouldn't be a concern.

[qwertydude]

The ripple voltage isn't a concern when on a lead acid. Reducing ripple in this case would be a bad idea because the regulator was designed around the voltage ripple, it was designed to peak at 14.6 in the example diagram I linked to, it's even higher in our regulators. The average voltage it was set to is 14.4 which means when you add the capacitor it brings up the voltage to the peak, on ours it's anywhere from 15-16 volts. Bringing the average voltage that high will boil the electrolyte from your battery and all the other bad problem associated with putting a cap on the system.

It seems the more I try to explain things the more people come out of the woodwork trying to be their own engineers. I say just do it then. Boil your battery and burn out your electrical system and report back on your success.

[prodigit]

on DC the frequency really does not matter. A cap evens out most frequencies; the higher, usually the better it gets evened out.

[qwertydude]

Sorta true on a pulsed DC system the higher the frequency the smaller a value cap is needed to smooth the wave, but it still won't be any help when smoothing the frequency entails raising the voltage to too high a level and risking burning up the regulator.

Unless you want to build your own actively regulated regulator that can work with a capacitor, one that can use the average voltage at low RPM's and then lower it high RPM's when the voltage smooths out more. Adding a capacitor does nothing more than needlessly complicate a system, so what would be the advantage? How is the ripple really affecting a scooter?

[prodigit]

The regulator regulates the voltage, not the cap.
That is the function of the regulator. It makes sure the voltage stays within limits, and excessive energy is rerouted.
It does not care whether there's a battery, or cap placed behind it. It's just there to control the output voltage.

[qwertydude]

The regulator does care what the voltage is because it's set to a higher voltage to compensate for the delay in voltage rise a lead acid battery has when a pulsed DC voltage is applied vs the near instantaneous voltage rise a cap would have.

Your understanding of electronics is very rudimentary. I was a nuclear engineer in the navy and specialized in electronics diagnostics and repair. Yes it does matter what is connected to the end of the regulator because if you don't understand how the voltage rises in an AC based system when it is connected to a battery vs a capacitor, including when it is half wave rectified, then it seems to you it won't matter. Those with actual advanced electronics and electrical design experience would beg to differ.

[prodigit]

I don't know what kind of delays you're talking about, but a voltage regulator regulates voltage several times per second.
It won't matter much. a cap will even out voltage spikes anyway, just like a battery would.

But I'm not going to discuss this with you, since you claim you know it all anyway (yet I don't see the connection to a nuclear engineer, and a voltage regulator).
Learning about electricity does not matter what you are. You can be a woodworker, and still know a vast amount of knowledge about voltage regulators...

[qwertydude]

You're thinking a linear voltage regulator. This is a silicon controlled rectifier which is based entirely on the AC side of the system. When dealing with AC it's not a simple matter of DC voltage. Let me ask you something, what do you think the voltage would be if you rectified your home 120VAC voltage and connected it to a capacitor. What is the minimum voltage capacitor you'd need? If you answered 120 volts DC you'd be completely wrong and you'd set your house on fire with an exploding capacitor. If you answered 169 volts DC then I'd think you'd have a place to start learning about AC rectified systems and how it affects the DC side. If you don't know how I calculated 169 volts you've got a lot more to learn.

If you don't see the connection to my job and voltage regulation, my entire job was to ensure the power distribution and electronics systems on board a nuclear submarine was intact and battle ready and if the crap goes down it was to be able to maintain and repair systems in events of power failure.

Submarines have multiple electrical systems. AC to DC motor generators, lead acid battery back up power. Diesel generator back up power. 440 VAC main power generators, 120 VAC subsystems, 250 volt lead acid battery banks. Numerous electrical control systems that we had to know the insides and outs down to the component levels and that includes basic things like SCR based rectification for DC conversion of AC, or vice versa, or AC to AC voltage converters. Port and starboard 2 and 3 phase alignment, load balancing between phases. And not just that as a nuclear engineer I have to know about the thermal side of things, not just because the reactors run damn hot, but the electronics used to control the power also need to be run within thermal limits so a thorough understanding of electronics and power theory is also a necessity, you have to know how the electronics generate and tolerate heat and how to be able to shed the heat without either frying the people inside the sub or frying the electronics. The list goes on.

I'm not saying a woodworker can't know electronics, but the fact that you keep assuming this is a DC only system tells me you don't have enough understanding about AC systems to know how these SCR controlled rectifiers which run on AC controls the DC voltage and how capacitors can throw off the voltage of the system.

If you think I'm just blabbing about electronics pretending I'm making all this up, I can assure you my years of electrical, electronics, and thermodynamics, does play into "making up" these facts.

[prodigit]

You calculate it with the square root of 2.
And that's why most 12V caps, are not 12V caps, but 16V.
Besides, there's always something using the energy from the cap, like tail light, brakes, blinkers, dashboard, and perhaps more stuff too.
All these things, every one of them, uses power as well, and prevents voltage rises, cause if the voltage rises, they just consume more watts.
So they essentially are doing the same thing.

[qwertydude]

The reason why 12 volt caps are 16 volts is for a safety margin. I've also seen in 12 volt power supply outputs using 14v caps. So you do have a basic understanding of AC. But why default to just looking at the DC side of the system?

Even though electrical components drain some of the excess voltage, it still doesn't stop the recharge of the capacitor from charging too fast. At high RPM's the peak voltage of the stator is something like 40 VAC. Which means the voltage rise to the 14.6 volt set point occurs nearly instantaneously with the capacitor vs being slower with a lead acid due to the natural opposition lead acids have to voltages above their natural resting voltage of 13.2. This means the amount of time the SCR's are shorted is significantly more than with the battery and the current it's draining out of the stator reaches much higher peaks.

SCR's needs to switch on and drain that current for longer, that's where the high peaks come from, longer than average shorting time. That very fast rise to the setpoint of the SCR's means very high peak currents which means very high I^2*R losses in the power transfer from stator to ground. And the load placed by the lights is not enough to drop the voltage significantly on a large capacitor to delay the voltage rise because of the very low internal resistance of the capacitor.

Sure you can try sizing the capacitor smaller to match the load and keep the DC voltage within spec. But either you'll be doing nothing at low RPM's and your regulator will go haywire (try running without a battery and it will have a similar effect as running with just too small a capacitor) or to compensate for low RPM's and prevent the voltage from cutting out by putting a larger capacitor, when you go higher RPM's and the frequency increases, then the voltage will rise too high because the regulator is set to a higher voltage. 15 to 16 volts in my own real life experiments.

You say the electrics will "just" consume more watts, well what does that mean? It means the entire electrical system will be under the strain of providing those extra watts. This means more heat on top of the high current peaks increasing heat due to I^2*R losses and then on top of that SCR's and diodes have a nasty habit of having a positive temperature coefficient so they'll in fact start going into the process of thermal runaway. On such a small system like this it can't tolerate 25% extra heat. The electrics are already shoddily built enough that they can't handle a brighter light bulb so why add to the electrical strain by adding a capacitor?

On top of that the light bulbs can't stand the extra voltage and are always burning out on many people's scooter. They burn out because a lot of times these regulators can't even regulate properly to begin with.

So tell me why are you so insistent on running a capacitor when there is no advantage and only disadvantage?