Somebody here has a signature line like, "One measurement from a real test is worth a hundred theories."
The way I put it is, "Electrons can't lie." Go make the electrons jump through the circuits and they will tell you the answer.

 

I think you are missing the basic problem. If the converter does not get enough voltage from the pulse it will not turn on. If this happens all of that energy is lost forever, so you want it to turn on at as low a voltage as possible. Once it turns on the capacitor makes no difference. It will stay on as long as it has enough voltage , then it will shut off. You cannot get more out of it than you put in. Please show me where I am wrong.

Thanks for your ideas, but I think you really wrong abut the capacitor.
Again, the power stored in the capacitor is C*V*V/2, so available power for the converter increases with the square of the startup voltage. The larger the capacitor, the longer the converter can deliver its intrinsic current-limited pulses before the capacitor is discharged below the .3V UV cutoff voltage. The converter operates a >.5MHz, so all these pulses are completed in less than a few mSecs, depending on the size of the storage capacitor.
Right now, I have 4500uf. So it works like this, energy is somewhat slowly built up in charging the capacitor, the capacitor is rapidly discharged but is able to deliver power until the cutoff voltage of the MCP1624 (.3V) is reached.
If no capacitor was used, the 150 source resistance of the generator would face almost a short circuit. At best, the current in the switching inductor could not rise above approx 20mA with at peak, .9V of voltage across it and the inductor current ramps up to >10mA in a few uSec.

 

Somebody here has a signature line like, "One measurement from a real test is worth a hundred theories."
The way I put it is, "Electrons can't lie." Go make the electrons jump through the circuits and they will tell you the answer.

Looks good, but I would have to assemble a PCB or carry a breadboard in my bike's breadbasket on my errands at least more than once. No room for the groceries. In this day and age, some worried people would think I am carrying some hi-tech bomb in my basket!. What do I do, disconnect it and put it into my purse before I enter Walmart? What might my girlfriends think?

 

OK. Still in the theory stage after 5 days...after at least 3 people have told you to try a different approach. This discourages us.

 

Looks good, but I would have to assemble a PCB or carry a breadboard in my bike's breadbasket on my errands at least more than once. No room for the groceries.

OK. Still in the theory stage after 5 days...after at least 3 people have told you to try a different approach. This discourages us.

That's what the Pope said to Leonardo D. and Galeleo G. , but one of them and his boyfriend carried the Mona Lisa around with them a good part of their lives.

 

That's what the Pope said to Leonardo D. and Galeleo G. , but one of them and his boyfriend carried the Mona Lisa around with them a good part of their lives.

And you think you're a match for them?

 

Not a match, nor a torch nor even a flashlight for them if I don't get this design finished.

But why resist the use of math and engineering when I should have the capacity to easily simulate what goes on between my legs as I roll down the road. I have no reluctance to do this and I much rather than being induced to riding around in my bonnet with haywire in my basket.

The answer can be found if I just simulate the output of the generator. Basically, it is a low power 8 PPS pulse source in series with 150 ohms of resistance with on and off times as approximately already described and having an amplitude between .65 and 4.5 volts, which can be easily varied at will. I don't have an arbitrary function signal generator, but I do have op-amps and MCU's and I can D 2 A myself silly to create a waveform to feed into a audio power amp..or I could fasten the coil of a 48V relay to my vise and with the help of a small motor with a brass rod soldered or superglued to its rotating shaft, attach thereon its end a niobium super magnet to twirl so close but not touch.

 

..or I could fasten the coil of a 48V relay to my vise and with the help of a small motor with a brass rod soldered or superglued to its rotating shaft, attach thereon its end a niobium super magnet to twirl so close but not touch.

I find a specially designed/modified generator a lot more practical than the options you've been studying so far. The trick would be to make it as frictionless as possible and avoid direct contact with the tire at all costs, since it would be too high a source of friction for you to pedal comfortably. Also, doing the math for a generator should be a lot simpler... but I suggest you do the math in retrospect, by first building the contraption and making actual measurements that could tell you the system's overall efficiency.

 

Thanks for your ideas, but I think you really wrong abut the capacitor.
Again, the power stored in the capacitor is C*V*V/2, so available power for the converter increases with the square of the startup voltage. The larger the capacitor, the longer the converter can deliver its intrinsic current-limited pulses before the capacitor is discharged below the .3V UV cutoff voltage. The converter operates a >1.5MHz, so all these pulses are completed in less than a few mSecs, depending on the size of the storage capacitor.
Right now, I have 1500uf. So it works like this, energy is somewhat slowly built up in charging the capacitor, the capacitor is rapidly discharged but is able to deliver power until the cutoff voltage of the MCP1624 (.3V) is reached.
If no capacitor was used, the 150 source resistance of the generator would face almost a short circuit. At best, the current in the switching inductor could not rise above approx 20mA with at peak, .9V of voltage across it.

Yes. of course you need to match the load to the generator in either case. It would not like to discharge the cap into a short either. You need to set the current so the input to the boost regulator looks like 150 ohms. Then you will transfer the most power.

 

The answer can be found if I just simulate the output of the generator.

That will be extraordinarily difficult. The generator output will depend heavily on the magnet field strength and shape, the air gaps (which will not be identical), the inductance of the coil, and probably a few more things I can't think of. The results you measure will depend on the properties of the rectifier and the load placed on the DC side. In fact to be at all useful, you'll need to map out the voltage and current over a wide range of load impedances. Everything you measure depends heavily on rpm, so you'll need good control and measurement of that.

I've done this sort of work and I wouldn't dream of using a bench simulation model in place of the actual device.

 

Actually, I was wrong about the capacitor. It makes it worse because of the higher current thru the diode to charge the cap increases the loss in the diode.
The simulation below shows the results of the 2 ways.

 

Here's an interesting link of an in-hub generator connected so as to power a bicycle's led light and recharge its batteries. It shows an extremely simple circuit for this purpose.

 

The TS has derided EVERY suggestion put forth, why is anyone still coming back for more of the same?

 

The TS has derided EVERY suggestion put forth, why is anyone still coming back for more of the same?

You're right... but I'm under the impression that the OP is still considering all viable options... I think she wants to fundamentally understand every aspect involved in her initial idea, with a complete registry of calculations and conceptual engineering... but there are situations in which doing this before building the actual device is more time consuming and expensive in terms of the effort involved... any experienced engineer knows that... I guess we're all waiting for her to reach the "aha moment" and move forward.

 

Those in-hub generators are sweet and I've drooled over them before. They lose out on the speed of a magnet that you get farther from the axis, but they are still able to extract as much power as the rider is willing to give up. The big advantage compared to the traditional bottle dynamo is that there is essentially no load on the rider other than the necessary amount to power the load. When the lights are off, there's no added drag.

Rats, now I want one again.

 

I should have the capacity to easily stimulate what goes on between my legs as I roll down the road.

I need to get my eyes checked again.

 

I find a specially designed/modified generator a lot more practical than the options you've been studying so far. The trick would be to make it as frictionless as possible and avoid direct contact with the tire at all costs, since it would be too high a source of friction for you to pedal comfortably. Also, doing the math for a generator should be a lot simpler... but I suggest you do the math in retrospect, by first building the contraption and making actual measurements that could tell you the system's overall efficiency.

Thanks for analysis, I studied and thought about what you've said carefully.
But I see that you are looking at the wrong end of the elephant, as one blind man said to another.
The idea in question is about energy harvesting of energy already delivered to the input of the boost converter, and duplicating the generator exactly on my workbench does not offer any way to optimize harvesting, which size capacitor is best for this, what startup voltage is optimal. I only have to create various pulse train scenarios to charge the capacitor to some value that could be studied for energy harvesting to optimally deliver the charge in the capacitor(P=C*V*V/2) at a particular Vstartup of the bo0st converter. By either of the two simulation methods I proposed, this task is quite easy to do.

 

Thanks for analysis, I studied and thought about what you've said carefully.
But I see that you are looking at the wrong end of the elephant, as one blind man said to another.
The idea in question is about energy harvesting of energy already delivered to the input of the boost converter, and duplicating the generator exactly on my workbench does not offer any way to optimize harvesting, which size capacitor is best for this, what startup voltage is optimal. I only have to create various pulse train scenarios to charge the capacitor to some value that could be studied for energy harvesting to optimally deliver the charge in the capacitor(P=C*V*V/2) at a particular Vstartup of the bo0st converter.

A valid point... but how exactly do you intend to generate the energy that you plan to harvest in the first place? If the method chosen for this purpose is not good enough then it won't matter if we're talking about trunk or tail...

 

A valid point... but how exactly do you intend to generate the energy that you plan to harvest in the first place? If the method chosen for this purpose is not good enough then it won't matter if we're talking about trunk or tail...

Thanks for your comments, I truly appreciate all of you for your patience with me and the gift of your comments and analysis.
Again, I wish to directly couple the output of the schottky bridge rectifier and 4500uf (value corredted) capacitor to the coil of the generator, which has a measured approx 150 ohms copper loss resistance. This is the way i am generating electricity to feed the boost converter to charge the battery.

 

Actually, I was wrong about the capacitor. It makes it worse because of the higher current thru the diode to charge the cap increases the loss in the diode.
The simulation below shows the results of the 2 ways.

Ronv, thanks again for your analysis. I have thought about the results you show, but you fail to try to examine the condition where there is no 150 ohm resistor in parallel with a 4500uf (value corrected) capacitor that simply wastes any power delivered to the energy storage capacitor.
I think you fail to consider the area of work, the area under the curve which is the available energy to be harvested and corresponds to the C*V*V/2 power to be delivered to the boost converter. Try your analysis again without any 150 ohm resistor to ground wasting power and plot available energy C*V*V/2 versus time.