Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Limiting alternator output with resistors in parallel

  1. Jan 6, 2010 #1
    Hello, this is my first post here! I am building a bicycle-driven generator using an alternator and battery (12v) pulled from a dead car. I can easily spin it up to the point where the alternator self energizes and produces output, but on a dead or low battery, the load on the alternator is so high that it is unsustainable for more than about 5 minutes. After reading various sites describing how to accomplish what I'm trying, I've found that by adding several resistors in parallel on the alternator output line, I am able to limit the production of the alternator to levels I can actually pedal for an hour, around 0.9a (the idea here is to capture power during my normal workout on my stationary bike). I had a few resistors in my bin, so what I've added in are two 10Ω 10w and one 8Ω 20w resistor in parallel. These total up to just over 3Ω. By decreasing the resistance I should be able to let the alternator produce a bit more power, but I honestly don't understand why this is so or how I can calculate the optimal resistance to aim for. I would appreciate any guidance or advice to help me understand why these resistors help keep the alternator pedalable and how to dial it in for optimal output.

    Here's my setup, except I cleaned up the giant roll of wire on the output line today:

  2. jcsd
  3. Jan 7, 2010 #2
    This is not an unusual problem. Amateurs have used Bicycle driven alternators have been used to power transmitters at Field Days for years. In order to get a usable level of power for a sustainable amount of effort requires and drastic gear down ratio. Something like using the lowest possible gear ratio on a ten speed and then stepping that down. Adding a fair sized fly wheel to the alternator should help keep things going.
    I would point out the the purpose of an alternator is to put out a constant voltage no matter the RPM applied. Stepping down the voltage, doesn't really gain you much.
    Gearing down the bicycle and removing the voltage limiters should make it work for you.
    I would also point out that 15 mins pedaling as about all most athletically capable men can do. Talk to your local Amateur club, or contact the ARRL (www.arrl.org). There is a lot of literature available.
  4. Jan 7, 2010 #3
    A world class cyclist can put out 400w for 20 minutes. You can probably put out 1 hp for 20 seconds.
    If I find my link for the all-day-work-output of a person I'll post it.

    Use resistors in series with, or a pulse width modulated voltage across, your alt's field winding to adjust the alt's output to a level comfortable for you.
    You have a speed/torque curve, just like an engine or a motor. You want to maximize the overall person/machine efficiency.
  5. Jan 7, 2010 #4
    A fairly common number tossed around by cyclists is 150w for sustained output. I'm a fairly strong rider so I anticipate that I should be able to get that much out of my legs. I know there's drivetrain loss in the system and other limiting factors. I'm basically looking to see how much of that I can capture and still get a good workout. So a key for me is reigning in the demand from the alternator. One of my basic requirements is that I can ride it for an hour.

    I tried putting these same resistors in series on the field line but it didn't seem to affect the level at all, which led me to the question, "How do I figure out what level of resistance will get it under control?" Placing them in parallel on the alternator's output line does bring it under control, just a little too much. I'd love to zero in on the best solution without throwing money at resistors and guessing. Whether that solution is placing them in series on the field line or in parallel on the output line doesn't so much matter to me, but I'm at a loss as to figure out how to come up with a theoretical level of resistance to try on either.

    I can measure that the field charging line draws about 3A when powered up. The output line with no resistors inline will put out something >5A and < 15A with me grinding as hard as I can against it for a few minutes.
  6. Jan 19, 2010 #5

    1/Rt = 1/R1 +1/R2 + 1/R3...
  7. Jan 19, 2010 #6
    My guess is that you would like a fixed alternator voltage (and power) output, independent of the bicycle RPM. This is easily done using a simple feedback circuit to the armature which will reduce the alternator excitation as the RPM increases. Then a single knob can adjust the alternator power output to say 50, 75, 100, 125, 150 watts, etc.
    Bob S
  8. Jan 19, 2010 #7
    I understand that resistors in parallel decreases overall resistance and I am familiar with calculating the combined resistance of them in parallel. I was talking about paralellel or series in two different places:

    1. The resistors in parallel I tried were on the alternator's output line, where lower resistance, seems to produce higher output. I've got that hooked up to a rheostat right now, and it's working fairly well, though I still need to fine tune it.

    2. Others have recomended as an alternative, placing resistors in series on the alternator's field charging line, where theoretically, higher resistance would decrease alternator output by limiting the field within the alternator.

    What I'm struggling with is figuring out how much resistance to try out in either of those arrangements. So far, empirically, I've found that very low resistance on the alternator output line achieves the desired effect (I have a cluster of resistors with a combined resistance of 0.2-2Ω (rheostat involved), though I wonder about how much power I'm wasting with those resistors. With this arrangement, I'm able to sustain output of 4.2A.

    I have had no luck with limiting current to the field line with resistors in series though, and this seems like it would be the more sensible approach to my mind. I'm curious about this idea of a feedback loop, what would I loop back to the shell? Not the alt output itself right?
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook