12V AC motorcycle regulator

Discussion in 'The Projects Forum' started by Suzukiman, Jun 11, 2010.

Thread Status:
Not open for further replies.
  1. Suzukiman

    Thread Starter Active Member

    May 1, 2010
    94
    4

    The capacitor may be used as a reservoir to help stabilise the dc voltage, but I am guessing here. I just used 12V as a ballpark figure, quite a bit of maths may need to be done to determine exactly what voltage that regulator would handle other than pysically testing it and measuring the output.
    In our case the only way is to get it wired into a a bike and measure the voltage or use a 24V transformer.
    The zener is definitely the main component that determines the regulating output and as far as I know it does not have to be 12v or 15v to regulate at 12 or 15, one could also have a 8.2v zener and together with a voltage devider made up of series resistors it can still regulate at higher voltages than the zener voltage. It mostly depends on the circuit design and the particular approach of the person who designed it. for example if you put a diode in series with an 8.2 V zener the voltage changes to 8.9V. Unfortunately I am not so well informed on all these tricks and application methods, so we can only hope someone who knows more about this comes along.

    It would be hard to determine whether it is for 12v or 15V unless we understand the design. In actual fact most likely the 6V AC regulator is exactly the same with just one component value change, which does not necessarily mean the Zener.
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,201
    1,807
    I ran a simulation using similar components; didn't have a model for your exact SCR (the 2N6507G), however the S4025L is a similarly rated (25A, 400V) SCR, so can be considered equivalent.

    The RMS output was just over 13v, which is just about what you want.

    Have a look at the attached schematic. Note that it is drawn in reverse of what was posted previously, as I like to conform to the convention of inputs from the right, outputs to the left.

    I've shown the coils of the magneto as two signal generators, 30v peak, 5 Ohms impedance, 1kHz output. This was just a wild guess. Your mileage may vary.

    Note that the previous schematic had the SCR drawn inverted; it won't work like that.

    The MURS120 is a rough equivalent of the 1N4003. The Zener I used is a 1W 8.2v unit.

    Keep in mind that the simulation results are approximate, but a reasonable estimate of what you might encounter with actual components.
     
    Last edited: Jul 29, 2010
    Suzukiman likes this.
  3. Suzukiman

    Thread Starter Active Member

    May 1, 2010
    94
    4
    Sgt Wookie,
    Thanks a million for testing that schematic! My electronic knowledge is very basic. That seems to be the best low component count schematic to date.
    In your opinion to lift or lower the voltage, would it be better to change the zener or the resistors. Eg for 6V AC and 14V AC?

    enduro250z, could you just double check the wiring to the SCR as per Sgt Wookies' comments above. The pins seem to coincide with your diagram, but he says the SCR will not work that way around.

    Thanks again to all!
     
  4. SgtWookie

    Expert

    Jul 17, 2007
    22,201
    1,807
    Sure. Y'all have been struggling on for quite a while with that, and I'm afraid I haven't been a whole lot of help. I simply haven't used SCR's/TRIACs very often, so I had to do some reading - along with locating reasonably close SPICE models just to get the thing running.

    Give yourself time. If you don't learn at least a couple of new things each day, you just aren't trying hard enough. ;)

    Well, it does have a pretty low component count. My main objection to things like this is that it is shorting out the mag, which creates engine drag and generates heat. If you don't mind the mag drag, then the simple way works.
    Ahhh, what are you trying to accomplish exactly?

    The input from the mag needs to be rippled DC, not AC. The regulator as shown will not work correctly if AC is input. I really don't know why they have a full wave bridge in the regulator, as if AC is input, the SCR won't fire during the negative half-cycle. It would need to be re-designed using a pair of SCRs or a TRIAC to work with AC.

    Just to make sure, I double-checked my simulation's SCR symbol vs the SPICE model that I downloaded earlier, and the anode, gate and cathode are correct in my model.

    Note that you cannot measure the output correctly unless you have an instrument capable of reading true RMS over a wide range of frequencies.

    Most inexpensive multimeters only show the average of the absolute voltage. Even an expensive Fluke meter might only report RMS accurately up to a few hundred Hz.

    If you will note the AVG voltage reading in the simulation I posted, it is significantly lower than the RMS voltage. If you tried to adjust the output using a cheap DMM, you'd be popping bulbs right and left.

    I tried changing the 8.2v Zener to 9.1v. The RMS output increased from 13.061v to 13.418v.

    Then I changed the frequency of the mag to 10kHz, leaving the Zener at 9.1v. RMS went up to 14.156v.
    Then I changed the mag freq to 25kHz, and the SCR went into lockup; it couldn't detect when current flow through it fell to zero, and stayed switched on. I don't know for certain what kind of rippled DC frequencies you'll get out of your mags. Count the number of coils in your mag, divide by two, multiply by your maximum RPM, then divide by 60.

    Changing the resistors is somewhat risky. Keep in mind that in the simulation, all components are exactly as shown; for example R1 is 2.2k. A real-world resistor might vary by 5%. If you decrease the current through the Zener significantly, you'll get into the "knee" where the breakdown voltage is much lower.
     
    Last edited: Jul 29, 2010
  5. enduro250z

    Active Member

    Jul 6, 2010
    69
    1
    SGTWookie, this is how alternator would most likely be. Either one coil like this, or 2 coils in series or parallel. No need for diodes/rectifiers to convert the input to DC for the regulator. The regulator accepts AC from the alternator and out puts voltage to what ever the reg design is uually 13-15 volts for a 12 volt system. Excess unused power is shunted to earth when not used or there is an over supply. Very simple, proven design thats been used for more than 30 years on 1000's of motorcycle, snowmobiles etc.

    Thanks for doing the simulation. I will check SCR connections but im im sure i got it right.

    [​IMG]
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,201
    1,807
    That regulator will only regulate on the positive portion of the cycle; the negative portion will not be regulated. As a result, you will have a wide variation of output voltage, and if you try to use it without a full wave bridge as shown, you will pop light bulbs very quickly when you rev the engine.

    And yes, I have tried it in the simulation.
     
  7. Suzukiman

    Thread Starter Active Member

    May 1, 2010
    94
    4
    If more people are following this thread, here is another AC regulator designed by Lou and he has a very detailled description on how it works. Unfortunately being a more complete design ( some posters earlier in this thread were already working in that direction) its component count is a bit high and it will be more difficult to fit into an enclosure on a dirt bike.

    Lous description gives a much better understanding on how these things work

    http://home.comcast.net/~loudgpz/GPZweb/RegRec/GPZacRegulator.html

    The original purpose of this thread was to find a schematic that has a low component count, but will just power the lamps and will have a small footprint.

    I thought just to add this to make this thread more complete.

    I was wondering what the circuit in a Chinese scooter looks like as its also quite small and includes the single phase regulation for the battery circuit in addition to the AC regulation.

    Thanks to all the respondants every contribution is appreciated!
     
  8. enduro250z

    Active Member

    Jul 6, 2010
    69
    1
    Well, re checked my SCR connections in my first drawing and I cant fault it. By the way I have drawn the SCR as im looking at the side with the numbers.

    The way you have drawn the power supply is not right. We are not dealing with a 2 phase power source (ignore my other topic for this) Most AC systems in bikes, snowmobiles etc use single phase power This is usally one or 2 basic coils mounted on a stator plate or it might be a stator with 6 or more poles all individually wound and connected together and this has a flywheel magnet rotating over it.

    [​IMG]


    *The photo above is to used for references only. The top 2 longer whiteish coils stacked on top of each other are ignition coils and are to be ignored in this project. The 2 light coils are not properly mounted or connected. The way these were connected was in parallel as below diagram, however the outputs were not linked together. One fed AC for lights and one fed a 6 volt half wave rectifier to trickle charge a small 6 volt battery with no regulation other than engine speed. I have also got another stator the same as above which has been modified and has the 2 coils wound in series with thicker wire and this makes the stator produce about 70 watts. If you try and put a brighter globe on, you get voltage drop and the light goes dim. I also have another stator thats priciply the same but with 2 much larger light coils mounted in a more efficient way in a 'L' or 90 deg shape rather than a 'V' or less than 90 deg shape which is how the stator is above but as i havent got coils mounted properly its hard to see this and they looks about 90 deg but when they are mounted properly they are not. Both coils are wound connected in parallel to AC regulator as shown in 'parallel' diagram below. This stator makes a usable 120watts at low-mid rev range.

    These are the sorts of stator systems we are mostly working with, so they are single phase, a couple of coils and mounted series or parallel. No center taps, no split/dual phases etc so you need to simulate that sort of alternator.




    You can also have an internal rotor set up. For ease of explanation i will just say we are dealing with 1 or 2 coils on a plate as shown in my photo. If there is 2 coils they can be wound in series or parallel as shown in my diagram.

    [​IMG]

    [​IMG]




    With AC one lead of the coil will always bee grounded but if you link the 2 coils in series then there is only 1 out put wire and the other end of the 2 coils which are linked at the stator is grounded in there, so its kind of like 1 big coil made from 2 small ones. I really dont understand which is better, but the general consensus seems to be that parallel gives you more current. I think its kind of like having 2 batteries in series or parallel. I think they wire stators in series if they cant get enough voltage from one coil. If there is enough voltage in one coil, then they wire the 2 coils in parallel to keep the voltage the same but double the current?

    Anyway, do you reckon you can do the simulation with only a single phase AC source, so basically just how my drawings below (still to come) indicate, which is how most AC applications are. I understand how with the 2 phase system you used that you need to rectifiy it first to get the 2 outputs in phase but we are not dealing with that sort of system these days.

    As ive mentioned before, 1000’s of different bikes, scooters etc use these simple AC systems and dumping the excess to ground does the alternators no harm. The drag, if any is not noticeable. in theory you get a horse power loss but it cant be much because no one ever notices this or cares. Its more important to have decent lights to see and be seen! and the regulators do the bulk of the work basically. This is a system manufacuturers came up with over 60 years ago or more and is ideally suited to when you need a lot of output to run anything that is not polarity conscious and you don’t have electric start or anything requiring a battery.



    By the way, im considsering purchasing some more regulators to sacrifice for this project and compare, but i think we really need to focus on getting out heads around these first otherwise we are getting no where. I would still really like a definitive answer on how/if the basic TO-3 tansistor, zenner diode and bridge rextifier type works in my first circuit i brought to this topic.
     
    Last edited: Jul 30, 2010
  9. SgtWookie

    Expert

    Jul 17, 2007
    22,201
    1,807
    There are now 149 replies to this topic.

    It would be most helpful if you would indicate the post number that contains the schematic that you are talking about. I really do not wish to dig through 15 pages of posts and guess which post you made first in here.

    There is not going to be a simple "one regulator suits all" solution that I can see.

    The simulation I posted would work for a split-phase magneto/alternator, but would not work for a 3-phase magneto/alternator, as the SCR would never turn off.

    As far as your Ducati, I started off thinking that it was a split-phase magneto. However, in your other thread I posted a link to a Ducati tech notes forum thread where a fellow claims that later model Ducati's use a 3-phase generator. I frankly do not know how accurate that claim is. I only suspect that it's a split-phase.
    Your other thread: http://forum.allaboutcircuits.com/showthread.php?t=41162
    The Ducati tech note alt failure thread: http://ducati-upnorth.com/tech/chargingfailure.php

    Virtually all automotive alternators are 3-phase. However, they are regulated by current through the rotor winding. Magnetos can't be regulated like that. So, you're left with either shunting excess output to ground, or using some type of series regulation.

    Something that has not been discussed at all is cooling of the regulator device. I have no idea where your regulators are mounted, but I suspect the location varies widely. My 1980 Kawasiki KZ-650 F1 had the rectifier and regulator behind one of the side covers, if I remember correctly. That meant the air flowing across it was pre-heated by the 4-cylinder engine just in front of it; not the most ideal location. In the Ducati article, the fellow moved his regulator from inside a very hot fairing to the license plate holder. That's certainly better, but I'm thinking that's not so good for a dirt bike; it would likely get covered with mud. Mud is a pretty good heat insulation, which means there would be no place for the heat being dissipated inside the regulator housing to go; so the regulator would burn up.
     
    Last edited: Jul 30, 2010
  10. enduro250z

    Active Member

    Jul 6, 2010
    69
    1
    Ok appologies, The post im refering to is pos 45 and 48 here
    http://forum.allaboutcircuits.com/showthread.php?t=39419&page=5

    Im not familiar with 3 phase stators left to be AC. Usually 3 phase is full wave brige rectified using 6 diodes and then regulated in the one unit. We are not dealing with that here though.

    Yes most car/truck alternators and road motorcycle bikes have 3 phase DC systems.

    I will answer the Ducati part in my other topic because it kinda belongs there and not to get off track here.

    We did discuss cooling to the regulator a little. Most bikes have the AC regulators under the tank or seat so they get little air flow. Ive had a bike with a 250watt AC alternator and the AC reg is under the seat, but since the regulator is not capable off dumping the full 250watts, the lights are left on. This set up has worked for 15-20 years or so. On most of my bikes now I am going to mount the regulators near the front of the bike where they can get direct air flow, just to be on the safe side, expecially if I have regulators that can dump all my available power which will allow having the lights turned off in daylight. So theres not really an issue there. Many regulator makers advise you to mount them where they can get a cooling air flow. Yes mud can be an issue im aware off.

    I will/can get my hands on a couple inline (series) AC regulators to see their circuits. They have 3 wires. AC in AC out and a ground wire.

    By the way, I am familiar with the 650/750 Kawasakis like yours. Very good reliable bikes and fun to ride. They sound good with 4 into 1 pipes.
     
    Last edited: Jul 30, 2010
  11. enduro250z

    Active Member

    Jul 6, 2010
    69
    1
    http://forum.allaboutcircuits.com/attachment.php?attachmentid=20917&d=1278573757

    on page 8


    This was the circuit I wanted to know a little more about.
    Even if you cant simulate it working, does it look alright and in theory would work ok? I can only find 12 or 15 volt zeners. 15 would probably be ok but I wouldn’t mind aiming for 14.8 volts. Can I just use 2 zeners in series to add up to 14.8 volts?

    Although the TO-3 trans is robust, Mounting it can be a pain so I wouldn’t mind finding some TO-220 or TO264 types but rated over 200watts. If I can get a 250 watt one then I know it should easily safe for up to 200w
     
  12. SgtWookie

    Expert

    Jul 17, 2007
    22,201
    1,807
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,201
    1,807
    OK, I take it back. I'd goofed in my simulation of the other circuit.

    Have a look at the attached. With the Zener being 15v, the RMS output is around 13.3v with the mag output at 30v P-P. However, if the magneto P-P output rises significantly higher than 30v, you'll get a higher RMS output.

    It needs 2 connections; one to the AC in, and one to ground.

    Power dissipation in the transistor will increase considerably if the P-P voltage of the mag output increases beyond 30v.

    Even though the TO-3 package is more difficult to mount, it is far more effective at transferring heat from the junction than TO-220 packages; about 4x better. You might consider using a copper heatsink from an old-type Pentium processor. Copper is much more effective than aluminum at transferring heat; nearly twice as good.

    Giving the heat sink a very light coat of flat black paint will help resist corrosion, and at the same time radiate heat more efficiently. Mounting a finned heat sink with the fins vertical helps a great deal with convection cooling when you are stopped.
     
  14. enduro250z

    Active Member

    Jul 6, 2010
    69
    1
    Ok so what exactly did you change? I see you added a resistor, changed the zener, and trans.

    Would the value of the bridge rect make much difference?

    Why wouldnt it work with the components i found in the old regulator? i know the guy who sold these sold many 25+ years ago so they surely would have worked and if not he would have got a bad rep pretty quick.

    Even if you cant really simulate the AC power source correctly did my circuit look like it would work in theory with the bridge, MJ802, 13 volt zener and no resistor? im just wondering if its because you cant properly simulate the AC power source that it is telling you it wont work?

    I havent rulled out T0-3s yet, i know they are kinda better.

    What do you mean by PP voltage? I reckon anything i got would be producing 40 volts un- regulated at peak RPM of about 7-8000.

    One of my rewound stators makes 120watts @ 3000 rpm. It would rise more after that, maybe to 150 at 6000? but its designed to have lights to 120watts. No point putting 150w of lights if you gotta ride everywhere at 6000rpm to make the lights run bright and then at 2000rpm you got dim lights!.

    While on the same subject i have found some 12v 20 amp (about 240w) AC regulators, made by an electrical company in Italy called Ducati Energia (sub division of the bikes i think) They make stators and regulators for all sorts of bikes. This particular regulator is 3 wire. Earth and AC in AC out. They are for Vespa and Lambrettas and other scooters/euro bikes.

    Thanks again for your time to do this.
     
    Last edited: Aug 3, 2010
  15. nerva

    New Member

    Jun 16, 2011
    1
    0
    Hi I'm a bit later and could be Suzukiman have already found the solution.
    Otherwise could be interesting to continue the investgation, having me the same problem as Suzukiman.
    In fact I found this forum because I was looking for the circuit schematic of the Ducati's Energia regulator (it is infact a subdivision of the well known Ducati motorcycle company), installed in the old Vespa and Lambretta scooters.
    I read all the posts and while the solution with a transistor seem to work the other with a triac not yet.
    I tried then to restart from the circuit in post #113, I modified it in order to reduce the components and I recalculate the value of resistors.
    The outcome is the attached schema: it should regulate the voltage at 14,3 V AC RMS.
    The advantage of regulating with a triac instead of a transistor should be less power dissipation in the regulator (should not heat).
    At this point should be interested to have a simulation from SgtWookie to have a confirmation that is working, if he is still reading.
    Regards to everybody
     
  16. beenthere

    Retired Moderator

    Apr 20, 2004
    15,808
    295
    And our Terms of Service make us close the thread.
     
  17. SgtWookie

    Expert

    Jul 17, 2007
    22,201
    1,807
    As Beenthere implied, the Boards' Terms of Service has been updated since this thread was last active; the relevant portion of it reads:

    So we will not be able to assist you with your request on this Board.

    If you wish to continue with the project, you can visit other electronic bulletin boards. You might receive some help over at electro-tech online.
    That board is here: http://www.electro-tech-online.com/general-electronics-chat/
    I am not a member over there.
     
Loading...
Thread Status:
Not open for further replies.