Announcement

Collapse
No announcement yet.

OT-alt issue

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • darryl
    replied
    Nice explanation, JT. As you surmised, D1 is there to allow TR2 to bring TR1 into the fully off condition without having to go into saturation, which would require even more current to flow through the zener. It's one method of improving the regulation action, or 'sharpening the knee'.

    That's an elegantly simple regulation circuit overall. The reason that pin 2 goes directly to the battery + is so that the actual battery voltage can be monitored, and not the voltage at the end of some length of wire feeding another component. Having that pin go to the coil will have an effect- unwanted- whenever the points are closed, the coil will be drawing current and lowering the voltage at it's terminal. The regulator will think that the battery voltage is low and up the charge rate. In many vehicles there's a ballast resistor- it's there to lower the coil current once the engine has started. Only when starting does the coil get full 12v (minus losses in the feed wire), and so then the regulator gets to see the full battery voltage when the points are open, then only about 8 volts or so when the points are closed. It probably also gets to see an inductive kick from the coil every time the points open or close.

    Even if the battery is fully charged, the alternator is going to be pulsing it with full rated current every time a spark is initiated, although this action is moderated by the capacitor in the pin 2 circuit. I would not want this to be happening in my vehicle. If you're intent on wiring pin 2 to the coil, wire it instead to the 'hot' lead of the ballast resistor. The fluctuating charge action will still be there, but won't be as drastic. You still risk boiling the battery, but to a lesser extent.

    If the pin 2 current is responsible for the battery problem, I would rather see that circuit fed through a relay so it can be disconnected whenever the key is off. The typical bosch relay for vehicle systems would be fine.

    Leave a comment:


  • J Tiers
    replied
    Explanation.........

    Transistor TR1 pulls current though the field coil. More current, more magnetic field, more charging voltage/current to battery.

    R1 and D1 feed current to the base of TR1, defaulting to "full on", or max charge current. Current can come from battery, in which case lamp will light, and current will be lower. or it can come from the rectifiers when alternator is running.

    TR2 "steals" base current from TR1 when on, to reduce charge current by reducing field current.

    When battery voltage is high, the output voltage of the "voltage divider" formed by R2 and R3 (and the potentiometer) is high enough to overcome the zener voltage of D2 plus base emitter voltage of TR2, so that TR2 turns on and reduces teh charge current as per the above.

    The "knee" is a little not-sharp" due to the need to supply not just voltage, but some current to the base of TR2, which has to come from R2 and the "pot", meaning that the voltage must be a tad higher than just the divider ratio indicates to really clamp down on TR1 base current (and charging).

    That no doubt accounts for the higher milliamp drain indicated for the unit, the lower resistance minimizes the error from the need to have base current.

    D3 prevents "flyback" pulses from damaging parts. C1 slows teh response a bit to prevent oscillation. R5 seems to provide a minimum load for lighting the indicator lamp when alternator is not running.. R4 keeps leakage from turning on TR1 (when TR2 collewctor has pulled down voltage), the same sort of thing I was mentioning with regard to TR2.

    Finally, D1 seems to be there to raise the voltage on TR2 collector, guaranteeing that it has full control of the base current of TR1.

    When alternator is running, indicator lamp will be off or very dim, since voltage will be pretty much same as battery voltage. if alternator quits generating, lamp will come on due to R5 and draw through TR1 and field coil (unless battery is so high that TR2 is "on).

    Originally posted by Willy
    JT, I didn't mean to imply that you were fundamentally wrong.
    And yes wide acceptance doesn't mean it's right, but after over 40 years of successful commercial application it does indicate that it is at least...good enough.
    And, the auto companies demand consistent parts.... they get 'em, too. If they spec a low leakage zener, they will get it, or else.

    I've had a contract manufacturer sub parts on me, when the original specified was guaranteed not to be able to cause a problem in that exact type circuit, but the sub was 50x more leakage and caused big headaches....

    it was simpler to add the base-to-ground resistor than to hold their feet to the fire after the fact. So I did.
    Last edited by J Tiers; 09-23-2011, 12:59 AM.

    Leave a comment:


  • CCWKen
    replied
    I've never used the resistor or a diode but I always use the idiot light when setting up custom applications. I use a #194 bulb in series on the #1 line. It should light when the key is turned to on and go out when the engine starts. The light will also indicate when the alternator isn't charging which is nice to know.

    The bulb and/or resistor act as a current limiter on the #1 line. The diode won't do that.

    Leave a comment:


  • Carld
    replied
    ecortech, I have both a Sun alt/gen, starter, 600A load tester and a Milton load tester and the battery appears to be good.

    Willy, I was not aware I needed the 10 ohm load in the #1 line. The alt seems to be charging correctly. When I start the Jeep it will charge from 20 to 30A and then in a few minutes gradually reduce to about a few amp charge.

    How much would the 10 ohm resistor drop the voltage and would it be the same as the approx .7v drop from the diode?

    I never have been good at using the ohm's law formula and the online calculators would not give me a voltage drop answer.

    Leave a comment:


  • Willy
    replied
    Carl I noticed in your charging system diagram that you do not show a 10 ohm resistance wire or indicator light in the circuit to the #1 terminal leading to the alternator.
    Not sure if this is an oversight on your part or you just decided to not show this, but this is a requirement on GM SI charging systems.

    Also your short driving cycles could be responsible for an undercharge condition.
    Another thing to keep in mind is that virtually all modern voltage regulation circuits are temperature biased, meaning summer time voltage output could be as low as 13.8, to as high as 15.8 volts at twenty below.
    Unfortunately I can verify this last point.

    Edited to add:

    Have a look here to help better explain the purpose of the indictor light.
    http://www.rowand.net/Shop/Tech/Alte...atorTheory.htm

    <H3>Dashboard Indicator Light

    If you have an alternator and are using the factory style indicator light on your dashboard, it is a pretty helpful thing. It helps kick-start the alternator into working at idle speeds when you first start the car, and it tells you if the alternator is putting out less voltage than the battery has in it, indicating a problem. The light is connected on one side to the field current system inside the alternator and to a switched ignition power source on the other side. When you turn the key on but have not started the car yet, the field acts as a ground and power flows through the light and out to ground - lighting the bulb so you know it works. Once you start the car, the voltage at the field is powered internally by the output of the alternator. If this value is exactly the same as the battery voltage, then you have the exact same voltage on each side of the indicator light and they balance each other out - kind of like a tug of war in reverse. If all goes well, the light never comes on, and you drive happily around knowing all is well with your alternator. If the output of the alternator should drop due to a slipping/broken belt or due to certain kinds of electrical faults inside the alternator itself, there will be less voltage on the field side of the light and more voltage on the switched ignition side of the light. The result is that some amount of electricity will flow through the light and into the field and the light will glow proportional to that voltage difference. This is how a slipping belt or an overloaded alternator will cause the light to glow very dimly, while a full-on failure will cause the light to glow very brightly. Note that if you disconnect (or forget to connect) the wire at the alternator, the light will never come on and the alternator will not charge properly.
    The dashboard indicator light circuit also typically has an extra wire with a calibrated resistance in it. This wire is run in parallel to the indicator light and has about a 10ohm resistance. It's purpose is to allow slightly more current to flow to the alternator field current system at initial start-up to make sure the alternator begins producing power as soon as the engine starts. About 1 amp total current is flowing to the field current between the light and the resistance wire, with the resistance wire supplying about 3/4 of an amp. This extra resistance wire does not affect the functionality of the indicator light in any way.
    NOTE: I've been informed by my readers that a Radio Shack 10 ohm 10 watt 10% wire wound ceramic resistor (part #271-132) has worked well on their GM vehicles. Use caution if you decide to do custom wiring work with resistors as they can get hot and melt stuff.
    Last edited by Willy; 09-22-2011, 01:56 PM.

    Leave a comment:


  • Willy
    replied
    Originally posted by J Tiers
    Wide use doesn't mean it's right....

    If you carefully select diodes it works..... and it's that little bit cheaper than using the resistor, which would make Delco bean counters very happy.

    However, it just may be (and probably is) true that a lot of cars with Delco alternators are charging less than others.


    Since the average car alternator beats on the battery fairly hard by overcharging it, maybe the ones that have a "problem" work better.........
    JT, I didn't mean to imply that you were fundamentally wrong.
    And yes wide acceptance doesn't mean it's right, but after over 40 years of successful commercial application it does indicate that it is at least...good enough.

    Leave a comment:


  • ecortech
    replied
    Just because your battery is a little over 1 year old is no reason to assume it's 100% good. I've seen batteries only weeks old that are weak, fully charge the battery and load test it, only way to be sure it's good. Even with a bad cell, and a recent charge from the alternator, it will probably quite easily turn that 6V starter.

    Ed

    Leave a comment:


  • Carld
    replied
    This just got interesting. I was hoping some one with more electrical knowledge than me would post how the alt reg works. I will have to read the last few posts several times to fully understand what was said.

    Remember, this alt is on a 1951 CJ3A Jeep with NO accessories, only brake lights, parking lights, headlights, one dash light and the ign coil. The battery is a little over a year old and it turns the original 6v starter. The electrical system is converted to 12v and retains the 6v starter which is common practice for these Jeeps.

    At some point in the future I may switch to a one wire Delco alt but for now it's run what I have for me. OH, and as of now there is no drain on the battery with #2 term wire run to the coil where the voltage gets turned off with the engine.
    Last edited by Carld; 09-22-2011, 10:18 AM.

    Leave a comment:


  • J Tiers
    replied
    Originally posted by Willy
    The schematic is indeed correct.


    There are literally millions of these alternators in service since 1969.
    The "error" was designed into this alternator by Delco-Remy's design team a long time ago and has not been changed for decades so it's effects must be minimal as the SI series alternator is probably one of the most universally used for a wide range of different OEM applications.
    Wide use doesn't mean it's right....

    If you carefully select diodes it works..... and it's that little bit cheaper than using the resistor, which would make Delco bean counters very happy.

    However, it just may be (and probably is) true that a lot of cars with Delco alternators are charging less than others.


    Since the average car alternator beats on the battery fairly hard by overcharging it, maybe the ones that have a "problem" work better.........

    Leave a comment:


  • ecortech
    replied
    Easy way to eliminate a lot of problems with a GM SI alternator, get yourself a self exciting regulator, one wire needed only. The #1 and #2 wires are not needed, I have used them many times, greatly simplifies the wiring.

    Ed

    Leave a comment:


  • darryl
    replied
    As I'm reading the diagram, pin 1 has no source of current when the switch is off. Pin 2 is a constant drain on the battery. If a diode added to the pin 2 wire were to stop current from the battery from flowing into the regulator, then it would also stop the regulator from receiving battery voltage to monitor the state of charge. The alternator would run 'full tilt'. If the diode were put in the other way, the battery voltage would have to rise an additional .7 volt before the regulator would begin tapering off the charge. The battery would get overcharged.

    Taking the diagram to be correct, the current flowing into pin 2 must be very little, since otherwise batteries everywhere would be draining down and dying early. Assuming a 30 ma constant draw on the battery through the regulator, it would take 33 hours to use up 1 amp hour of battery capacity. Assuming a 50 A/H battery, technically it would take a little over 2 months of non-use for the battery to deplete. Most batteries in vehicles have higher capacity than that.

    This suggests two things- the current flowing into pin 2 is probably less than 30 ma, and secondly if it's the drain from the alternator causing the battery charge to deplete, then the battery is shot or the rectifier bridge in the alternator is leaky and draining much more than 30 ma.

    The field winding has no source of juice when the engine is not running except through the indicator light and the switch. There is no source of voltage to T2 unless the rectifier bridge is leaky, aside from a small current which may pass through the zener from pin 2. Only if the battery voltage is fairly high will any current normally flow through the zener. With the engine off, the battery voltage would drop off from the peak, essentially to the point where no current would be flowing through the zener. At any rate, current flowing into pin 2 is at a low level, unless the regulator is screwed. The current cannot increase in any way because of the transistors being in there. If some base current flows in TR2 it's of no consequence as there's no voltage coming in to the rest of the regulator circuitry for it to have any control over.

    If there's nothing wrong in the alternator, it isn't causing the problem. What could be happening is that it's not producing enough charge to keep the battery up.Your battery could be sitting at 5% charge, and then a couple of days sitting would kill it. I've had a vehicle sit unused for a year or more, and still had enough charge to crank the engine, so the normal drain from the alternator must be pretty low- low enough as to be inconsequential under almost every circumstance.

    Interesting though- if the alternator isn't putting out what it should, it could be because it's dirty inside. Enough dirt, add some moisture, and it could then be drawing more current that it normally would. I got brave one day and decided to hose out my alternator when I noticed that it was getting intermittent- I could see the brightness of the lights changing sort of randomly. I went to a car wash and basically blasted the inside of the alternator with soap, then again with rinse. I figured I had nothing to lose, since I'd be faced with replacing it anyway. That fixed it right up, and it was good for years. I did that again a few years later and got the same improvement. I did eventually have to repair it, not for bad bearings but because the slip rings were worn out. After getting over sticker price shock, I rebuilt the slip rings myself and used the standard bearings/brushes/rectifier kit. This is the original alternator in my Toyota with over 400,000 miles on it.

    Leave a comment:


  • Willy
    replied
    Originally posted by J Tiers
    If that schematic is really correct, there is a bad error in the design....

    .............
    The schematic is indeed correct.

    Although I obtained this copy online for the sake of this being the easiest route, I do have several factory GM service manuals with the exact same illustration in them.
    There are literally millions of these alternators in service since 1969.
    The "error" was designed into this alternator by Delco-Remy's design team a long time ago and has not been changed for decades so it's effects must be minimal as the SI series alternator is probably one of the most universally used for a wide range of different OEM applications.

    Leave a comment:


  • J Tiers
    replied
    If that schematic is really correct, there is a bad error in the design....

    D2 is connected to the base of TR2.... and there is no base -to-ground resistor shown.

    it is common for zeners to have some leakage, there is a spec for it on any datasheet.

    In this case, if there is leakage, TR2 can "steal" current from the base of TR1, just as if the voltage were higher than it is.... That can cause the battery to be always undercharged, which may cause it to have low capacity and act as if it has a bad leak, not to mention gradually sulphating up due to undercharging. With a resistor, teh leakage current will not end up as base current, and the problem is reduced or eliminated.

    That will not be the problem here, most likely, unless the battery does have a low charge state, and the 32 mA is enough to drain it..... A clean start actually requires very little in terms of ampere-hours, so a battery can be pretty dead before it won't start the engine in warm weather.

    Leave a comment:


  • Carld
    replied
    The way the diode is installed the voltage can get to the regulator in the alt but when the engine is off it keeps any voltage from coming out of the alt into the vehicle wiring.

    I have been a mechanic for a long time, retired now, but I have never had this issue. On the other hand I worked on road tractors and you hardly ever see them again.

    When I first got this Jeep the battery would get low enough over night to make it hard to start. After doing a search I found the wiring diagram I posted in post #1 and put one diode in #1 line. I still had a slight discharge over night so I put the other diode in #2. It seemed to fix it but when I let the Jeep set for a few days the issue was back so it never really went away.

    That's when I posed the problem here and some statements led me to search for an internal wiring diagram for the alt. When I saw it and read the associated text I realized that instead of the #2 diode stopping reverse flow it allowed flow from the alt output stud to the internal resistors and then to ground as someone said may be happening.

    Then I decided to put the #2 wire on the coil post that is fed by the ignition sw. and was going to remove the diode. I got to thinking as I looked at the internal diagram and realized if there was any bleeding of voltage to the #2 sensor term it would also bleed into the coil primary but on the other hand it would also bleed into the resistor to the ground. So, for the time being I will leave it in but like you, I don't think I need it.

    Right now there is no current draw between the (+) battery post and the cable so I will call it cured. If there were something shorting to ground in the alt I would get a voltage between the cable and the post.

    Leave a comment:


  • J. Randall
    replied
    Carl, I have wired a lot of alts up over the yrs., and worked on a lot of others. Yours is the only one I have seen with a diode in the #2 sense wire, there is just no need for it. Although the #2 wire will work from the battery post, or the battery stud on the back of the alt., the ideal place is a remote junction block in the harness, or the hot side of the starter solenoid, that way while running it senses the needs of the whole system, not just battery voltage level. If you had to isolate #2 from constant voltage to cure your problem, you have a problem in the alternator.
    James

    Leave a comment:

Working...
X