Fried Rectifier

[wambiker]

Hi all, my 2000 naked has been a little slow turning over when starting for the last week or so, which as it has a hawker odyssey battery, less than 12 months old was a bit puzzling. So as I got off work handy this afternoon, I thought I'd check things over. So started with battery connections and cables to starter, no problems, so I checked the battery voltage, about 12.8v,. I would have expected there or there abouts. But when I started up and checked alternator output I was horified to watch the voltage reach 17.5 volts at 4000 rpm. Shut the bike off and stood back, typical, as an advanced instructor I needed the bloody thing for the first course of the year this saturday, but will not run it again till I've replaced the regulator/rectifier. I suppose I should be lucky at least the ECU survived, but I suspect the battery is toast as well, so thats Xmas overtime spent!

Bl@@dy motorcycles.

Cheers Gary

[raz]

But when I started up and checked alternator output I was horified to watch the voltage reach 17.5 volts at 4000 rpm.

Some people say if the alternator has bad grounding it could give too high voltage at the battery, since the alternator has a different point of view. I'm not sure it could have that effect though. A bad grounding would equal to a voltage drop, and a voltage drop from an alternator giving 14V should produce too low voltage at the battery. On the other hand, a bad grounding may result in more complex scenarios than a static resistance, so maybe they are right.

[callison]

Some people say if the alternator has bad grounding it could give too high voltage at the battery, since the alternator has a different point of view. I'm not sure it could have that effect though. A bad grounding would equal to a voltage drop, and a voltage drop from an alternator giving 14V should produce too low voltage at the battery. On the other hand, a bad grounding may result in more complex scenarios than a static resistance, so maybe they are right.

 

Almost right. If the regulator has a bad ground, then the regulator output voltage may rise out of limits because any resistance to ground establishes a regulator case reference voltage that isn't actually ground. This isn't much of a problem on any Guzzis where the regulator is solidly bolted to the frame such as the V11 Sport series. On the Sport 1100i's/Daytona RS's where the regulator was mounted on the fairing sub-frame it is a legitimate concern.

[raz]

 

Almost right. If the regulator has a bad ground, then the regulator output voltage may rise out of limits because any resistance to ground establishes a regulator case reference voltage that isn't actually ground. This isn't much of a problem on any Guzzis where the regulator is solidly bolted to the frame such as the V11 Sport series. On the Sport 1100i's/Daytona RS's where the regulator was mounted on the fairing sub-frame it is a legitimate concern.

Regulator is what I meant

 

Still, the flow is from a nominal 12V at the regulator, to the bike, with a voltage drop in between. How could that result in too high voltage anywhere?

 

Anyway, if it's bolted to ground this is not an issue for Gary of course.

[Guest Gary Cheek]

Where did you take the reading?

[callison]

Regulator is what I meant

 

Still, the flow is from a nominal 12V at the regulator, to the bike, with a voltage drop in between. How could that result in too high voltage anywhere?

 

Anyway, if it's bolted to ground this is not an issue for Gary of course.

 

Okay. The regulator wants to put out 13.2V or whatever referenced to its case. If there is a resistance between the case and the actual ground of the vehicle, as the regulated current rises with rpm, a potential will be developed across the resistance of the ground path. The Daytona RS/Sport 1100i and maybe the Sport 1100 were notorious for this. The regulator on those bikes is mounted on the fairing sub-frame. As the bike gets older, corrosion built up in the mounting bolts and those mounting points were the only real ground provided for the current flow. What happens with this potential is that regardless of what the regulator thinks it's doing, the case has to be at ground to be correct in the system. If there is a 5V potential across the ground connection to the regulator case, the regulator will see that potential and not actual ground. That voltage will be added to the regulators output voltage and if it exceeds 17V or so, the zener on the old WM16M would fry. To add insult to injury, a corroded/loose/painted joint used as a ground path vibrating is rarely a constant resistance, so the joint would cause spiking in the system. For this reason, I've been posting to anyone with one of the aforementioned bikes to ground the regulator case directly to the engine anytime I've heard them mention problems with the electrical system. It's going on 8 years now, we should have run out of unmodified versions of the bikes by now, but apparently, the problem is random enough that not everyone runs afoul of it in the first years of ownership. I'd say it's inevitable if the bike every gets ridden or wet though.

[Tom M]

Where did you take the reading?

 

I'm guessing England?

 

 

Sorry...

[wambiker]

Across the battery terminals, with a digital multimeter, and the engine off then running and gradually raising the revs. Till as I say 17.9 at 4000 rpm. In England

Cheers Gary

[raz]

If there is a 5V potential across the ground connection to the regulator case, the regulator will see that potential and not actual ground. That voltage will be added to the regulators output voltage and if it exceeds 17V or so, the zener on the old WM16M would fry.

I was humble enough to make that extra grounding anyway and I'm definitely not saying you are wrong (I mean, you made my schematics ), but I still don't get it

I think of it this way: From the alternator to the rectifier/regulator is separate AC wiring not connected to ground. At, or in, the regulator this current is rectified and regulated to a reference voltage. Is that reference sourced from or compared to something outside the regulator? I don't think so and maybe that's where I'm wrong. If I'm not, it means the output from the regulator will be correct (at its terminals) regardless of voltage drops anywhere later. Which in turn means the voltage after that point can only be lower.

 

So, is the reference compared to something external to the regulator? Looking just now at your great schematics (btw, thanks!) I can see there are two more wires, one for the charge light and then one more, that could very well be some input for the regulator. Hm. Maybe I get it now. So should I submit this post or just shut up. Nah, I can't shut up

 

I'm not sure I can spell to inquiring minds, but you know the phrase

[callison]

I was humble enough to make that extra grounding anyway and I'm definitely not saying you are wrong (I mean, you made my schematics ), but I still don't get it

I think of it this way: From the alternator to the rectifier/regulator is separate AC wiring not connected to ground. At, or in, the regulator this current is rectified and regulated to a reference voltage. Is that reference sourced from or compared to something outside the regulator? I don't think so and maybe that's where I'm wrong. If I'm not, it means the output from the regulator will be correct (at its terminals) regardless of voltage drops anywhere later. Which in turn means the voltage after that point can only be lower.

 

So, is the reference compared to something external to the regulator? Looking just now at your great schematics (btw, thanks!) I can see there are two more wires, one for the charge light and then one more, that could very well be some input for the regulator. Hm. Maybe I get it now. So should I submit this post or just shut up. Nah, I can't shut up

 

I'm not sure I can spell to inquiring minds, but you know the phrase

 

Forget anything about the regulator for a moment except the case (ground) and the output +12V (which we know is more than 12 volts, but 12 is an easier number to deal with than 14.3 or somesuch).

 

The regulator does exactly two things. It turns the AC from the alternator into DC, and it regulates the DC to a specific voltage. Period. The alternator can kick out as much as 90VAC at redline on a Guzzi so you can see that there is enough input voltage on the regulator that if the regulator has a problem, it has the "oomph" to fry everything downstream. The stated current capacity of the V11 Sport alternator is 25 Amps. That's enough figures to start doing some math, so now we need something to do the math with: Ohm's Law.

 

I = E/R I is current (amps), E is voltage and R is resistance (ohms).

 

Under ideal conditions, the regulator case (ground) is solidly and electrically at vehicle ground - typically the engine/transmission block and hopefully the frame as well. That would be zero ohms resistance. As the regulator does it's job, any voltage beyond the desired output voltage is shunted back to ground as current through the case of the regulator. Since the case has zero ohms resistance to ground, no voltage potential exists between the regulator and vehicle ground.

 

However, if there is a resistance between the case of the regulator and the vehicle ground, a potential will be developed between the case and ground as the current load increases during the voltage regulation process. This is a bad thing.

 

Re-arranging ohms law for I*R=E means that for every amp the regulator passes back to ground through the 1 ohm resistor a voltage drop of 1 volt will occur across the resistor (unless I have all of this horribly messed up). Well, there isn't an actual resistor between the regulator and ground, but on my Sport 1100i, there are four bolts holding the fairing sub-frame to the bike frame and when those points get correded or loose, they do become in effect, resistors. I don't know how much because I don't want any resistance there at all, so remedying that problem is more important than knowing actual values. But I digress. The regulator can handle 25 amps and if 1 amp equals 1 volt, than 25 amps = 25 volts. And your response is, it's only a 12 volt regulator - and right you are! But - that's twelve volts referenced to ground and our ground for this problem (Sport 1100i) is NOT vehicle ground, it's the potential across the resistance of the fairing sub-frame mount points. Remember, the regulator is rectifying as much as 90VAC and it positively has to be able to bleed that extra voltage back to ground to keep it from being passed on to the vehicle electrical systems. Fine, if the ground is really ground, but if there is that 1 ohm resistance between the regulator case and the vehicle (real) ground, then at one amp of regulation, the regulator output is not 12 volts but rather, 13 volts. As the current being shunted to ground at the regulator increases, the output voltage of the regulator will rise with the current because the regulator is not seeing the vehicle ground (engine), but rather, the "false" ground of the fairing sub-frame electrical resistance to the actual frame. I'm not sure, but I think that at under worst conditions - high resistance to ground, full rpm and no load on the regulator, it would have the potential to pass through the highest voltage of the alternator minus the regulated output which would be around 78 volts. More than enough to fry ECU's and everything. Fortunately, it IS loaded and there IS a good ground so the regulator can do it's thing without damaging things downstream. It's just that it's not a perfect world and occasionally, something goes wrong. For the Daytona RS's and Sport 1100i's, it was a bad design decision to mount the regulator where they did and not provide a good ground path for it. I've been harping on the addition of a heavy gauge wire from the regulator case to a point on the engine or transmission for these things ever since I investigated it in 1999. A friend of mine in SoCal burned up several tachs and ECU's that would probably never have experienced a problem if we had known at the time about the grounding path problem for that series of vehicles. That problem really does not exist on any of the other Guzzi models that I'm aware of but if your regulator mounting bolts become loose for any reasons, you can expect to have problems.

 

I hope I don't spot any egregious errors in this and have to edit it as that would mean that I'm not a good instructor (that's probably a given). However...

[raz]

Remember, the regulator is rectifying as much as 90VAC and it positively has to be able to bleed that extra voltage back to ground to keep it from being passed on to the vehicle electrical systems.

I hope I don't spot any egregious errors in this and have to edit it as that would mean that I'm not a good instructor (that's probably a given). However...

With all due respect I still have the feeling you miss my point. The regulator bleeds 0.001 amp or 24.995 amps or whatever to heat, just to regulate the voltage to a target value. But it does that before the grounding problem. It does that internally. It couldn't care less what we call ground or not. Anyway, that is my assumption. This is my whole point and the reason I don't get this. Now I could be completely wrong and I probably am, but you still haven't convinced me you understand what I mean.

 

OTOH, I will re-read this tomorrow when I've burned those beers to heat. MMMV. Cheers!

[Guest Gary Cheek]

From what I gather at this site:http://www.euromotoelectrics.com/EDLMG.html the regulator in late Guzzis is a series type rather than a shunt device. The alternator rectified on the positive leg by a simple diode while a thyristor in the positive leg is gated into conduction by the regulator. There is a simple diagram under "Click HERE for Voltage Regulation and Current Rectification information in Adobe Acrobat PDF© Format" at this site:

 

 

The ground provides the path to ALLOW charging to take place. As the battery and system loads increase the voltage drops, the regulator gates the thyristor allowing the output to rise. The diode conducts during the positive going cycle when the thyristor is gated. There is a diode and a thyristor in each leg to allow for full wave operation as the voltage swings through a full cycle The alternate legs conduct as the regulator calls for output.

The AC voltage actually runs inverse to the DC output voltage. In a shunt system (like the old Lucas zener system) the regulator merely shunts excess output to ground, creating heat in the process. In the series system lighter demands equals a lighter load on the regulator. With the shunt (Lucas) system, light loads increase the work done by the regulator since it must convert the surplus electrical energy to heat.

 

BTW I am using this system on my CalII and have been very pleased with the results. The parts are reasonably priced replacements for the Sport.

[callison]

With all due respect I still have the feeling you miss my point. The regulator bleeds 0.001 amp or 24.995 amps or whatever to heat, just to regulate the voltage to a target value. But it does that before the grounding problem. It does that internally. It couldn't care less what we call ground or not. Anyway, that is my assumption. This is my whole point and the reason I don't get this. Now I could be completely wrong and I probably am, but you still haven't convinced me you understand what I mean.

 

OTOH, I will re-read this tomorrow when I've burned those beers to heat. MMMV. Cheers!

 

 

I'll guarantee you that the electronics on your bike care what the regulator calls ground.

[raz]

The ground provides the path to ALLOW charging to take place. As the battery and system loads increase the voltage drops, the regulator gates the thyristor allowing the output to rise. The diode conducts during the positive going cycle when the thyristor is gated. There is a diode and a thyristor in each leg to allow for full wave operation as the voltage swings through a full cycle The alternate legs conduct as the regulator calls for output.

The AC voltage actually runs inverse to the DC output voltage. In a shunt system (like the old Lucas zener system) the regulator merely shunts excess output to ground, creating heat in the process. In the series system lighter demands equals a lighter load on the regulator. With the shunt (Lucas) system, light loads increase the work done by the regulator since it must convert the surplus electrical energy to heat.

Thanks, that diagram is pretty near what I had in mind. OK so what will happen when there is a really bad ground connection? Let's say the battery has 12 volts. The regulator will see little less than 12 V battery voltage (the voltage drop is very small now since there's virtually no current). It gates some current through the thyristors. Now this current results in far more voltage drop over the bad ground (lets say 5 volts) so the battery (and the rest of the bike) doesn't get much off it, but the regulator doesn't know this. It thinks the battery immediately raised to 17 volts so it will lower the current. Repeat.

This will probably result in really nasty ripple and possibly ugly spikes. And it will definitely result in bad charging or none at all. But I can't picture it resulting in high sustained voltage on the battery side. Now spikes can of course fry stuff too.

 

I'll guarantee you that the electronics on your bike care what the regulator calls ground.

I'm sure it does, I just try to figure out why. My woman is a black box to me, not much else

[callison]

Thanks, that diagram is pretty near what I had in mind. OK so what will happen when there is a really bad ground connection? Let's say the battery has 12 volts. The regulator will see little less than 12 V battery voltage (the voltage drop is very small now since there's virtually no current). It gates some current through the thyristors. Now this current results in far more voltage drop over the bad ground (lets say 5 volts) so the battery (and the rest of the bike) doesn't get much off it, but the regulator doesn't know this. It thinks the battery immediately raised to 17 volts so it will lower the current. Repeat.

This will probably result in really nasty ripple and possibly ugly spikes. And it will definitely result in bad charging or none at all. But I can't picture it resulting in high sustained voltage on the battery side. Now spikes can of course fry stuff too.

I'm sure it does, I just try to figure out why. My woman is a black box to me, not much else

 

Okay, let's try this. The regulator determines its output voltage according to the difference between it's output and it's case. It will always do this, regardless of what kind of ground the case has. There is circuitry internal to the regulator that is always in action doing this, independent of battery state, alternator voltage - anything. That circuit is the one that absolutely needs the case of the regulator or regulator ground if it is running through wiring, to be the real ground of the vehicle. The simple diagram a manufacturer shows you of a product is not going to show you the details that make it work. That would be bad business.

 

If you took a 12 volt battery and placed that in series between the regulator and ground (with correct polarities of course, ground of the regulator to the +12V battery terminal and the - terminal of that battery to the real ground), you would now have a regulator that would think it's putting out 12 volts but with its case ground at +12 volts, it's output would really be +24V. This is pretty much what a class H audio amplifier does, so there isn't anything magical about the theory. In effect, the bad ground of the Sporti type spine frames is the same as that series battery example. The resistance of the fairing sub-frame to the frame of the bike causes the regulator to think it's putting out +12V when in reality its the regulator output plus the voltage drop across the bad connection of the fairing sub-frame to the frame proper. Since the bike is a mechanical piece in motion, the fairing sub-frame mount points exhibit motion (microscopically perhaps, but there nonetheless) and that motion changes the resistance in accordance with the quality of electrical contact. If the continuity is interrupted in it's entirety, the regulator will be really high for a moment until the regulator action can bring it back to normal. In this kind of system, there aren't supposed to be spikes so it's unlikely that the regulator circuitry has a short enough time constant to completely prevent passing that spike to the output. Make it high enough and poof, sensitive electronic circuits start having problems. The battery can level off a voltage surge somewhat, but it's impedance is high enough that it won't do too much for a voltage transient. After all, the battery responds at the rate of a chemical state change (really slow). That's just the spiking effect from motion, if the mount points exhibited a poor connection constantly, then the regulator reference point is shifted likewise and the output (referenced to the sub-frame but perhaps higher than the vehicle ground reference by several or even many volts) can have an output high enough to cause downstream electronics fry. The WM16M ECU has a 17 volt zener diode on it's +12 volt input on its circuit board. In nearly every instance of a failure of these ECU's, it is that zener that has died. That means that the zener has had at least 17 volts applied to it long enough for it to burn. If the voltage was higher than than 17 volts, it just means that the zener has to try an pass that much more current to ground with the result that it burns out faster. A momentary over-voltage that is not unreasonable for a vehicle system will just get regulated by the zener with no harm done to it or the ECU. Sustained over voltage will kill the zener. I don't think the Veglia tachs are even protected to that point. I think the IC chip in the tach is a low power Schottkey though, and they're typically good to 17 or more volts anyway so perhaps the designer of the tach circuitry would assume that the tach would never see more than 12 volts or so and would not bother to add addtional circuit protection. He really shouldn't have to, the regulator system of the vehicle charging circuit is supposed to be doing all of that. Ignition noise can be filtered out at the input of ECU's and tachs and so forth with simple RC networks and inductors. These same circuits won't do a single thing to an over-voltage, so a regulator with a bad reference point really does cause problems downstream. I'm less certain about the thyristor model of the Ducati Electronica regulator because my electronics days are back about 35 years or more and I haven't thought about this stuff at this level for at least as long. Either way though, a bad ground confuses the regulator and the results are not good for the electronics of the bike. That's electronics. Electrical stuff, battery and bulbs and so forth, suffer a whole lot less initially.