Topic: Headlight fuse blowing, battery hitting 15.5v, regulator bad?

[jorwesflow]

I've recently been blowing a lot of headlight fuses, and noticed that charging is hitting between 15-16v at idle to 4k rpm.

I'm about to go through TT's tests http://forums.sohc4.net/index.php/topic,1012.msg8345.html#msg8345 but wanted to see if there are any initial thoughts from anyone. Are the fuse and overcharging related? Bad regulator?

Gonna go through those tests and report back.

[jorwesflow]

First off, I’m pretty far removed from stock.

• One of these combo reg/rect from Dime City which I’ve had 5 years https://www.dimecitycycles.com/vintage-cafe-racer-caferacer-bobber-brat-chopper-custom-motorcycle-electronics-parts-honda-cb350f-cb400f-cb500-cb550-cb750-rectifier-864218.html
• Pamco ignition
• Magna coils
• Antigravity 4 cell lithium battery
• Starter and magnetic switch is bypassed

Begin with problem verification and characterization with recorded data.
A -- Fully charge a known good battery.  Let it rest for 2 hours, off the charger, and measure the battery voltage.  (Target is 12.6-12.8V.)

I'm getting 14.7v on a rested battery  This is a fairly new—probably less than 10hrs riding—Antigravity 4-cell LIFEPO4. It's my second one. The first lasted me almost 4 years!

B -- Start the bike and measure the battery voltage at idle, 2000, 3000, 4000, and 5000 rpm.

This is where I'm getting 15v and higher. I know that Antigravity does states they do not want their batteries charging at anything higher than 14.6v. So hopefully I haven't fried this battery already.

C -- Repeat the measurements of B with lighting off.

Part B was with no lighting, since I have the headlight disconnected.

Worth noting: i've got the same voltage reading between the battery terminals, and then between the black wire and ground with the ignition switch on.

The above tests identify charging system success, failure, or degree of "faulty".  The success voltages are listed in the Shop manual.

D -- Assuming the above indicates faulty, do check the RECTIFIER diodes with a diode tester or ohmmeter capable of testing diodes and uses more than .7 volts to make the test(s).
Of the twelve test made in D, six must read low ohms and six must read very high ohms.

Not totally sure how to do this with my combo reg/rect

E- assuming no faults were found in D,  Measure the white and green wires disconnected from the REGULATOR.  CB750s should 6.8 ohms - ish,  CB550s/350s/ and 400s should read 4.9 ohms- ish.

I’ve got 4.2ohms here with the multimeter switch on 200.

F- If there are no bullet holes or road rash/divots on the alternator case, the stator is probably good.  But, you can check for yellow to yellow continuity (.35 ohms) among all the wires, and that no yellow wire has continuity to the engine case.

I’ve got 1ohm between all the yellows. That too high?
No yellows have continuity to ground.

G- Assuming no faults found in D, E, and F, measure the disconnected terminals of the REGULATOR.  The black and white terminals should measure zero ohms (subtract meter error if there is any).  Higher than Zero ohms, indicates internal contact contamination needing cleaning and attention per shop manual.

I’m getting 9 on the 20k setting on the multimeter (sorry, not sure how to express this in the proper way). I assume this is not good? Again, this is an aftermarket reg/rect that cannot be opened and serviced.

I'll just stop there since I seem to have hit a bad result. I'll wait for any feedback.

[scottly]

Lloyd's procedure is more geared towards under charging, not over charging. You have a regulator problem.

[scottly]

G- Assuming no faults found in D, E, and F, measure the disconnected terminals of the REGULATOR.  The black and white terminals should measure zero ohms (subtract meter error if there is any).  Higher than Zero ohms, indicates internal contact contamination needing cleaning and attention per shop manual.

I’m getting 9 on the 20k setting on the multimeter (sorry, not sure how to express this in the proper way). I assume this is not good? Again, this is an aftermarket reg/rect that cannot be opened and serviced.

I'll just stop there since I seem to have hit a bad result. I'll wait for any feedback.

This assumes a stock mechanical regulator.

[TwoTired]

If the Vreg is not getting a true report of  actual battery voltage, it can overcharge, even if it is working correctly.

When using a multimeter, set to lowest range and note the display when the probe touch each other.  This reading must be subtracted from all future resistance readings.

I recommend you measure the voltage lost between the battery terminals and the Vreg voltage input terminals.  Normally this is the black wire and the green wire and both paths back to battery should be measured.
Voltmeter probes on Batt POS terminal and Vreg Black input.  (With ALL Lighting ON).  Write down V reading.
Voltmeter probes on Batt NEG terminal and Vreg green connection.  (With ALL Lighting ON).  Write down V reading.
Sum the two readings and you have the error voltage the regulator is seeing compared to actual battery voltage.

You can repeat the test with lighting off, which will likely show less error.  But, voltage report errors over 0.5 V indicates wiring , connectors, fuses, or switch contact issues.

My test procedure was intended for stock vehicles with a mechanical regulator.  The B/W resistance test results will not be the same for aftermarket components, as silicon devices behave differently than a switch contact.

Cheers,

[Don R]

 According to Murphy's law a headlight will usually protect a fuse by blowing first. You got lucky.

[TwoTired]

According to Murphy's law a headlight will usually protect a fuse by blowing first. You got lucky.

Murphy was proven to have poor engineering skills.

But really, headlights are more prone to filament failure with inertial shock application than over voltage.  Though over voltage will make shock vibrations more effective at filament breakage.

The stock headlight mounts had vibration isolation for reliability reasons.

Cheers,

[bryanj]

McGinties 3rd law is the real #$%*e

"If something can physicaly be fitted backwards, no matter how stupid it looks, some idiot some where will do it"

[PeWe]

Very often problems with the aftermarket (hot?) running regulator/rectifier combos.
I have never heard about problems with stock charging setup except for corroded connectors, broken/shorted wires, damaged battery and not serviced charging system according to Honda CB750 shop manual ch 8.

[Spanner 1]

Meh, I know my regulator is allowing about 15.5volts to the battery on my K1 750, don't care . My headlight is nice and bright  Some day I'll clean all the connectors on the bike and try to reduce the  1.5v voltage drop at the Black terminal on the regulator. Common on these old bikes  but not a ride killer unless you fell victim  to an ill advised ' electronic' 'update ' where a steady 14v is necessary for ignition function 

[scottly]

Meh, I know my regulator is allowing about 15.5volts to the battery on my K1 750, don't care .

That is batterycide.

[PeWe]

My stock CB750 charging system gave occasional too high voltage ( LED supervision alarmed 1 minute or less). I adjusted the screw on the regulator and it was sorted. About 1/4 turn out, maybe a little bit more. I had adjusted it to maximum before due to 3 ohm coils I later replaced with 5 ohms.

Constant voltage supervision with a LED is very convenient. My latest build got one too.
https://www.ebay.co.uk/itm/LED-Motorcycle-Battery-Alternator-Charge-monitor-level-indicator-warning-bike-G/200826330214?hash=item2ec22e9c66:g:QWIAAOxyY9VROzRz:rk:25:pf:0

[pjlogue]

I would make sure all your wiring connectors are giving solid connections and make sure the wiring harness is in good condition.  A bad corroded connection between the battery and regulator sensor wire can give a high voltage output because the regulator is seeing a false low voltage. 

-P.

[TwoTired]

On the original mechanical regulator, the adjustment is a safety level to prevent the battery from overcharge.

It is NOT, I repeat NOT a system level adjustment for voltage.  The system voltage level is the domain of the battery charge condition.  The Vreg is adjusted so that the battery will never get above 14.5 14.7 V, and that saves the battery from cooking, losing large amounts of electrolyte, which would in turn allow alternator voltage spikes to cook diodes, etc.

If you can't resist mucking with things you don't understand and must tweak the Vreg, do so with minimum load on the electrical system and have the alternator revved to make as much power as possible, with a known fully charged battery.  Under these conditions, adjust for 14.5-14.7 V at the battery.

Now, if lighting is on and and the battery voltage goes too low, you either have placed too much load on the system, or something is broken besides the regulator.  If the voltage is too high and the regulator does not see actual voltage of the battery, you have a reporting path to fix on the bike.  The vreg can't do it's job with lies reported to it.

The Vreg adjustment is the point where the alternator is told to back off it max ability to charge the battery, which varies with RPM and electrical load.  Tweaking that screw changes that safety level.  It does not make the alternator put out any more than ever capable.   It is the wrong way to compensate for system connection losses, and has the potential to damage batteries, and the diodes in the charging system.

I still think the "Voltage Regulator" was misnamed, as it is, in reality, a battery protection device.  The Battery charge status is what actually controls the system voltage.  And even in stock form, the alternator cannot meet max system demands at its lowest RPM.  And, the Vreg can't do anything about that condition but ask the alternator to do more (which the alternator cannot do until it's revved up).

FYI

[jorwesflow]

Thanks, TT. Got some readings.

I recommend you measure the voltage lost between the battery terminals and the Vreg voltage input terminals.

Voltmeter probes on Batt POS terminal and Vreg Black input.  (With ALL Lighting ON).  Write down V reading.

0.39v

Voltmeter probes on Batt NEG terminal and Vreg green connection.  (With ALL Lighting ON).  Write down V reading.

0.23v

Sum the two readings and you have the error voltage the regulator is seeing compared to actual battery voltage. voltage report errors over 0.5 V indicates wiring , connectors, fuses, or switch contact issues.

That would be 0.62v 

I assume this should more or less be equivalent to the difference between these readings:

Battery reading with lights on, bike not running, switched on = 12.52v
Regulator Black and Green reading, bike not running, switched on = 11.92v

So the reg is getting an underreported voltage. I know there is a process to tracking down the culprit with a multimeter—I remember doing it years ago—I just need a refresher.

[TwoTired]

Consider a path from battery to regulator, there are a number of wires and devices in that path, fuse switch wire, etc.  The voltmeter probes placed on each side of the any chosen device will report the voltage difference in to out on the display.

I guess I am most surprised to see that much loss on the  ground return path.  It's double what I would accept.  Can you define that route?

You will never get zero loss in any pathway, as all components have some resistance.  Everything usually works ok if the total error report to the Vreg is  under 0.5V.  But, if your battery needs 14.6V max to survive or keep warranty intact.  You will have to beef up connectors, switches, fuses, and wiring to get to those engineering specs.

Cheers,

[Bodi]

The ground drop is surprising, the battery ground cable should be rather heavy gauge (roughly 6AWG I think) and, since it's a short cable, resistance and voltage drop should be very very low. Have you relocated the battery and run long cables? Once the ground connection is in the frame, resistance is also quite low. So the drop is probably coming from the wiring harness... where is it grounded, how much wire and how many connections are between there and the regulator?
For the "+" voltage drop, an easy fix is to add a relay switching directly (through a fuse!) from battery + to the regulator power input. Relay gets switched on via a black wire - switched power from the keyswitch.
But: with a lithium battery that's not to be fully discharged or be forced into overvoltage, your regulator - designed for lead-acid batteries - may be a problem. I think that models designed for lithium batteries are available. The problem of voltage drop - making the regulator think the battery voltage is lower than it really is and allowing overvoltage - affects any regulator. I would solve that first and see what your max battery voltage is (fully charged battery, lights off, ~4000RPM).
There is no good way to prevent deep discharge in these bikes (other than making sure there is no electrical draw when off, and avoiding use of the parking light). Well designed devices that use lithium batteries will have controllers that disconnect the load when voltage gets dangerously low: that is not an easy modification to our bikes' electrical system and I have not heard of any home mechanics doing it or anyone marketing a protection system to do it.

[BenelliSEI]

Just read this. In response to your very first question: “ blowing head light fuses”, have you checked the fuse box? In my experience, these are the worst culprit. Fuse mounting clips weaken, overheat, melt the actual case, get looser, overheat more, blow constantly..... an endless circle of events. They can be the cause of intermittent misfires, sudden death shut offs and flashing lights! Watch out. They get hot enough to burn fingers, before they pop.

[jorwesflow]

I know it's been a couple months, but I wanted to close this out. I finally got around to working on this and just finished buttoning everything back up.

I threw out the old fuse panel, which I had attempted to salvage and re-solder many times through the years—but finally just tossed it. I used some inline ATM style fuses instead. I also rewired a few things, and removed some unneeded lines since I don't use the starter and I'm kick-only (my 4cell Antigravity battery can't handle the cranks). I also redid a bunch of bullet connectors.

No more overcharging, and the headlight fuse has been fine. The regulator is operating as it should as well.

I don't know exactly wha the culprit was causing the increased resistance—although it could have been a sum of several bad connections. Regardless, all good now!

[Deltarider]

I know it's been a couple months, but I wanted to close this out.

Yeah... and all that time you've managed to conceal what bike it was about. Anyway, owners of a CB500 that happen to have the airfilter case removed, can easily access the regulator from under the seat (see pic) and perform precautionary maintenance. The second pic shows what state of the connectors you can expect to see on a 45 year old bike.

[rotortiller]

The voltage regulator is called such for a reason. It seeks to maintain a specified system maximum voltage. Anytime the battery is being charged (when current is flowing into it) means the alternator is supplying all electrical demands-lights, ignition and battery replenishing. Easily proven with an amp-meter. The battery is simply a reservoir called upon as intended when alternator rpm is low or nill. The only difference with this system over say that of a car is the under-sizing of the alternators lower rpm output. So basically the regulator  maintains system voltage while it has the capacity to do so. So at that time the battery simply goes along for the ride like a good load bearing device would, drawing what it wants. My stock CB750 alternator typically supplies all load demands after 2800 rpm. People try to make it out to be something else for some reason.

[TwoTired]

The voltage regulator is called such for a reason. It seeks to maintain a specified system maximum voltage. Anytime the battery is being charged (when current is flowing into it) means the alternator is supplying all electrical demands-lights, ignition and battery replenishing. Easily proven with an amp-meter. The battery is simply a reservoir called upon as intended when alternator rpm is low or nill. The only difference with this system over say that of a car is the under-sizing of the alternators lower rpm output. So basically the regulator  maintains system voltage while it has the capacity to do so. So at that time the battery simply goes along for the ride like a good load bearing device would, drawing what it wants. My stock CB750 alternator typically supplies all load demands after 2800 rpm. People try to make it out to be something else for some reason.

Philosophical difference, I suppose. 
The battery (if it is big enough), and as long as it is well charged, will run the bike safely forever and maintain a system voltage all by itself.  No voltage regulator needed, as never will it provide more voltage than the system can safely handle.  The smaller the battery the faster it will decrease its voltage as the chemicals complete their conversion.
On the other hand, the alternator CAN make more voltage than the system can handle, especially the battery.  The Voltage regulator only limits the alternator output when the battery voltage gets too high in voltage.  It does so by sensing the system voltage, which is dominated by battery charge status.  In it's depleted state, the battery limits the system voltage, no matter if the alternator is going full output.  The battery absorbs all the alternator power it can make, protecting any device from over voltage.
Yes, the Alternator can supply all the bike needs if it is spinning fast enough and has the field excited.  It doesn't do that at idle, and the battery state still dominates system voltage level.  The regulator can't do anything about that, no matter what it tells the alternator to do, it can't make enough power to keep the instantaneous voltage at the set-to-limit state...until the alternator revs.  Then, and only then, can it limit the alternator output to maintain a safe battery charge state and therefore system voltage level.  In practice, it is far more a battery protection device, and nothing else on the stock bike needs a precise voltage operating parameter to survive.  By design, the battery can never make components burn out from over voltage.  And voltage regulation is not needed for that purpose.

The voltage regulator, imo, is rather misnamed.  It is an over voltage protection device.  As that is the only condition where it's activity is needed.   But, it's probably a distinction lost on the average garage parts changing mechanic. 

Cheers,

[jorwesflow]

Yeah... and all that time you've managed to conceal what bike it was about

Ha oops. 1977 CB550K

[jorwesflow]

Sorry to dig this one back up, but I've been traveling and have had little time to work on/ride the bike. As of my last post, everything appeared to be working fine in my garage—i.e. charging system working right with 14+ V up around 4000rpm (sitting stationary in the garage). So I took it out for a 20 mile ride and it all seemed great, but as I pulled back up the driveway, she started sputtering out and died and I took the remaining momentum up to the garage door. Seems I barely made it back.

I checked the battery and it read 9v. Looks like something happened during the ride and the charging system stopped working, and drained the battery out. Luckily it charged back up and is holding nicely.

So I went through TT's A to L test ( http://forums.sohc4.net/index.php/topic,134836.msg1515431.html#msg1515431 ) and got hung up on D. Here are my results from the rectifier test using the diode setting on my meter (which uses ~2.4v to test diodes)—I also am not sure what the number value is being displayed on the diode setting, so I'll just write it as I see it... Reminder, this is an aftermarket combo reg/rect I bought from Dime City https://www.dimecitycycles.com/vintage-cafe-racer-caferacer-bobber-brat-chopper-custom-motorcycle-electronics-parts-honda-cb350f-cb400f-cb500-cb550-cb750-rectifier-864218.html

POS lead to Red, NEG lead to:
Yellow 1: beep, 0
Yellow 2: beep, 0
Yellow 3: no beep, 1

NEG lead to Red, POS lead to:
Yellow 1: beep, 0
Yellow 2: beep, 0
Yellow 3: no beep, 516

POS lead to Green (ground?), NEG lead to:
Yellow 1: no beep, 487
Yellow 2: no beep, 487
Yellow 3: no beep, 511

NEG lead to green, POS lead to:
Yellow 1: no beep, 1
Yellow 2: no beep, 1
Yellow 3: no beep, 1


It would appear I've got some bad diodes? But any help deciphering these results is appreciated!

[TwoTired]

The reported results say you have one bad diode.  If you can't replace that one, you have to replace the whole unit.

[jorwesflow]

The reported results say you have one bad diode.  If you can't replace that one, you have to replace the whole unit.

Well here i am reviving this godforsaken thread again a year later.

I’ve since replaced with a new regulator/rectifier combo unit a few months ago.

After that, I tested it in my garage at varying RPMs and voltage at the battery stayed around a healthy 14v. So i went out for a ride and about 20 minutes in, the bike crapped out at 60mph. i pull over and see smoke billowing from under the seat. i pull the seat off (aftermarket, with a battery tray underneath for a thin antigravity 4-cell) and see the battery swollen and oozing chemicals everywhere. by gods grace it didn’t catch fire like my first 550 http://forums.sohc4.net/index.php/topic,98440.0.html

So i walked it home 3 miles (slight downgrade, thank god) I then went ahead and bought another Antgravity 4-cell.

Since then I’ve rewired 90% of the bike and reduced what was once a 0.39V loss in the ground line (green) to now 0.02V with lights on.

The 0.23V loss in the switched power line (black) is now 0.18V with lights on.

So i’ve taken it out for some test runs the last few days. I hooked up a digital voltage meter to the battery terminals and stuck in on the top triple tree so i can monitor voltage as i’m riding and avoid any..... potential fires.

Every ride starts out great for a few miles. Barely topping out at 14v. But then it’s almost like something changes around the same time in each ride and it starts spiking past 15v.

So i disconnected the black regulator wire and jumped it straight to the battery positive terminal so the reg would get full battery signal. But the charging was still spiking up past 15v.

So i can only conclude this (somewhat) new reg/rect is trashed. Is it possible it heats up to a certain point and then just stops doing its job?

Perhaps that last time I fried the battery it also fried the vreg? Or maybe the vreg was just shyt from the beginning, combined with too much voltage loss at the black wire.

I don’t know. Am i missing anything else? replace the vreg again? i feel like i’ve been through these charging system tests a dozen times already.

[scottly]

What reg/rect are you using? The voltage spiking past 15 volts is what is smoking your lithium ion Antigravity 4-cell batteries.

[Deltarider]

Personally I am more and more convinced an analogue voltmeter is a better tool to diagnose - dynamically that is - than a DMM.
Many years ago, out of curiosity, I did some testing en route with a DMM and was surprised to read spikes of above 15V. It never worried me as I didn't have any bulbs blown and the battery (conventional lead-acid) has continued to serve me many years ever since.

[jorwesflow]

What reg/rect are you using? The voltage spiking past 15 volts is what is smoking your lithium ion Antigravity 4-cell batteries.

It’s one of these combos from 4into1 https://4into1.com/regulator-rectifier-honda-cb350f-cb400f-cb500-cb550-cb750/.

I had a similar one from dime city (link posted further up in this thread) that worked with my first Antigravity lithium 4-cell for almost 4 years perfectly fine, which was well past it’s expected life, then the battery petered out eventually. So i replaced it with a new one, same Antigravity, and ran it with that same reg/rect for at least a year fine.

But then this thread was started when i blew a headlight and realized i was overcharging. then (as prev posted) i followed twotired’s instructions to find the egregious voltage loss in the ground circuit, which must have been contributing to giving the reg an underreported voltage, and thus causing the overcharging.

But I also (see prev) found that this original dime city reg/rect had a bad diode! That’s when i replaced it with this latest reg/rect from 4into1.

After that is when I fried the Antigravity.

Now i’m on a new Antigravity, and i’ve fixed the voltage loss in the ground circuit. But i’m still hitting 15+

Yeah Antigravity are adamant about not charging past 14.6v. Although my newest one says don’t go above 14.4v. i don’t think the specs changed, so they probably lowered the recommended limit to play it safe.

Sorry for the novel.

[jorwesflow]

Personally I am more and more convinced an analogue voltmeter is a better tool to diagnose - dynamically that is - than a DMM.
Many years ago, out of curiosity, I did some testing en route with a DMM and was surprised to read spikes of above 15V. It never worried me as I didn't have any bulbs blown and the battery (conventional lead-acid) has continued to serve me many years ever since.

I noticed this cheap amazon digital meter I put on the triple tree shows at least .5v higher than my multimeter. Hard to say exactly, because it’s only to the 1/10th. So 12.9 could mean anywhere from 12.85 to 12.94, and my more precise multimeter consistency showed slightly lower readings. But either way, seeing it climb to 15 was enough for me to pull the clutch in and coast at 1000rpm. pyrophobia.

But regardless, you’re saying an analog gauge might be more appropriate. Why is that? I would guess it kind of averages the reading because the mechanical movement is  naturally slower than a digital meter.

[Deltarider]

Personally I am more and more convinced an analogue voltmeter is a better tool to diagnose - dynamically that is - than a DMM.
Many years ago, out of curiosity, I did some testing en route with a DMM and was surprised to read spikes of above 15V. It never worried me as I didn't have any bulbs blown and the battery (conventional lead-acid) has continued to serve me many years ever since.

...
But regardless, you’re saying an analog gauge might be more appropriate. Why is that? I would guess it kind of averages the reading because the mechanical movement is  naturally slower than a digital meter.

Yep, that's it. The mechanical movement of an analogue dampens where the DMM is hysterical  .
The same goes for dwell meters. Analogue is a joy to work with compared to digital, especially when you switch briefly from 1+4 to 2+3 and back to compare. And the same goes for tachometers. Who wants to interpret digits when with analogue you can see from the corner of your eyes what is happening? Not to mention exhaust gas analyzing! Now on long rides I always have my automotive DMM with me, because it's relatively small and is packed with functions. In a workshop it's another story however...
Note that a conventional bulb blown by overvoltage usually will show this typical - what shall I call it - white smoke inside.

[scottly]

I've only seen spikes on a DMM when using the stock mechanical regulator, and are more extreme with a fully charged Lithium battery, but with a solid state regulator the output is much more stable.
To the OP, there are regulators specifically for Lithium batteries as well as adjustable regulators. I personally would consider the Oregon Motorcycle Parts adjustable regulator, set to 14.4V.

[Deltarider]

I've only seen spikes on a DMM when using the stock mechanical regulator... but with a solid state regulator the output is much more stable.

Am I right to assume these spikes are due to induction?

I personally would consider the Oregon Motorcycle Parts adjustable regulator, set to 14.4V.

I keep that in mind for if my OEM regulator would fail. Any idea where one could adjust such an Oregon regulator? I mean, do you have to open it, or can it be done from the outside?