Topic: charging system producing too much juice when HL is on ?!?

[grumpy]

I've had some problems with my charging system that seemed to be coming from the VReg (the battery was seeing 15V above 3,000rpm)
So I replaced the VReg with a solid state unit from OMP. It works perfectly: constant 14.5V to the battery - with the lights turned off.
With the lights on the output jumps 15.25V as measured at the output of the rectifier and across the battery.
What the heck?

HL is a 45/45W halogen
LED tail/brake, no compensator
all other lights stock
CB750 K3

[TwoTired]

We already explained that if the bike tells the vreg that the battery is a lower voltage than what it really is, the vreg will try to overcharge the battery.

You must compare the voltage seen at the vreg Black and green terminals with what the battery voltage is at the battery's terminals.

The headlight current draw will drag down the bike voltage on that black wire, due to interconnect losses or switch contact losses, making the vreg think the battery is a lower voltage than what it really is.

YOU have to stop your bike from lying to the Vreg.

Cheers,

[grumpy]

Yes, you gave a very good (and helpful) explanation of the vreg function. Thanks, I really do appreciate it.
I'm starting to understand how the whole thing works but I don't quite get how to fix it. The black wire off the vreg directly feeds several components and switches, any of which could be bad. The obvious place to start is the HL switch but after that I'm lost.
Is there a wiring diagram that gives the normal operating voltages of all the wires? If I knew what normal was I could test for drains. (old guitar amp schematics do this - they give the expected voltages of all wires in the amp which allows you to isolate a bad component or path). The manuals I have don't show these specs.

[TwoTired]

What is it about this test do you not understand?

You must compare the voltage seen at the vreg Black and green terminals with what the battery voltage is at the battery's terminals.

After that test, and assuming there is a disparity between the two measurements.  You have two electrical paths to scrutinize.
A - The electrical path from the battery POS terminal to the vreg black terminal.
B - The electrical path from the Battery Neg terminal to the Vreg Green terminal.

Path A consists of the following intra-path connections:
1a - Battery post the battery cable.
2a - Battery cable at the solenoid to the red wire for the fuse.
3a - Red wire to Fuse connection input,
4a - Fuse connection output to another red wire
5a - red wire connector to to red wire connector at the key switch block connectoin.
6a - Key switch solder terminals where the red wire is soldered.
7a - The internal key switch contacts.
8a - The black wire solder connection on the back of the key switch.
9a - Inside the harness is a black wire distribution interconnect, presumably soldered.  (You either take the harness apart to physically examine these.  Or, you measure the voltage drop from the key switch terminal to Vreg connector.
10a - The Black wire connector lug to the spade terminal on the Vreg.

Path B consists of the following intra-path connections:
1b - Battery post the battery cable.
2b - Battery cable to frame connection.
3b - Frame connection of the Green wire.
4b - Inside the harness are Green wire distribution interconnects, presumably soldered. (Check these in similar fashion as the black wire distribution.)
5b - The Green wire connector lug to the spade terminal on the Vreg.

So far, I've identified 15 connections that can contribute to voltage loss.  If, for example, each one contributed a 0.1 v drop across it, the voltage at the vreg would be 1.5 volts lower than actual battery terminal voltage.

Not specifically identified in the list above, are the wire to terminal crimps at each connector.  These can also corrode and induce loss of voltage.

I used the Honda shop manual wire diagram found on page 172 for the above connection identification.

Good luck.

[TomC]

Hi TwoTired
     How much of a voltage drop between the battery positive and the black wire? Say the black wire at the voltage regulator. This would be with the engine running, the headlight on, the charging system working. 0.00 volts would be ideal. How much is realistic?
          TomC in Ohio

[TwoTired]

I could tolerate 0.5V to .75V.  More than that, and something needs fixing.
Note that this translates to 5 to 7.5 watts of system power lost as heat through high resistance connections.  (Assuming the bike is drawing 10 amps through the black wires.)

Cheers,

[TomC]

Hi TwoTired
     That is about what I was thinking.
          Thank You! TomC in Ohio

[grumpy]

OK guys, I'm going to go after this thing starting tonight so stand by for more noob questions.
I'll replace every frigging contact if I have to.
OMP says to check the ignition and light switches, too.

heh - I just thought of something. I could totally cheeze it by setting the vreg output to 14.5v with the lights on and then just never shut the lights off while running. 

 

[grumpy]

Here's some test data

bike running, lights off:

pos battery to vreg black = .32v
neg battery to vreg green = .06v

can this be right? seems like very low values. I repeated the test several times.

bile running, lights off:

vreg black to vreg green = 14.13
across battery terminals = 14.38

bike running, lights on

vreg black to vreg green = 14.09
across battery terminals = 15.09

MORE DATA:

I think I found a drain on the ignition switch:

ignition on, bike not running:
battery voltage = 12.78
black wire at vreg to ground = 12.02

[TwoTired]

So, the vreg is seeing a volt less than what the battery actually has.

At a 10 amp draw the loss of 1 volt means that there is .1 ohms in a circuit that had 1/10 of that when new.

I think I found a drain on the ignition switch:

ignition on, bike not running:
battery voltage = 12.78
black wire at vreg to ground = 12.02

You haven't proved that yet.  (Oh, and you failed to state if your headlight was on (probably not, though, if the battery was that high)).
It could be your key switch contacts are worn or dirty or burnt, etc.  You can check this by putting the meter from the red to the black terminals when the bike is drawing full current load (lights on).
The headlight on causes more current to be drawn through the circuit and magnifies voltage losses through any resistance.

You're getting there.  But, still no smoking gun, yet.

Cheers,

[grumpy]

(Oh, and you failed to state if your headlight was on (probably not, though, if the battery was that high)).

light was off

You can check this by putting the meter from the red to the black terminals when the bike is drawing full current load (lights on).

black & red terminals at the ignition?
should the motor be running or not?

Thanks!!

[TwoTired]

You can check this by putting the meter from the red to the black terminals when the bike is drawing full current load (lights on).

black & red terminals at the ignition?
should the motor be running or not?

Back of the switch where you put the key in.

If you do the comparative tests quick enough, you can get meaningful results without the engine actually running, but with all the switches enabled as though it were.  Of course, you will be chasing a moving target, because the battery voltage will go down as the battery drains.
It's not much more stable when the engine is running at idle, and when you rev it, the voltage will climb as the battery recharges from the alternator power.

Another complication is that given a fixed resistance, more current is draw when a higher voltage it provided at the source.  The more current drawn through a resistance, the more voltage is dropped across it.  It's a mathematical formula. I = E/R, E = I*R, and R = E/I. (all the same formula actually)

Assuming there is resistance in your delivery circuit from battery to black wire, you will see the greatest voltage loss in the path when the system voltage is at its highest, and the current demand is also at it's highest.  Since your original complaint was that the battery is overcharging during cruise with the lights on, it would seem like test ought to be performed under those same conditions, or as close as you can make it while sitting in your garage.

By the way, you claimed your headlight was 45/45 watt.   Probably should look again.  Headlights have high and low filaments and they aren't additive.  To get different intensities the two watt ratings are different.  Could be 45/55, 45/65 or even 45/95.  But a 45/45 would be considered, odd.

Cheers,

[grumpy]

OK TT

data from back of key switch:
battery 12.75v
all were done w/ engine off

key on, light off
red to black           .02v
red to brown      11.75v
black to brown   11.75v

key on, light on
red to black           .03v    what the heck is going on here?
red to brown          .54v
black to brown       .50v

ignition off, key out
red to black         12.75
red to brown          0.0
black to brown     12.75

the HL is a 45/45
similar to this one http://store.candlepower.com/4512vh4hebu.html

[TwoTired]

Looks like the biggest voltage loss occurs in the path between the battery and the key switch.  You've got more components in that path to measure voltage loss across under load.

Cheers,

[grumpy]

why did all the voltages drop off with the key on, light on?

does it appear the switch is wired wrong? in your previous post you expected full voltage across the red-black with the key on...

[TwoTired]

I thought I explained that in post # 11.   

Let say you connect Ten- 10Ω resistors end to end.
If you apply +12v to one end of the first 10Ω resistor, but do not connect the other end of the chain, the other end will also measure 12V.  In fact 12V will be measured at any junction of any of the 10 Ω resistors.  This is because there is no current flowing through the resistor.  Voltages are only the potential for electrical flow.  Current is the actual flow measurement.

Now connect the other end of the resistor chain to the power source Neg terminal.  Current now flows through the resistor chain and each resistor will drop a voltage according to the total current drawn through the resistor chain.

Ten - 10Ω resistors total 100Ω.   Using one of the formulas posted earlier we can determine that 12v divided by 100Ω is 0.12 Amps, which is the total current drawn through the entire resistor circuit.  Each resistor in the series is passing 0.12Amp.  Using another of the formulas, we can find that 0.12 Amps times 10Ω is 1.2 volts.  This is the voltage dropped by each resistor in the chain when 12V is fed to the entire chain.  At the junction of the first resistor to the second resistor, you will measure (12v minus 1.2V) or 10.8 V, the next resistor junction in the chain will measure (10.8V minus 1.2V) or 9.6V, and so on for each resistor in the chain, with the last one consuming the last 1.2 v available.

In your bike, between the battery and your Key switch, is a resistor at each connector junction.  I outlined where those were in a prior post (path A, 1a to 10a, and path B, 1b to 5b). Each of these junctions is a potential resistor with magnitude unknown.  But, they will only drop voltage across the junction when current flows through them, and the amount of the voltage drop is related to the magnitude of the current flowing though the resistances.
You are losing one volt from the chain of resistors in this path, when you have the light and key switch on.  I don't know what the current is, but you can find the junction(s) that have too much resistance by measuring across each junction to determine which one has the most voltage loss. ...When the key switch and lighting is on.

I hope this helps.  Because, I'm just about out of explanations to make this any simpler.

Cheers,

[TomC]

Hi Grumpy
     With the engine running at say 2000 rpm and the lights on. Measure the voltage across the main fuse at the clips that hold it. Do this so that your probes are touching the clips and not touching the fuse it's self. Might as well check the other fuses as well. There are as TwoTired pointed out other places that may have resistance but I would check the main fuse first.
          TomC in Ohio

[grumpy]

TT -
Man, it's been 25 years since I had to think about electronics theory. I thought I remembered how it works. Guess not.  I think it's starting to sink back in. Thanks

TomC-
Good call on the fuses.
Half of one of the clips that holds the tail light fuse in place was broken off and stuck between the headlight and the main fuses. It wasn't making contact w/ any metal just then, but it sure looked like it had been bouncing across the other contacts.
I pulled the holder. All the solder joints on the back were pretty corroded & a couple looked like they were about to come loose.

I'm looking for a new fuse box...