The International CBX Owners Association

There’s no doubt; the original CBX Z alternator is not brilliant.
It is mechanically complex (that clutch!), not particularly powerful, and will not begin to charge the battery until quite a few RPM above idle ...

But what bothers me most when I ride (main lights always on) is observing the voltmeter dip to unnerving lows, usually below an indicated 11V, every time I slow down or do any city riding.
And yes, the needle will always creep up, well past 12V when I wind the throttle or drop a gear.

All the above make me glance at the voltmeter more often than not, always wondering about the battery and alternator.
With this behaviour how am I to know when I will get a real alternator failure?
As we all know, when this fails, it is time to switch off all non critical electrical loads and head home at the speed of light (no pun intended). Or else …

But is the puny CBX alternator to blame for the low reading of the voltmeter, or could there be another reason for this behaviour?

For example, why do I get an indicated 10V or so, on my stationary bike (engine kill switch to OFF), with ignition & lights on (low beam)?
Surely my brand new, fully charged monster YUASA does not go from a healthy 12V to 10V in 2 seconds flat?
And yet, at least in my bike (UK spec Z), this is what I have been observing every time. Oh, and before anyone shouts foul, the electrical system of the machine is bog standard down to every detail.

Well, there is one way to find out. So I took the Z instrument cluster out of the bike, and connected the voltmeter alone to a quality power supply and multi meter.

First, and as an aside, I found its current consumption to be 105mA at 13.5V. Wow, we are dissipating here 1.25W just to move a tiny needle! A modern LED digital voltmeter I used later on, never exceeded 10mA, or less than 1/10th of the power consumption of the above ☺.

Then I plotted input voltage (X axis), versus indicated (Y axis), covering its entire operating range, at room temperature: The results point to a highly non linear instrument.
However if we look closely and check its indicated error (absolute value) plotted against input voltage as below, we observe that it is surprisingly accurate, at less than 2% around the all important 11.5 to 13.5 voltage range, and very accurate indeed, at around 1%, between 12 and 13 Volts! So it should be adequate for the job, and yet I was getting those silly 10V readings I mentioned earlier.

Trying to get a feel for the power consumption of the bike with various loads, I disconnected the battery, replacing it with a quality, heavy duty HP power supply. I then turned the ignition to ON (engine kill switch to OFF) and incrementally brought in various loads to see what happens. The results are as shown below ..

Load type Total Incremental
Quiescent: 2.5A - Ignition to ‘On’, Engine kill switch to ‘Off’
+ Parking lights On: 4.5A 2.0A
+ Drive lights On: 8.8A 4.3A 60 W high beam on
+ Brake lights On: 11.5A 2.7A
+ Kill switch to On: PS Out of Range

Interestingly, and despite my rock steady 12V power supply, with ignition on and high beam on, the CBX Voltmeter was showing just over10V!!!

To ensure I was not making a mistake anywhere along this line of thought, I temporarily connected a digital voltmeter across the power supply. At the time the CBX voltmeter had dived to ~10V, this was sitting nicely at 12.0V!

This was simply not right and it was time to dive into the CBX’s circuit diagram, where all was revealed! The way the voltmeter is wired at the very ‘end’ of the puny CBX wiring loom, means its less than stellar accuracy is further seriously compromised by the various loads, connectors and switches along the way. This is true in BOTH the positive AND earth paths, as shown in the diagram above.

Encouraged by the more than acceptable, intrinsic accuracy of the voltmeter in the 11.5V to 13.5V range, I removed the positive wire from the instrument and connected it directly to the power supply (aka battery) positive. And presto! With the same previous load, the voltmeter jumped from an indicated 10V to an indicated 12V!

So, where does that leave us? Well, in my case at least, I will proceed to correct this by way of a minor wiring change to the instrument cluster plus a small relay. Equally important, I have worked a way to make this 100% reversible.

Then, with much more realistic voltmeter readings I will begin enjoying the road rather than worrying about low indicated voltages. And if I observe low readings again (heaven forbid), then I will immediately know there is real trouble on the way, and time to switch off the lights and run home.

Summary of the above:
Various electrical loads, e.g. the bike’s headlamp, acting on the less than perfect wiring of the original CBX Z loom result in erroneous, lower voltage readings of the voltmeter, accentuating the effects of the weak CBX alternator.

Best, Aris

Very nice information, Aris. I never rely on the voltmeter for accurate readings, but I do use it as a relative indicator of whether or not the charging system is working properly. Your observations give me ideas about modifying my CBX voltmeter circuit too. Thanks for all the information.

I like your ideas and I can tell you are much sharper then I am. I am wanting to do the same as you because I am tired of my bogus readings also. Can you tell me exactly how you are going to do it and the relay you are going to use. I am very interested in this. thanks

Hi Aris,

Yet another well documented and detailed explanation of one of our Bikes many Nuances.

Cheers
Chris J.

Perhaps Aris would submit this for publication in the CBX press?

Aris - Your posts are the best !!

Aris: Very nice informative article on the CBX charging system. You are way ahead of me, so I sure appreciate your excellent articles.

Excellent, Aris!

Dave

Thanks guys.

Having spent quite a bit of time analysing the voltmeter issue, actually doing the mod was relatively easy.
So first and foremost, a simple drawing depicting what needs be done: Parts needed are a relay, a 1A inline fuse holder and fuse, plus of course some crimp terminals and wires. First were the mods to the instrument cluster, best shown in the attached pic: With this out of the way, the instrument cluster was reassembled, and ready to put back on the bike.

Then two wired spade (male) crimp terminals were inserted in the 2 empty slots of the main loom side of the 6 slot connector, and these the were routed into the headlamp cowling.

In there, I connected the relay, and finally routed its +12V wire, suitably protected by a 1A in-line fuse to the ‘hot’ side of the starter relay terminal (just for convenience), as shown below. Finally the instrument cluster was put back on the bike, everything reconnected including the battery, and Bob’s your uncle.

The weather was absolutely not right for a proper road test. But firing the bike in the garage and revving the engine as it was warming, was enough to show me that this was a very worthwhile modification. My Z voltmeter will no longer act like a cheapo rev counter ☺, and is for the first time doing, what it was originally designed to do!!!

A few notes regarding the mod:
1. I used what stuff I had around, all automotive grade, of course.
2. The relay, I had sourced some time ago for my Goldwing and I like it because it is long but narrow, therefore easy to accommodate in the crowded headlamp cowling.
3. The only non reversible part of this mod, was cutting the voltmeter terminal connector, as documented in one of the pictures above. You DON’T HAVE TO CUT IT, of course. I did it in a vain attempt to prove to you guys, I have not completely lost my mind.

With this opportunity I would like to wish you all my ICOA friends, Merry Christmas and a very Happy New Year!
Aris

What an elegant solution to a nagging problem. My hat's off to you, Aris.

What a great article! Thank you very much, Aris. I know, this will be appreciated by those who favor the original alternator in their CBXes and have long expressed this preference.
While I am not one of them and have the alternative on all of my CBXes, I think this is an awesome piece of work and will help those who keep the CBX part quite a bit.

Hi everyone and Happy Christmas!!!

Earlier on today the sun finally decided to show its face to a cold-ish Athens.
After all the festivities, food and excesses which go hand in hand, I thought; what better opportunity than to take my Z out for a test ride. Find out in the field, how well our little mod above has fared ...

So even before starting the bike from cold, with ignition 'on', the voltmeter needle settles a tad below 12V, a far cry from around 10V before, and a good initial indicator!

Riding in town, the voltmeter needle hovers between ~12V and ~13V. More important, the span is now narrower. Hence more reassuring than before.

On the motorway, again, initial observations were much - much better.
At an average engine speed of around 4K RPM I was consistently observing:

~14.0V with lights off.
~13.5V with the position light (and hence taillights + instrument lights) on.
~13.2V with all above + low beam on.

Clearly, the readings were much tighter than original. More important, good or bad, I know I can now trust these reading.

So concluding, very happy overall.
I would recommend this modification to anyone with a Z or A model (not sure if it applies equally to the B & C models).
Clearly common sense and care & attention should be applied while performing the modification, as we should always do when dealing with power electrics.

All the best, Aris

Thanks Aris for all your time posting this article it will help alot of people out for sure. I am going to do this mod to all my CBX'S in the future.

EMS wrote:I know, this will be appreciated by those who favor the original alternator in their CBXes and have long expressed this preference.
While I am not one of them and have the alternative on all of my CBXes ...

Mike & everyone, thanks.

What I have set out to do with this research and simple mod, is uncouple the actual operation and behaviour of the CBX alternator itself, parameters which were previously, at least in my case, significantly distorted, as a result of the poor design of the original wiring loom and a less than perfect measuring instrument.

In fact, much more certain that what I see now, is very close to reality (battery volts), I can begin to look deeper into the ageing design of the alternator, see if this can be improved at all (no idea).

And a strictly personal point of view: I absolutely love and adore the looks of the original CBX alternator. It so wonderfully in line with the rest of this magnificent engine. I for one would hate to have to replace it with the undeniably, much superior Kawasaki alternator + adaptor plate. But, this is me.

Best, Aris

Thanks Aris,

I am working my way through some charging issues at the moment and like the idea of making sure my volt meter is telling the truth.

Now I am no electonic whiz at all (far from it), but why do you need to have the relay in this fix? What purpose does it serve? Why can't the hot wire go straight to the + side of the volt meter as long as it is fused to protect it?


cheers

Tony

Hi Tony

The relay is actually essential as the OEM voltmeter has a dreadful current consumption which, if permanently connected to the battery +ve, would drain it in a few short days when standing in the garage.

As I say in my original post:

" ... I found its current consumption to be 105mA at 13.5V. Wow, we are dissipating here 1.25W just to move a tiny needle! A modern LED digital voltmeter I used later on, never exceeded 10mA, or less than 1/10th of the power consumption of the above ☺ ..."

Best regards
Aris