• Enter the March CB750 Supply gift certificate giveaway! It's easy... Click here, post something, and you're entered into the drawing!

Stator Resistance way Below Spec?

Gergify

CB750 Member
Messages
8
Reaction score
0
Points
1
Location
Illinois
So my battery isn't getting any charge (<12.5 above 3k rpm). The regulator/rectifier tested ok. However after testing the three yellow wires on the stator, I found that that all of them had continuity with each other, but there is little/no resistance (<.000 ohms). The manual says there should be between .41-.51 ohms. None of the yellow wires are shorting to ground.

Can some one explain what's going on here?
 
It appears the ohm reading doesn't matter much on the stator. Learn something everyday!

The problem actually seems to be the rotors. I'm barely getting .5ohms on the rings and the readings are jumping all over the place
 
Hey Greg...your stator rotor should be reading between 4-6 ohms...anything less than 4 is telling you ,the rotor is on its way out...if your only reading .5 ,it is totally scrapped....you need a new one ...just some advice to save you some cash....do it yourself , don't buy the honda removal tool...just go to any decent fastener place and buy an M18 bolt about 40-50 mm long....that's all it is... Also don't skimp and buy a Chinese cheapo rotor for $90 ...spend twice as much and get a "Ricks rotor" , you won't regret it...np
 
Stator and rotor are definitely NOT the same part. On these it's 90% chance of being the rotor, they just don't last at all. Many will pass a static ohms test and still be bad as well. How you get so many bad Ebay ones. They pass then spinning at higher speed the wires then short out, slow down and they quit to run OK again.

Run bad rotor long enough and it then takes out the regulator from forced overheating so don't wait. Running bad battery takes out the rotor early, see a trend here?
 
I just got via UPS today my newly rewound rotor, from Custom Rewind Inc. Cost was $125. I was a nice guy and shipped him my old rotor as a core.

The rotor arrived in a box with Essex's branding all over it - obviously he buys magnet wire from Essex. The wire part was Thermalex 200, here is the blurb on this wire:

THERMALEX 200® is class 200 when measured in
accordance with the ASTM D2307 test method. Heat
shock resistance meets 220°C.

Let's just say that seeing this - he's won my rewinding business from now on. He could have used plain magnet wire which would have burned up after a year.
 
The quality of the insulation is everything on that rotor. The glue holding it all together figures real big in there too. I've given some thought to how to open up the case to flow some air through there before, the sealed environment really cooks parts there, common to pull cover to find cracked and shed pieces of the coil coatings.
 
The insulation on the magnet wire he used has a softening/melting point way above that of electrical solder. In fact they recommend copper welding for terminating this particular wire.

The easiest way to get cooling on it would be to mount a 1U 12v computer fan on the bike, (this has about a 1inch radius and very high CFM) then duct it with a piece of rubber radiator/coolant hose about that size to a large hole cut in the side of the stator. But, the problem is keeping this water sealed would be a PIA.

The reality though is that use of a modern high temp wire on the rotor is all that's needed. We can produce electrical parts today that run hot enough to burn your thumb in a fraction of a second, so there's no point in Rube Goldberging forced air cooling mechanisms.

The problem here is the design. The rotor is too small for the power output desired. Honda used the same size rotor as their permanent magnet rotor for those engines probably to save on the engine case redesign. But smaller means the windings have to be all next to each other with no gaps for cooling.

With alternators, the faster they spin the lower the field voltage (the rotor voltage) so as the bike accelerates the rotor gets cooler.
 
Are we talking about the glue holding the assembly together when we talk about 'insulation'? If not then you can insulate all you want and still have fails, these are glued pretty much all the way down to the core to lock the windings to each other for zero movement. Movement at all is death to the part. The glue eventually cracks loose to allow the wire loops to abrade at running rpm to eat through any insulation there to short. Or crack wire to an open. So abrasion resistance just as important as temperature. I've cut the wire off down to the core casting and they glue the loops to each other all the way down.

Some rotors will check fine on ohms and run fine until at higher rpm the short happens from centrifugal forces then touching two bare loops. Slow down and it goes right back to charging, it drives you crazy.

Electrical parts today? Just changed another 6G Ford alt today for a bad diode plate, they are absolutely horrible about cracking the diode connections, the missing diodes then take out the regulator. So much for today's technology, that design does that left and right. 4 out of 8 were cracked loose on this one. I've soldered one to get a free alt again but with 4 to solder I have to do too much heat damage to the diode plate, the diode connector plate is way too much of a heat sink and I can't fix without severe overheating to melt the plastic overplate. Got enough of both diodes now to rig a free diode plate but trying to figure out how to assemble it to stop vibration cracking, it's horrible in that alt.
 
I don't know what Custom rewind does, the coil wasn't saturated with enamel or vanish, though. He may have used tape, though, in between layers, Essex sells specific polyester tape for this purpose. In looking at the spec sheet for the Thermalex 200, abrasion resistance is the second quality they have listed in the blurb for it so the magnet wire manufacturers are not unaware of the issue.

My personal feeling is saturating the coil with varnish or enamel is worse then leaving it dry, since as the coil expands due to heat, cracking of the varnish is inevitable. The best choice I think is insulation that slides easily so the wires can expand and move against each other.

The very first rotor I ever pulled off my other CB750 I stupidly used a puller with a slide hammer, one of the legs caught the top rotor plate and fractured it. I put a lot of JB Weld on that to epoxy everything back together, and a year or so later when I pulled that rotor, the epoxy had already cracked, clearly due to heat expansion.

The rotor I sent back as a core was also clearly a rewind job, no glue on the windings either.

We'll see how long this one does either.
 
Dunno and just saying...........the OEM coating glue apparently does crack eventually to let the windings short since the wire itself is only paint insulated. Once the winding gets loose it abrades to short. I'd think at least the outside OD would be coated with glue simply to hold the outside windings in place. The up to 9500 rpm motor redline wreacks havoc with those windings. People have reported that ones that were not glued on the very ends of the loop that comes out to the solder contact on the front plate have ripped loose the last one inch or so of wire from centrifugal forces.

I wonder as well what keeps inertial forces from sliding parts of the coil if say one is doing stupid things like popping the clutch or burnouts that suddenly greatly change the motor speed. Heck, even hard racking of the throttle could probably do it. The winding coil needs to be pretty much locked to the outside casting and to itself internally. I would think anyway. Been wrong before though.

There's probably a reason why Glyptal winding paint/coating is so high priced.

How thick is the wire insulation? Thicker and you lose actual wire amounts to lose charge too.

Longer term project but I have one rotor pressed apart in halves and all cleaned up, playing with idea of winding one myself the slow way. Got the front plate off clean too and the screw holes all cleaned out and threads clean to put new screws in. Thinking about either milling the back off to emulate an 1100 rotor or getting another dead one of those to have the lighter rotor with same charging ability. Been looking at magnet wire brands and a way to vacuum the whole thing down to suck the glue down into it. Wondering how much of an issue balancing will be.
 
I don't think the inertial forces are the issue, I think the magnetic forces are the issue. If the motor goes up to 9500 rpm the regulator should just about completely kill power to the rotor with barely any magnetism being spun around inside the stator. A normal good rotor should have the highest magnetic field on it at 1500rpm because that's when it's turning the slowest (and still producing power) and so the magnetic field must be the strongest to get electricity out of the stator. I think the forces from the drag of the strong field on the wires at slow RPM are far greater than inertial forces at high RPM

Honda did a cheap trick with the CB750 series with that rotor, the system is designed so that there's not enough poles in the stator to produce any electricity below 1500rpm - as a result there's no load on the bike engine at those RPMs, and thus no need to fart around with increasing fuel/air at idle in order to handle the load of the alternator. That engine has a lot of trouble producing enough power at idle just to keep itself running, probably because it's air cooled and it's dangerous to have an air-cooled engine produce high power output when the bike isn't moving and thus air is not cooling the engine.

I know that it's non-intuitive but it goes to the fundamental of power generation. The stronger the magnetic field that passes over the wire the more electricity is produced. A lower RPM field passes fewer times per second over the stator wire thus to get the same power output as you get at high RPM the magnetic field must be stronger.

Tomorrow I'll take a gauss reading off my rewound rotor at different RPM using a smartphone app, to see if the theory is borne out.

If I had an arbor press I would seriously consider rewinding my own rotor. I do think the expense/quality of the coating on the magnet wire is the key thing here.

I also don't think the insulation thickness is of any consequence. The wires in the coil are generating a field that extends several inches beyond the rotor into the stator. An extra thousandth or so of an inch of insulation thickness on the wire is nothing. Keep in mind back in 1907 all they had was cotton-wrapped magnet wire and they built electric motors that could do quite a bit of work with that stuff.

Unfortunately, a lot of the troubleshooting advice dealing with this system is plain wrong. For example, in this system you have a varying output alternator frequency. However, voltmeters are typically calibrated to read voltage at 60Hz. I very much doubt this system produces AC power of 60Hz at any time. Thus, a voltage reading of the wires from the stator, before rectification, is worthless.

To get a true reading of what's going on with the stator, you need a calibrated oscilloscope. For the rotor, voltage input at a given RPM is also worthless since it's dependent on system load which is dependent on the headlight bulb used, what the ignition system needs, etc. However, relative voltage measurements at the rotor terminals would be useful.
 
Several Hondas use that same trick, or not charging at dead idle. My CB550 does the same, it doesn't go to charging until around 2K or so. The alt rotor there is cast iron with no windings, the field is a separate insert coil that fits up inside the hollow rotor poles. No inertia since they don't turn whatsoever and about 500% more reliable even though they have reduced output. The 750 SOHC has the same setup.

I think they don't charge at idle to help with a problem, or the fact that motorcycle engines have a much wider range of useable rpm than say a car. They are above idle so much of the time percentage wise that it is largely not a problem to not charge at idle. And overcharging due to excess rpm is a problem, almost every Japanese bike I've had (a bunch!) has a problem with boiling the water out of the battery in longer term use. From constant high charging rate. The not charging at idle may even let parts cool off a bit.

I have to disagree on the inertia thing, these about 90% are always failed rotor, the stators last even when well cooked. Honda worried about the inertia load enough to not use a crank key to locate them so in extreme load the rotor could slip on the taper without destroying the crank. As I said not gluing the outside lead down all the way to the solder point can fail it lickety-split, several have reported it, the inertia tears them loose. The rotor checking ohm number is so close to zero that it can read in spec and rotor be junk, or internal short across windings but not to ground. A lot have run into that too. Pretty rare to short to ground.

By insulation thickness I was thinking of .010" rather than .001" which would be more of a coating difference. I measured 11 layers of .028" diameter OEM wire when unwinding mine. I have no idea of the insulation you praised earlier whether it was a coating or actual plastic insulation which could quickly add up to lose a coil layer of wire. BTW throwing the rotor into a sealed container of acetone to sit for a couple weeks greatly weakens the epoxy so you can chip/cut it off. The actual pressfit of the two pole sections is not that great, a normal gear puller easily splits them with the windings off already. Glue holds the two halves together easily as much as the pressfit does.

Don't have an oscilloscope however, if I get proper charge output with nil ripple does that matter in the real world? It pretty much tells me the rectifier and regulator are working right and seems to work every time I rebuild an alt, it then goes to working perfectly until more parts wear or give out. The quality of the regulator used figures bigtime in that, or a difference of years. The low dog regulators are utter junk and why so many of the parts store rebuilds break in a week or a month.

Been wondering on these how hard it is to get to the neutral connection of the stator wye, simply adding a couple diodes and another wire could have an easy 10% or more alt power output but the regulator might not like it, they fold up pretty quick already.
 
Last edited:
A scope will tell you if a diode failed, a voltmeter won't. Good trick on the acetone. The insulation I praised earlier appeared to be a coating. My guess is the lack of a key was more because they were concerned about shock if the bike fell over and something bent in the stator cover.
 
If you have known good numbers you can use a voltmeter all day long to pretty much show at least indirectly a missing diode. For instance, on Focus cars with the small 6G alt anything under 14.0 volts at idle is a strong indicator of at least one gone. I use the 13.6-13.8 range as a one at least missing point. Been there way too many times and the broken part so far has always been a diode. Of course other issues can exist but it WILL show. A check of A/C volts in the system is all that a parts store test machine does as well. Or ripple volts. The amount of A/C is the amount of ripple and a direct lead into diodes working. And even the oscilloscope won't show one bad that only drops out under certain conditions.

http://www.venselenterprises.com/techtipsfromdick_files/updatingyourripplevoltagetutorial.pdf

The important part to me is the last two very small paragraphs, one is how to easy check for diodes working right.

The bike falling over destroys the brush holding plate before anything else. Common.
 
A voltmeter works in those auto systems because they use a belt to spin the alternator quite fast even at idle so there is always enough mechanical energy available at the alternator to generate full voltage even at idle. Since the engine computer is so dependent on consistent readings from a multitude of sensors you want full voltage available whenever the engine is running.

But on these CB's of this year, as I mentioned Honda has designed the system so the alternator has no output at idle, the system must be carried by the battery at idle. So it is a lot more difficult to tell with a voltmeter if you have lost a diode or not with a simple voltage level check. That's why a scope is useful.
 
Alligator clips on the VOM leads and meter on lowest A/C range you can get and then spin the running motor up to 3 or 4K just long enough to take a reading off battery terminals............helpful to have previous known good readings to compare but for sure ripple should be under .5 volt. Running right will probably be much lower. FYI the Focus cars read around .015-.025 volt at idle. The ripple is a scope pic turned into a useful number instead, it reflects volts leaking in either direction and the amount. Exactly how the parts stores test them too either outside on car or the alt-out-of-car tester.
 
I really doubt a cheap volt meter would have the resolution to make any sense out of 3000Hz ripple on it's A/C setting, these voltmeter chips are designed assuming 50-60Hz power input for A/C. A Focus alternator at idle - maybe 700rpm with an alternator having a stator with 6-8 poles in it - that's much closer to the frequency the meter was designed for. That's why I think a scope is better - and nowadays small digital scopes are really cheap.

Spinning up to 3K rpm is not conclusive. The bad rotor in my CB750 output 14.2v at 3000rpm. From a voltage reading you could hardly tell the difference between the new good rotor and the old bad one - but, the battery stays charged with ordinary riding on the new rotor, it did not with the old rotor.
 
You measure amp as well as volt-there is a spec on it. And saying the test is inconclusive when I can get a number on a/c? Let's just throw the voltmeter away is what you seem to be saying yet I get what appears to be good result that backs up my rebuilds of stuff all day long and no expensive tools used. I've done it umpteen times and it has always worked at least until crap quality parts broke down again to quit. My 800 pound gorilla. Best of the best parts are crap nowadays. I change basic parts only; I do not buy complete alternators and haven't in 40 years.

No one test is ever conclusive on charging issues, you add several things together to get some kind of a handle on it. I never make a decision based on only one thing.

Of course there are better quality tools but if 100% of what I'm doing is working with a voltmeter I see no need to step up my cost. When I start getting misleading numbers or things not working right I'll look into it. I personally view scopes as nice but not needed and so far I have no reason to change that thinking. I certainly don't seem to have the misdiagnosis issues everybody else on the planet does. One part and fixed in 99% of the cases.

My voltmeter was a $100 part back in the mid '70s. The parts store voltmeters were about $200 each now and out of hundreds of times using them I can't say that even once they mislead me on a diode fail, the number of failures of people trying to fix this or that based on their own conclusions though was astronomical. Meaning I will trust my meter way before any person until I find something that tells me otherwise. I got no other choice.

A simple one minute OD check of say a Focus pulley size and multiply (or divide if needed) it out to direct drive of one of these and you know the speed that according to you you have to be close to. They are 8 pole magnets.

As I have said rpm can make them easily quit charging out of the blue so the flat 14.2 @ 3000 could mean nothing there. Speed up to 4000 and zero volts. Slow back down to 3000 and 14.2 volts. I've seen it. The bad rotor as well ohm tested as a good one. New rotor fixed it. The bike driven sedately never ran the battery down at all, only when it got put out on the highway at 70+ mph did the problem show up. I slowed down (100 miles away from home) and took a slower indirect route and made it back and the battery which had run down enough to not crank before was back up enough to start it again. I actually used the bike for a while longer as my car at the time was down, I was repairing it. I just drove it at lower rpm and battery stayed charged fine and always started it up.

They can flaw in more than a 1000 different ways, it depends on exactly where in the coil the short or open happens, you can have what appears to be good power but not quite enough and runs battery down at slow speeds and you can have just the opposite like I detailed higher up. Or fine at slow and bad at higher. Any open will have more of a marked effect than a short which just charges less, it can be slightly less or a lot less. Unless it is a short to ground and not common on these, they almost always short layer to layer of wire. An open when it happens makes the charging quit. It could even mimic the partial short if the open were only say part way across the wire to be cracked.

The Focus car diodes do that all day long, vibration makes them contact then not for reduced charge but not all the time, they often test good repeatedly until finally problem gets bad enough to show up on a volt test. What my last one did, I knew something was wrong with it but it didn't show up solid for like 2-3 weeks after an initial incident. Testing it through there showed it to be good several times.

Somebody else who rewound his own DOHC rotor.......................

http://www.cb750c.com/modules.php?name=Forums&file=viewtopic&t=15474&start=0
 
Last edited:
I'm sitting here wondering why I even need more resolution on a voltmeter when reading higher rpm a/c, the volts rise by both frequency and amplitude, either one increases the voltage since it is accumulative based on frequency. Drop the regulation off and the volt meter easily reads the 75-100 volts the a/c can build to. Are you saying those numbers are not real as well?
 
Yes, I am saying those numbers are meaningless - on a DVM. A DVM works by sampling then averaging the samples, I don't know what your average DVM samples at but here's a discussion about a fairly newish DVM from Radio Shack that uses a serial port outputting at 4800bps

http://stackoverflow.com/questions/18473869/reading-data-over-serial-port-from-voltmeter

If the meter outputs at 4800 bps then it's probably got a sample rate around 4K at max - and since the meter is sampling the AC asynchronously your not going to get enough data points for the CPU in the multimeter to average them to calculate RMS voltage accurately unless your running at a frequency of perhaps no more than 1000Hz, that would be 3 data points per transition.

Of course this assumes the DVM chip is calculating RMS - my guess is in practice they run the AC into a resistor and capacitor that averages - which is worse because now you have a basic frequency filter - so your going to get voltage reading swings as frequency increases.

An interesting test would be is if you have a frequency generator to set it to generate a 3v sinewave and slowly run it from 1 Hz up to 20K Hz and see in what range the meter reads 3 volts. I'd bet there is only a narrow range of frequency it's accurate.

This kind of thing is why I still keep an analog DVM around the garage. A needle is a lot more responsive than a DVM for changing voltages.

Now I am NOT saying to throw away the voltmeter because it's still usable on COMPARITIVE measurements, that's what your missing. Meaning, you can take a perfectly working, functioning, system, connect a voltmeter, do your RPM tests, record the voltages, then take that same meter to a malfunctioning vehicle and the meter will show the fault. None of the sampling rate error matters since it's the same error introduced on both systems and is thus canceled out.

What you CAN'T do is say that on YOUR meter that at 4K rpm someone should see XYZ amount of AC voltage at the stator output because they may be using a different meter model which has different characteristics. This is why I say a scope is better - since your scope and my scope can come from different manufacturers but they will readout the same on a frequency measurement.

Now, as for the rest of it:

I agree there's a 1000 ways these things can fail and you bring up some good scenarios. We could talk about alternator failures all day long. But, limiting it to -this- forum specifically about the CB's of this year, the weak point in the system appears to be the rotor. And in my opinion, the root cause is heat, and inadequate materials selection.

I agree you can have chafing and rubbing of wires so sure, immobilization with varnish or epoxy might have some benefits. But, using cheap varnish or epoxy just means that once the rotor heats up the immobilization goes to hell because the immobilizer material disintegrates - since it can't withstand the heat.

Honda probably used the same marginal varnish on the stator coils but got away with it because the stator coils don't move. It's the old story of engineer it to be as cheap as possible so that it fails right after the warranty expires. But the rotors are clearly a design failure. Sure you can argue this is a 30 year old bike and the rotor is just old. Of course that completely ignores that the stators don't die, and it ignores that there's tons of 30 year old electric motors out there that are perfectly fine. The reality here is that Honda just used too-cheap materials. As I posted, there's magnet wire with insulation far more heat resistant, Honda should have used that wire in the rotor. I discount wire chafing as the root of the problem, I see it as more of a symptom. If the insulation isn't heat resistant then when the rotor gets hot, the insulation softens so any wire chafing or movement in the spinning rotor is going to saw right through the softened insulation like it was butter.
 
Back
Top