You seem to know more about this stuff than I do so the only thing I can tell you is that there is no diode in the charging circuit aside from those in the rectifier.

 

I had this last month after fitting an Ignitech. My stator windings tested ok for résistance on the charging circuit, but it would barely put out 10 vac on any of the three wires. Put in a g8 stator and now the charging circuit works.

 

This is interesting.
I forgot to mention that I measured the output of the 3 phases open circuit which were about 40vac at fast idle but when I plugged in the reg/rect they dropped to 10vac at any RPM. I figured this was because of the normal load but I have no data to support this.

 

I also recently suffered a charging failure with all usual tests passing. Internal shorting of charging coils is the explanation I’m accepting, one of the few things difficult to test for. 1980 cx500 Deluxe, 20k miles. I’m installing a RaySan 3.5 Full kit and a G8 charge only stator as the bike is in pretty good condition and low mileage. This type of failure seems to be being reported more frequently lately. I’d like to see a definitive diagnosis on postmortem examination. I think I remember that an old fashioned test for internally shorted coils on motors/generators was called a “growler”.

 

My failed charging mentioned above also only registered under 40 vac at 5k rpm, yellow wires disconnected. I would have expected 90 vac or more.

 

The next day I pulled off the Reg/Rect for some tests. All the rectifier diodes measured good.

Did you test all 6 diodes in both directions? With the diodes connected as shown in the FSM I would expect to see full wave rectified output but your scope shows half wave.

Then I explored the sense wire. It disappears into the wiring harness and is not really assessable anywhere convenient. Surprisingly, at the regulator it measures about 11.5vdc with the key turned on (running or not), which is about .7volts below the battery voltage. Is there a diode in this circuit? I don't see one in the FSM wiring diagram. I measured about .6 ohm from the the regulator end to the + battery cable (disconnected). What causes that?

43 years of oxidation in the connections between the battery and the sense wire (including the keyswitch) could easily add up to 0.6Ω

These measurements indicate that there's over an amp flowing in the sense wire. This doesn't seem right to me.

It is my understanding that the sense wire merely measures the system voltage which shouldn't need more than a mA or 2 so that is more than I would have expected too. But I've never measured it so it could be normal.

I'm inclined to jump it over to the fused battery feed which is nearby to see if that helps.
I'd love to hear any additional ideas.

The sense wire's purpose is to determine the system voltage so that the regulator's IC can adjust the regulation threshold depending on the load on the system in real time. In theory the voltage on the red wire (fuse to keyswitch) and the black wire (keyswitch to everything else) should be the same when the key is turned on.
When resistance in the connectors &c causes the voltage on the black wire to be lower than it should be the regulator usually compensates by raising the threshold, which results in a higher than normal voltage at the battery and a number of people have moves the regulator's sense wire to the red wire to correct that condition.
Note: The fact that their batteries didn't drain when they are connected like this would also indicate that the sense wire current should be very low.
I can't see any way that this condition could cause the regulator to decrease its output voltage.

I think I would start by checking that both ends of both battery cables are solidly connected (it wouldn't hurt to loosen each nut and work the lug back & forth a bit to displace any oxidation between the lug and what it connects to) and if that doesn't help measure the voltage across the cables to see if one of them has developed resistance (usually between the lug and the conductor).

The 3 stator windings all measured around 1 ohm and there were no shorts to ground.

It still could be a stator problem. I would expect at least 60-80V between each pair of yellow wires with the reg/rec unplugged.

 

Did you test all 6 diodes in both directions? With the diodes connected as shown in the FSM I would expect to see full wave rectified output but your scope shows half wave.
View attachment 215665


43 years of oxidation in the connections between the battery and the sense wire (including the keyswitch) could easily add up to 0.6Ω


It is my understanding that the sense wire merely measures the system voltage which shouldn't need more than a mA or 2 so that is more than I would have expected too. But I've never measured it so it could be normal.


The sense wire's purpose is to determine the system voltage so that the regulator's IC can adjust the regulation threshold depending on the load on the system in real time. In theory the voltage on the red wire (fuse to keyswitch) and the black wire (keyswitch to everything else) should be the same when the key is turned on.
When resistance in the connectors &c causes the voltage on the black wire to be lower than it should be the regulator usually compensates by raising the threshold, which results in a higher than normal voltage at the battery and a number of people have moves the regulator's sense wire to the red wire to correct that condition.
Note: The fact that their batteries didn't drain when they are connected like this would also indicate that the sense wire current should be very low.
I can't see any way that this condition could cause the regulator to decrease its output voltage.

I think I would start by checking that both ends of both battery cables are solidly connected (it wouldn't hurt to loosen each nut and work the lug back & forth a bit to displace any oxidation between the lug and what it connects to) and if that doesn't help measure the voltage across the cables to see if one of them has developed resistance (usually between the lug and the conductor).


It still could be a stator problem. I would expect at least 60-80V between each pair of yellow wires with the reg/rec unplugged.

Thanks for your input Bob,
I just went out and checked the open circuit AC voltage of all 3 phases at 5000 RPM. They were ~ 57, 58 and 50. Is that too low? As I added more RPM they didn't move up much.

Yes, I did check that all 6 rectifiers were good in both directions as in the FSM.
I have to admit I was a bit perplexed by the lack of a full-wave rectified display too. But, applying the signal across each of the 3 combinations of inputs individually resulted in the same display - so I'm pretty sure nothing was shorted. I may go back and re-visit this test.

I checked the Battery Cables' resistances which were less than .1 ohm from terminal lug to engine, fuse and Regulator output terminal (R/W). This bike has almost never seen any rain - washing yes. There's very little corrosion anywhere.

Somehow I'm going to isolate the sense wire and make sense out of that (pun intended 😉). I intend to make a current measurement through it. I totally agree that it should only be a couple of ma. I'll report back with my findings.
If I need to get a new stator or reg/rect where do you recommend I source them?

 

Water doesn't cause metal to oxidize, oxygen (20% or the atmosphere) does (water acts as a catalyst).

I have no idea where to get them. When I had a stator fail I rewound it myself (I wanted the experience) and when the reg/rec failed I had a spare good used one on hand.
Since you have already converted to Ignitech and no longer need the CDI windings I will recommend changing to a "G8" type stator as used in the TI models. They have the same footprint but without all that space taken up by the CDI stuff the whole space can be used for charging windings, increasing the power available from 150W to 252W.

 

Water doesn't cause metal to oxidize, oxygen (20% or the atmosphere) does (water acts as a catalyst).

I have no idea where to get them. When I had a stator fail I rewound it myself (I wanted the experience) and when the reg/rec failed I had a spare good used one on hand.
Since you have already converted to Ignitech and no longer need the CDI windings I will recommend changing to a "G8" type stator as used in the TI models. They have the same footprint but without all that space taken up by the CDI stuff the whole space can be used for charging windings, increasing the power available from 150W to 252W.

I'm not inclined to rewind my own but that's ambitious and brave that you did! That's a feather in your cap. I will definitely use a G8 type if it comes to that.
When we were kids in the 1950s my dad used to take us into his office at the MIT Instrumentation Labs where he designed motors for guidance systems. We would watch his technician, Marie, winding these tiny little gyro motors in the lab. I tried my hand at rewinding slot car motors as a teenager but never really got hooked.

I just removed the regulator again and measured the resistance from the black sense wire to its green ground wire at 32 Kohm. So, that limits the sense wire current to much less than a ma. There also appears to be an internal diode protecting the sense input from going more that .72V below ground.
I measured the resistance between the black sense wire on the harness connector to the red power wire on it's harness connector. With the key off that's open circuit. With the key on it varies from .3 to .5 ohm. This leaves me scratching my head as to where the voltage drop (.7v) is coming from when everything is connected. I wish I could get at the ignition switch wiring.

Update:
I just discovered that my main fuse (still glass) is measuring .2 ohm and dropping about 1.4 Volt across it.
Could this be causing the issue? Time to upgrade to a blade type I guess.
Who knew?
As I started the process of changing to a blade fuse I discovered that there's an OEM spare 20A glass fuse in the plastic cover of the main fuse enclosure. I installed that. It certainly reduced the voltage drop across the fuse but didn't help the charging issue one bit.

 

A few people had mentioned rewind their own stators on the forum, I thought it would be interesting to try and I had recently retired so I had the time. I learned a few things and the guy at the local electric motor repair place thought it was an interesting thing to try so he sold me the wire and dipped & baked it when I was done all for $35. For that price it was worth trying.

I think there may be 2 separate issues here. At one point I had enough resistance between the battery and the places where the voltmeter I installed in the fairing connected to make the meter read something like 2.5V less than the voltage at the battery when everything connected to that ground was turned on and that had no effect on the battery charging.

It might not be a bad idea to have the battery tested before you start ordering parts.

 

I essentially test the battery every time I hit the starter button. It cranks with enthusiasm and starts right up. It also stays about 11.2 volts while cranking with kill switch engaged. It readily charged up above 14.5 volts with the battery tender in a few hours after it went almost dead. Its a fairly new Yuasa sealed battery and acts like a good battery should.
Speaking of going dead - I'm wondering how the Ignitech kept the bike running so well with the battery so low. Doesn't it use the 12V for power? It was so long ago that I installed it that I can't remember how it's connected.

 

Yes, the Ignitech is powered from the 12V charging system.

 

This morning I measured the three stator phase currents at ~2k rpm with the current clamp:
8A, 9A, 0A !
I think that tells me the stator is toast?
I've ordered a new G8 style stator from Caltric and a used (but tested) OEM rect/reg off eBay. Looks like a winter project.
Is there any place to order a new rear cover gasket w/o having to order a whole new set?

 

https://murrayscarbs.com/product/honda-cx-gl-rear-engine-cover-gasket/ for one.

 

Gaskets – Murrays Carbs

 

you beat me to it... LOL

 

Thanks guys!
Bob, can you say "The problem is probably" three times fast? 😉

Since I already measured ~1ohm between all the stator connector terminals I don't think there's a broken wire on the stator side. And, I measured forward diode drops on all 3 of the reg's stator connector terminals so I don't think there's a broken wire on the regulator side. The connector and wires look great too.

But, I love your idea of switching the wires around to see which side the problem sticks with. I'll report back on that next week. We're about to head off on a 3 -day ride to Burlington, VT on the Ultra Classic.

I did a quick test by pulling all the terminals out of the reg's stator connector and flipped a couple of wires. Now I see about 8 amps ac on all 3 wires. I put the wires back in their original positions and measured again. Still 8 amps on all 3 wires. I guess I must not have had the clamp all the way closed on the original test. Anyway the charge current to the battery remains at 5-6 amps dc. Still too low. Battery holding about 12.4 vdc while running.
Does anyone know what the stator currents should be?

 

I meant this

Note that the 5A is just the current to the battery and does not include what 1s powering other parts of the electric system such as the lights and Ignitech and power that is being dumped by the regulator.
And speaking of the regulator, note also that it works by dumping current to ground to pull the voltage down to the required 14-15V. Even your low reading of 50VAC would rectify to approx. 65VDC so it needs a good bit of current to pull that down. But don't think that the current the regulator draws will significantly reduce the available power because any current used anywhere will contribute to the regulation so when you draw more somewhere else the regulator will need to dump less to achieve its goal.

I did a quick test by pulling all the terminals out of the reg's stator connector and flipped a couple of wires. Now I see about 8 amps ac on all 3 wires. I put the wires back in their original positions and measured again. Still 8 amps on all 3 wires. I guess I must not have had the clamp all the way closed on the original test.

It is possible that you didn't have the clamp closed but it could also be that when you moved the wires around the bad part hidden inside the insulation made contact again.

Anyway the charge current to the battery remains at 5-6 amps dc. Still too low. Battery holding about 12.4 vdc while running.

5-6amps should be sufficient to keep the battery charged. 12.4V at idle is acceptable but at 3000 RPM or more it would be low.

 

Most of us test here for open circuit voltage ac. Probably should be 90 vac @ 5k rpm give or take. 30 vac and it won’t charge.

 

No idea what the current on each leg of the stator should be but IIRC the charging system test in the FSM says somehting about 5A on the DC part.

 

I saw that 5A spec in the FSM. I frankly don't know what that is referring to:

I was led to believe the original stator is good for ~170 Watts. 5A at 14.5 volts is only 72.5 Watts. I'm confused. 170 W should be 11.7A.
I measure 5A charging current in it's present condition. But, the current draw out of the battery is 8A. Is something else drawing too much power? (12.5V x 8A = 100W) Aside from my Zumo and the Ignitech there's nothing else hanging on the electrical system.

 

In industrial 3 phase ac calculations there is a 1.73 factor that comes into play, V x A x 1.73 = total watts if I remember correctly. This should apply here too.

 

It took two months to get to it but I finally replaced the stator with the G8 type I bought from Caltric (PN ST-340). I was pretty nervous about removing all the unused wires - especially ones that had a connection to the Ignitech. But, it started right up. However, now the battery voltage is 15.25 volts instead of 12.4 which seems high. It holds that voltage from idle all the way up to 5k rpm. Battery was fully charged when I started the test.
I swapped my original voltage regulator back in which gave similar results. The only thing I can think of is that the sense wire is disconnected. I'll have to dig into that. As I remember, it was exceedingly difficult to access.

 

Run the sense {black} wire direct to the positive battery terminal and see if that brings the charge rate down.

 

I tried a bunch of measurements this morning and made these significant findings:
1. The voltage at the sense terminal connector is about 14.5 volts - indicating the regulator is doing its job.
2. The voltages at the lighting fuses are about a volt less than the battery voltage indicating there is a high resistance somewhere in the feed from the battery.
3. There is 0.0 ohms between the regulator connector's ground terminal and the battery (-) terminal showing there is no ground problem.

I shorted the black current sense wire to the battery (+) terminal through a current meter and all the lights came on drawing about 6 amps. This indicates a good connection of the sense wire to the lighting circuit.
When I started the bike in this configuration the battery voltage was about 14.8 volts. This seems to have bypassed the problem area.
My conclusion is that there is a high resistance point somewhere between the battery (+) and the lighting circuit causing the sense voltage to be low at the regulator. The regulator has to put out more current to overcome the voltage drop in that circuit.
I believe the sense wire is connected through a switched power connection so that it does not drain the battery when the bike isn't running. It does draw a couple of ma through that connection. A direct connection to the battery would drain it slowly.
Years ago I did repair a break in the feed wire to the ignition switch. I'll look there first.

 

I have a few bikes wired with the sense direct to the battery.

It's not an issue and you can just leave it there. I have a few bikes . They have AGM batteries and start straight up even if not started for months. I don't have them on tenders.

Checking all of the wires and connectors is still good idea.

 

Good info! I'm hesitant to cut into my wiring harness though.
I have traced the problem to the ignition switch which has .2 - 1.4 ohms resistance on those contacts. Just ordered a new one on Ebay.

 

That resistance is probably caused by oxidation on the contacts. Spraying electrical Contact Cleaner into the switch and working it back & forth vigorously a bunch of times will usually fix that.

 

Too late!
I tried to take it apart to inspect it but one of the locking catches broke. Looks like it could have been resurrected as you suggest.

 

Here's the final resolution to this no charging / over charging problem:

1. The stator failure was the root cause for an undercharged battery and replacing it with a G8 stator allowed the battery to start charging again. However, this G8 stator uncovered a new problem.
2. The battery was being overcharged at 15.25 volts after its installation.
3. The reason for overcharge was that the sense wire on the regulator is connected through a circuit path that caused it to see voltage that was about a volt lower than the actual battery voltage.
4. That current path goes from the battery, through the main fuse, through the red wire to the ignition switch connector, through a pair of contacts in the ignition switch, back out through the ignition switch connector to the black wire which supplies power to all the lighting circuits, and finally back to the voltage regulator connector's sense terminal.
5. It was no wonder there was a 1 volt drop in that path!
6. I traced the main culprit to the ignition switch which varied from a few tenths to over 1 ohm.
7. After breaking my factory ignition switch trying to disassemble it I ordered a Chinese made one from Ebay.
8. I installed the new switch and the problem became worse not better. There was an intermittent connection at the ignition switch connector now in addition to the high battery charging voltage. Wiggling the harness connector on the new switch would cause sparking and complete loss of power.
9. Moving on to the connector, I bought new terminals and a crimping tool to replace the main power terminals in the ignition switch harness connector. This gave no improvement what-so-ever.
10. The only thing left was the new ignition switch. I returned the first one and bought an Emgo one made in Taiwan from Parts Giant. I installed that one today. The intermittent problem went away and the battery now charges at 14.8 volts instead of 15.25!
11. I seldom question Honda's engineering. After all, the bike did go 43 years without this becoming an issue. But, I understand that newer Hondas have the sense circuit internal to the regulator.🤨
12. Even with the new connector terminals and a new switch there is still about a .3 volt drop in the sense circuit which is perfectly understandable with the lighting load in those (16 gage?) wires. (It comes out to about .03 ohm) I wonder if Honda took that into account when they set the sense control voltage to 14.5 volts? I'll bet they did!

 

Wow! That was quite the adventure and some good troubleshooting on your part. Thanks for sharing as this is good info for others. It also spotlights some new cheaply made parts are worse than old original ones!