Regulator/Rectifier

Bike is a 1981 KZ750E. I tested two RRs using the diode mode of a multitester. I followed the ElectroSport charging system flow chart. My original 27 year old RR indicates approximately 600 millivolts in one direction for all three readings on both the white/red and yellows and all three testing black and yellows. In the opposite direction for all connections the meter displays "1", according to the flow chart the RR is good.

I test my less than two-year old aftermarket replacement RR and the meter displays a "1" in both directions on one of the yellows, according to the flow chart the RR is bad.

Here is the interesting part. Using an analog multimeter set to ohms following the procedure of the Factory Service Manual, the old RR gives me the same no continuity reading in both directions on one of the yellows. My new RR tests good using the FSM resistance test.

Is the diode test procedure the more reliable test? Can I be certain of the results enough to use the old RR and toss the newer RR?

I have charging issues, charge the battery it shows 12.6 volts with the digital meter, check with the hydrometer all four balls floating. At 4000 RPM I read 12.8 volts. I have 50 volts ac on all three phase out of the alternator. Any suggestions?

You need a test meter that has a specific diode test range. It should be marked on the dial with a diode symbol. Other meters (especially old analog meters) may or may not give the true reading. It depends on their internal impedance.

A good diode will read the diode drop in one direction and open circuit the other direction. Reading open both ways means the diode is blown open.

drkaw wrote:

I have charging issues, charge the battery it shows 12.6 volts with the digital meter, check with the hydrometer all four balls floating. At 4000 RPM I read 12.8 volts. I have 50 volts ac on all three phase out of the alternator. Any suggestions?

That would indicate the rectifiers are not rectifying the AC and sending it to the regulator. Or, the regulator may not be working and sending a false signal that tells the system the battery voltage is too high.

The system voltage should come up to about 14V at 4k rpm. The battery staying at 12.6V means it is not getting any charging current from the alternator/rectifier assembly.

What wires (wire colors) are on the reg/rec you are trying to test? A newer reg/rec will have 3 yellow, a black/yellow (just black on aftermarket), a white/red or solid red on some aftermarket. The 3 yellow wires are for AC legs coming in. The black/yellow or black is the ground and the white/red or all red is the power coming out. The newer reg/rec also sense how much power is needed by feedback from the white/red. Older units also have a brown wire on Kaw reg/recs. This brown wire is the feedback source. If you have issues with this type reg/rec, unplug the brown and check voltage on the brown wire coming off the harness to make sure it has DC voltage on it... if the brown wire coming off the harness has a break in continuity somewhere (brown subcircuit not working), then you won't get feedback from the main harness so the voltage won't rise when you rev the motor. If you have the newer type with the white/red or red being used as the voltage sensing source, keep in mind that this wire will need to be hooked to the main fuse box AND the battery in order to work as it should... check connections.

My 1981 KZ750E has 3 yellow wires, one black, one brown, on white/red.

My post had more than one question so I think I will restate:

1. Is a digital multi-tester in diode mode (diode test is as described in ElectoSport flow chart) a more reliable test of R/R than testing resistance as described in Factory Service Manual? I am testing both directions on white/red wire against three yellows and both directions on brown against three yellows. A total of 12 tests. Diode test is as described in ElectoSport flow chart.

2. Using either test method, diode tester or ohms, original 1981 R/R indicates good. Aftermarket R/R indictes good testing brown wire and yellows but no good using either test method on white/red wire and yellows.

3. The DC voltage at the battery with engine running is the same with either of the two R/Rs, 12.2 volts @4,000 rpm, not enough! Makes me wonder, are both units bad or good exactly the same or not able to tell?

4. I ran new 12 ga. wire from battery through blade fuse holder to ignition switch and disassembled and cleaned ignition switch. With engine running DC volts between brown wire and positive on battery is 0.65 volts. According to Electrosport flow chart this should be less than 0.2 volts meaning I have a problem in brown circuit between ignition switch and R/R. I will look for a problem in this circuit and then go back to checking R/R.

drkaw wrote:

My 1981 KZ750E has 3 yellow wires, one black, one brown, on white/red.

My post had more than one question so I think I will restate:

1. Is a digital multi-tester in diode mode (diode test is as described in ElectoSport flow chart) a more reliable test of R/R than testing resistance as described in Factory Service Manual? I am testing both directions on white/red wire against three yellows and both directions on brown against three yellows. A total of 12 tests. Diode test is as described in ElectoSport flow chart.

No, because you are also testing the regulating circuit, not just the diodes.

2. Using either test method, diode tester or ohms, original 1981 R/R indicates good. Aftermarket R/R indicates good testing brown wire and yellows but no good using either test method on white/red wire and yellows.

Sounds to me as if your aftermarket r/r is defective (New = shiny; not working)

3. The DC voltage at the battery with engine running is the same with either of the two R/Rs, 12.2 volts @4,000 rpm, not enough! Makes me wonder, are both units bad or good exactly the same or not able to tell?

Assuming your battery with both wires off reads approximately 12v and not 10 (1 bad cell) or 8 (2 bad cells), I would look at the following.

It is starting to sound to me as if you have a high resistance (voltage drop)in:
1. your battery ground
2. your battery positive to the solenoid lug
3. the connection at the solenoid lug, OR
4. the wire from the regulator to the solenoid lug

4. I ran new 12 ga. wire from battery through blade fuse holder to ignition switch and disassembled and cleaned ignition switch. With engine running DC volts between brown wire and positive on battery is 0.65 volts. According to Electrosport flow chart this should be less than 0.2 volts meaning I have a problem in brown circuit between ignition switch and R/R. I will look for a problem in this circuit and then go back to checking R/R.

This is not the immediate problem. The brown wire is the sensor wire to test the "end of loop". A low voltage reading here would cause the regulator to allow more voltage to the red/white wire to compensate.

I do not think your problem is r/r related. I think your problem is a voltage drop (high resistance) in either a wire or a connector.

drkaw wrote:

My 1981 KZ750E has 3 yellow wires, one black, one brown, on white/red.

My post had more than one question so I think I will restate:

1. Is a digital multi-tester in diode mode (diode test is as described in ElectoSport flow chart) a more reliable test of R/R than testing resistance as described in Factory Service Manual? I am testing both directions on white/red wire against three yellows and both directions on brown against three yellows. A total of 12 tests. Diode test is as described in ElectoSport flow chart.

No, because you are also testing the regulating circuit, not just the diodes.

2. Using either test method, diode tester or ohms, original 1981 R/R indicates good. Aftermarket R/R indicates good testing brown wire and yellows but no good using either test method on white/red wire and yellows.

Sounds to me as if your aftermarket r/r is defective (New = shiny; not working)

3. The DC voltage at the battery with engine running is the same with either of the two R/Rs, 12.2 volts @4,000 rpm, not enough! Makes me wonder, are both units bad or good exactly the same or not able to tell?

Assuming your battery with both wires off reads approximately 12v and not 10 (1 bad cell) or 8 (2 bad cells), I would look at the following.

It is starting to sound to me as if you have a high resistance (voltage drop)in:
1. your battery ground
2. your battery positive to the solenoid lug
3. the connection at the solenoid lug, OR
4. the wire from the regulator to the solenoid lug

4. I ran new 12 ga. wire from battery through blade fuse holder to ignition switch and disassembled and cleaned ignition switch. With engine running DC volts between brown wire and positive on battery is 0.65 volts. According to Electrosport flow chart this should be less than 0.2 volts meaning I have a problem in brown circuit between ignition switch and R/R. I will look for a problem in this circuit and then go back to checking R/R.

This is not the immediate problem. The brown wire is the sensor wire to test the "end of loop". A low voltage reading here would cause the regulator to allow more voltage to the red/white wire to compensate.

I do not think your problem is r/r related. I think your problem is a voltage drop (high resistance) in either a wire or a connector.

drkaw wrote:

My 1981 KZ750E has 3 yellow wires, one black, one brown, on white/red.

My post had more than one question so I think I will restate:

1. Is a digital multi-tester in diode mode (diode test is as described in ElectoSport flow chart) a more reliable test of R/R than testing resistance as described in Factory Service Manual?

It is as long as you are only testing one diode. Some dides get a higher voltage rating by putting two in series and the tester won't have enough forward voltage to light them up. If the diode range tester is showing a "good" reading, I'd wager that diode is good.


drkaw wrote:

2. Using either test method, diode tester or ohms, original 1981 R/R indicates good. Aftermarket R/R indictes good testing brown wire and yellows but no good using either test method on white/red wire and yellows.

sounds hosed

drkaw wrote:

3. The DC voltage at the battery with engine running is the same with either of the two R/Rs, 12.2 volts @4,000 rpm, not enough! Makes me wonder, are both units bad or good exactly the same or not able to tell?

Are you sure they are correct units? Some alternators have a single phas eoutput and some have multiple phase. They require different rectifier assemblies. is it possible the wires don't connect right to your alternator? Sometimes things fit the connector but are not connected the same.

drkaw wrote:

4. I ran new 12 ga. wire from battery through blade fuse holder to ignition switch and disassembled and cleaned ignition switch. With engine running DC volts between brown wire and positive on battery is 0.65 volts. According to Electrosport flow chart this should be less than 0.2 volts meaning I have a problem in brown circuit between ignition switch and R/R. I will look for a problem in this circuit and then go back to checking R/R.

Ig switch contacts may be flaky and have some voltage drop there, but I am not sure how that would keep the alternator from workin?

Post edited by: bountyhunter, at: 2008/02/26 13:40

The brown sub-circuit will power your horn, any auxiliary lights, the turn signals (as memory serves). You can also connect ANY switched sub-circuit wire to the brown IN to the reg/rec. I also don't like to continue commenting on the Electrex troubleshooting routine. If the bike were mine, I would check for AC on the output of the stator/generator or whatever you have. This is done by disconnecting the yellow wires from the in to the reg/rec and putting a meter on AC volts and see what you get by putting the probes on any two of the three yellow wires. I suspect you should see between 40 and 50 VAC on any combo. If you get significantly less, you have found the problem. While testing diodes is fine I guess, I would prefer to just test the reg/rec. Connect it back up and measure the voltage on the white/red with the bike at idle and then at 4K. This will verify your reg/rec. I would first find a good feedback wire if you are having problems with the brown circuit and also make double sure the black/yellow or black wire is grounded good. Scrape some paint and ground it to the frame. On some KZs it goes to the wiring harness ground and that may be an issue.

OK... if the reg/rec is about 12 VDC at idle and about 14 VDC at 4K rpm, you are in good shape so far but if not, the reg/rec is the problem. While it seems hard to believe there may be two bad ones out there, you may own two bad ones.

IF your reg/rec works out OK, then test your battery. the white/red splits and one leg goes to the battery to charge it and the other leg to the main fuse. TEST THE BATTERY CELLS for a dead cell using a small and very inexpensive hydrometer. If you have a dead cell, no point in doing more work on this problem till the battery is replaced. A hydrometer is an eyedropper with colored balls in the clear, glass (or plastic) part. You draw in some battery fluid and if all the colored balls float, that cell is good and if not, that cell is either dead to getting to dead.

If the battery tests good you have pretty much fixed anything that would cause a charging problem at this point. Sounds like you may have issues with the ignition switch. You can disassemble and clean but take care not to lose any small springs or BBs lurking inside. You can use your meter to do continuity checks on the various switched subcircuits....