81 XJ650.... The DIODE IN THE HEADLIGHT BUCKET Problem.

Hey Gang,
I have been slowly progressing through a BOBBER project. But the only issue I have (fingers crossed) is my Diode for the headlight melted.
What I know is the Diode sends power to the headlight relay. so right now i have no headlight.

I know that i can jumper the relay and get my headlight to work, but i'd like to fix problem.

So in my tests i replaced the diode block, started the bike but noticed right away that the diode started to get hot....REALLY HOT. I shut the bike off before any damage

???Does anyone have a idea as to why its getting hot???

???is it too much voltage doing that???

???how would you test for that???

I have tested for resistance in the White wires from Reg/Rec, one goes to diode. and I checked Voltage and it was 6-7V

any ideas

 

test the charging system.
diode block only carries the current to close relay
also test the headlight relay.
at what rpm were you testing the alt output?
also the test would be ac voltage.

Make sure your headlight relay is a headlight relay
The Ultimate Relay, Switch, Sensor, and Diodes Guide

Identification:
- small metal "cube" relay, inked 3H5-00 on the top face.
- has a yellow (yellow) paint mark on the bottom terminal connector block.
- normally open; has an internal diode.


headlight relay has a latching diode inside so no current should be pulled through the diode block after the headlight turns on

 

The Headlight relay is the correct one...checked and double checked'
I have tested the Headlight relay and its good
any voltage test have been done at an idle

 

the charging system diode is that the reg/rec???

 

sorry for bad punctuation

test the charging system.

diode block only carries the current to close relay



did you do any wire mods?

did you test the diode in the headlight relay?

AC voltage comes from alternator, diode block rectifies it into dc voltage to close relay
the latching diode inside of relay then keeps relay closed. if it has failed the diode block continues to keep relay closed drawing current constantly through the diode block​

 

No wire mods, just repositioned everything.

Ok i am starting to get this Diode thing lol...

I go and test the diode in headlight....

I as thinking about a SHORT TERM FIX.

So I know you can JUMPER where the relay connects to get your Head light on, but its not IDEAL. Could I put a switchable jumper to be able to turn the headlight off while starting, thewn switch on for riding???

 

the diode in the relay is GOOD

 

Since you moved things around this doesn't quite apply, but perhaps you were mislead by the service manual as to the headlight relay location.

https://www.xjbikes.com/forums/threads/the-ultimate-relay-switch-sensor-and-diodes-guide.27543/

XJ650 Maxim and Midnight Maxim models:

Flasher: near the steering headpipe, on right side frame rail, under the right side ignition coil.
Flasher canceller: under the rear of gas tank, behind ignition coils
Starter motor solenoid/relay: on right hand side, attached to and behind the battery box.
Diode block: inside the headlight housing.
Starter/Ignition cut-off relay (no color): behind the turn signal self-canceller. (note: the drawing and the description in the service manual is wrong).
1980-81 Sidestand (kickstand) relay (blue): did not use.
1982-84 Sidestand (kickstand) relay (blue): frame bracket outbound of LH frame tube, just behind the left ignition coil.
Headlight relay (yellow): behind the left frame side cover, on the plastic regulator-rectifier mounting plate. (note: the drawing and the description in the service manual is wrong).
Neutral switch: in a hole in the bottom of the crankcase, just "outside" of the oil pan, on the bottom of the engine, up inside a recessed "well".

Verify the wire colors where the headlight relay is plugged in, they should be:

Harness connector wire colors:
- on all XJ550 all models, XJ650 all models (except Turbo):
* Blue wire with black tracer stripe
* solid White wire
* Red wire with yellow tracer stripe
* solid Black wire

Note that the starter cutoff relay also has an internal diode - here is a diagram of the headlight relay to illustrate pin assignment.



And the starter cutoff relay pin assignment:

Harness connector wire colors:
- on all XJ550 all models, XJ650 all models, XJ750 all non-X models, and XJ1100 models:
* Red wire with white tracer stripe
* another Red wire with white tracer stripe
* Black wire with yellow tracer stripe
* solid Light blue (sky blue) wire

 

@Rooster53 thanks for your reply. I do have the correct Headlight relay (yellow). and its correct socket. but now its nice to see a diagram of the relay.
thinking out loud....
So I should be getting 12V "from Generator" which passes through the DIODE in head Light Bucket to activate the relay.

Why Would my Diode in headlight bucket be getting hot and melting?

Am I producing to much Voltage???

I did a AC voltage test, at idle it was aprox 21v at 3-4000rpm (i have no tach) it was 44v
when I did an resistance test on stator
across the 3 white wires I got 1.1ohms
across brown and green wire I got a reading of 12.8omhs

 

Just means too much current flowing through the diode and obviously the key is to find out why - that is why you are here.

Try checking the white wire at the headlight relay to chassis ground - try it with the headlight relay in and out. You should get about 100 ohms with the relay in and open with the relay out.

Those number look about right if you are checking with the AC Generator disconnected (3 pin connector) and measuring the AC voltage from phase to phase - white to white.

Also go ahead and check the diode block to see if the diode for the headlight relay is now shorted. This is the XJ750 Seca diode block, but I believe they are the same for these two bikes.

 

if you want to bypass the diode block you just need a momentary switch to close the relay like in the test listed in the relay links


For an on-bike test, you can:
- disconnect the White with blue tracer stripe wire at the headlight relay.
- turn on the ignition - the headlight should be off.
- momentarily jumper 12V to the pin on the relay where the white/blue wire was connected, and the headlight should come on and stay on.

 

Try checking the white wire at the headlight relay to chassis ground - try it with the headlight relay in and out. You should get about 100 ohms with the relay in and open with the relay out.

I got 110.5 ohms

Try checking the white wire at the headlight relay to chassis ground - try it with the headlight relay in and out. You should get about 100 ohms with the relay in and open with the relay out.
I get .05 ohms even when I swap test leads

momentarily jumper 12V to the pin on the relay where the white/blue wire was connected, and the headlight should come on and stay on
This works

 

that is how the relay works the latching diode brings power from the output side of the contacts to the input side of the coil in the relay keeping the headlight on ending the current flow through the diode in the diode block. this set up is a government required safety rule.

you could get a cheap relay an on switch and run the headlight relay with it this would keep headlight on if you stalled bike .

you could also use a momentary switch powered from battery to activate headlight relay but risk headlight shutting down in a low voltage situation.

this still leaves you with the unknown issue of why the diode block melted and your replacement gets so hot.

what are you running for a headlight?

 

Clarify this please - so 110 ohms on the white wire with the relay in and .05 ohms on the white wire with the relay out - that wouldn't really make sense. If it is .05 ohm on the white wire with the relay in then the internal diode in the relay is likely shorted and the low resistance is from the headlight bulb.

 

I agree with this because that would make it so that the headlight is drawing amps from the shorted diode up through the white wire overheating the diode block.

 

New Glass and new bulb

 

Sorry guys. The .5 not ,05 was when I checked the diode inside the diode block and i am not sure if Im doing this correct

When I check the diode in the relay I get 501 using the diode setting ->+

 

Using the diagram provided above it is a fairly simple check - W/G to W for the diode of interest. As a sanity check test the other two diodes in the block with the meter set to the diode setting - that should give you some confidence that you are testing it correctly. You can also test the resistor as a sanity check just to know you are probing the correct pins.

Was this replacement diode block new or used?

 

Used diode block

 

did you check the diode in both directions on diode setting in the relay?

you could remove the red white wire from the relay and see if you get 12 volts on the output side of the relay disconnect headlight

 

Yes

So would i run a new relay to the white leg of the head light relay. to give it power engage it.

 

you would no longer use the white wire from alternator /diode block set up. and yes the relay would supply power to that pin

 

you could just add a switch light would shut off with key just cut the white wire from diode block and cap it


no added amp draw and if internal diode failed light would still work
and you only buy a switch

 

I will try this. Thank you.

 

You could do that but if the switch works then a good diode block should work. Did you ever check it to see if the diode is shorted in the diode block? Yamaha could have just put a switch in for the headlight, but they made it that way so you don't forget to turn in on, particularly in the daytime.

 

they made it that way due to DOT regulations. many of the euro bikes have an on off switch for the lights.
lights on . cars are drawn to motorcycles with lights on and can not help hitting themfor safety program when it all happened "studys" came out saying turning on a light on your motorcycle caused accidents. moth to a flame thinking cars are drawn to motorcycles with lights on and can not help hitting them was the nay sayers claim. gvt said better to see you with

all the safety circuits were due to regulation. automatic lights, starter safety circuit. had to protect us from ourselves

 

Yes I agree with safety regulations. I not 100% positive but I think it's the same in BC Canada.

I want a quick fixso I can ride. But also will order a diode block to make sure I have a working head light system

Thanks for your help

 

US dot it still would have spilt over to our friends north of the boarder just for ease of production for yamaha

 

So to post an update.

I was able to find another reg/rec to see if that was my problem as I have basically replaced all other components. And I believe a bad reg/rec was the issue.

Funny thing is. I believed that it tested Good.

Anyway I am going to get some insurance and go for a rip.

 

how did you test it?

 

Followed a video on YouTube. But then I could have done it wrong.

 

2 tests one is to check the diodes in the rectifier second is when bike is running see what the voltage climbs up to. if it goes over 14.8 volts at or above 2k rpm is a problem on the airhead xj bikes

 

from what i remember all my tests were good.

I thought I figured out my problem, but my Diode still get hot.

Does anyone know off hand what the wire colors are on their diode block.

81 xj650. Mine are:
white,
white/green,
blue/white,
green,
black/red

I also know that this diode block is not the original one from my bike.

 

measure your diode block the same way you do a rectifier diode check plus 1 resistor.

did you ever try the headlight relay bypass to see if your block still gets hot?

diode blocks are known to have different colors the wires you want to id are the ones from main harness that connect to the diode block.

another question if you remove headlight relay does the block still get hot?

 

I going to check these in the morning

 

So I removed relay and block remained cool
On to test 2

I bypassed relay and diode block remained cool

relay installed block gets hot,
So does that mean relay is no good? To me it seems that the relay is doing its job.
I think some how the diode block is receiving to much voltage or amperage???

For now will run with a bypass till I have time to put in a switch to bypass relay

 

you can open up the relay and clean the contacts. maybe resolder the internal diode that latches the relay in the closed (on) position.

it sounds like you are still drawing amps through the diode block. which could be a bad diode in the relay or voltage drop in the relay causing the white wire to continue drawing amps through the diode block to keep relay closed.

 

you could measure the voltage drop across the relay contacts . measure input side and output side of contacts.