turn signal self cancelling mod.

hi all,
I've just bought another xj600 euro model, (my third now) ,I am going to turn it into a bobber type bike ,sometime in the near future. I am looking to do some mods on the electrical circuits, involving changing the turn signals to led ones , and changing the clocks to digital.
at the moment I am working on the turn signals, I have them all working now with the original bulbs in.
I am well aware of the led situation ,with the flasher relay , and I am hoping to get the self cancelling working from an electronic relay (with the mod write up), but disposing of the reed switch in the speedo.
my question is how can I "simulate" getting the self cancelling working whilst the bike is partially built in my garage. at present everything is stock.
I believe the self cancelling requires approx. 250 metres of forward travel plus 20 seconds of time to have elapsed, the time bit aint a prob, but how can I simulate the forward travel.
would putting a battery drill on the speedo cable give me the "travel " and how does the flasher relay "count" the revolutions.
i've tried grounding the white and green wire that goes to the clocks, then open circuiting it , but cant get them to self cancel in the garage.
anyone know precisely how this system works.
thanks
stu


p.s. ive swopped the relay for a spare, and cleaned (well, sprayed) the contacts in the handlebar switch.

 

you could try the speedo cable thing because the tire rotation will not work
self canceler should just time out but I have never had that happen.
it will distance cancel but I think it has to be above a certain speed to do that.

the reed switch picks up a signal when speedo is moving then cancels out.

there is no drawing I am aware of that tells what is in the little black box.

 

It does not just time out. It must have both time AND distance to cancel. The point is to not cancel when you're sitting at a long light, but also not to cancel too quickly when you're changing lanes at highway speeds.

The reed switch periodically grounds the white/green wire. I have no idea how many pulses it takes to indicate 150 meters to the canceller. I think you would have to measure that, and, if you're not using the stock speedo, you would instead need some other output from the speedo you do use, or a magnet/pickup on the front wheel, to provide a reasonably correct number of grounding pulses.

 

when your thumb gets tired it self cancels

 

, that's brill , pollock.
thanks guy's , i'll do some more investigating

 

I think he has a post some where on that mod

 

The self cancelling unit works on frequency to voltage conversion coupled with an RC time constant. The frequency of the pulses at 30 miles per hour is approximately 74 Hz. The frequency is obtained by the gear ratio of the wheel to speedometer cable drive, and quadrupled as each revolution of the speedometer cable results in 4 cycles on the reed switch. It would take something more than just a magnet / sensor on the wheel to approach those numbers as the wheel revolution at 30 miles per hour is only about 6.5 Hz.

 

Interesting. Seems overly complicated. I wonder why they didn't just use a resettable counter instead....

Anyway, sounds like the simpler solution would probably be a replacement for the self-cancelling unit that did take a pulse from a wheel magnet. A programmable divider and a 555 timer could do it.

 

early 1980's?? I agree, I always thought a simple counter circuit like you mentioned even for that era was doable in a small package.

Oh, and for those that have tried spinning the wheel in theory that should work, however, the circuit has some built in decay in it so it may be hard to stay ahead of that. If you sit at an extra long light or just stop for a prolonged period (couple of minutes) you can notice the point of cancel is further out than what would have been with a quick stop.

 

yes, thanks guys, I did some more investigating today ,
I put a wooden matchstick, in the battery drill ,and spun the speedo, with a buzzer/meter, coupled to the reed switch and ground.
and ,as "rooster" says , the buzzer sounds 4 times for each revolution of the cable.
so I set the signal going on the bike , and put the battery drill on the speedo again , and it cancelled every time , after 1/10 of a mile.
even turning the drill as slow as it could go , which didn't even register on the speedo, it still cancelled after 1/10 of a mile.
the buzzer pulses , even at 5 mph, were indistinguishable , and sounded like a continuous low tone.
I tried to simulate the pulses by setting the signals going, then grounding the reed switch wire as fast as I could ,repeatedly , but I only managed to cancel them once , after several times of trying.
i'm just uploading some pictures , be back in a mo.

 

The stated spec is 150 meters or 490 feet, so 1/10 of a mile (528 feet) is quite close. The distance will be the same up to about 30 MPH at which point the frequency to voltage converter saturates and the cancel time is set strictly by the internal RC network, which is about 10 seconds. And of coarse, the faster you are going above 30 the greater the distance traveled before cancelling. So, the general logic is 490 feet or 10 seconds, whichever occurs last.

 

All of my self-cancelling boxes on all my bikes are defunct, so I'm quite used to manual cancel.
Occasionally I forget though, and remind myself that I've got a simple 10 second timer to install, sometime.

 

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
All of my self-cancelling boxes on all my bikes are defunct, so I'm quite used to manual cancel.
Occasionally I forget though, and remind myself that I've got a simple 10 second timer to install, sometime.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------
yes ,I looked into doing the same , but I ca'nt figure out how a timer would work.
it would only work once. then after its timed out (or open circuits ) then the signals wont work again. its a complicated setup and not easy to replicate
stu

 

I picked up the STS Smart Turn System for my FJR 1300 when they had a sale a few weeks ago. It's always annoyed me that my 1982 Yamaha has self-canceling turn signals while my 2013 does not.

 

A timer, which would be better than nothing, but not nearly as good as original, would need a relay and a 555 circuit. It would reset and start timing when given the enablement pulse that comes on the Y/R being pulled to ground when the turn signal switch is pushed to one side or the other. Once running, it would disconnect the Y/G that runs to the original flasher. When the timer expired (or when the timer first gets power) it needs to come up in a mode that applies 12V to the Y/G wire. So, you pretty much need a relay that includes normally closed contacts to run that 12V through to Y/G.

But, if you're building such a timer circuit, you could also add a pulse counter tied to a front wheel magnet and sensor, so you could implement similar logic of time AND distance before shutoff, just like the original circuit.

And, if you're building all this, you could also just build your own with a modern electronic flasher as well, and not have to disable it quite the same way.

 

OH,

 

I did this a couple of years ago to satisfy curiosity. Here is what is inside the self cancelling unit:



And for those interested PDF schematic attached

 

Hi there, this is just what I have been looking for - for AGES! I really miss my self canceling indicators on my XJ650 and a replacement module also appears to be 'dead' I have checked the wiring and switch thoroughly a few times and have used buckets of contact cleaner! The reed switch is good as well so all that I can conclude is that these units die after 40 plus years!
So I would like to build a replacement myself as I am reasonably proficient at circuit building. Any idea what the the transistors are?
By the way how did you get all the potting off? When I tried to open one of these up it was solid - no way to get at the components! (My plan had been to replace the components on the PBC)
If you had any other photos they would be greatly appreciated.
Many Thanks in advance , Steve Plant

 

I used a heat gun in a well ventilated area with a fan blowing at my back as I am sure those fumes were toxic. Amazingly, the board still worked after all that heat.

The labeling on the transistors was very sketchy, and the only one I could identify for sure was a A720, which should be a 2SA720. That's the output transistor that drives the Flasher Relay.

I have found them to be very reliable, the only one I have that didn't function seemed to have a blown Zener, but I never attempted a repair so there may be more bad parts.

Here is a picture of the back, the transistors are marked with the red ovals



I would think it might be easier to do as SQL suggested and design / build a more modern version than to repair one of these. It is a shame Yamaha used such an unfriendly encapsulating material, but I suspect they didn't want them repaired or copied. As a side note, the encapsulation on the Rect / Reg is much worse and is everything but bullet proof.

I attached a pictorial of the components. Note the reference designators are not marked on the board and I just made those up to keep track of the components.

 

--------------------------------------------------------------------------------------------------------------------------------------------------------------------

Thanks so much for getting back to me so speedily! That is very helpful info. I have been having no luck chipping away so far so I will try a heat gun as you suggest on my remaining non functional unit. Just two more final questions though.
Firstly VR1 puzzles me (Next to C5 and above D7) as I cannot see it in the photo of the stripped off unit or is it just hidden by C5? - and is this a variable resistor or a zener ?

Finally, (and sorry if this is a dumb question!!) but what is page 2 of the attached PDF as it doesn't seem to relate to the photos of the PCB or cleaned components ?

Many thanks for your help, Best regards, Steve P.

PS Now I have two dead units I wonder if the heat from the engine kills them? On my XJ it is mounted directly over the engine.

 

Yes it is hidden, and it is a 6.8V zener

Yea, that shouldn't be there, it is one of the tachometers I was working on that must have gotten merged into that file when I created the PDF

I think that location is pretty common throughout the early XJ series, and with properly rated components it should be OK. Plus most of the time the bikes are moving, which sends lots of cool air over the electronics under the tank.

Did both of your units exhibit the same failure symptoms?

 

OK thanks for the clarifications. The failure symptoms for my original one were that it has gradually become more and more unreliable at canceling the turn signal, only canceling it occasionally now. So I bought a used one but that doesn't cancel at all! (Leading to my theory that they all fail)
I am considering getting a 'New old stock' one but they are £75 (ouch!!) This might clear up the problem once and for all. If it fixes the issue, great, if not then I know the problem is somewhere else in the switch/wiring and I can sell the unit on. However I have checked this side of things pretty rigorously already.
Anyway thanks so much for your help and I will let you know how I get on in due course. Regards, Steve Plant

 

I'm also having problems with indicators. They were working fine until I left the bike alone for a few weeks, now they won't work at all and I can't find out why. Am I right in saying that the reed switch could be responsible? If so, how are they tested? Is it possible to repair?

 

Do you mean the self cancelling function is not working, or the flashers are not working in the left or right position?

 

The whole of the indicator function is not working. The obvious things like fuses are fine. Does the reed switch have the ability to do this?

 

No, the reed switch it is only used to count wheel rotations for the self cancelling unit.

Do any of the flasher illuminate? Is the bike running? Is it connected to a charger?

The self cancelling unit can disable the flashers and in this case they are not illuminated. Unplugging the self cancelling unit rules this out.

Wires can break on the control switch causing no illumination of the flashers on one or both sides, but in this case if one side illuminates it should flash if all else is well.

Insufficient load can cause no flashing, but they will be illuminated - this can be caused by low voltage, dirty sockets, or incorrect wattage bulbs installed.

The flasher relay normally closed contact can open, and in this case none of the flasher illuminate

 

Thanks for the reply. The indicators were working fine until I left the bike for a few weeks in the garage. when I returned, nothing! With the ignition on but engine stationary the whole indicator system is totally inactive. From what you say it seems that the relay could be the issue. Is it possible to test it? It's 5 pin. Cheers.

 

What bike year and model, a 5 pin would likely be aftermarket if it is an XJ series?

 

It's marked Denso FB257H. 5 pin 6mm flat spade conns. Bike is XJ600 1989. I'll send some pics when I figure out how.

 

So you can pretty much ignore what I said as I had assumed earlier version of XJ. And, I can't help much here as I don't have prints or knowledge on that model

 

This is the bike and the part in question. Thanks for your help so far. let me know if you can direct me to how I can test the unit.

 

So I did find this online, but I still have to guess a bit.



The test they have you do appears to check power to the unit (Br) and then power out (Br/W). This would imply like most flasher relays there is a set of normally closed contacts inside the flasher relay, so that makes sense that both pins should have near 12V. If they do both have 12V, then that would point to an issue with the turn signal switch, which directs that to either the left or right flashers and subsequently placing a load on the relay that causes it to operate at a predetermined rate. It's also possible it's bulbs or wiring, but not as likely with both sides not working with no illumination.

Another easy test would be to disconnect the plug from the flasher relay and apply 12V to the Br/W wire on the plug. Then when the turn signal switch is moved to the left or right the applicable flashers front and rear should both illuminate.

The W/G wire is the self cancelling signal that goes to the reed switch. You should be able to toggle that signal by rotating the front wheel and watching the voltage at that pin - not sure of the value but likely toggling between 12V and zero, but could be a lower number and zero depending on the internal circuit.

 

Got it going again!!!
My first 'melt out' was something of a disaster - heat gun to too high! (see photos) But I was much more successful on my remaining duff flasher canel unit. I found that gentle heat made the potting compound go 'rubbery' and more easily removed but it was still a painstaking process.
Once it was all clear I swapped out all the electrolytic capacitors as they were most 'suspect' but oddly they all tested to spec' one removed!
I checked the zener and that was fine as were the resistors and diodes I could get to - but all to no avail when I re-tested it on the bike (using a power drill on the speedo cable)

Finally I was going to move to the transistors when I noticed what appeared to be two dry-ish solder joints. Re-did these joints and bingo! It now seems to work fine and reliably.
I now just have to figure out how to repackage it for permanent fitting back on the bike. I'm a bit reluctant to re-pot it!!!

As a side line a friend of mine says a modern microprocessor version of this is 'quite easy' . I will let you know if he comes up with the goods!
Regards, Steve P

 

Working on this, should I keep going?

 

Soooooo-- Following on from this thread i decided to make my own self canceller just using a 555 timer and a relay as SQLGuy mentioned.
my bikes are all XJ600s and use the flasher relay Rooster posted the diagram to above. as i could'nt find any "how to" on here i scribbled out several circuit diagrams, and ended up with this one , that seems to work (on paper)
there may be different solutions to this, especially as there are different bikes with different diagrams, SQLGuy mentions a Y/G wire for the 750 with the separate flasher and canceller, my bike uses the Y/R yellow /red wire to initiate the self canceller.
heres the diagram i made


so i ordered the 555 adjustable timer and a miniature 10amp relay and an electronic adjustable flasher relay

 

so i raided my parts cupboard, and dug out a spare wiring harness i refurbed last year , a spare ignition switch a spare L/H handlebar switch, and 4 working indicators, and hooked the harness up to an old car battery i had

 

so i started to wire up the parts as per my diagram, all the connections you need are available at the original flasher plug, so using that plug to access the harness, i started wiring in the components

 

the N/O 555 timer contacts interrupt the supply from the brown feed wire to the new electronic flasher relay.
the new flasher rely then feeds the Brown/White connection in the flasher harness plug
i then needed to access the Y/R wire to feed that from the timer coil circuit, back to the harness plug that goes to the handlebar switch initialising finger contacts,
mine was Y/R wire (some may be Y/G) but whichever bike you have , its the wire that feeds the momentary finger contacts in the handlebar switch
shown here with the red straw

 

next i soldered the correct colour wires to the relay using the N/O contacts on the relay, the Brown wire feed to the relay coil, is tapped off the Brown indicator feed wire AFTER the 555 timer contacts, the other coil wire goes to earth

 

the electronic flasher relay i ordered only has 2 spade contacts, cos thats all that is needed, this takes its feed from the original brown wire at the harness plug, and then provides the flashing feed to the indicators on the B/W wire (Brown /White)
the relay is marked B and L, the B connector is the brown wire connection , the L connector is the B/W feed to the indicator handlebar switch Brown /White

 

here it is all boxed up ( need to find a smaller box)
i set the 555 timer to 10 secconds, and the flasher to match the original time of the indicators.

all set up and works every time

 

if the above video does'nt work for you , its on the Test Zone on here on (just testing) thread, till i find out why

 

see if this link works

 

Excellent.
The ten second timer is all you need, as the extra sophistication of the distance travelled is, while nice to have, a frippery.

 

thanks Dave, yeh just need to transfer it to one of my bikes, then build another 2, total cost about £6 GBP for each one

by the way did the video work for you?

 

The video worked for me since around mid day here, so I never saw a busted link

I like the "frippery" of distance traveled as time turning is not always consistent - E.g. stop lights and signs

 

yeh, im new to the video thing, i found that my settings on you-tube were private, and the video was a blank screen at first, once i changed the settings to public the video came up,
you live and learn, ----but i also forget,, i turned 75 today so its excusable.
its also fitting that "haircut 100" Fantastic Day "was playing on my radio in the background

 

Yep, works fine.