How to convert to LED's and still have everything work

I am making a bit of an assumption here, but I am fairly certain the brake and tail circuits will operate just like the head light circuit that also uses the reed relays. The internal CMS circuitry (output side) will set a fail when not illuminated unless there is a filament present to apply a ground. Since the LED bulbs do not apply a ground, then the filament check must be disabled. The good part of that is it takes very little wattage to disable the filament check. When the brake is applied, and 12V is routed to the CMS (input side), then the CMS relies on the reed relay to set a safe condition. As mentioned before the tail circuit is always on with USA bikes, so the filament check is not really used or needed.

Yea it looks like it is fairly proportional with the chosen reed switch. One coil winding is about 1 amp, two layers is about .5 amps. I suspect the linearity of that might fall off a bit if more layers were added, but not for sure on that.

 

OK I'm unclear on how the CMS work exactly or why LED doesn't "apply a ground." I was assuming (with no actual reference) that when the computer sees 12 V on the brake or tail light, it'd expect to see 12 V on the output of the reed switch, and if not then throw a fault. But now that you mention it, I couldn't see any circuit connected to the output of the reed switch for the tail light (small coil) so I guess that doesn't make sense.

How exactly does the CMS work then to check the tail and brake filaments? Are the reed switches only used to tell the CMS when to check a circuit?

 

I don't have a diagram for the tail and brake, but using the head light reed relay logic you can get a good mental picture of what is going on. You can basically split this one in half, with the top being the tail and bottom brake, or vise versa, and also eliminating the indicator bulb. The logic would be either the reed relay switch is closed controlling the state of a transistor, or the filament pulling the same point to ground to control the transistor. Resistor values and configuration would change so instead of the balanced circuit below that requires two points to go low, the node A or node B pulled low would be sufficient to turn the transistor off. So, this either or is disrupted when an LED bulb is installed as there is no path to ground for the condition of filament check; brake light not illuminated.

Sidebar - I don't currently have a Turbo CMS, and it is possible that the configuration is different. Note the Seca $685, Maxim $382, and Turbo $267 indicates simplified circuitry, and possibly function. That said, the presence of the filament check can be noted by the existence of a non-illuminated voltage at the respective pins. For example on the Seca with the head light H4 socket removed and no head light supply voltage, you will measure approximately 4.4V at the high and low terminals. Likewise, on the Maxim tail and brake terminals with no supply voltage, a voltage of approximately 3.8V appears.

So here is how the head light warning works in the Seca CMS using the reed relays. The tail and brake circuit I suspect are similar on CMS assemblies that use the reed relays for the tail/brake BIT functions. Similarly, the head light circuit for the Maxim and Turbo CMS likely operate on the same design as the Seca below.

Note: This was originally evaluated back in March regarding the subject of installing an LED H4 head light. I had suspected the filament check could be disabled with 1K resistors, but I had neglected to note how the High Beam indicator bulb was wired utilizing the low beam filament as a source for ground. Subsequently, to disable the low beam filament check using the 1K resistor, the high beam bulb must be re-wired to move one side to ground and the other side to the H4 high beam terminal. With the High Beam indicator bulb (incandescent) re-wired as stated it disables the high beam filament check and requires only one resistor on the low beam leg to disable the low beam filament check. Subsequently, the LED bulb must also draw enough current on the illuminated side to activate the reed relay, which to my testing to date is just above 1 amp for both the high and low beam reed relay.

So a brief description of how the circuit works, UK bikes excluded for simplicity.

Q1 must be turned off to go high and indicate a safe condition.

Low beam illuminated:

Low beam illuminated reed relay closes switch S1 and grounds node "A" reducing base current to Q1
High beam filament grounds node "B" through D2 - Q1 turns off indicating a safe condition.

High beam illuminated:
High beam illuminated reed relay closes S2 and grounds node "B" reducing base current to Q1
Low beam filament grounds node "A" through D1 - Q1 turns off indicating a safe condition

So, it is important to note it always takes two to create a safe condition. The LED bulb electrical characteristics will defeat the filament check and consequently requires a load resistor (1K or so) for the non-illuminated low beam side as well as a reconfiguration of the high beam indicator bulb wiring.

Ground though two reeds would also work, but not a normal condition and jumping the reed switches will disable the CMS headlight monitoring for all conditions.

 

Does anyone know how to get the Turbo Seca computer to run through it's checks without having the motorcycle running? I assume that the white wires from the alternator that go to the diode are involved? I'd like to play around with magnets tripping reed switches without having the bike running at the same time.

BTW, for the Turbo Seca, I had to stack two of the magnets to trip the switch from outside the gauges but it was possible. As Rooster53 mentions this is *just* disabling, not getting it to work. What's also nice is that there is a metal bracket right behind the tail light reed switch. It was annoying at first because the magnets would pull off the velcro, but with a steel shim in there I'm confident the magnet wouldn't move.

 

The white wire has to go to 12V on the CMS 9 pin connector, very easy on the Maxim and Seca through the AC Generator main harness mating connector, but the Turbo is quite different.

You could try this, hopefully the wiring diagram can be trusted for accuracy.

Remove the starter cutoff relay or unplug the starter solenoid so the starter will not engage
Or, put the bike in gear with the side stand down to disable the starter circuit
Push the start button (key and kill switch on)
The fuel pump relay should close and the fuel pump should run, hold the start button for approximately 3 seconds and the CMS should initiate a check sequence.
You will have to hold the start button as once you let go the CMS goes back to standby mode.

If that doesn't work you could temporarily remove the white wire on the CMS 9 pin connector. With the key on, kill switch on and 12V applied to the white wire the CMS will stay in check mode.

 

Excellent Rooster53, you were exactly right. I ended up disconnecting the fuel pump, the putting it in 1st with the kickstand down, and holding on the start button. This usually only spins the fuel pump, but I never noticed the computer going through it's checks after 3 seconds.

I also realized that grounding the extra connector off the fuel pump relay could do the same thing! (Not on the diagram). I knew it primed the fuel pump by triggering the fuel pump relay, but didn't realize it ran through all the tests without activating the starting or safety circuits.

As for the LED taillight, the magnet worked perfectly to disable the test. It only took one with a 1/4 inch metal shim, placed between the reed switch outside of the gauge because the gauge has a steel mounting bracket that kept attracting the velcroed magnet. The shim also strengthens the magnetic field from the one magnet. Testing without the shim required two magnets in two stacks, stacked side-to-side over the switches.

The 5 lumen LED looks good for the indicators and I used the 20-30 lumens ones for the back lighting. I went with a bluer white, unfortunately it looks cool everywhere except the speedo and tach because their paint is a little yellow white.

 

Well I do have one interesting consequence from converting the dash to LED. When I pull the clutch in to shift gears (when the bike is running), my Neutral light blinks once rapidly both when I grab and release the clutch. I'm guessing this has something to do with the engine cutoff relay that is looking for either clutch or neutral, and it must ground the line ever so quickly. I think the filament in the old bulb would take time to come on, so it wasn't noticeable, but with the LED it's really clear (and a little annoying).

What do you think I should do? Put the old bulb back in just for the neutral light? Wire something in that takes time to build up voltage at the LED? Ignore it?

 

It really should happen only when the clutch switch opens, but it is not that unusual for switches to have a bit of bounce, so it could occur both ways. What is causing it is the collapsing field in the starter cutoff relay coil that produces a large voltage spike. It is enough power to momentarily illuminate an LED but would be insufficient to light the incandescent bulb.

The easiest solution would be to use a directional LED - hopefully they can tolerate the short burst of voltage (approximately 1 msec at 300 volts) from the collapsing field in the starter cutoff relay, or just change back to the incandescent bulb.

Edit: The directional LED eliminates dampening of the collapsing field from the starter cutoff relay applying that voltage to the diode block diode. Not likely a problem, but reverse rating on the diodes can not be determined as the diodes utilized by Yamaha can not be identified because of lack of markings.

 

Rooster53 you must be an electrical engineer! I thought all LEDs only flowed in one direction, I wonder how they make the bulbs work. Where am I going to find a directional LED...

 

Directional LED's were quite common, and used to be all that were available. I think now they are getting harder to find. To create a bi-directional (non-polarized) LED a full wave bridge rectifier is used that actually adds to the cost and complexity of the device. The higher power LED's get even more complicated, utilizing constant current sources and adding even more to the number of components and complexity of the devices. The advent of the bi-directional LED simplified the installation and allowed them to work without having to consider the polarity of the circuit design.

The only instrument cluster LED I have in house right now is from Walmart. It is made by Sylvania (marked for off road use only), but it is a directional LED so it must be installed observing the correct polarity. It is also impossible to find any detailed specifications for it. I did a quick check at Superbright LED's, and the directional LED's appear to be getting scarce. I did find this one, but don't really know if it would be suitable for brightness. It's almost like you just have to try them, and the choices of beam angle, color, and lumens are all going to affect the perceived brightness. Given what you chose as acceptable, I suspect this may be too bright.

https://www.superbrightleds.com/mor...e-wedge-retrofit-rvb/2493/#tab/Specifications

There are a couple of other options, but I did not mention them because they require modifications to either the existing harness or the starter cutoff relay.

1. Make the bi-directional LED directional with a series diode
2. Add a suppression diode to the starter cutoff relay

Edit: Option 2 making the LED directional eliminates dampening of the collapsing field from the starter cutoff relay applying that voltage to the diode block diode. Not likely a problem, but reverse rating on the diodes can not be determined as the diodes utilized by Yamaha can not be identified because of lack of markings.

Edit: Option 1 will work. However, Yamaha did a nice job of designing this circuit such that shorted components will not cause damage. If the suppression diode were to short, a blown ignition fuse would result.

 

Thanks Rooster53! Man you really have all of the answers!

Also thank you for looking up the LED for me, I wasn't sure what I was even searching for. I definitely agree that 30 lumens would be far too much as a dash indicator light.

It's ironic how bidirectional LEDs are supposed to make the installation simpler. Well I'm in this deep already, I could easily solder an diode into the wire to the neutral light in the gauge. Figure that'd be easier than disassembling the relay. A diode pointing the right way shouldn't hurt anything right? Even if an incandescent is replaced later?

 

Well, my apologies but I am going to change my mind a bit and err on the side of safety. It is true that the pulsing light is caused by the collapsing field from the starter cutoff relay coil, and replacing the bi-directional LED with a directional LED will cure that problem, or adding the diode. But, when I tested this earlier I had a defective incandescent bulb, and misinterpreted how that bulb would actually suppress the spike from the starter cutoff relay coil. Now, it may not be an issue, but doing the mod with the directional LED or the separate diode eliminates the suppression of the starter cutoff relay coil voltage (300V). What now needs to be considered is the reverse voltage of the diode in the diode block and the consequence of it becoming damaged if it is not properly rated. Odds are, its reverse voltage is well above the 300V and likely Yamaha would have selected a diode knowing that bulbs burn and consequently would subject the diode to these spikes. However, I would not recommend risking damaging that diode, which if shorted would allow the starter to engage with just the side stand up while in gear. Therefore, the best logic here is to keep the incandescent bulb for the neutral light and maintain the integrity of the safety circuit.

And yes, the diode would work fine with the incandescent bulb but would also block the suppression of the starter cutoff relay coil voltage.

 

I need to look more carefully at the wiring diagram to understand how the diode block plays into all of this. Is the bi-directional LED ok because it is a load going through the circuit.

Just spit-balling here, but what about two diodes and a resistor, one that allows current to flow from the LED bulb to the coil, and another in the opposite direction into a resistive load, to the LED. That way there's extra resistance in one direction <---Probably overkill, I should just put the incandescent back

EDIT: Or maybe just a diode and a resister in parallel?

 

Yes the bi-directional LED is dampening the voltage on the starter cutoff relay, that is why you see the light flashing when the clutch switch is opened. That is so little power, LED's are amazingly efficient.

You could put a diode and resistor in parallel with the LED, but you have to go down to about 47 ohms (LED efficiency once again). The drawback here is considering worst case (diode shorts) then the 47 ohm resistor sees a full 12V when the neutral switch is closed, which puts it up around 4 watts. You could also just use the diode, which covers dampening the spike from the starter cutoff relay. However, considering worst case again (diode shorts) then you lose your signal fuse in neutral.

I did de-pot an old scrap diode block in hopes of finding some markings on them and determining their specifications, but no such luck on markings.

I can say with a fairly high assurance that the series diode or directional LED would be OK. I don't have the equipment to test the diode block diodes, but I can verify that it is not clipping and therefore rated higher than the voltages being produced by the starter cutoff relay collapsing field. I see this as the best alternative (next to just keeping the incandescent bulb) as the failure of this diode or the diode block is less severe than the other options worst case.

Sidebar: Never noticed before but the original design defeats the safety circuit if the signal fuse blows or if you just remove it. The current path for the starter cutoff relay and side stand relay is now through the neutral indicator bulb to the various low impedance high side signal circuits. The directional LED bulb will eliminate that anomaly as will the bi-directional LED with a series diode.

 

I have a question about the function on the self-canceling unit.

From the original post under item one:

What will happen if the momentarily connection between Y/R and B to the canceling unit is permanent? Will that constantly reset the canceling unit?

Asking because I'm changing my controls on the handlebar and the new ones don't have the function of conduction permanently between one pair and momentarily between two others.

 

Yes it will hold the cancelling unit in reset, and the flasher relay will always be enabled so the flashers would have to be manually turned off.

If you are set on going with those controls you will have to give up the auto cancel feature. The best way would be to just remove the auto cancel unit and the flashers will operate in the manual mode.

 

Hmm Ok, what I suspected...

I'm not entirely set on changing them, this is sort of the deal breaker..
And how complex would a circuit be that only sends an impulse even though constantly on?
Fill up a capacitor? That then blocks DC once loaded? Bad idea?

 

My initial thought is it would get fairly complicated and component heavy. It would be easy to come up with a power up reset, but gets more complex since the flasher relay is in a state of applying and removing power every 1/2 second or so. A simple power reset would then hold the cancel unit in a constant state or reset.

 

It probably will.
Think I came to the same conclusion as you, keep it in constant reset mode, unless indicating R/L using a normally closed relay maybe and steal some other wire to open it once indicating.

 

Put together kits all parts needed and for us dummy's picture s real life and you will sell a lot of them I will gladly pay$$$$ to upgrade my head lights getting old like the bike lol need more light to see and be seen

 

I've been playing with this idea for years but only got to the point where I had to add more resistance there by defeating it's real purpose I've got three bikes I could do it too right away

 

This post is great and I'm installing an LED H4 on my 81 750 SECA. I don't fully understand the note for two reasons: first, I'm not that smart. second, my headlight doesn't function as described. The low beam does not throw a warning at all (which is great), but the high beam does. I understand that the high beam indicator (blue light on dash) is grounded through the low beam (which won't work with the LED). Can I just ground it direct and have it work? Which wire do I ground (don't know how to read the electrical chart)?

 

Nice work along with the write up. Any pics or drawings of the led conversion? Some of us are more visual than over written instructions.

 

The photos are gone. Unless SQLguy decides to become active again and get his photos hosted somewhere that allows third party viewing we won't have access to them again. Keep reading the thread, there are supplemental diagrams and photos throughout.

 

Hey gang,
Long time member but have been away for awhile. It looks like I have at least the pictures from page one of SQLs post pertaining to the seca turn signals and rear led lights. I could try to post them but it seems this thread has evolved beyond that specific topic so I’m not sure if it would help anyone currently. Let me know if it would help any one and I will do what I can this weekend.

 

You cannot post other people's photos without their permission.

 

alrighty then

 

start a conversation with photo owner he may respond

 

That's a nice write up...thanks for doing that!
Cheers.

 

Ok here goes I just installed front and rear led turn signals by just installing an two prong electronic flasher. I just cut the brown and white wire and the brown wire from the factory relay. FYI this was on my 1985 Maxim 700 with 29,000 kil. Also just to let you know my cancelling still works as to most people saying it wouldn’t work.

 

Later model Maxims use a different canceling system. The earlier bikes self-canceling units are driven by the proprietary OEM flasher, and will not work with a standard 2-prong relay.
Please keep in mind that this thread was written specifically for the '81-'83 Seca 750 and Maxim 750.

From:
The Ultimate Relay, Switch, Sensor, and Diodes Guide

Turn Signal Self-Canceller Relay:

NOTE: all XJ700 models and XJ750-X models use a consolidated relay assembly (mounted under and to the rear of the fuel tank, on the wishbone frame triple-tube joint) which combines the turn signal flasher, the turn signal canceller relay unit, and the starter circuit cut-off relay.

 

Very cool that's great but my bike is a seca 750 81 82 83 and has the computer built into it that registers and tells you whether or not a bulb is burned out and I have to fool that computer or modify that computer so that the headlight and turn signals don't tell it it has a lesser load which makes it thinks it has a blown bulb. And then flashes in my face all the time but great work thanks

 

It would certainly take more than just cutting a couple of wires to keep the cancelling unit operational. I assume you wired the new two prong electronic relay to the brown and brown and white wire, which would be appropriate to bypass the consolidated unit to run LED lights. No way doing that is the original cancelling unit going to work; perhaps you could provide a link to the two prong electronic flasher relay that you used and clarification on how it is inserted in the wiring?

 

Sorry it's upside down this is the computer I'm talking about

 

Lovingly referred to as the ATARI.

 

From Amazon and it’s adjustable

 

Do you happen to have a pic of the set up? Looking into the same thing myself on a 700. Thanks

 

Me neither.

 

Yes pictures make all the difference.

 

Having this issue, I converted all to LED, as you see on the video with the engine not running the flashing ( turn signals) works ok, but as soon I turn on the engine all of them goes nuts, flashing faster and the last LEDs on the light are not turning on, any thoughts about it?
and thanks in advance, I really appreciate it.

 

for more info I'm using one of this: Dewhel 1 pc CF-13 CF13 EP34 3 Pin Electronic Flasher Fix 12V 0.02A-20A For LED Turn Signal

 

You need 2-pin flasher relay. The third pin on the stock relay is proprietary, and for the auto-cancel function.
You will also need to add diodes to the dash lamp to prevent voltage from crossing over to th opposite pair of turn signals.

 

Thanks, @k-moe where I can find an electric diagram explaining where and how add the diodes? any link here in the forums? The 3 pin flasher works ok I have the 3rd ping grounded.

 

The diodes are only required when you have a single indicator light, which causes all 4 flashers to blink when either the left or right mode is commanded.

That device uses an internal transistor to enable a 555 timer and it is tied to the output load signal (L). Anytime there is a sufficient load the 555 is enabled and the turn signals will flash. So, I would first double check the ground for the "E" pin and be sure you choose a good point - at the ignition coil mounting bolt is a good spot.

Another consideration is the quality of the 12V powering the device. There is a good chance the ignition coils are inducing voltage into the power line to the CF-13, which would cause the control transistor to switch on and off resetting the 555 and causing the berserk flashing. A .1 uf 100V capacitor from "B" to "E" should cure that if that is the issue.

And one other thought is the load - can you provide a link to the flasher that you are using? The CF-13 is likely to become a bit more unstable if the load is nearing the minimum current spec.

Might be nice to know the year and model of the bike also??

 

Hey, @Rooster53 Thank you man
The bike is a 1981 XJ550 Maxim, here is the link for the CF-13:
https://www.amazon.com/gp/product/B0786CX8LP/ref=ppx_yo_dt_b_asin_title_o02_s00?ie=UTF8&psc=1
The ground is connected from E pin to the ignition coil mounting bolt as it shows in the photo.
I had my thought on this "There is a good chance the ignition coils are inducing voltage into the power line to the CF-13", unfortunately, we don't have a RadioShack here in Manhattan, KS so I can test this today.
thank you for taking care I appreciated it.
I will keep posting after I review everything you wrote here

 

I meant to ask for a link to the actual indicators as I was curious if they specified current / wattage in the link.