Car vs. Moto Vibration - Coil Relay Mod?

[markdjr]

So I just had another relay go out.... This one was nice as it happened right after I hit a wicked rut in the road, so it must have been the vibration/bump. But what I don't get it how could the vibration be that different from a car to the bike? I have used tons of these relays for auto alarms, ziptied loosely to a wire coil under dash, and I have done similar for the bike, ziptied loosely near the battery, plenty of wire slack. It shouldn't be seeing much vibration. At this point I've read where everyone else has placed theres and how long they've worked for. But is there an actual, measurable way to understand how the vibration in the bike is so bad that it kills such a high percentage of these relays?

[posplayr]

Well there is usually always a way to measure

the vibration with instrumentation (accelerometers being the most direct) but in lieu of that please consider:

1st: MC have much higher power to weight ratio than a car.
Compare 100 hp/500 lbs v.s. 200hp/4500 lbs that is close to a 5:1 ratio

The weight of the live mass compared to the rotating mass or perhaps hp might be a resonable measure to compare vibrations levels of the frame (all else being the same which it is not of course).


2nd: MC engine is virtually hard mounted to the frame. All cars have rubber motor mounts . Rubber mounts probably start to impede any vibration above 10 hz (a guess) 80 Hz is then almost 100:1 (40 db approx)

These two factors alone would explain, much higher high frequency (50 + hz; 5K RPM=> 83 Hz) frame vibrations levels than any car.

I'll stop at this level. Mount your relays in rubber; not hard fixed to the frame.

Posplayr

[simon_C]

wrap the relay in some semi dense foam (like seat foam) and shove it into a piece of pvc pipe. mount pvc pipe to frame
voila. no more vibe issues

[focus frenzy]

check your voltage, something is not right. you shouldn't be going through relays like this.where do you have it grounded to?

[markdjr]

the vibration with instrumentation (accelerometers being the most direct) but in lieu of that please consider:

1st: MC have much higher power to weight ratio than a car.
Compare 100 hp/500 lbs v.s. 200hp/4500 lbs that is close to a 5:1 ratio

The weight of the live mass compared to the rotating mass or perhaps hp might be a resonable measure to compare vibrations levels of the frame (all else being the same which it is not of course).


2nd: MC engine is virtually hard mounted to the frame. All cars have rubber motor mounts . Rubber mounts probably start to impede any vibration above 10 hz (a guess) 80 Hz is then almost 100:1 (40 db approx)

These two factors alone would explain, much higher high frequency (50 + hz; 5K RPM=> 83 Hz) frame vibrations levels than any car.

I'll stop at this level. Mount your relays in rubber; not hard fixed to the frame.

Posplayr

This is a great reply, I appreciate it. Although I am I feel accomplishing the same as rubber mount or foam with my loose fit zip tie, essentially it's floating as it's in between other wires (ie no tension or hard fit to frame). But the resonance itself is a great point, is it resonance similar to sound waves? Could a particular rpm range or db or sound or what not be tweaking the inner coils of these relays just such as to degrade them at a higher than usual rate?

[rokacook]

What effect does the relay going out have on the bike?? No Start?? Poor performance?? I am stupid or at least ignorant!!

[salty_monk]

The relay powers the coils after this mod... therefore when it goes the coils don't produce a spark so the engine won't run.

Dan

[SqDancerLynn1]

the bike shuts down, like turning the switch off

[Matchless]

If you have a look at the flasher unit, you will see that it is just a relay in another configuration, but mounted with a rubber mount and mine is 28 years old already and still working! Are they better built, designed to work in such an environment, is there less vibration on the side of the battery box - possibly, just think about it.

Have you opened it up to see what actually failed, such as a wire broken off or has the coil burnt out? If the relay coil is burnt out it may be caused by something else. Is it actually still testing faulty?

If your original problem/reason for installing the relay is still present it could
likely be related to this problem. An intermittent wire connection that makes the relay lose power and could correct itself as you move the wires to look at the relay would also cause someone to immediately suspect the "new" relay. Not saying this is it, just that this sort of thing is known to happen and I do not know the complete background.

Easy way to limit some vibration to the relay is mounting it with some rubber. Cut a small piece of reinforced rubber and fit it to the relay tab with a small bolt then bolt the rubber to the frame so that the frame does not touch the relay case. You could also use a rubber grommet, just enlarge the jole in the relay tab, push the rubber grommet on and use a bolt with a lock nut.

[Jaguarjoe]

If the lead wires to the coil are breaking due to vibration you can easily fix that by potting those wires in some silicone sealer to prevent them from moving.

[Sandy]

I don't understand why there have been so many relay failures with the coil mods unless the relays these days are real crap. It sounds like they are failing on the coil (the one inside the relay) side do to vibration. The only time I've run this style of relay on a motorcycle was with a '77 Yamaha 750 triple where I used it for a horn relay. It was bolted directly to the frame an last for 4 years and 80,000 kms without failing, so obviously vibration wasn't an issue. When the wire to the coil breaks look at it closely and try to see if it is a clean break or does it look like it melted like when you look at a blown fuse with a small ball of melted material on the end. If the coil has a good ground there shouldn't be any excessive heat on that wire unless it's mounted in an excessively hot spot or you are having too much voltage or possibly a voltage spike. One thing you might try is to pop the cover off the relay and put a dab of clear silicone on the coil wire to stabilize it and also check to see if there is any movement at the coil itself and maybe silcone that too.

[Skateguy50]

You all got me worried now about my relay... its mounted off the frame cross brace under the seat. I have it zip tied on but not hanging that loose.

Also the idea of possibly failing and having no backup isnt great, should I get another relay to keep in my tool kit or is there a short term fix you could do on the roadside if it failed?

[bakalorz]

Relay mod reliability increase ideas

Since it would kind of suck to have your ignition fail during rush hour on the highway, here are two ways to make the relay system more failure tolerant.

Both involve paralleling the relay with something to take over when it fails.

The simplest concept is two use two relays in parallel.
Both share the load.
If either fails, the other keeps the bike running till the failed one is replaced.
Since the bike keeps running fine even if one relay has failed, we need a way to check that both are working.
So you put an on/off switch in the control line to each relay. (these switches are normally both left on)
Each switch turns off one relay, so that the other one has to keep the bike running all by itself.
When you start the bike, you turn one switch off and then back on, then do it to the other one.
If the bike stops during this, one of the relays is bad.
If it continues to run, both relays are good; so you confirm that both switches are on and drive away.
Any pilots among you will recognise this as being the same as the dual magneto check procedure on (some? all?) piston engine planes.

The other (simpler) way to do it would be to use a diode to provide a parallel current path instead of another relay.
The diode would go from the positive control terminal of the relay (the one coming from ignition/cutoff-switch) to the output terminal of the relay (the one going to the ignition coils).
The band end of the diode would be toward the ignition coil terminals.
If the relay fails, the diode would basically re-establish the OEM connections (albeit with an extra .3 volt drop) and let the bike keep running.
When the relay is good, the extra drop of the diode keeps it effectively out of the circuit.
You need to use a diode instead of just a wire, because otherwise the relay would never turn off once it was on, since it would allow power to go "backwards" to its own control terminal once started.
You would want to use a schottky diode (since it has a lower voltage drop) of at least 3 or 4 amps ratting. If you are willing to live with a .7 volt drop, you could use a regular power diode.
(The drop would only exist if the relay fails)
To check for correct installation (and/or a burnt out diode), pull the fuse supplying power to the relay...the bike should continue to run.
To check for a burnt relay, there would be an SPST switch in series with the diode.
It is usually turned on to keep the emergency diode path available; to test the relay you turn it off, if the bike keeps running the relay is good.
Some have said that they can tell that the relay mod was installed because the bike runs noticably better. If this is true for you, you don't need a switch in series with the diode since you will be able to tell that the relay is bad when the bike suddenly runs roughter.

If any of you are interested in doing either of these but can't understand it from the written descriptions, I can eventually make schematics, but it will take a while to get to

[bwringer]

I don't understand why there have been so many relay failures with the coil mods unless the relays these days are real crap.

Yes, most commonly available electronic and electrical components you find these days are, in fact, badly made cheap crap. Some of it is quite frightening crap.

Even if you're perfectly willing to pay more, it's very hard to find components that weren't glued together in a third-world sweatshop by the very lowest bidder.


Some good info (and a good source) about quality relays:

http://easternbeaver.com/Main/Faqs/Relay_Info/relay_info.html

[bakalorz]

So I just had another relay go out.... This one was nice as it happened right after I hit a wicked rut in the road, so it must have been the vibration/bump. But what I don't get it how could the vibration be that different from a car to the bike? I have used tons of these relays for auto alarms, ziptied loosely to a wire coil under dash, and I have done similar for the bike, ziptied loosely near the battery, plenty of wire slack. It shouldn't be seeing much vibration. At this point I've read where everyone else has placed theres and how long they've worked for. But is there an actual, measurable way to understand how the vibration in the bike is so bad that it kills such a high percentage of these relays?

I'll agree with the other stuff said about vibration in this thread, but add four more points.

Whenever you have a mass mounted on a flexible mount, it will have a resonance at one (or more) frequencies. If the forcing frequency is the same as the resonance frequency the amount of vibration will be much higher than if the item was hard mounted. The resonance will be affected by the mass and the stiffness of the mounting. The point being that although the spring/rubber mounting is usually better, its not always so.

When something is soft mounted, it is more free to move. Under some conditions it will be able to move until it hits some kind of constraint, at which point it can't move any further in that direction.
When it hits the constraint, it may be subject to a shock load, which is much worse than the vibration would have been.
Again, the point is that although the spring/rubber mounting is usually better, its not always so.

The vibration sensitivity of mechanically complex parts (like relays) can be very different in different orientations. If yours keep breaking, try turning them sideways, or on end, and see if that helps.

The frame of the bike seems "rigid" to us, but at the level of the vibrations, some parts will be nodes, and some anti-nodes. So hard mounting to the frame at one point may kill the relays, but hard mounting the exact same way 6 inches away from there on the frame may be fine.

The whole point of all this is that with regard to vibration it is almost more art than science to get it working.
(well, it really is science ... but we don't have millions of dollars worth of lab equipment to analyze everything ... so the best we can do is take (hopefully educated) guesses at what might work)