Relay burning out and fuse blowing

[bwringer]

I see this mentioned above in passing, but the other thing that can "fry" a relay is rapid cycling. There are many ways this can happen -- as noted above, vibration, a bad connection somewhere, a bad ground, accidentally using AC from a stator wire to trigger the relay, etc.

I'd suspect there's an error somewhere between the diagram and the reality...

 

[glenwill]

I see two possible solutions here, one should be considered "temporary" until you can implement the other:

1. Move your sense wire to the switched terminal in your Eastern Beaver fuse box. That will eliminate any voltage drops caused by dodgy connections because it is VERY close to the battery.

2. Get a SH775 R/R from a Polaris and eliminate the sense wire altogether. This has the added benefit of prolonging stator life on your smaller, higher-revving engine.

Thanks. I was actually just reading up on posplayr's SSPB, and he is also recommending the SH775. I'm thinking maybe I just rewire with both of those, and eliminate the risk of a failing relay or fuse.

 

[posplayr]

I see this mentioned above in passing, but the other thing that can "fry" a relay is rapid cycling. There are many ways this can happen -- as noted above, vibration, a bad connection somewhere, a bad ground, accidentally using AC from a stator wire to trigger the relay, etc.

I'd suspect there's an error somewhere between the diagram and the reality...

You are correct, this had not occurred to me that it may very well be that a low control voltage is cycling the relay. It is common practice to use a snubber or flyback diode on the relay to protect the coil when it is opened.

http://electronics.stackexchange.co...ck-diode-with-an-automotive-relay/56323#56323

https://www.youtube.com/watch?v=c6I7Ycbv8B8

While the flyback diode will probably protect the relay, I suspect other issues because many people install relays without fly back diodes (as far as I can tell). But repeated cycling will probably hasten the relays life.

You get get a fat 5 amp diode from radio shack and put that across the coil of the relay as indicated. It provides a short to the negative going pulse of current that occurs when the coil is opened. This saves that current from going through the coil but instead chooses the diode parallel path.

 

[glenwill]

You are correct, this had not occurred to me that it may very well be that a low control voltage is cycling the relay. It is common practice to use a snubber or flyback diode on the relay to protect the coil when it is opened.

http://electronics.stackexchange.co...ck-diode-with-an-automotive-relay/56323#56323

https://www.youtube.com/watch?v=c6I7Ycbv8B8

While the flyback diode will probably protect the relay, I suspect other issues because many people install relays without fly back diodes (as far as I can tell). But repeated cycling will probably hasten the relays life.

You get get a fat 5 amp diode from radio shack and put that across the coil of the relay as indicated. It provides a short to the negative going pulse of current that occurs when the coil is opened. This saves that current from going through the coil but instead chooses the diode parallel path.

Recalling my original post, I had 2 problems where I was not sure which was causing which, and still don't. My son reported steadily dropping voltage before the bike shutdown, and when we went to diagnose we found the relay and fuse were blown.

So I am not sure if the relay going was causing a drop in voltage by steadily shorting out, or if the steadily dropping voltage reached a point where it caused the relay to fail and short out the fuse. If it's the latter, it would mean that the relay failed from 1-2 minutes of low voltage at most.

I have not been able to reproduce the steadily dropping voltage running in the garage, so it's time to get out on the road with some spares in case it shuts down again.

Glen

 

[posplayr]

Recalling my original post, I had 2 problems where I was not sure which was causing which, and still don't. My son reported steadily dropping voltage before the bike shutdown, and when we went to diagnose we found the relay and fuse were blown.

So I am not sure if the relay going was causing a drop in voltage by steadily shorting out, or if the steadily dropping voltage reached a point where it caused the relay to fail and short out the fuse. If it's the latter, it would mean that the relay failed from 1-2 minutes of low voltage at most.

I have not been able to reproduce the steadily dropping voltage running in the garage, so it's time to get out on the road with some spares in case it shuts down again.

Glen

The only cause and effect I'm concerned with is what could burn out a coil in a relay. Solving that will get you into a more stable situation to find out what the root cause if of your charging problems. Most likely if you just followed the GS Charging health thread it would go away, but obviously there is the possibility of something pathological not covered by that guide.


Words of advice when trying to solve a complex interrelated problem.

Fix what you know is wrong First without trying to rationalize away what it's effects might be. The cause of the symptoms is not always that obvious or may manifest itself in ways you did not expect.

You know the relay should not be burning up, but it obviously is. The diode will probably stop the relay from burning up. You will probably them find out that the relay is cycling for some reason, but that would be prejudging the outcome.

By fixing what you know is wrong, you may or may not solve the problem, but it is not something that also needs to be removed later.

 

[glenwill]

Starting with fixing what is wrong first, I cleaned up my grounds, and switched to a single-point ground, using the frame underneath the seat. RR, Battery, Black/White and Engine ground all meet there.

I also replaced the R/R with an SH775. I still have the relay disconnected, and I'll ride it like this for a bit and see if I have any further trouble. Until I put the relay back in, or put in a SSPB, I have a voltage drop.

One question. After putting in the SH775, my charging voltage at the battery is >14 volts all the time. At idle it's around 14.2. At 5000, it's around 14.3. At idle, turn on high beams, it drops to <14 briefly, and then rises back to just under 14.2. If I turn the headlights off, it increases to around 14.25, and then drops back to around 14.2.

Is this normal for a series R/R?

Glen

 

[posplayr]

Starting with fixing what is wrong first, I cleaned up my grounds, and switched to a single-point ground, using the frame underneath the seat. RR, Battery, Black/White and Engine ground all meet there.

I also replaced the R/R with an SH775. I still have the relay disconnected, and I'll ride it like this for a bit and see if I have any further trouble. Until I put the relay back in, or put in a SSPB, I have a voltage drop.

One question. After putting in the SH775, my charging voltage at the battery is >14 volts all the time. At idle it's around 14.2. At 5000, it's around 14.3. At idle, turn on high beams, it drops to <14 briefly, and then rises back to just under 14.2. If I turn the headlights off, it increases to around 14.25, and then drops back to around 14.2.

Is this normal for a series R/R?

Glen

Fixing what you know is wrong first is the best approach.

Without knowing what your voltage drops are I would be very hesitant about even speculating about +/-0.1V changes as you turn on and off things. I don't recall the setpoint on the SH-775 it is probably either 14.25 or 14.5V so you should be very close and have very low drops.

 

[glenwill]

Fixing what you know is wrong first is the best approach.

Without knowing what your voltage drops are I would be very hesitant about even speculating about +/-0.1V changes as you turn on and off things. I don't recall the setpoint on the SH-775 it is probably either 14.25 or 14.5V so you should be very close and have very low drops.

Thanks. I was mostly asking about having the idle charging voltage over 14 volts. With my previous R/R, idle charging was low/mid 13's, and would rise up with RPM to mid 14's. With this, I am at 14.2 even at idle. I'm assuming this is a benefit, and reflects a stable charging voltage at all RPMs, but I just wanted to confirm this.

Glen

 

[glenwill]

While doing some searching for wiring, I came across this post, a little too late:
http://www.thegsresources.com/_forum/showthread.php?236972-DB-Electrical&highlight=db+electrical

My SSH775 was from this vendor. I never saw a conclusion on that post whether what they are selling is actually a series R/R, but to avoid it as a risk. Too late for me, but is there a way I can test it to confirm it is a series R/R?

Glen

 

[Grimly]

While doing some searching for wiring, I came across this post, a little too late:
http://www.thegsresources.com/_forum/showthread.php?236972-DB-Electrical&highlight=db+electrical

My SSH775 was from this vendor. I never saw a conclusion on that post whether what they are selling is actually a series R/R, but to avoid it as a risk. Too late for me, but is there a way I can test it to confirm it is a series R/R?

Glen

If a vendor doesn't show the end plate so that you can see the Shindengen logo with the type number on it, you can assume it's a MOSFET and is being sold in a sneaky fashion which makes it a fake in my view. Even at that, we have seen fakes with fake logos and numbers on, but the typeface/spacing was wrong.
By this time, there are likely hundreds of owners out there who think they bought an SH775 series R-R and are treading the path to yet another burned-out stator, without realising it.
There is only one guaranteed way to buy a NEW genuine series SH-775; get it from a Polaris dealer. You can be sure if any of them start passing off crappy fakes as the real deal, they'll be in deep crap with Polaris, so that's some reassurance. You can buy a genuine used one, if you tread carefully and have a good look at the end plate. Breakers who won't show the end plate are dodgy, and worth avoiding.

 

[posplayr]

While doing some searching for wiring, I came across this post, a little too late:
http://www.thegsresources.com/_forum/showthread.php?236972-DB-Electrical&highlight=db+electrical

My SSH775 was from this vendor.1.) I never saw a conclusion on that post whether what they are selling is actually a series R/R, but to avoid it as a risk. Too late for me, 2.) but is there a way I can test it to confirm it is a series R/R?


Glen

1.) The vendor admitted it was not an SH-775. The email response is in teh thread.
2.) If you have an AC clamp meter you can see the change in current flow in the stator as you change loads liek headlamps.

 

[posplayr]

Thanks. I was mostly asking about having the idle charging voltage over 14 volts. With my previous R/R, idle charging was low/mid 13's, and would rise up with RPM to mid 14's. With this, I am at 14.2 even at idle. I'm assuming this is a benefit, and reflects a stable charging voltage at all RPMs, but I just wanted to confirm this.

Glen

Ideally you want full voltage all the time. My understanding is that the first commercial series R/R were for Harleys so they would charge when the engine loaped at 300 RPM.
You voltage could be a little higher at idle because of a slightly higher idle RPM.

 

[glenwill]

Eureka! I think I figured out my relay/fuse burning out problem! My schematic was accurate, but I wasn't seeing what the problem was until I looked at it again today, and it hit me. I'm not sure if I want to burn out an other relay to prove it, but here's what I think is going on:


Here's the relay wires:
Coil +: Orange - return from keyswitch to trigger relay. Was originally feeding the fuse panel.
Coil -: Frame ground
Com: Red - battery side of main fuse
N/O: Orange - feed to fuse panel to power bike


This setup means that both the charging current and operational current are traveling through the main (15A) fuse, since the R/R is one side of the fuse, and both the battery and Relay Common are on the other side. Per other posts on this site, this arrangement results in >15 amps, and will blow the main fuse.


Here's where things get interesting. Once the main fuse blows, the relay coil is still powered by the R/R, and the relay Common still has power from the battery, so the bike keeps running on battery power. Unfortunately, this means the relay coil is the only draw on the R/R. I suspect this alone might burn it out, although I don't really know how this kind of load works. However, the R/R sense wire was connected to one of the lighting feeds, and so was being fed from the battery, and not the R/R!


I don't know how high the R/R voltage will go, but no matter how high it went, the sense will still pickup a lower voltage from the battery, which also would have been steadily depleting as the bike was ridden. When the relay finally burned out, the bike would die as the battery power was no longer feeding the rest of the bike.


One of the symptoms was the battery voltage going down right before it died. Since my mounted volt meter picks up from the ignition circuit, it would have been showing the voltage from the battery. I assume running on just battery with headlights on would deplete voltage somewhat quickly, and when the relay was burning out, I think it would begin to chatter and pass partial voltage through until it died?
I'd love to hear if this makes sense. I know the only proof would be hooking it back up this way and seeing it fail, but I've already replaced the R/R and disconnected the relay, so I don't know if I want to bother just to prove this.

Glen

 

[Guest]

That doesn't sound like your setup is correct. The relay should be triggered by the kill switch, passing 12V+ directly to the coils and ignition from the battery.

 

[glenwill]

That doesn't sound like your setup is correct. The relay should be triggered by the kill switch, passing 12V+ directly to the coils and ignition from the battery.

You're describing the coil relay mod, but I did a system relay mod. Similar idea, but this bypasses the keyswitch for the whole electrical system.

It's wired as I described, but I should have connected the common to the R/R side of the main fuse.

 

[Guest]

Sorry about that, running on little sleep lately. One question, what would be the advantage of a system relay mod?

 

[glenwill]

The difference is that the coil relay mod eliminates voltage loss only for the ignition circuit, but for both the key switch and the kill switch. The system relay mod eliminates voltage loss for the entire electrical system, but doesn't eliminate voltage loss in the kill switch on the ignition circuit.

I'm not sure if one is automatically better. I may depend on your current situation. I could not find a way to disassemble the key switch to clean it, and have about .8 volt loss through it. I was able to get the kill switch cleaned up pretty well. So I decided to do the system relay mod to gain .8 volts through the whole system. I do understand that some loss in the lighting system has less effect than loss through the ignition system, but for me this seemed to work fine.

The trouble started when I replaced the fuse box with a new unit, and rewired the system, introducing this problem.

I'd love to hear if anyone thinks my theory on the relay/fuse problem is valid or invalid, as I plan to reintroduce the relay now that I understand what is going on.

Glen

 

[Guest]

I see the point of bypassing problem areas to provide direct battery voltage from the battery to the coils/ignition system or headlight but how do you do an end run around all of your plugs or switches and other connectors. I really have to point out the benefits of DeoxIT D5. I had cleaned up my harness originally when I put my bike together some years ago using CRC contact cleaner and dielectric grease. It's ready for another cleaning after six years or so. I had used the CRC on a set of computer speakers with a noisy volume knob and it lasts for a short time. I used the DeoxIT D5 on it fairly recently and it is still perfect long after regular contact cleaner would have started to fail. There is a good tutorial on cleaning ignition switches available unless your switch is different. This time around I intend to replace all my harness plugs with modern weatherproof plugs, go through all the switches and bullet connectors with DeoxIT D5 and dielectric grease and hopefully won't have to ever fool with it again. The coil relay mod will remain. I made up a set of jumpers that will turn it back to the stock wiring configuration in case of relay failure at the relay.

 

[glenwill]

Yes, of course. I have cleaned up all of my connectors. They are not the problem.

How did you get your key switch cleaned? That was my specific problem. I could not find a way to disassemble it, and spraying into it closed up didn't do much to help.

 

[Guest]

I haven't done my switch yet because of all the stuff I have to remove, but will be doing it shortly, however I just took the switch off of my '79 GS1000E frame and it is really simple to figure out. The switch housing is held to the underside of the top fork tree with two 5mm Allen head screws on either side of the switch. The whole assembly with wires and plug comes off as one piece. Turn it over and remove the dust cap over the wires. You will see the four wires crimped to the connectors on the bottom. This is definitely a DeoxIT D5 cleanup point because those wires are riveted to the internal contact points and are subject to corrosion. Remove the screw that goes through the wire strain relief clamp and tease the switch out of the key housing. Take a small flat blade screwdriver and pry under the strain relief tab to start separating the white cover from the brown part where the wires are attached, keeping the white part down. Go to the opposite side and do the same and the cover will slide off. Be careful, there are two copper plates one on each side in the white cover that are on springs. They did not come apart. The three bumps on each plate and the corresponding contact points riveted to the wires need cleaning until bright and then apply dielectric grease and re-assemble. Your switch may be a little different but I doubt it's much different. If you need a visual, I could post pictures but it is so simple I'm sure you will figure it out right away. Do not rotate anything in the switch or mark the other side to line it up and everything will drop back together as it was when you removed the assembly.