| Author | Message | ||
     Gemini |
normally open relay. when relay is energized, you get 3 legs of charging(pin 30 is connected to 87 of internal to relay). at @2000 rpms to 4000 rpms, the relay is de-energized to open the circuit and allow 2 legs of charging. this is why the ecm needs to be programmed with the harness install. no relay control means only 2 legs of charging only | ||
| Gemini |
G7L-1A-TUBJ-CB-dc12 this should be the part number you would want. single pole(one set of contacts) that are normally open, rated for 30 amps at 220v AC that is controlled by 12 volts dc. | ||
| Timebandit |
My G7L-2A-TUBJ-CB is not a current model number. It's a NOS part. It will work fine in this application, because with a DPDT relay, it doesn't matter if the relay is NO or NC, you just wire it up as needed. To confirm things then, you're saying that the relay is NO, which is exactly like the Omron in the picture. the only difference is that you're recommending a single pole. that's right, but you could use a DP without any problems. am i correct then, that the truth table involves 3 switched states? a. IDLE: energized / closed; 3 legs charging b. 2k-4k:relaxed / normally-open: 2 legs charging c. > 4k: energized / closed: 3 legs charging the nice thing about the DPST version is that it allows you to do interesting things with the other circuit...  edit: the part I've listed is a DPST not a DPDT relay. (Message edited by timebandit on January 03, 2012) | ||
| Gemini |
your part number is a good part number. the last digits that i have listed are the operating voltage and voltage type for the coil. i can't make out the small print on your relay. as for your understanding of relay state, you are correct | ||
| Timebandit |
yes, you're right -- the part number is still good. it's just that when i bought this part a few years ago, it seems that they weren't appending the part number with the voltage spec. instead, they used a pictogram of an open rectangle for 12V coil and a rectangle with a diagonal through it for a 24V coil. i tried to get the numbers in focus, but my digicam's macro capabilities are limited. in case you're squinting: old part number: G7L-2A-TUBJ-CB new part number: G7L-2A-TUBJ-CB-DC12 As you mentioned earlier, the G7L-1A series (single pole) would work fine. I just happened to have a flange-mount style DPST relay on hand, so the TUBJ part of the part number could be changed for anyone who prefers the other mounting styles. the J reflects that my model has the test button. you could get by without that as well. (Message edited by timebandit on January 03, 2012) | ||
| Rkc00 |
Well after having two of my relays go bad and a third one go bad and burn up the harness I hope one of these will solve that problem. Will it plug into the harness connector that is on the bike? | ||
| Timebandit |
I mentioned on the previous page that nobody manufactures a relay with the proper specifications in the ISO-automotive footprint because a relay with the proper specifications is not useful for automotive applications. Our only option is to adapt a non-automotive, power control relay to an automotive application. I provided the photo of the suitable part to demonstrate that it wouldn't fit an ISO-type automotive relay connector. If you go back one page, I mentioned some mounting options to help people figure out how to do their own adaptation. Unfortunately, this means that the answer to your question is no. If you want to try installing a better relay, then your only option is to do the custom adaptation. Another way to skin the cat would be to buy a large supply of automotive relays, monitor their status, and keep replacing them before you suffer a burnout. (Message edited by timebandit on January 03, 2012) | ||
| Reepicheep |
Are there plug compatible solid state relays available you could upgrade to? (Message edited by reepicheep on January 03, 2012) | ||
| Timebandit |
Could be. I don't know. I didn't find any when I did the parametric search at Omron. Of course, that doesn't mean that a suitable SS relay isn't out there somewhere. I just haven't found it. | ||
| Rkc00 |
Timebandit, Thanks for your time working on this. I just love both of my 1125's. I would hate to get rid of them because of this problem. I am now out of the warranty on both of them. It is out of pocket at this point. | ||
| Avc8130 |
Timebandit, Got any suggestions for an easy way to monitor the function of the relay? ac | ||
| Timebandit |
yes. we're all in the same boat. i have an 09 with the harness update, and i don't like the idea of having a cheap relay take out an expensive stator. there's got to be a better solution. in the big scheme of things, i'm viewing the charging harness update and relay as a short term band-aid repair. HD did this to get them past the warranty period, and then drop the problem in our laps. i'm interested in finding a better solution for the charging system. just not there yet. i think a relay upgrade would be a good temporary fix. i've got all the parts on-hand, so i might end up doing the relay conversion with a set of jumpers when i get around to doing winter maintenance on my bike. | ||
| Timebandit |
i think Gemini mentioned how to do the voltage drop test. that will tell you when the contacts go bad, but it won't tell you whether switching is taking place. the problem is that you have to perform multiple tests to adequately assess relay health, and they all involve pulling off the seat. i think the best standard would be to monitor current flow through the switched leg. i can't say that i'm too excited about putting an ammeter on the dash, though. | ||
| Gemini |
As for switching function, get a 12 volt led. Mount in on the dash and ground leg to ecm side control of relay and the positive side on the 12v feed from the active intake solenoid feed. Led lit bright indicates relay is being energized. Note: ecm may show a trace ground even when not commanding relay to close. This trace ground will cause a led to light very dim. | ||
| Gemini |
As for sensing when relay is carrying current, how about a winding of wires around the leg that goes to the relay and monitor induced voltage?.... | ||
| Timebandit |
if all you want to do is monitor the control signal from the ECU, then it's simple enough to just use that other pole on the Omron relay that I recommended. switch an LED with that. i wouldn't recommend that though. the problem with using an LED to indicate relay control signals/switching is that the LED doesn't provide a true indication of deactivation of the stator legs. it only shows that the control signal is being sent to the relay by the ECU. If you have the contact side of the relay fail, then your LED will continue to flash on and off happily, and you'll blissfully unaware that the stator switching circuit isn't working. the status light will be telling a lie. building an inaccurate indicator isn't all that helpful. to verify that you are protecting the stator, you have to monitor current changes through the legs. monitoring anything else creates the possibility that the indicator tells you that everything is fine while the stator gets incinerated. | ||
| Timebandit |
>"As for sensing when relay is carrying current, how about a winding of wires around the leg that goes to the relay and monitor induced voltage?...." that could be made to work.  | ||
| Gemini |
coil of wires(not sure of number of windings or gauge of wire) around leg of charging that goes to relay. both leads of coil wires go to pair of leds wired in parelell. each led put it opposite direction. when relay is closed and charging load is present, leds would light up and flicker due to ac power circuit induced voltage. thoughts? i got some leds at home that i am gonna play with windings when i get home from school | ||
| Timebandit |
this reminds me of the way that some old Tube TVs were designed. they powered some tubes' filaments with a single wrap of wire around the high voltage flyback transformer. a potential downside to your idea is that if you just coiled a new wire about the stator leg, it wouldn't work unless you properly aligned the B-fields of the coils. off the cuff, i think you'd need to design a more complex sensor. let us know how it works out. | ||
| Reepicheep |
I doubt it will be that simple... an LED has something like a 1.8 volt forward bias that must be overcome before any current will flow through it. I don't think you would be able to induce that much voltage inductively without more or less building a transformer. That's not to say you couldn't do something... you could, but you would need an op amp. You might be able to pull it off with just a microcontroller. 4 analog inputs to measure lots of stuff, 12 bit resolution with only a couple of bits of jitter. And it could then drive the LED's (or for that matter, an LCD just as easily). Under $30 worth of parts, and you could program it pretty easily. Just building and programming a raw PIC is pretty easy as well if you have help (and I would). Especially if you started with something like an audrino, or other already sorted platform. For what that's worth... | ||
| Zac4mac |
Omron makes inductive pick-ups for board mounting. Not sure the output but there's room for a lot of windings... I have a bunch of "thru-hole" style I'm sure they make them for SMT also. If Gemini, TB or Reep want one - send me an sase and it's on its way. Z | ||
| Bextreme04 |
When mine went bad the voltage was noticeably lower. It went from steady 14.3v all the time to about 11.8V at idle and 12.8V max above 5000RPM. It was very noticeable almost immediately. Seems like kind of a waste to build some complex monitoring system when it is pretty obvious when it does fail. Of course it is winter and I'm all about complex solutions due to shear boredom but it just doesn't seem necessary to me. Although I am in SoCal so I ride year round as my daily driver. I've only had the relay fail once right after I switched to the FH0012A regulator because the stock regulator had failed. I have a couple thousand miles on the new regulator and relay now and no more problems so far. | ||
| Timebandit |
i think you're missing something. the purpose of the "complex monitoring system" isn't to tell you when the stator is failing. it's to let you know when the relay goes bad so you can prevent the stator from ever failing. it's to prevent the unrecognized wear-out of a $5 part from causing the silent destruction of a $400 part. 2009 bikes that have the charging harness update should be able to avoid stator failure altogether if the charging harness update works as it's supposed to work. if the charging harness update works as intended, then stators should never fail. the problem is that people have had stators fail after the charging harness update because end-users have failed to recognize the need to replace the relay that performs the business of the harness update. when that happens, it's like the charging harness update isn't even there, so the stators fail. personally, i don't think it's a bad thing to design a "complex monitoring system" to tell you when to change a $5 part so you don't blow up a $400 part. if I could buy one of these monitoring systems I'd be all over it. but since nobody makes them, it looks like I'm going to have to roll my own. Zac, thanks for the offer. As it turns out I've just ordered some sensors to tinker with. | ||
| Alaskacr |
Ok, I admit, I am missing something. Isn't the relay normally open - and so then doesn't it fail open too? If it fails open, then why does leaving a leg of the stator unused cause premature stator failure? THanks for you patience. | ||
| Gemini |
it does normally fail open, however, the contacts can weld shut and fail leaving a condition present that all 3 legs charge all the time. running 2 legs all the time is just as bad because that is a great recipe to have slow cranking or no start due to low voltage. i will tell you from experience, that really sucks in the summer in stop and go traffic to stall the bike and be unable to restart without a jump start. | ||
| Timebandit |
> the contacts can weld shut this is something that's not inherently obvious so it's worth talking about. automotive relays are rated to switch 12VDC, and the stator leg can be producing around 120 VAC at 10k RPM. we're dealing with higher voltage than the insulation of the relay was designed to handle. then you have to consider that the shunt regulator draws maximal current out of the stator at all times, so the relay is always switching maximal current. the combination of high voltage and high current means high power. high voltage and high current causes arcing across the switching terminals. a better relay, like the omron relays that are rated for switching high power, should be designed with better contacts and better isolation and shouldn't be as likely to fail. nobody's mentioned this yet, but it's probably a good idea to place a small high voltage cap across the switched contacts to prevent arcing. i'm kind of surprised the engineers at H-D didn't think of that. but then I'm equally surprised that they only decided to switch one leg of the charging system. (Message edited by timebandit on January 09, 2012) | ||
| Timebandit |
> why does leaving a leg of the stator unused cause premature stator failure? that question brings up my doubts about the efficacy of the harness update in preventing stator failure. i don't think it prevents failure at all. i think that it only delays failure. it buys us time. but since it's all we've been given by H-D, it makes sense for us to keep an eye on the relays to make sure the harness update continues to function properly. if we let it fail because of a relay problem then it can't possibly help us. the way that the H-D harness update works, it switches off some of the coils in the stator by using a relay to interrupt the circuit. those coils in the stator get to cool down when they are not in use. the problem is that the harness update doesn't rotate which coils get turned on and off. it switches the same coils every time, and leaves the other coils running all the time. those coils that never get switched off continue to operate at a 100% duty cycle just like they did before the relay came along. if you're using a shunt regulator, those coils that remain unswitched continue to work at full voltage/full current (full power) output at all times. they never get a break. their situation never improves, so they're going to fail. the only hope that the harness update provides for those coils is the theoretical ability for the coils that are switched off to conduct heat away from the coils that remain on. obviously, this is nowhere near as good a heat dissipation solution as the direct oiling upgrade from EBR. by itself, I don't think the charging harness will be good enough and we'll still need the oil cooling mod. we have to take it on faith that the harness update actually does provide useful heat dissipation for us to pursue the relay problem. if it doesn't provide useful heat dissipation then there's no point in this thread. | ||
| Bextreme04 |
I wasn't missing anything. I understand exactly what you are saying but it is a matter of probabilities. Is it possible for a normally open relay to fail closed... yes. Is is likely... no. When mine failed it failed open and was very obvious. If it fails closed it is also obvious because you won't hear the relay give that nice, loud, audible "click" when you turn the kill switch to run with the key on. Which brings me to another good point, the relay engages primarily before the bike starts which means you have no current to arc on the relay contacts. Plus I have never seen the relay disengage at anything other than idle.. isn't the stator output at idle ~23VAC per leg or less? My diagnostic manual shows it as ~45VAC per phase at 3000rpm so I think it is definitely less than 24VAC on that one leg. Standard bosch style automotive relays are rated for 24V usually so unless i'm missing something..... Also I have only seen the relay disengage when stopped at idle for more than 30 seconds while the bike is REALLY hot. As soon as I rev it up to take off from the light the relay clicks back on and my charging goes back up to ~14.3v. Just seems to me if you know how it works you can check it for proper operation every time you start the bike just by listening and looking. Listen for the click when you turn the kill switch to run and check your voltage once started. ~Alaska- I think it's possible that dealership mechanics have improperly diagnosed electrical problems and replaced VR's and stators that may not have been bad for a low charging voltage condition that could have been caused by a failed relay or harness. Having the relay disconnected will kill your battery pretty quickly but I don't think it will kill the stator. | ||
| Bextreme04 |
~Timebandit Just to clarify I'm not saying switching to an AC relay isn't a good idea just that it isn't necessary. I did find a solid state 120VAC@25A or @45A load with a 3-32VDC control circuit but it would require cutting apart the harness and mounting the relay somewhere. Not sure if it would cause more potential problems than it solves. Also it is 29.00 for the 25A and 39.00 plus shipping for the 45A compared to $5-10 at any autoparts store. Here's the link http://www.opto22.com/site/pr_selector.aspx?cid=3& qs=100610161053 | ||
| Timebandit |
Based on what Gemini told us, the relay setup is more complicated than just changing states at key-on and at idle. It also switches at higher RPM. I don't have ears in my ass, so I can't hear a relay clicking under the seat when I'm riding the bike through the 4k or 5k RPM switching threshold. I need status lights to tell me what's going on. by 4k or 5k PRM the relay is switching moderate voltage (maybe 60 VAC) but the stator is already making full current output, which is what, 30A? 40A? that is a lot of current and it wouldn't surprise me at all to see arc'd contacts. that's the failure mode I'd expect to see. i don't think a 12/24VDC bosch relay will hold up -- not because of the voltage rating as much as the failure to include a cap on the switched contacts to prevent carbonization. i agree with you that it's hard to be confident in everyone's ability to properly diagnose these electrical problems. i think that a fair number of mis-diagnoses could be being made, and that we're trying to sift through a lot of dis-information as a result. |
