Announcement

Collapse
No announcement yet.

IPF 900XS Relay Buzz

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • IPF 900XS Relay Buzz

    Anyone experience both the 20 amp and 2 amp relays buzz when the engine is running? It's a pretty loud buzz. Works fine with the engine off. Wondering if it's a grounding problem. Looking for suggestions.
    Mongo

  • #2
    It is the pulsed electrical system in the JK if you tie the lights into your headlights for a trigger (same thing happens when doing a headlight upgrade with a full wiring harness). To correct this you can run a diode on your trigger wire, and a cathode from terminal 86 to 85.

    The other option is to just wire your trigger wire to the battery.
    Jon

    2008 Rubicon Unlimited
    4.5" AEV lift, 37" Nitto Trail Grapplers, AEV front/rear bumpers, AEV Corners, Factory Ten Axle Shafts.

    Comment


    • #3
      Buzzing relay in Jeep Wrangler JK Light system

      The buzzing is a known problem here since we saw the CAN Bus vehicles in 2005. That's not saying that CAN has anything do with it (it doesn't), it's just that the CAN vehicles also introduced PWM lighting in Chrysler vehicles.

      PWM is Pulse Width Modulation. Where instead of feeding the lights a steady diet of battery voltage (10-14V) they can extend the life of filament, conserve power, use smaller/lighter wiring, and better regulate the amount of light produced by powering the line in pulses of variable (modulated) width. The duty cycle (or time the line is powered per cycle) is constantly monitored and adjusted as needed. As the battery voltage gets to a nominal of 14V the pulse width can be shorter than it was at 12V when the car was first started. PWM for traditional lighting is more expensive then steady feed but has a ton of benefits.

      So... the problem you are having is you have a relay's magnetic coil being fed from this PWM. The relay designers were expecting a 9V+ source to power the coil, creating a magnetic field to pull the armature from the open poll to the closed circuit poll. You have a magnet that is being powered and unpowered several times a second, this may be enough to release and re-catch the armature every time causing a chatter, may release just enough to vibrate the armature causing a buzzing, or which is most often the case there is no buzz at the relay but there is potential for electro magnetic interference on sensitive electronics like fast digital communication streams or radio signals.

      What I find interesting is that with the duty cycle of the lights and the capacitance of the coils in the relays that the IPF lights come with, it hasn't been a problem when we do the 900SX lights. There is no noticeable buzz or interference problem on CAN (which is designed to handle EMI) or the analog AM/FM radio (which is not at all designed to handle EMI since analog radio waves ARE EMI).

      We have noticed the problem with the IPF 900HID lights. Major interference on the AM radio when the ballasts were on. In this case, it was the ballasts which are step-up coil driven and cycling rapidly. It didn't take long to determine that it was the HID lights causing the radio to fill with static.



      I have to disagree with John on the solution. While a diode across the coil is appropriate to protect solid state electronics from a the discharge of un-powering a relay, and for blocking the buzz that comes from powering a relay with AC voltage; a diode across the relay shouldn't do much to help here.

      You have a PWM and there aren't many options to fix that. One trick is to use an electrolytic capacitor across the coil to try and fill in / smooth out the ripple in the PWM. The disadvantage of the electrolytic cap is that they have a nominal life at room temp and rated voltage of around 2000 hours. They always last longer than that but they will wear out, I have seen worn out caps disconnect and I have seen them short together as well (bad). Without knowing the pulse width and current relative to the relay I'm not sure what size, but I think between 470uF and 4700uF would be a start. There is also a small matter of the ESR (equivalent series resistance) that will allow a small amount of current will leak to ground, but the ESR in this circuit shouldn't be too much of an issue.

      There have been other posts on this, but I am curious because we haven't had a problem with the 900SX lights using the supplied relays (some relays will be better than others because the coil will act like a capacitor and not all coils are made equally.

      We've been looking into the issue and are examining options. A similar project been on our whiteboard here in the office for awhile.

      Comment


      • #4
        Would a SS relay work, or would the lights just dim with the PWM?

        Comment


        • #5
          A solid state relay or a quick switching FET (field effect transistor) you'll wind up with exactly what Chrysler/Jeep is doing in the first place. A PWM light setup.

          I can't say it would 'dim' as that is a relative term. Is it dimming if you can't notice it? A PWM will never drive a light as bright as a same voltage steady on, you may be able to get 99% of the light with 85% of the power required. So technically, it would not be as bright, but it could be very close.

          Since the stock bulb's filament isn't the same in rated wattage, resistance, or capacitance, I would say it's uncertain how much light would be put out. If I couldn't really guess off hand if it would be noticeably dimmer or not. I do know it's tough to find solid state relays that are cheap and rugged enough to handle this application at these switching rates. I never had good luck find rugged let alone cheap.

          Comment


          • #6
            Thanks for the thorough explanation. What's interesting is I have upgraded the headlights to Cibie' H4 E-codes with a heavy duty relay and that is working perfectly without any buzzing. I think I will get the same relay I got for the headlights (which appear to be heavy duty Hella relays) and swap them for the IPF.
            Mongo

            Comment


            • #7
              Originally posted by Mongo View Post
              Thanks for the thorough explanation. What's interesting is I have upgraded the headlights to Cibie' H4 E-codes with a heavy duty relay and that is working perfectly without any buzzing. I think I will get the same relay I got for the headlights (which appear to be heavy duty Hella relays) and swap them for the IPF.
              If you ordered from Daniel Stern with his wiring harness, or the Susquehanna wiring harness they use the dieodes and relays as was mentioned earlier to solve the issue.
              Jon

              2008 Rubicon Unlimited
              4.5" AEV lift, 37" Nitto Trail Grapplers, AEV front/rear bumpers, AEV Corners, Factory Ten Axle Shafts.

              Comment


              • #8
                Thanks, Jon. I purchased the harness from SUV lights. http://www.suvlights.com/index.php?cPath=24_73
                Mongo

                Comment


                • #9
                  Originally posted by Mongo View Post
                  Thanks for the thorough explanation. What's interesting is I have upgraded the headlights to Cibie' H4 E-codes with a heavy duty relay and that is working perfectly without any buzzing. I think I will get the same relay I got for the headlights (which appear to be heavy duty Hella relays) and swap them for the IPF.
                  The buzzing will have to do with the amount of power required to move the armature to active the relay. It's possible an AC signal on the 87 side could effect the coil but with DC is considerably less likely.

                  The coil will have both a resistance and capacitance (difficulty the electricity will have getting through the coil, and the coil's ability to retain that electricity for awhile).

                  I don't know because I haven't hooked an oscilloscope up to the headlight line in awhile but, I have very very little confidence that a diode across the could will be doing anything here. I suspect the heady duty harness and relay John is talking about are the right combination, the same thing we are seeing with the IPF stuff.


                  A diode across the relay is to block AC signals or to protect from a momentary back feed surge when the coil is discharged. This is neither of those things, my quick recommendation is to try a different relay if it's just a mild buzzing. If it's actual interference to something like the radio, that's something else.

                  Comment


                  • #10
                    I am probably not describing it very well, so I added a picture I found of the fix I was trying to describe.
                    Attached Files
                    Last edited by JeepinJon; 12-02-2009, 07:40 AM.
                    Jon

                    2008 Rubicon Unlimited
                    4.5" AEV lift, 37" Nitto Trail Grapplers, AEV front/rear bumpers, AEV Corners, Factory Ten Axle Shafts.

                    Comment


                    • #11
                      Jon, you were explaining it fine. Although, you never mentioned a capacitor which is the only thing that will help in that diagram.

                      The diode shown could prevent current from draining backwards during the 'off' part of the duty cycle, but only if it's resistance was less that that of the ground circuit.

                      In the diagram that diode is not doing anything. In fact, it is redundant because the cap will drain into ground, and professional electronics like the TIPM will already have a diode or some other protection circuit in place inside the unit. On all high side drivers there is protection in place for a short-to-ground or short-to-voltage event.

                      All adding a diode on that line will do is lower the trigger voltage down 1.0-1.6V per typical diode. This is referenced as the Forward Voltage Drop. This means the relay will be missing ~1.5V of coil voltage relative to no diode. That 1-1.6V will be used up as heat, noise, and light, this effect is what makes LED's work.

                      This diagram would be ideal for using a DC relay on an AC circuit which I'm sure is where it came from. For DC, skip the diode.

                      Comment


                      • #12
                        Hey Jordan I drew that diagram up 3 years ago when I upgraded my lights to Cibies with H4 bulbs. Have had no problem with the cap going bad so far or any other problems with that set up. At the time I added the diode just to be on the safe side. Was worried about some type of signal getting back to the TIPM.
                        Doug
                        Last edited by dsy; 12-03-2009, 12:14 PM.
                        AEV 5.7 kit, 35"KM2's, AEV 3.5 premium lift, AEV snorkel, AEV tire carrier, AEV corner guards.
                        My DIY Hemi swap http://forum.aev-conversions.com/showthread.php?t=485

                        Comment


                        • #13
                          Ah, makes sense.

                          Chrysler was way ahead of you on protection I would sooner worry about a wet tipm than I would an electrically effected one.

                          As for cap life. Unpredictable at best. I know I rarely see them at more than 2000 hours of life at rated voltage. Thing is, it's not very clear how long a 35V cap will last on a 14V extended temperature application. Could be 2001 hours, could be 20,000. One could go crazy with double 100V caps but in reality it doesn't mean you are any safer from an eventual short. If I were to do it, I would add the cap inside the relay, making a fix as simple as swapping to a standard part.

                          Comment

                          Working...
                          X