Fuse #5 fighting with the central locking system

[Tinkwithanr]

Need some input from the group. Currently chasing an issue that seems to be tied to my central locking system (1988 M5), which of course decided until my vintage prep marathon to rear it's ugly head. Every time I go to unlock the car, fuse #5 is blowing. Oddly enough the system will lock the doors fine. The fuse is only blowing when I go to unlock them. And they do unlock successfully, but the fuse goes with them.

I tried a second central locking control unit and got the same results. This is making me lean towards a wire being shorted out somewhere with the locking motors, as I think they reverse polarity when operating, so a short would only show itself on one operation.

Recent work I've done to the car would be:
-Performed the Holy Grail Labs blade fuse retrofit in the fuse box. Obviously this was the first place I looked for an issue, but after 4 times of going over everything I am fairly confident this isn't the issue. Note that it did come with 7.5 amp fuses instead of the stock 8, but I'm not betting on that being a big enough difference to cause the problem.

-Adjusted door strikers and alignment. Again, maybe a wire got pinched weird somewhere in the this process, but it really doesn't seem like it.

-Replaced the starter. Definitely don't think this had anything to do with it.


At this point I'm going to run the car with the central locking unit disconnected and make sure nothing else is acting weird.
Any one else have similar issues? It seems like everything I can find relating to fuse 5 is a random issue with one of the many things that runs off that circuit.

[willyv124]

I've actually been dealing with the same exact symptoms, although mine sometimes blows when locking as well. But I have tied it always to the central locking system. Like you I also have recently done the holy grail labs blade fuse retrofit and I've also tried a new central locking control unit with the same issue. I'm currently experimenting with only using the trunk to unlock/lock, I've been thinking about trying a different sent of door lock actuators as well. So no help, but just wanted to chime in that i'm trying to figure it out as well ha

[Tinkwithanr]

Hmmm, interesting. Did your issue start shortly after adding the fuse upgrade? I was also thinking of disconnecting one actuator at a time and seeing if the others worked better then. I can't help but wonder if the ATO style fuses are a little quicker to blow than the originals. I did measure the draw across fuse 5 while idle and it's only .026 amp, so the act of unlocking alone is enough to burn it up (whether its a short or just really hard to operate actuators). I couldn't get a clean reading from the multimeter while trying to unlock sadly. I might try that again with the oscilloscope just to see it is 8.5 amps or 80.

[willyv124]

I don't remember having issues before the fuse upgrade, at least not the frequency I've had them now, where in one day I could blow the fuse like 4 times. I've also measured current with a multimeter while actuating and I do see that current draw peaks past 8 amps while actuating, i think i saw around 9.5 but i couldnt get it to blow while measuring, which should generally be fine since its within a certain tolerance/time at that amperage, but clearly when it does blow it must be going higher/longer past that.

I've been wanting to do the actuator test you've mentioned but its annoying work that i have yet to commit to. I suspect that if the actuators are sticky, then it could end up keeping the peak amperage for long enough to blow.

I have experimented with using a 10a fuse for a bit and had no more issues, but didnt want to risk burning any wiring.

[Tinkwithanr]

Yeah, my guess at this point is system as a whole is just taxing the fuse harder, and the old style fuses just didn't care as much.

I did notice a few other people in the holy grail labs install thread mention a similar issue. The only difference is they mentioned running aftermarket key fobs. So maybe anything extra on the system is enough to push it over the edge.

I might try disconnecting the trunk actuator and only using the central locking for the doors. See if that is enough to hold a fuse. I normally lock the trunk with the key when I'm done in it anyway.

I thought about trying a 10a fuse, but a lot of the wiring off fuse 5 is only 0.75mm which is recommended for ~6 amp max. Obviously that's more of a continuous rating and these are more momentary loads, but better safe than sorry.

[willyv124]

Hmm, I might experiment with disconnecting the trunk actuator as well. Also considering some universal actuators like these https://www.amazon.com/Zone-4-Pack-Univ ... EBSO&psc=1

[Galahad]

...I can't help but wonder if the ATO style fuses are a little quicker to blow than the originals. I did measure the draw across fuse 5 while idle and it's only .026 amp, so the act of unlocking alone is enough to burn it up (whether its a short or just really hard to operate actuators). I couldn't get a clean reading from the multimeter while trying to unlock sadly. I might try that again with the oscilloscope just to see it is 8.5 amps or 80.

My suspicion is that the ATO fuses blow meaningfully faster than the originals (which is an "improvement" from the perspective of fuse design in this case). On top of that, most people with this issue - including my car - have wireless fobs. Notably with fobs, you're tripping 5 actuators at a time instead of 4, since if you use the key the system doesn't have to un/lock the door you did with the key (it tries but the current used is much lower since there's no mechanical thing to actuate). Combine more sensitive / accurate fuses with a higher current demand and potentially sticky lock mechanisms and you've exceeded your safety factor. I'll think about a good way to test this, I've got some reasonably sensitive equipment that should be able to detect the difference.

I put a 10A fuse in mine and haven't had issues since, but I'm probably more willing to accidentally burn something out.

[tig]

FWIW, fuse #5 started blowing more frequently on Maytag when unlocking after installing the blade style kit. Maytag also has a keyless / remote thing installed. I put a 10A in ass well and it addressed the issue (with some additional burn to ground risk).

[willyv124]

I've been locking/unlocking exclusively with the trunk for 3 or 4 days now, no blown fuses. This morning I just removed the trunk actuator and am going to go back to using the driver's side lock and see if it stays resolved.

[Mike W.]

I put a 10A fuse in mine and haven't had issues since, but I'm probably more willing to accidentally burn something out.

The difference between an 8 and a 10 is so negligible I wouldn't give it a second thought. I wouldn't be surprised if the old fuses especially were made with 20% tolerance.

[Galahad]

I put a 10A fuse in mine and haven't had issues since, but I'm probably more willing to accidentally burn something out.

The difference between an 8 and a 10 is so negligible I wouldn't give it a second thought. I wouldn't be surprised if the old fuses especially were made with 20% tolerance.

I agree with you, but I don't like making definitive claims without some sort of data to back it up. I'm going to try to free up some time this weekend to do some testing and see what conclusions can be drawn.

[Galahad]

I came up with a way to measure meaningful data about what's going on (at least from the fuse side of things) and ordered some equipment and a zillion fuses to be sacrificed. I should be able to determine what sort of tolerances we're working with and how quickly the fuses blow, at a minimum.

[willyv124]

Looking forward to see what you find. After removing the trunk lock actuator, locking/unlocking from the driver side has been fine. So i'll probably stay with this for a while.

[1st 5er]

Same symptom on the return from Vintage which is where the HG fuse panel was installed.

I wish I could remember which lock I used which blew the fuse.

Tach not working was the first noticeable symptom.
Later in the drive non working flashers while driving thru an intense rain event was the second item on the list to address once we arrived back home.
Third item was central locking no workie when we stopped for dinner. That is where I replaced the 7.5 with a 10 and never looked back.

Now to address the other issues after the 2,400 mile excursion.

Edited to add: No fob was used during this exercise.

[Galahad]

This is VERY early data (I've only blown up a handful of fuses while trying to figure out equipment settings), however it's meaningful enough I'd figure I'd share it.

At 40 Amps:
7.5A ATO fuses (at least the ones I ship with my kits) consistently blow at 50ms (0.050 seconds)
8A torpedo fuses ("borrowed" from one of my project cars) consistently blow at 500ms (0.500 seconds)

I was expecting a difference, but not an entire order of magnitude. Given how nice both numbers are I suspect the official spec for ATO is 50ms and torpedo is 500ms, but I need to do some research to confirm the suspicion.

[willyv124]

That looks to roughly match the spec sheet here https://m.littelfuse.com/~/media/automo ... asheet.pdf

Heres a spec sheet for torpedo fuses https://www.littelfuse.com/media?resour ... eramic-36v

[Galahad]

That looks to roughly match the spec sheet here https://m.littelfuse.com/~/media/automo ... asheet.pdf

Heres a spec sheet for torpedo fuses https://www.littelfuse.com/media?resour ... eramic-36v

Thanks! That's exactly what I was looking for.
I'm shipping Littelfuse in my kits so the ATO graph is spot on.

[Mike W.]

Wow, you guys are getting into it, my complements. RonW would be the one who would really know, but I wonder about a capacitor in the circuit. They can kind of act as electrical shock absorbers and might take some of the impact so to speak off the fuse. I haven't looked at the wiring diagram so I can't say where it would go, and most of my electronics experience was back in the 80s, but I do remember some. Paging RonW.

[Galahad]

Wow, you guys are getting into it, my complements. RonW would be the one who would really know, but I wonder about a capacitor in the circuit. They can kind of act as electrical shock absorbers and might take some of the impact so to speak off the fuse. I haven't looked at the wiring diagram so I can't say where it would go, and most of my electronics experience was back in the 80s, but I do remember some. Paging RonW.

I did think about that. I can see some issues - if you put the capacitor before the fuse you're not actually changing anything, and after the fuse you risk adding energy to a situation the fuse would otherwise protect. If RonW has some input I'd be happy to hear it

[Galahad]

Took some data on my car lock system today, and it's enough to start making conclusions (assuming data from my car can be extrapolated to other cars).
Testing setup: I replaced fuse 5 with a 1 ohm (1% tolerance) resistor and measured across it with an oscilloscope while I triggered the lock system with my fob. Using a 1 ohm resistor means that voltage readings from the scope are the same as the current through the circuit in amps. This did significantly increase the total system resistance (and therefore reduce current), I'll address this in a bit.

Graph 1:
lock then unlock e28 with a fob by Holy Grail Labs, on Flickr

This is a graph of the car attempting to lock, then immediately unlocking afterwards. My driver's door lock sticks sometimes and causes this behavior. I did compare this to locking and unlocking separately, and the two curves next to each other are not meaningfully different than if the car had successfully locked, then I made it unlock (I'm using the combined graph to save space in the post). Both lock and unlock take roughly the same amount of time (~575ms) but unlocking takes substantially more current. With the inline measurement resistor, locking peaked at around 4.9A and unlocking peaked at 6A. The system is a bunch of motors and therefore nearly entirely inductive (for the purposes of this discussion) which is why it takes some time to climb up to the peak current.

I measured battery voltage to be 13.4V, giving a total system resistance while unlocking (battery + wires + resistor + locks) of around 2.2 Ohm, or 1.2 Ohm subtracting the extra resistor (confirmed later).

I am not sure why there's a current difference between lock and unlock - maybe because unlocking you're lifting everything against gravity?

Graph 2:
lock trigger across 5mOhm by Holy Grail Labs, on Flickr

This is an unlock* curve measured directly across a 5mOhm fuse without the measurement resistor in the system. *I didn't write down if it was lock or unlock unfortunately, this is the conservative assumption and accurate to what I remember.

Now we can make some interesting observations. First off, the process takes about 700ms instead of 575ms. I do not know why this would be the case, I would have expected it to take less time since you've got more voltage across the locks and can actuate them faster. Second, this is stabilizing at 10A (this confirms the ~1.2 Ohm unlock system resistance, taking various fudge factors into account), which is way more than the "8A" fuse original to the BMWs. However, if you look at the time vs overcurrent ratio curve for the torpedo fuses, you'll see they are rated to last over an hour at 12A for an "8A" fuse, so 10A for less than a second is perfectly fine if you have a good condition torpedo system.

I believe whoever decided how to label the torpedo fuse current ratings needs to get slapped upside the head.

More relevant to this discussion is how the blade fuses behave in minor overcurrent situations. According to the spec sheet, the 7.5A fuse should last 10s at 10A, so we're good right? The problem is if you bump the current the time rating starts dropping extremely quickly (0.5s at 13A). Additionally, those values assume no temperature derating which is relevant (losing 5+%) in a hot engine bay, and we're completely ignoring any sort of manufacturing tolerance on the fuses. The math agrees with people's experience that we are operating pretty close to the edge of what a 7.5A blade fuse will handle.

Littlefuse (the manufacturer of what I currently ship) recommends using a fuse large enough that your operating current is less than 0.75 * temp derating factor * current specification of the fuse. In the blade fuse spec sheet they precomputed everything and have a chart for maximum current vs temperature, it looks like anything hot you'd want a 15A fuse to cover 10A for the door locks.

One final important point to make is comparing the steepness of the overcurrent vs time curves for the two types of fuses. Unfortunately the spec sheets do not provide this data in exactly the same format, but it's close enough to make comparisons. Some additional context may be important, namely that blade fuses were designed to save computers from melting while the torpedos were designed to keep your wires from melting - the former has significantly stricter requirements than the latter. An ideal blade fuse would conduct perfectly fine until you cross the rating and instantly break (yes, I'm ignoring cases where you'd want a slow blow fuse for simplicity), while a torpedo fuse may not have to do that. Your circuit design would change depending on which fuses you use: you'd want to avoid throwing a bunch of inductive loads on top of other stuff all at the same fuse if you have blade fuses while that wouldn't matter so much for torpedos, to pick an example.

At 12A, an 8A torpedo fuse is rated to last at least 3600s. A 7.5A blade fuse will last ~0.6s, and a 10A blade fuse will last about 4s.

At 16A, an 8A torpedo fuse is rated to last at least 3s and at most forever. A 7.5A blade fuse will last ~0.3s, and a 10A blade fuse will last about 1s.

I'm considering switching to shipping 15A for fuse #5 to avoid this issue - anyone have input?

References:
blade fuse spec sheet
torpedo fuse spec sheet
more info about fuses

TL;DR: if you blow fuse 5 while locking / unlocking your car and have a blade fuse kit, throw at 10A or 15A fuse in there - it's still more protective of your car than the original 8A torpedo.

[1st 5er]

Might be time for new fuse box covers.

[willyv124]

@Galahad. Thanks for doing all this research. It's great to see you stand behind your product and figure out if it was pre existing or related to your product with hard data. I'm all for a 10 a or 15 a, deferring to you and whatever you see fit. This product is already aftermarket and an upgrade. I don't see a need to keep this aspect "stock/oem" if there is data to back up the reason for the change and prove its okay.

my 2 cents

also definitely on board with new fuse box covers as well, lol

[Ju@n]

I know it's an old thread, but wanted to thank you guys
Just installed the Holy Grails Fusebox upgrade and had that behavior. I'm going to buy a 10A fuse! (Burnt 2 7.5 ones trying )

[Galahad]

I still plan on doing some in-depth ATO fuse testing, I have about a thousand fuses to blow up. I've been waiting for the semester to be over so I have some semblance of free time but I'm very interested in seeing what we can find out

[Tinkwithanr]

Wow, I wasn't expecting this to have gotten so in depth. Very nice work.

I'm very surprised to hear the rating difference between torpedo and blade fuses is so large. I'll probably start with a 10a fuse in place of the 7.5 and see how that goes. It doesn't look like that should cause any issues with the stock wiring.

That also explains how the original fuse boxes could get so hot/melt from the larger fuses without blowing.