This morning I installed LED headlights and fog lights with a relay on the passenger side of my 2001 C5. I first put in the fog lights and then the relay and low and high beams on the passenger side. At that point, I decided to try the lights before going any further withlout starting the car.. They look great but there was a clicking and the fog lights were flickering and I turned off the headlights but may not have turned off the fog lights. Because I had read that could happen when LEDs are installed I installed the driver side low and high beams tried the lights again . Now the clicking was louder but the lights were not flashing. I left it on to try and see (hear) where the clicking was coming from (I think from the passenger side way down or by the engnine). At this point, I am totally without power in the car. No lights, no starting, no messeges - nothing. Battery tests great with the negative terminal disconnected but after installing the negative terminal the voltage drops to about 7.5 volts. I have been looking for any wires I may have crimped, I checked all fuses (not sure how to test relays) associated the ignition, headlights, fog lights and any other light on the car. I have removed all the LEDs and relay and put the old halogens back in and when I connect the battery there is very dim light from the hallogens and then that goes off also. I have disconnected the battery and have a battery tender on it for the night.

Any thoughts, help, suggestions would be greatly appreciated.

 

Moved to C5 Tech for assistance.

 

Good luck

 

More AI vomit.

 

You might have shorted some wires when you added the new lights, the clicking sound could be a thermal circuit breaker cycling from a short. Check the positive feed wires going to the lights for a short to ground with the lights disconnected.

 

Thank you . I will try and repost.
A little more info. I let the car sit all night with a battery tender. This morning when I reconnected the battery I heard clicking and the headlight motor activated and they retracted. I then tried the headlight switch and they came up and the lights went on. I turned the car on and found a ton of DIC codes all history. I tried starting and it started right up. I am still concerned because although the battery tender is still connected and battery voltage checks a little over 13, the tender charging light is still on. I believe there is still a draw.

 

It's cringe to have the stance "AI bad"
I've solved countless problems with my c5z build (and b58 build for that matter) using AI. Don't be one of those boomers...

 

I got news for you slick, the "boomers" grew up using technology as it was introduced/evolved from the first computer to what we have today. When AI is mature enough to sort through the 50% misinformed/crap suggestions/answers, I will be on board. Right now, only trust it to give me answers on information that is not subject to interpretation...or will not change.

 

Hey thanks for the news! Just because you're unable to harness it properly doesn't make it useless, that's a you problem

What model(s) have you tried using?

 

 

It's cringe that you keep posting AI drivel.
If it's useful information then fact-check it and post it here.
People who care to try AI will try it. So far it's laughably bad in so many cases that it's next to useless.

 

I'm posting a link. You don't have to read it, no one is forcing you to. Most advice cites the GM manual.

Also, what do you mean by this "So far it's laughably bad in so many cases that it's next to useless"

Provide an example so I can explain why you're doing it wrong lol.

"People who care to try AI will try it."

I'm using a custom engineered model specifically for trouble shooting C5s that I built. This is an order of magnitude better than going to chatgpt and asking a question, but again, you don't understand the technology.

"If it's useful information then fact-check it and post it here."

This troubleshooting, not verifyable information. How do you propose I "fact check" troubleshooting steps? If there was a known answer then then this thread would not need to exist.

 

What I mean is the current crop of LLMs hallucinate often and even when called out on that and asked to fix it they often claim they have but haven't.
Examples are all over the place in hundreds of news articles. One recent example was in Paul's Security Weekly, the segment about Aardvark being used to clean up AI code.

I think the point you are attempting to make is an expert can get better results out of AI than a non-expert. No one would likely dispute that.
The vast majority of AI usage is by non experts who then go on to trust the information and pay the price when it's wrong.

 

I previously worked as an AI dev for a surgical robotics platform and recently founded an AI company in environmental compliance. The consequences of hallucination or incorrect information can be as bad as killing people/billion dollar projects being litigated and/or cancelled respectively. For example failing to flag a robot arm with anomalous test data can lead to unexpected behavior during surgical procedures, or failing to notice that your project proposal misses a compliance measure can lead to a developer being sued and losing a housing development contract.

I've spent countless hours developing protections against nonsensical answers. My current platform embeds citations which upon clicking highlight the exact text in original documents which they pull information from to generate answers. Unfortunately that isn't possible for generating troubleshooting advice, which is fundamentally speculative in nature. Testing, tweaking according to feedback are the ways that these models can be made more accurate and useful. Refusing to engage by virtue of something being AI generated isn't going to progress the technology.

Forum posters are not immune to giving incorrect advice, it's always going to be up to OP to collect information from whatever sources are available and decide accordingly what makes sense to proceed with, you just have to use your best judgement.


On another note, the podcast you linked isn’t really critical of AI. Its main points were: 1. AI is dramatically increasing the number of apps being built, but many “vibe-coded” AI-generated apps are insecure by default through exposed keys, misconfigured APIs, and missing access controls. Treat AI like a junior dev and use tools (like OpenAI’s Aardvark-style agentic security scanners) along with real threat modeling and audits to harden security. 2. Threat actors are starting to use AI (tracked by efforts like GTIG’s AI Threat Tracker), but most ransomware still doesn’t rely on AI, and much of the “AI-powered” attack chatter is hype or sloppy research that needs scrutiny. Meanwhile, AI platforms themselves (like ChatGPT) can have vulnerabilities like any complex SaaS, and “shadow AI” inside companies, employees quietly using unapproved AI tools or copilots, creates new data-leak and governance risks if it isn’t formally managed.

 

I get it friend. You want others to love your creation. They don't. Results are all that matters. Most posters here have little interest in helping you beta test your model.

Totally agree with you on this point.
You have an uphill battle convincing people your computer generated advise is better than even an average poster here.
Even though everyone who has spent any time at all on pretty much any forum knows a lot of the advice there is just bad. I can count on one had the people who consistently give good advice and maybe a couple dozen others that are generally useful.

 

So..... who cares? No one. But because use of AI is as random as the platforms themselves, let's keep their unfiltered BS off the forum to both avoid the commonly encountered misinformation and to retain investment in individuals with direct experience sharing that experience which is the lifeblood of this forum.


>>>BACK ON TOPIC<<<


As for the LED lights. Why did you have to install a relay? I have not encountered an LED source yet which required more draw than even the weak factory wiring harnesses can supply. I have done relays and upgraded harnesses when I was running HIDs with ballasts though. The headache work would be to chase each line with a voltmeter and that is probably advisable now that you believe there is an unwarranted draw from an unknown location. Myself, I would see about the draw and then unplug the relays / work I did so see an instant result. From there you can try running the LEDs without the new harnesses and using just the OEM harnesses but running resistors inline to keep the buckets closing when on low beams. Or you could rebuild your harnesses as well. Not expensive. Just time consuming.

Always check your grounds.

The issue could also be entirely unrelated to the lights, but be the result of something else being loosened or somesuch while doing the work.

 

Update on my LED issue.

First , thanks Tusc for your thoughts and for getting this thread back to the LED Llights. To answer your question, I put the relay (rather than a resistor) on the low beam passenger side because that is what I read on this forum.

As I said before, the car is now starting with no issue. There is, however, a parasitic draw of 1.44 amps after removing the relay, fog lights and hedlights and putting the old halogen lights back in. After removing every fuse, micro relay and maxi fuse that could in anyway be associated with the headlights and fog lights, and then some, there is still a a parasitic draw of 1.44 amps. Each time I test the draw after removing a fuse, etc., there is a clicking sound coming from the passenger side headlight well and another sound, like the hatch relase sound, coming from the back passenger side of the vehicle. Finally, I turned on the headlights without starting the car and they work fine. I then tried the fog lights after turning on the ignition and got no light. I think I am gong to remove the fog lamp switch and test that. Does anyone know if there is a relay in the passenger side headlight well, or any other reason, that could be the cause of the clicking sound I hear when I test for the draw after removing fuses?

 

This edges beyond my electrical expertise. I can make solid wires and relays and custom wire a Haltech to run parallel with the oem ecu.... But diagnosing problems is not where my experience lies.

To that, I always say one dumb but incredibly repeatable effective rule of thumb which is it check but also refresh every ground on the car. Those exposed to weather I refresh and then cover the connections with no-ox-id.

 

OK here is an update. First, I cleaned all the grounds and Splice Packs in the engine compartment which no effect of the parasitic draw. There actually was little to no corrosion found except the ground near the battery. Been spending a lot of time looking for the parasitic draw and still no result. I do know that the fog lamp circuit is terribly compromised. The fog lamp switch is shot. There is power to the switch and if I jump the hot wire to the output to the BCM and relay I get nothing. Removing the fog lamp relay and the old halogen bulbs does not affect the parasitic draw. The ground wire at the switch harness does not have good continuity to ground (but the trunk release which is part of the same switch works fine). Lastly, the ground socket at the fog lamp relay (pin 85) which should only be a ground shows 12.5 volts when probed to chassis ground.

What is truly curious, if I remove the fog lamps, the fog lamp relay, the fog lamp fuse and the fog lamp switch there is no effect on the parasitic draw.

Thanks for taking the time to listen to my woes.

 

Did you remove all the fuses and any circuit breakers then check the current draw? A 1.44 amp draw is substantial, either something is shorted, burned out or slightly on.