What's new
CK5
Register an account today to become a member! Once signed in, you'll be able to participate on this site by adding your own topics and posts, as well as connect with other members.

Electronics Gurus? (resistors, diodes, etc)

rampage said:
By the way, from the 6V drop across the LEDs I think they are arranged in 2 rows of 4 LEDs.
Click to expand...
I had the same conclusion. As for an oscillating LED, I don't know. That circuit shouldn't really be sinking any current and in fact Dorian's current measurements show that it hardly is - at least DC. I was sort of thinking the opposite, that the 5V is coming from the driver circuit acting up.

Yeah, for a parallel resistor, you want the biggest value that works. But we need something as proof of concept. If he has a bag of resistors, he can also make a 1W 1k out of 4 1/4W 1k or something like this and it should still fit.
 
Maybe I'm not explaining it right or it's just a far fetched analysis. By oscillating I mean going from looking like an 'open' circuit (the load across the ECM ) when the LED is off to a 'closed' circuit when the LED is on at it''s threshold on value of 5V. Then when the driver sees the load or reference voltage it's trying to drive it to the opposite state that it should be in. Then the cycle repeats. I agree, the driver is acting up, it's confused. Wish he had a scope. Wish I could find the original datasheet for the driver.

I don't think the parallel resistor value he uses is going to matter much considering that the original bulb was so low (6 ohms off and 51 on). The driver can handle 750mA so I'm not too worried about him damaging it by using too low of a value. I think the driver (or if you look at the equivalent circuit, the transistor collector) needs to see a reference voltage instead of floating.

I guess it's a moot point trying to figure out why's it's behaving this way. If the resistor works it works. Looking forward to results ..... :popcorn:
 
Appreciate all the thought you guys have put into this. As I said before, weekend is shot, but beginning next week I should be able to test the parallel resistor idea.
 
rampage said:
By the way, from the 6V drop across the LEDs I think they are arranged in 2 rows of 4 LEDs.
Click to expand...

:doah: WRONG !! :doah:

White LEDs have about a 3V drop. So 2 LEDs in series would make 6V. So there's 4 of those in parallel.
 
Geniuses! The lot of you! Ok, well not Mark, his contribution was far too easy to implement (just kidding of course lol).

The 1K resistor across the LED legs fixed the problem! ECM bulb check works as it should, and no dim light when it's running! I'm going to call this one case closed as soon as I figure out how I'm going to add the resistor to the wiring. Wish I would have figured this out before I epoxied the bulb together...I may re-do the bulb holder assembly.

If anyone was just replacing bulbs with LED's, this would be a quick and easy fix. As little as the CEL should be on, not much point vs. the stock bulb, but if you like to make things difficult like me, this should help if you run into the dim CEL.
 
Congrats!!
As I mentioned earlier, you should see if you can up the resistor. I'm sure you have 1/8 watt resistors and with the CEL light on that resistor will see almost 1/4 watt. You might be able to run as high as 10k ohm or even higher which would lessen the chance of releasing the magical smoke that makes your resistor work.
By the way, did you check to see if you could see any AC voltage at the ECM when the LED was dim? Just curious.
 
Did not check for AC voltage. I can, but as usual, next week or so is shot. I can test with different resistors, looks like I've got a nice selection.

Assuming it possible, if I had "too much" resistance, what should I see? Dim light?

That way I know what to look for as I work my way up.
 
Yes, you'd get a dim light again. That's the worst that would happen. I think the only thing the resistor is doing is giving the driver some feedback to that transistor in equivalent circuit you see in the datasheet. At some point the resistor may get too big and the circuit would see it as an 'open' so it'll start flickering very fast again (assuming my theory is correct). As it is now with the 1k, if you leave the CEL on how warm does the resistor get?

I'd take the LED assembly apart again and put the resistor inside the case instead of past the light socket holder. That way, in the off chance you want to (or need to) go back to an incandescent bulb it won't cause any problem. Not that I think it'd interfere with an incandescent but just to be on the safe side.
 
Unfortunately I don't think the resistor will fit inside the housing anyway. The bulb holders are pretty small, the resistor pretty large. I'll probably just use the resistor as a "jumper" at the instrument cluster connector, soldered to the terminals for power and the CELwire. If I did ever go back to incandescent, I'd just cut the resistor legs.

I didn't notice that the resistor was getting warm, but with our temperatures being what they were yesterday, I'm not sure touch was an accurate indicator. I'll try and do a better job checking that. Worst case, instead of getting different resistors, would just be to use two in parallel, no?


As much effort and time as this cluster conversion has taken, I better not have to go back to incandescent! Kind of the point with the LED's, I shouldn't have to replace any of them the rest of the time I'm alive. Even if they last half of what they should (5000hrs I believe that would be) thats ~six months of continuous run time, and I doubt the truck will see that before I'm gone. :)

But it will be a mysterious mess for whoever inherits it.
 
Two equations that I think everyone should know:
V=IR
P=IV

And for those of you that don't know:
resistors in series --> add their values together.
resistors in parallel --> total resistance = (R1 x R2 x R3 ...)/(R1 + R2 + R3 ...)
It's explained here (although there is a slight typo in one of the parallel equations):
http://www.dummies.com/how-to/content/electronics-components-combine-resistors-in-series.html

When you have two resistors of the same value in parallel the effective resistance will be cut in half, the power in each will be the same (voltage across each is the same).
When you have those same two resistors in series the effective resistance will double, the current in the entire circuit will be cut in half (so will the power) and each resistor will dissipate half the power.
Better explained in this video:
http://www.youtube.com/watch?v=WNPFyqkDn6Q

So, if you put your resistors in parallel you'd only be cutting down on the resistance, the power in each would remain the same. Put them in series and you'll double the resistance and your resistors will each dissipate 1/4 the power (half the current and each resistor shares half the power).

If you need larger resistors I could send you some. Thought you were trying to make it look neat and not all jumbled together.
 
Did you see the back of my cluster? Jumbled together is an understatement. The connector is large enough that they would be well protected/out of the way as I suggested, but one would be simpler.

I'll see what all I've got. I think I still have the package that listed their values.
 
I actually thought it looked kinda nice with all the wires neatly bundled together. :dunno:
 
Thanks! But compared to that printed circut "board" GM used, it looks a lot more messy to me. It certainly looks more complex.

If someone really wanted to do a bang up job they COULD all be routed inside the cluster, but for something you will hopefully never have to look at again, why bother?
 
All my resistors are indeed 1/4 watt. I tried a 10K resistor right off the bat, and the light works as it should, so I'm going to stick with that one!

Interesting note that has little to do with this project, but in case someone searches at a later date...the headlight switch must be grounded to turn on the interior lights properly. It's a horrid setup, since it grounds through the nut that clamps it to the dash.
 
:waytogo:
If you wanted to you could get an 1/8 watt 10k resistor from Radio Shack. From the pics it looks like it'd fit inside the LED case.
 
Yes he could, but I was thinking of something that was readily available. He mentioned something about the possibility of others using this idea.
 
Smaller might fit, but I went ahead and used the resistor as a jumper between the wires at the IP connector because I was there and trying to make a bit more progress lol.

I'm going to have to do some research, the metri-pack 150 connectors that I'm using do not appear to be readily depinned, as if they are a one time use kind of deal.

On the plus side, everything appears to still work like it should. Project should be coming to a close as soon as I put the truck side cluster wiring into the matching metri-pack 150 connector.

I appreciate all of your time and help, the resistor trick should help anyone else that wants to go LED lighting who runs into the same issue.
 
You must log in or register to reply here.

Latest Posts

Top Bottom