Cypress, the term polarized threw me. The articles read stated that if inserting the led in the connector socket did not illuminate the light, then rotate 180 degrees for the bulb to work. That implies polarization although they didn’t use the term. Until I physically shorted out a led, I didn’t totally understand.

So, I assume you took out the fiber optic cable with using led’s. Cool. I’ll see if I have to do the same. In duplicating what you did, what characteristics did the component led’s purchased locally have, such as luminous intensity & viewing angle, if you remember. I assume they where 5 mm leds.

I’ll do some checking around work to find out what kind of light source is used in fiber data switches & routers. Be interesting to know!
I’m going to setup the prototype tonight on the ebay cluster panel, which just so happens to have a burned out wiring trace to the rheostat. How convenient! I can only guess that who-ever sold this had the same issue and didn’t fix it, but replaced the cluster. I’ll post pic’s.

I found the same thing about the LED's and the fiber optics. I went back to filament bulbs for the general lighting.

On the general lighting, one guy installed a multi LED bulb (about $17 each,) and says he got good light transmission through the fiber optic. Seemed a bit too expensive for me.

I found another problem with LED's in my conversion. There is a bigger filament bulb, I forget the number but not the T10, that designates the battery charge/discharge. I changed that to a LED along with everything else in the cluster. When I reinstalled I noticed the warning lights would not go off until the r's were up to about 1500 or more. In one of my many cluster removals chasing an electrical gremlin, I changed that battery light back to filament and the warning lights worked as intended. Put the LED back in and it went back to the delay. I believe that has something to do with the alternator requiring some power to function properly.

Bob338, I noticed the same thing with the charging light. Upon starting the charge light comes on. After increasing the idle speed the light goes out. I’d assume the same situation as you, that is the alternator requires a certain amount of power to start working and prehaps rpm. If memory serves correctly, generators will produce power without an initial 12 volts, but alternator require voltage to start working.

I was very curious with his claim & asked for pics with the dash switches lighted/lit. I am still waiting, maybe you should ask him to post pics too?

He uses the 9 led (T10/194), you have the 5 led (T10/194). You could very easily find out for yourself, just connect your T10 5-led to +12V & point a fiber optic cable directly at it & see how bright it is. I know when I did it, it was still dimmer than stock.

The 9-led T10 is $12.95 at superbrightleds, most likely even cheaper on eBay.

That uses a T10/194 style bulb, see my pic on post #6...it shows I use a 5-led T10/194 installed. I don't have issue you mentioned.

Yes, I took out the fiber optic cables completely. 5mm leds are way too big...I am not even sure if 3mm will fit. Try to find ones that are "high intensity". My led is just like high intensity led shown here, except miniature version of it.

Attached Thumbnails

 

Hmmmm, looks like I'll try that 5 LED for the battery. What I have is a larger single LED so that's obviously the difference. I'd rather have uniformity. Thanks.

That 9 LED might be worth trying for the illumination. I used the same single LED for the illumination and that obviously didn't work. Not worth asking that guy, could have been testosterone talking, as well as personal opinio. Likely he'd not admit a mistake. Is that 9 LED the one you compared and found lacking for the illumination?

When I did mine 1-1/2 years ago...I had to borrow a friend's spare cluster just to measure what the allowed depth was. At that time, the 9-led was too tall for my cluster..only the 5-led one would fit.

When you point a high intensity led at it, it was dimmer than stock. So 5-led or 9-led, it does not matter...you'll still have this issue, just my opinion.

You know I don't light my cluster using 5-led bulbs anymore, I use high intensity led strips now.
Have you seen my two threads on this.

Your other choice would be to add a bypass resistor to the circuit, to carry the field current the alternator needs. That's what VW did on their cars that used an LED for the alternator light. I don't know offhand what value to use, but you could figure it out with Ohm's law based on the rated wattage of the original bulb.

__________________
1981 Mercedes 300TD, 1994 Honda Civic Del Sol
http://mefi.us/images/fuelly/smallsig-us/67195.pnghttp://mefi.us/images/fuelly/smallsig-us/103885.png

Thanks, and yeah, I remember your cluster illumination and really like it. It's just a bit more than I want to get into to do something like yours. Maybe, if my frustration gets more than I can handle down the road.

Hahaha, you know you want it...did it 3 times b4 I was happy.
Just trying to save you precious time.

Attached Thumbnails

     

Don't do that to me!

Those LED's in you A/C control I really want! Had mine apart and started to order parts but then reality set in. Had too many little nit-picky electrical problems all over and a malfunctionion climate control, so decided to straighten out the problems first before tackling upgrades. I'm on the short strokes of fixing the last one in my coolant temperature gauge and once done with that you may have lead me astray.

If you still have the climate control unit still apart, resolder every solder joint...you'll be surprise that even if the solder joints looks good, it could still be a cold solder joint...causing intermittent failures.

Ok, here are the pic's. The conductor(s) on the led needed to be flipped around so they where polarized for the MB connector, that is one side to 12 vdc and the other to ground. I put the prototype cluster back together to see what the light was going to look like.
Also, Cypress I started to think about something you said about the switch lights not looking correct through the fiber cable using a led light source. Not sure as to why this is, but I found the following information. First, the original bulbs are incadascent bulbs. I found a graph of what wavelength of an incandascent light source and compared to that of natural sun light. The next graph is that of a cool white led. Since the incadascent light wavelength of yellow and reds are reduced in the spectral analysis might explain why the light didn't look correct. I have found led's which emulate natural sun light. Might have to try these out.
So since this worked, now I need to commit and do this to the car, plus fix the odometer.

Attached Thumbnails

         

Interesting, but it looks from his pictures that Cypress was using yellow LED's whose intensity is on par with the daylight. I'd be interested to hear from him to see if that's the case, or the color of the LED he used for comparision. I WAS using the daylight LED though it was a single LED, and since the light is directional from a LED and the light has to reflect to the fiber optic in the cluster, I figured that to be the reason for the dim lights. Still, it seems like the investment ($$) is hardly worth the return at something like $30 for the two LED's, with the shipping. Hardly seems worth it.

Cypress, I reassembled my CC unit and it's working fine, but I have another board for the cluster which I'm going to resolder. Has anyone heard of baking the boards in an oven to melt and resolder one? Not sure where I read it, possibly here, where putting a board in an oven at 365° for 6 minutes, takes care of those crystalized solder joints. I've stripped a board taking off gauges and their internals and I'm going to try it unless someone has a better suggestion. I have a fear that with a small soldering iron and so many connections, some of the solder will run to an adjacent joint and short the thing.

Looks pretty bright, good job.

Cool white should have the most luminous intensity & have all the color's wavelengths. But yet, it's dimmer...

Incandescent 360* light dispersion, while LEDs have directional light dispersion. This would make a difference in how much light is reflected to the fiber optic cable.

Will be interested to your findings on the natural sunlight led.