Cluster Lighting mod

[Caddyguy]

Hi, I just wanted to document my process of upgrading the burned out incandescent bulbs as well as get opinions / input along the way. At this point I don't have any final design decisions made. At the expirimental phase. Anyhow, here are pics of where I am so far. Just some breadboard testing.
I'm fond of the blue and green so far. The pink kind of looks nice too.

[harofreak00]

Been there, tried that. Unless you're able to change the color of every instance in the interior, it will look out of place.

I ended up just changing a few to blue: http://www.pontiacbonnevilleclub.com/fo ... 10210.html

[Caddyguy]

Most of it is gutted. No radio. Climate control converted to wire and relocated. Going to take a look at the lighting in that. Also going to take a look at the DIC lighting, but with the right main color choice that being red could be worked into a scheme.
Go fast beater / cold weather camper conversion.
Also in the dash will be controls for the diesel heater, and various color switches, wideband, maybe phone mount or tablet.

[nos4blood70]

Red back light is at it was intended to be. As Andrew said, once you start changing colors it looks no longer factory and tacky.

[Caddyguy]

Technically the warm white mimics the factory incandescent the best.
The red gives it an extra pop, especially in the needles. The car is extremely far from factory.
Agreed the red film in the overlay is kind of a pain for color change, but doesn't really bother me.
I haven't looked at the HVAC lighting. I'm pretty sure nothing will be able to be done with the vacuum florescent display. At least not without a lot of datasheets and sourcing.
one reason I'm leaning towards white or green or blue. I do like the enhancement of the red though.
IMO, blue works pretty well with red.
Green with red, while it could look christmas like, it could also be open to interpretation for other themes.
White just kind of makes it pop like something more modern.
Pictured are all just traditional LEDs. To use them will require Modding the circuit on the board to convert it from parallel to a combination of parallel and series.
I do have some of the 3 SMD Through hole lights, but only white and blue.
i do want to retain a functioning dimming circuit, so have do do some playing around with that. I'll likely have to underdrive them so the pwm that controls them remains effective.
I've read for some they seem to still dim, but not really.
I'll also likely be changing blinkers, PRNDL, and deleting unnecessary warning lights.

[Caddyguy]

It is looking like everything should be doable. I have to source some SMD LEDs.
Climate control doesn't appear to be an issue.
I think I should be able to peel the diffusion layer and red filter off the DIC and Climate displays pretty easily, then replace with a white diffusion layer.
The type of LEDs I have at the moment for testing, using a white one, looks white through the display head on but slight tilt and it turns red.
I'm not sure why I was thinking the display in the climate was vacuum florescent. It is not. That was a brainfart.
I am going to be stupid and experiment with RGB LEDs and an RGB controller.

Pics

[Caddyguy]

Progress Video: https://youtu.be/3tzIiBZN2_Y
Update: I have some SMD samples to work with. It's looking like I'm going to take the RGBW route. The white in the mix can really help the red tinted areas show through while still having a bit of accent color. This can kind of be seen later in the video.
In experimentation to keep the factory dimmer working without messing up the color mix I have been using a PWM fan controller to simulate the PWM from the IPM. I'm not sure how close of a comparison it is though without a scope.
Interestingly with an Alpena RGB controller from Wally world, running the output of the fan controller to the RGB controller, the dimming works quite well and maintains color mix until they all shut off. I suspect there to be very little voltage regulation / protection in the Alpena controller, if any. What you put in seems to be what you get out on the 12v LED strip rail.
The controllers I plan on using are WB5 ones off of Amazon.
They seem to have substantially better protection and filtering. Powering the WB5 with the fan controller, current draw while trying to dim pretty much stays the same until the controller cuts out.
This is actually what I was expecting the Alpena one to do. I guess that is a testiment to how cheaply the Alpena one is made.
The plan is looking like mounting the RGBW SMD LEDs to some Breakout daughter boards. Affix them to the cluster in location and wire up as needed. I have some test boards on order. Should get them in a week or 2? I would like to make the boards myself, but with how little it costs to get them made now a days, it is kind of hard to justify it.
Down the road I would like to do an Audrino or Raspberry PI touchscreen head unit with built in RGB controls and maybe borrow a steering wheel button to cycle colors.
Just planning on the WB5 controllers to get 2 zone up and running asap. 2nd Zone probably being some RGBCCW dome lighting just because.

[Caddyguy]

I did just take a look at the issue I am going to run into with the buttons.
I was thinking they were red plastic on the bottom with a clear layer, then white layer, then the top grey layer.
After digging into one that is not the case.
As I'm sure some are aware, they are fully molded with translucent red then a coating of white print, then the top grey coating. I suspect the grey is then laser engraved to produce the label.
I was hoping I could use a soldering iron to get through the red layer, but broke through the front of the "Auto" button.
I think the white coating made it look kind of red then clear when it is really all red.
I do have a plan to resolve this so I am going to continue with the full color change.
That said, the plan is to make new buttons. Doubleshot keycap style for those familiar. Translucent white with as close to the factory grey exterior that I can manage. Maybe a clear core or clear core with white factory like coating then do the exterior color pour.
I'm going to have to get a small CNC for that first though, so that will be a secondary project after completion of the color change. Also, some silicone and resin.
As for the ruined Auto button, no biggie. Part of R&D. Never a cheap process. Also, I have a parts car and there is a decent amount of climate control modules available in not too far away junkyards ranging from $25 to $50 if it came to that.

[Caddyguy]

If anyone has any idea what type of coating the top grey button layer is it may be a simpler process to duplicate them and have them laser engraved than doing a double shot casting. Also no CNC needed.
I suspect it is a polymer coating on the mold before they are injection molded. Thinking they spray it to a thickness, then spray the white, then injection mold. The laser tuned right would stop at the white layer.
Edit: Doing a little digging, I'm pretty sure the process used is called FIM, Film Insert Molding or Film Backed Molding. Now to figure out if there is a way to duplicate it with silicone molds and syringe pressures instead of injection molding machine pressures.

[harofreak00]

The forum used to be filled with content like this, but most users have moved on from the forum and from Bonnevilles in general.

You will appreciate this post from ddalder: http://www.pontiacbonnevilleclub.com/fo ... 15462.html

[Caddyguy]

I like his craftsmanship. Good to see I'm on the right path of how the buttons are manufactured normally. I can't just repaint them though as I need to switch from red plastic to clear.
As far as most moving from Bonneville's, I'm not most people. I'm shooting for 1 million miles on the odometer, well see what happens when the digits max out anyway.
Anything worked out on this project or abilities gained can also apply to any other vehicle on the planet.
As an aside, I was always planning on adding the heated washer fluid back to my DTS after removal due to recall, but it never made it high enough on my priority list. Also in the back of my mind I didn't want to risk the window cracking on a really cold day.

[Caddyguy]

I cracked open one of the WB-5 controllers to work on the dimmer switch issue.
The input rail is actually directly connected to the output +12.
I tried removing one of the caps to see if that was smoothing the output but made no difference. I didn't really expect it to work as the chips inside won't be happy with PWM.
I then tried running the PWM + to the 12V of the lights, bypassing the controller. I hooked the PWM - to the controller in - and ran a constant 12v to the 12v in. This behaved in the same way.
Testing this following part on the Red channel.
Finally I thought to just power the controller with constant + and -. I left the PWM + bypassing the controller going straight to the lights. For the PWM - I desoldered the Source pin on the control transistor and raised it from the board a bit. I touched the PWM - to the raised source pin and I now get a nice dimming effect on the red channel with the fan controller.
I believe all I should have to do is desolder the control transistors. Mask off the Source pad. Solder the transistors back down. Add mod wire to the source pins and add a PWM - input. Likely a PWM + passthrough to keep it clean and on one connector for input one connector for output.
This actually isn't a bad solution because it keeps the PWM from the IPM from having to power the electronics and only has to power the lights.
I will post pics of the modded controller when done.

WB-5 RGB+CCW controller




P.S. with the buttons. It may work out to paint new clear castings and laser them then. That could potentially reduce the effort of doubleshot. I have to read through that thread in detail.

[Caddyguy]

WB-5 controller modified for Overall brightness control using PWM without messing up the color mix.
Inputs: +12 V, Ground, PWM +, PWM -.
Outputs: Red, Green, Blue, White, PWM +.









Controller problem appears to be solved. Just need to make a little circuit to turn the controller(s) on only when the PWM is active. otherwise the electronics would always be on with the ignition even if the lights aren't, unless I come up with a better plan looking at the diagrams.
I did damage one of the MOSFETs a tiny bit during removal. It likely still works, but I just put it on the channel I'm not using. I warmed the package up a bit too much and lifted on the middle cut off pin a little too hard. Took a little chunk out of the epoxy. It wasn't even the first one. It was the 3rd or 4th. :-\ Little rusty on my hot air work.
Another reason for the project is I kind of miss some aspects of the manufacturing world.
I ordered some replacements just in case and I want to have some various MOSFETS on hand for other projects too. The datasheet doesn't match exactly but I think they would be close enough if I need them. To get direct replacements required a minimum order of 4000 from China. A new controller would be like $25.
The second controller should go smoother. I could have gotten it done sooner, but I need to take less breaks and stay off YouTube.

[Caddyguy]

A little info about the PWM Signal.
I measured with my meter and it appears to be 100 Hz.
With the dimmer switch in the minimum position it is 40% duty cycle.
in maximum position just before it clicks the interior lights on it is at 90%.
It appears to stay at 90% when the interior lights click on too.
working on the current hurdle. Not really a necessary one, but peace of mind.
I thought about running ignition voltage to power the controller, but after thinking, that will result in the controller being powered on when the lights are off during the day.
Not a big deal, but with no A/C I want as few heat sources as possible in the summer. also longevity of the controller.
I am working on a circuit to trigger a relay when the PWM signal is present.
I don't have a scope but was able to use an online schematic simulator to generate the PWM to play with. I have some component values to test out.
I have old radio shack components on hand for testing, but for the final build I'll likely order new, probably SMD.

[95naSTA]

This is a bit out of my current swim lane (no pun intended) with electronics but it's cool to follow along either way.

[Caddyguy]

I finally broke down and ordered an oscilloscope. Not the one I was planning on ordering for my first one, but it should help get me on my feet.
I used to repair license plate cameras and build and repair Library self checkout and return systems.
I'm very much a noob when it comes to designing from scratch.
I like learning though and my endgame is building repairing cars for people to get away from the corporate world.
I'm currently a refrigeration / trailer / truck mechanic with more of an emphasis on the electrical side of things. I do love welding though. :-)

[95naSTA]

Nice. I begrudgingly do things that aren't mechanical but I'd rather learn whatever I can to avoid having others do anything for me.

[Caddyguy]

https://youtu.be/X_OInW3hLRk

Breadboard testing appears to be working.
I probably couldn't have gotten it working without the oscilloscope. At least not on the bench and without danger to the car.
The fan controller was not an appropriate substitute at all. It doesn't appear to be an actual square wave at all and it was running about 25 khz. I suspect being it is a motor driver there may be some inductors involved? Haven't opened it. It is another part of the build to control 2 - 6" ventilation fans mounted in the rear 6"x9" location. I had it laying around and am impatient.
I played with trying to amplify the signal generator on the scope with some success, but I think I kept killing transistors so I just ordered the PWM generator.

The PWM signal passes directly through the controller to the 12v wire of the LEDs. The PWM is also tapped off to a low pass filter which is fed to a comparator. When the PWM is present it triggers a MOSFET to drive the coil on the relay which powers up the controller.
The controller handles the color mix and the vehicles PWM handles the brightness.
Being incandescent from factory I don't think the car will have any problem powering this instead.
I'm not so sure I had to reroute the ground paths of the mosfets in the controller as I have both grounds tied together anyway? I did that before I had the P-channel / N-channel thing somewhat solidly grasped. Pretty sure the factory computer must use P-channel or something like that. That is how they get a single signal wire and can ground a load directly to the body. If an N-channel was used the ground has to come back to the computer or whatever controller to be switched and the load gets constant voltage. that is why all these controllers have PWM + and -. N-channel are cheaper.


This is just the chunk to simulate the PWM for bench testing.
I had to add a P-channel mosfet with the blue breadboard to the PWM generator to simulate the PWM as the vehicle uses. Due to this just turning the controller off makes the output go to 12v, so as a workaround just to bench test I just plug and unplug the wire. An additional side effect of this is the settings on the controller. Adding that MOSFET flips the duty cycle. The factory range seems to be 40% to 90% Duty @ 100hz. My equivalent settings are 60% to 10% Duty @ 100hz. 10% being the brightest. Basically controlling the duty cycle for the off time instead of the on time.


This is the lowpass filter / comparator part. The red breadboard is the driver MOSFET for the relay coil. (Relay is massive overkill, but I like to keep systems separate and I like those style relays.)
The LED I'm covering up was just an indicator if I had signal going to the relay coil driver.
Next on this part will be to generate a BOM of automotive rated components and layout a board.


Overall picture, mainly focused on relay to controller power.


LED assy. Equivalent to an LED strip, just with a custom layout instead of actual strip.
Next step is to get the layout and wire routing figured.


Picture of my test batch of boards. I haven't tested them out yet. I should have put in spots for resistors, but we'll see how it turns out.
I am going to have to get the capacity to make boards I think though. I want faster turnaround for the prototyping phase.

[CMNTMXR57]

Warm red = Pontiac.

Going any other color is blasphemous.

[Caddyguy]

Guess I'll just set it to red then if there comes a time to not be blasphemous. :-)
Pontiac also probably didn't intend it to have a diesel heater or a refrigerator. Lol.