Toyota GR86, 86, FR-S and Subaru BRZ Forum & Owners Community - FT86CLUB

Toyota GR86, 86, FR-S and Subaru BRZ Forum & Owners Community - FT86CLUB (https://www.ft86club.com/forums/index.php)
-   Cosmetic Modification (Interior/Exterior/Lighting) (https://www.ft86club.com/forums/forumdisplay.php?f=36)
-   -   The proper (and I mean PROPER) turn signal flasher mod for LED bulbs. (https://www.ft86club.com/forums/showthread.php?t=114284)

catsspat 01-05-2017 10:15 PM

The proper (and I mean PROPER) turn signal flasher mod for LED bulbs.
 
Believe or not, I removed the TapTurn from my BRZ and put the stock flasher back in. TapTurn has this annoying glitch at the beginning, possibly coming from when the turn signal switch contact first connects. All physical switches, including the turn signal stalk, will glitch when it connects and disconnects. Module should damp that away, but it doesn't seem to do a good job with that. My TapTurn module is one of the earlier ones, so I don't know if the newer ones are better.

I sort of wanted the assurance of dead bulb detection (hyper-blinking) anyway, which TapTurn doesn't provide.

The hack method was to grind down the shunt resistor until the hyper-blinking stops. This isn't the most elegant method, but it works, and as long as you don't grind it down too much, dead bulb detection will continue to work. Here's the grinded flasher from my 2013 BRZ that is currently in my 2015 Impreza:

https://c5.staticflickr.com/6/5176/3...7c4f4538_b.jpg

I've now modified the stock flasher from the Impreza and put it in my BRZ. They have the exact same part number anyway.

First, I took the shunt resistor off and measured it. Its resistance is extremely low, so I couldn't directly measure it. Instead, I ran two 21w bulbs (the stock turn signal bulbs) through it and measured the voltage across it. With the supply of 12.6V and 3.62A, voltage across the shunt resistor was 0.068V. With the Ohm's law (V=IR), you can figure out the resistance as 0.0188. My instruments aren't perfect, and neither do they need to be. I'm pretty sure the shunt resistor is 20 milliohms.

https://c3.staticflickr.com/1/394/31...3013e5ae_b.jpg

Since I'm going from 21w x 2 = 42 watts to somewhere around 5w to 5.5w x 2 = ~10.5 watts, it's basically 1/4 power.

What the flasher module is measuring is the voltage across the shunt resistor. It's calibrated for 2 bulbs (front + back). This voltage is current*resistance, and if one or both bulbs burn out, current will be reduced in half or to zero, which will also reduce the voltage across the shunt to half or zero. The goal is to make this voltage across the shunt the same with the lower power LED bulbs.

Power=IV, linear relationship, so it's trivial to figure out the new resistance value you need.

For 21 watt incandescent, it was 20 milliohms, so for 5.25 watt LED (1/4 power), resistance needs to be 4x (80 milliohms) in order to maintain the same voltage across it (remember V=IR, I is 1/4 now, so R needs to be 4x in order to keep V the same).

If you have 4 watt LED, then you'll need 105 milliohms.

Note dead bulb detection logic is disabled when you run hazards because now you're running 4 bulbs, even though the detection logic was calibrated for 2 bulbs. However, even during hazards operation, all power goes through the shunt resistor. Therefore, it has to be able to handle the current running all 4 bulbs. In my case, 5.25 watt x 4 is 21 watts. At 12.8V nominal, that means 1.64A is going to flow through it. 0.08-ohm resistor will be dissipating I^2 x 0.08 = 0.22 watts. So even a 1/4 watt resistor is sufficient in this case. However, if you put incandescent bulbs back in, we're now talking about 84 watts total. That 80 milliohm resistor now needs to dissipate 3.45 watts! Even if you had a resistor capable of handling that much power, it will heat up too much inside the flasher module. This is why you should NEVER put the incandescent bulbs back in if you modded the flasher module for LED bulbs.

I wanted to be extra conservative, and I had a hard time finding a satisfactory 0.08 ohm resistor, so I used two 0.16 ohm 1 watt resistors in parallel to form a single 0.08 ohm 2 watt resistor. This is way overkill, and in theory, it might even handle incandescent bulbs for a little while since the turn signal bulbs aren't constantly on, but I would never do that.

Final result:
https://c8.staticflickr.com/1/474/32...76f6d620_b.jpg

I tried it and it worked as expected, and when I pulled one of the bulbs, it immediately went into hyper-flash mode.

The nice thing with this mod is that I can easily undo it, as long as I don't throw away the original shunt resistor. Well, I can always buy another one. I'm pretty sure it's 20 milliohm 1 watt type. If I ever sell the car, I'm pulling my Philips Ultinons out, so I'll need to restore the flasher module as well.

Remember, NEVER use the modded flasher module with incandescent bulbs. Generically speaking, too high resistance in the shunt or too much bulb power will overload the resistor and can lead to bad outcome. Some as minor as burnt out resistor, to as major as a fire.

tCsonfrs 01-05-2017 11:21 PM

Yeah thats a lot of work for LED turn signals, hard pass.

catsspat 01-06-2017 08:01 AM

Quote:

Originally Posted by tCsonfrs (Post 2826226)
Yeah thats a lot of work for LED turn signals, hard pass.

How about this?

New Resistor Value = Stock Resistor Value * Stock Bulb Watts / New Bulb Watts.

NewR = 0.02 * 21 / LED_watt

NewR = 0.42 / LED_watt

That's it. The only thing you need to know is the LED bulb's power consumption value (watts).

1) Get a 1-watt resistor of the value you need (0.42 / LED_watt). For example, for 3 watt LED bulb, you need 140 milliohm resistor. 4w->105mOhm. 5w->84mOhm.
2) Get the flasher module out.
3) Desolder the shunt resistor.
4) Solder the new resistor in.
5) Put the flasher module back in.

Kodename47 01-06-2017 08:38 AM

http://www.ft86club.com/forums/showthread.php?t=96999 ;)

Good to see more of the math on here. Easy enough to do solder the resistor in/out too.

Stang70Fastback 01-06-2017 08:58 AM

I haven't had any of the issues you described with my Tap Turn module. IMHO, all the added features you get more than justify the loss of dead bulb detection. If you run LED bulbs, and install them properly (and they aren't cheap Ebay bulbs) you shouldn't really have to worry about dead bulbs anyway. Besides, I have approach and departure lighting enabled on my Tap Turn module, which turns my signals on solid for 60 seconds when I lock or unlock the car, so I see them pretty regularly anyway.

catsspat 01-06-2017 10:10 AM

1 Attachment(s)
Quote:

Originally Posted by Stang70Fastback (Post 2826350)
I haven't had any of the issues you described with my Tap Turn module. IMHO, all the added features you get more than justify the loss of dead bulb detection. If you run LED bulbs, and install them properly (and they aren't cheap Ebay bulbs) you shouldn't really have to worry about dead bulbs anyway. Besides, I have approach and departure lighting enabled on my Tap Turn module, which turns my signals on solid for 60 seconds when I lock or unlock the car, so I see them pretty regularly anyway.

It's hard to describe in words what the glitch is, so here's a slightly exaggerated illustration.

First image is the expected normal behavior. Turn ON-OFF-ON-OFF...
Second image is what I experience, and only sometimes. It's pretty sutble. Some people might not even notice it.

Anyway, nothing against TapTurn. I think it's a cool device. Mine is like first gen, so the only features it has are the tap turn (3-6 blinks), slow hazards, and wig-wag hazards. No approach lights, no strobes, etc. Of all these, tap turn is the only thing useful in real life, but I'll survive without it. I did for multiple decades, after all. LOL

Chimera 01-06-2017 11:18 AM

Wouldn't a simple debouncing circuit 'fix' this.
its prob not noticeable with halogen bulbs which ae more analog and 'warm' up into their brightness. LEDs are discrete in there state, they are either on or off, there is no middle state.

Stang70Fastback 01-06-2017 11:48 AM

Quote:

Originally Posted by catsspat (Post 2826381)
It's hard to describe in words what the glitch is, so here's a slightly exaggerated illustration.

First image is the expected normal behavior. Turn ON-OFF-ON-OFF...
Second image is what I experience, and only sometimes. It's pretty sutble. Some people might not even notice it.

Anyway, nothing against TapTurn. I think it's a cool device. Mine is like first gen, so the only features it has are the tap turn (3-6 blinks), slow hazards, and wig-wag hazards. No approach lights, no strobes, etc. Of all these, tap turn is the only thing useful in real life, but I'll survive without it. I did for multiple decades, after all. LOL

I am insanely OCD, so I can promise you I would have noticed that. It's definitely a problem that has either been resolved, or was perhaps the result of a defective unit you have. I wasn't sure how useful some of the features would be, but I've actually found them quite useful. The approach and departure lighting is actually really useful at night. I've used the wig wags when I was the first person on an accident scene and had to block the leftmost line so someone didn't t-bone the crashed car. The strobed hazards I like (along with strobed lock/unlock) because they draw a bit more attention to the car.

Your writeup is really, cool by the way. As an engineer, I like the initiative you took to get what you want. I'd probably try this if Tap Turn didn't exist!

catsspat 01-06-2017 01:00 PM

Quote:

Originally Posted by Chimera (Post 2826442)
Wouldn't a simple debouncing circuit 'fix' this.

Exactly. For all I know, it was added to later revisions of TapTurn. It has evolved quite a bit from the early days, after all. I'd like to think that people like me (early adopters) helped to make that happen. *shrug*

I did open up my TapTurn just to look and found high quality components all around. Good stuff. If anything, the plastic case is the lowest quality component of it.

SkAsphalt 01-08-2017 02:08 AM

My tap turn works great... kthxbye

Wil 03-25-2017 06:10 PM

Not like I want to steal all the credit from you, but this is a solution I demonstrated some time ago :-)
http://www.ft86club.com/forums/showthread.php?t=96999

ermax 01-22-2018 06:55 PM

My front bulbs are 4w when flashing and the rear bulbs are 3w when flashing.

Does this math look right?

(.42*2)/(3+4)=.12

I ordered a 1w 120mOhm but I'm still getting the fast blink. I guess I need to go bigger. I should have ordered several values rather than a single 120mOhm. Sucks to pay $4 shipping on a .33 resistor. Hahaha. It sucks not being able to buy stuff like this locally.

Edit: Thinking about now. Maybe I should only apply the math to the smaller bulb: .42/3=.14
I guess worst case I go to big on the resistor and I loose the burnt bulb detection which I am fine with.

ermax 01-27-2018 06:16 PM

To follow up. There aren’t that many options with values this low so a picked up a 150 and 240. The 150 was still too low. The 240 was high enough to kill the hyper flash but too high for it to detect a burnt LED. I’m fine with that though.

catsspat 02-05-2018 06:08 PM

Quote:

Originally Posted by ermax (Post 3031920)
My front bulbs are 4w when flashing and the rear bulbs are 3w when flashing.

Wait, so you have different front/rear bulbs? If so, your math is right. You can also just use (front+rear)/2 for the wattage. However, what do you mean by whatever-watts "when flashing?" Flashing means nothing to this equation. All you care about is the bulb wattage when it's ON.


All times are GMT -4. The time now is 06:23 AM.

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.
User Alert System provided by Advanced User Tagging v3.3.0 (Lite) - vBulletin Mods & Addons Copyright © 2024 DragonByte Technologies Ltd.


Garage vBulletin Plugins by Drive Thru Online, Inc.