A cure/replacement for the self canceling signals

[Wallowgreen]

Wallowgreen has already developed a µP controlled replacement, but the cost to build might end up being on the high side.

Exactly!

The cost to me for development was real cheap but that's because I have the luxury of being in a position where I can obtain sample components, development suite's etc.

Even ignoring set up one time expenses, I did a costing to produce the unit I designed for the use of others and the cost of commercialisation became prohibitive to the point I felt no one would pay that much. Consequently, I didn't take it any further.

[rustybronco]

And that's the reason for my trying the el' cheepo route using latching relays.

http://www.thegsresources.com/_forum/attachment.php?attachmentid=15195&d=1323780825

Digi-key 8A dual coil general purpose latching relay. one coil turns it on the other turns it off. $3.82 each

http://parts.digikey.com/1/parts/752928-relay-gen-purpose-dpdt-8a-12v-rt424f12.html

[rustybronco]

This problem keeps eating at me.

So far I'm thinking of using a 556 (dual 555) timer in bistable mode to operate the relays and using a second 555 as a coundown timer when reaching a speed of 15 miles per hour. Then using a lm2907 as an over speed latch to start the 555 timer. Possibly a nand gate used so you can't operate both signals at once.

Any comments?

[chuck hahn]

Has anyony tried soaking an OEM board in something that would dissolve that hard black sealer to expose the undisturbed components and curcuit routings??? Kinda use that as a master of sorts. And Martin..If it worked for many many years and was totally reliable I would buy 3 right now regardless of cost. Reliable and WORKING turn signals are way cheaper than a life at any cost.

[rustybronco]

I opened one up by grinding away the outer plastic covering then using a heat gun to remove the potting. A few of the caps exploded while I was heating the potting material.

Chuck, I'd rather use later components then the originals components. It would make repairing it a bit easier with newer components down the road.

[chuck hahn]

For sure..just was thinking about the getting to see the pathways and numbers on the components to reference off of..thats my thinking here.

[robertbarr]

This problem keeps eating at me.

So far I'm thinking of using a 556 (dual 555) timer in bistable mode to operate the relays and using a second 555 as a coundown timer when reaching a speed of 15 miles per hour. Then using a lm2907 as an over speed latch to start the 555 timer. Possibly a nand gate used so you can't operate both signals at once.

Any comments?

Boy, you're really aiming at an elegant solution. Mine has all the subtlety and sophistication of a ballbat -- almost insultingly primitive.

A pair of PICAXE's -- one is dedicated to watching the speedo reed switch and counting. The other handles everything else, which includes bar controls and the actual flashing. Output is to a pair of SPDT relays, one for left and one for right, with a simple Op-amp relay driver (like a UL2003) handling the relay coils. Two microcontrollers because the very instruction I need -- COUNT -- hogs a single unit down.

(Current PICAXEs provide a simulated "parallel processing" feature. OK, it's a polling routine, but who cares? It's great for programmers who are either unsophisticated or lazy, or both. Hey, I'm comfortable with either! But that 'COUNT' command screws up my flash rate. Can't have that!).

I'll be using the bar control from a '79. I have one that's almost new. All momentary switches (for the signals, of course. I'll have to adapt the switches for Hi/Lo beam and horn. No big deal there).

The relays will be socketed and mounted external to the 'box', which can now be really compact -- since there's almost no guts! Socketed relays mean that they can be replaced easily and quickly. I've found Bosch (which no longer makes relays!) on eBay for about four bucks a copy.

My concerns are of robustness (is that a word?) and maybe switch bounce. Other than that, the possibilities are endless. I need a robust unit because I haven't been able to figure out a "limp-home" mode for this. Either it works, or you have no signals (other than your hands...).

I'm also a bit concerned about power conditioning, but probably don't need to be. I'm planning on something really basic like an LM7805 linear regulator to feed the processors and an LM7812 to feed the relays. Protect it with a cap. Maybe a zener, but I'm not sure what to use there.

I should have this operational by early May or so.

[rodman101]

this all seems very complicated, what exactly is wrong with the two latching relays schematic, other than self canceling signals, and no colors to the schematic lol, hopefully not a stupid question, i just prefer works as opposed to not working

[rustybronco]

Mr Barr, I don't know if mine is all that elegant of a solution or not.

One of the considerations I had when I thought of making a replacement TSCU, was to make it from commonly available parts driving a pair of relays, only because I don't care to add learning a programming language to the things I already have on my plate.

I started out with the idea of using a pair of two coil SR relays to direct the current to the flashers and adding a means of 'time out' signal cancellation in addition to the cancel button. That evolved into using a bistable flip flop circuit to drive a pair of replaceable automotive SPST relays in sockets for ease of servicing as you have indicated you are doing with yours. One of the things on tomorrows agenda is to measure the coil current needed to operate the relays I already have on hand and see If I need to add a driver circuit or just go ahead and find a relay within the 555's requirements.

I'll 'probably' throw in a dedicated means of voltage regulation, but I'll make the decision after I see how much trash on the line those components can handle.

I have to admit, during the process of figuring a solution to this thing, a µP was starting to look like the easiest way around it.

[TheCafeKid]

Orrrrrrrrr

You could buy a simple $7 flasher relay and remember to turn off your damn signals on your own

I've had GSes with the self canceling feature. Some worked, some didnt. I found them annoying, particularly when one has to restart the timer in a long left turn que. I just did it my damn self so I knew they were on, or off....

Much simpler than fancy gizmos. Call me old fashioned I guess.

[rustybronco]

Josh, on the earlier model TSCU's with the factory provided turn signal switch it isn't quite that simple.

The early model turn signal switch has momentary contacts so you would need to hold the switch in the right or left turn position while its flashing, or remove the centering spring so it will not return to center when you release the knob AND rewire to eliminate the TSCU. They are nothing like the 80~up units.

[rustybronco]

what exactly is wrong with the two latching relays schematic, other than self canceling signals, and no colors to the schematic...

Nothing wrong with using the two latching relay method of getting around the problem.

Here's another solution -> http://www.thegsresources.com/_forum...3&postcount=21

AND, vee dough nee no steenking colors...

[rustybronco]

Mr Barr, Instead of using two picaxe's, what about these?

http://www.ti.com/tool/msp-exp430g2 at whopping $4.30!

http://www.ti.com/tool/ez430-f2013 $20.00

[robertbarr]

Orrrrrrrrr

You could buy a simple $7 flasher relay and remember to turn off your damn signals on your own

There's a trouble-maker in every crowd...

[Dogma]

There's a trouble-maker in every crowd...

Not only that, a hypocrite too! As I've learned from much bitter experience, he doesn't cancel his turn signals, automatic or not. Just pokin' ya, Josh.

I don't actually disagree. The self-canceling system isn't really smart enough to take care of it's own job. You have to monitor the signals anyway, and for the wrong reasons. I wouldn't mind going without. I'm so trained to cancel turn signals that I do it in the car now, too.