what's a good shiftlight that would directly connect into the tach signal without pills?

 

I'm working on one right now for my G....

Its designed to be user programmable through the VDC button. If I ever get it working I'll post more info. Its actually a pretty cool ideal - instead of using a big ugly light I'm using a small, 3mm super bright blue LED with a water clear lens. The LED will be mounted in the A pillar about eye level. When not on you won't see it, when on you won't miss it.

I've been working on the project for over a month now and I'm just about finished with the programming. Making the entire system user programmable has turned out to be a pain in the butt, but will be well worth it in the end.

In my search for shift lights I haven't been able to find anything that is special for an affordable price....

 

keep us up to date, I started using my Gtech for now

 

Hey guys, I don't have an opinion on anything speial on a tach light but I do have a heads-up on the install.

The tach signal going to the Tach itself is useless as this is a CAN system with only about five wires running everything in the cluster.

I used an Autometer and it is looking for a square wave output of around 12 volts.
the Coil on plug on the Z has a constant 12 volts negative and positive present on the connector at all times; The signal comng from the ECU (which is the middle of the three wires on the connector) is a less than 1 volt signal that gates a power transistor inside the Coil which is what fires the coil.
This signal was not strong enough for my tach light or my Dynojet Wideband commander Datalogger.
For the tach what I discovered is that when the coil is fired even though the 12 volts to the coil is not turned on and off it dips low enough due to the sudden load of the coil being fired that it DOES create a signal that will work. The 12 volt hot on the connector is the one furtherest down when it is plugged in.
The Injector signal can also be used but what I have found with it is that at higher rpms the injector duty cycle is so continous that the on-off signal to it almost becomes constant and the signal becomes unstable.
I also tried the greddy RPM adapter across al the ECU coil signals and the car would not start at all.

Hope this helps.

 

I was planning in using either the fuel injector signal or going directly to either the crank trigger or a cam trigger. The module I'm working on, however, should swith on ANY voltage over 0.7VDC. Its strange that its only 1v though, did you scope it or meter it? It might be switching so fast that your meter isn't reading it right. I'm finishing up the code right now and once its done I'll get my scope in the car and see what I can find

 

No I didn't scope it; just used a fluke 87 so I may not be seeing peak. Thing is it just gates a transistor without supplying the actual "firing" voltage so it may be a very small signal-not sure

good luck

 

I know when using my Fluke 189 (logging meter) on the door lock pulses the voltage doesn't read right because of the shortness of the pulse (< 0.3S). If I leave autoranging on it misses the pulse all together. It will be interesting to see what the scope shows. The one thng to keep in mind is that you arn't necesarrily looking for a square wave. Only the fuel injectors, at 50% duty cycle will produce a square signal. Everything else (I assume) will just be spikes (like t.he cam/crank signals and I would imagine the ignition pulses). I'll know more once I get this damn software done.

 

Update:

I successfully tied my shift light into the ignition signal only to find that at WOT I get erros as the timing is advanced and it 'tricks' my shiftlight! As the timing advances it actually speeds up the pulses enough to appear as if the engine is moving faster then it really is. So, my conclusion, is that niether the ignition signal nor the fuel injector signal will provide an accurate tach signal for a shift light.

My next step is to move directly to the cam or crank signals. I haven't done it yet due to the changes in pulses per revolution. Right now my shift light can only handle roughly 28,000 pulses per minute which would provide accuracy to only 4666 RPM (6 pulses per revolution). I"m waiting on some hardware upgrades that will increase my throughput by 10, giving me enough resolution to track all the way to 46,666 RPM

I better upgrade my rods and heads so I can get ready to test the upper limits!!!

I integrated the shift light into the VDC switch and that part works flawlessly. If you hold the traction switch down for 5 seconds the LED flashes (for 30 seconds then it times out). Then you rev the car to 1/2 of where you want the light to come on (3500 RPM for a 7,000RPM setpoint). Once the tach is where you want it you push the traction button again and it does all the math processing internally to determine the proper shiftpoint. For the time being I just ran a superbright LED up the A pillar and its sticking out the the top. Once everything is working I'll drill a small 3mm hole in the a-pillar and permanently mount the LED. I used superbright LEDs that come in a clear package so when they are off they are almost impossible to see.

My other setback right now is that my engine is being pulled out of my car right now, so testing will have to wait until I get my new engine installed!

 

I have the shift light prototype installed in my work truck and everything seems to be working perfect. MrTomCat, are you interested in testing one out?

 

I've been using the Gtech which has been doing pretty well but I am still interested.
Only issue now is that I am currently Z less since I totalled it in the rain. Once the insurance finally pays up I'll get another one and will get back at you.
In the meantime can you post some pics?

 

Sorry to hear about the accident. I don't have any pictures of the unit, but I'll try to take some this weekend as I am installing one in a Honda.