With all of these turbos going in I would think a few people would be interested in turbo timers, but I haven't read anything about them, yet.

Does any one have one? What features should you look for when shopping for one? Are they even worth having?

 

I think they are a good idea if you want your turbo to last. I had one on my turbo Honda, but never had one turbo on my 18g'ed DSM and my turbo was in good condition (I let it cool down only when I was ragging on it).

The main two options of turbo timers I know of are the plain jane turbo timer, and turbo timer + boost controller.

Either one shouldn't be all that hard to install for an alarm shop.

 

If you use water cooled turbos (like APS), good synthetic oil and change it regularly (<5k), the need for a turbo timer is unecessary (unless your really into lots o' gadgets. If you do none of the above , it's a good idea.

 

Going Deep and myself have a greddy turbo timer...
I feel better knowing my car can cool down for a few minutes after i drop a whooping on some poor defenseless viper, or porsche!

 

I don't think they are a bad idea at all, they may not be totally necesarry, but I don't think they can hurt.

I'm trying to do a little research on them before I waste any time making one.

 

Funny you'd say that, I've acutaly witnessed were they can hurt.
I live near SF , lot of hills.
Make sure you "curb you tire when parked on and incline" and set that EM brake hard

 

Was that with a manual tranny? I can't imagine having issues with an AT. But, since I have a 6mt I am planning on incorporating the PNP switch in to the system, so it won't let the car run if its not in Neutral.

 

It was a stick (not my car). The issue wasn't leaving it in neutral , it is relying upon the EM brake alone to keep the car from rolling on a grade.
Car rolled into a busy intersection , "nasty sceen". When parked on a grade it is safest to leave in 1st gear or reverse, Set EM brake and curb the wheels. With a TTimer, you loose the extra keep it in gear CYA option.

 

turbo timers are in my opinion a waste of money. on my evo i only let it cool down if i have been getting on it pretty good, otherwise just keep the rpms low for the last minute or two before you get to your destination and they will be fine. also make sure you warm it up fine in the morning.

 

Yeah, that would suck. I would NEVER rely on my parking break to actually hold my car on a hill. Like I said, I would take advantage of the PNP switch so if the car is in any gear the turbo timer would not activate and you can shut the car off like normal.

As far as them being a waste of money, it will take less then a week (most likely) to develop one and then cost me almost nothing (relatively speaking) to build it.

 

Now if you wanted to design something really cool, the Z and G could really benefit from "launch control". One reason the M3s have such impressive 0-60 times. The sensors from VDC are already in place , just need to conrol boost in a ramped fashion taking into accound rear wheel slip.

 

Thats an interesting idea. Compare front wheel speed to rear wheel speed and integrate that in to a boost controller....

 

The Only Downside To The Turbo Timer Is That When You Do Park It, You Will Always Have Someone Telling You...'hey, You Left Your Car On...' Then You Have To Explain To Them The Process. Also It Sucks When Or If You Have The Car Valet Parked And The Valet Does Not Speak English So You Have To Go Over And Tell Him Before He Tries T Start Your Car While Still Running, And Explain To Him The Process Again....lol

 

It can be done on NA autos as well by controlling the TB since it's "Drive By Wire" (the way and M3 does it). Nissan really should have offered it stock , since all the expensive components are already already there to serve the VDC function, and the traction control algorithm doesn't exactly act in "shade of gray" as it is.

However; contollling boost would be a better choice for an aftermarket FI product. Since dinkin with the DBW could lead to serious liabilities if it forked up.

 

A 'safer' approach to that idea would be to intercept the gas pedal and acutally modify its values, which will in turn modify the throttle body at an inverse rate of the slip. I'm going to be busy tonight I have a feeling....

 

Yep , as long as you can "guarantee" a stable "inverse rate".
Don't want no positve feedback in that loop.

 

The way I look at it is I would cut the signal and run it in to the module, run it through the processor and output it back through another wire.

Under normal circumstances there would be no change in the signal. But, as soon as slip is detected it would inversely change the signal at a rate that will hopefully counter all slip.

I'm going to probably need a LOT of help on something like this. I can do the electronic stuff no problem, as you can probably tell I am a fairly quick thinker, as I have this whole module laid out already

However, what I don't know, is what % of 'slip' is detrimental. Also, I know that when you turn your front tires they don't spin at the same speed, so that will have to be taken in to consideration (as it can be seen as slip). That part alone will take some clever engineering (maybe RPM base it - if you are less then 2000 RPM don't modify the signals). At a high rate of speed (above 2000RPM) I doubt the wheels will turn enough to create enough difference in wheel speed to be detectable.

Later tonight I will look in to the wheel sensors and see what kind of signal they give. I would imagine its a pulse per revolution. If that is the case, it would be very easy to monitor all four wheels and the tach signal and make all calculations based on that.

I am also toying with making all variables laptop programmable. This may add a huge delay to the development though, because I would have to work side by side with my brother on the GUI and the downloading of variables. I know together we can do it (he is a really good computer programmer) but we are both VERY busy and rarely have time to get together to work on stuff.

Variables I can think of off the top of my head are:

Slip % (how much faster does one wheel have to be spinning to be seen as slip)
Max Close % (The max % the module would be able to close the TB or boost)
Change % (The 'slope' of change)
RPM setting (Above what RPM to start watching for slip)


I'm thinking this could be a VERY cool project and I should probably start a new thread.

 

When you turn your front wheels will effectively speed up, as they have to cover a bigger circle in the same amount of time. Therefore, the difference in wheelspeed between your front wheels and your spinning rears will be reduced a bit. I think.

 

Yes and No. The tire on the Inner part of the circle will spin faster then the rears, and the tire on the outside of the circle will spin slower then the rears. If you think of two circles with a 3 ft different in radius you will see what I'm talking about (the 3 ft radius sybolized the wheel base - the bigger circle will have a larger circumfrence.)

 

With repest to turning.
I beleive that in a luanch control system , you want to limit the systems operation (at least to begin with) as long as the side-to-side differential is within a tolerable window. Basically if the car is turing or slipping sideways you either want to defeat the system of act aggresivley , like the OEM VDC.

Now another variable , less easy to monitor, but makes gain adjustment easier is to adjust feedback) gain based upon gear selection, since the slip reaction (thottle to TQ output) at the drive wheels will drop as higher gears are selected.

The really tricky part of this type of system is that the transition between Static and Dynamic friction at the tires is a very non-linear relationship WRT throttle input, and thus is difficult to optimize with a simple linear PID algotihm. You'll need to set conservative activation "thresholds" to bound the problem space.

I would expect that a good proto start , would be a micro controller monitoring you variables and a few dash mounted Pots to adjust aspects like actvation threshold , gain , etc. At least it would accelerate empirical testing, and not require a laptop during the many intertive test cycles.