Rob Edwards

Tuomo-

Perhaps it'd be helpful to better explain what you mean by:

Otherwise this all sounds like you want a blingy LED on your dash to show you what gear you're in, which you already know by resting your hand on the shifter.

Asking out of complete ignorance, what's the benefit of tailoring boost to rpm in different gears? Is this some way to limit boost in lower gears so you minimize wheelspin? You can just mash the pedal and the controller takes care of the rest? Or ????

Hilton

Makes good sense to me - easier to deal with than a manual interface to an electronic boost controller like a push button (a la Mr Kuhn)

Might be easier however (and possibly less failure prone) to stick in a small circuit, as ken says, based on engine rpm and wheel speed?

You could source one of the later front-wheel speed sensors (can't remember which year they changed it - 90 perhaps?) to get signal from?

PorKen

All the motersickle displays must be waterresistant, so I could put it out on the hood in a wee pod.

All the gear indicators I see from a quick Google are programmable, rpm vs. speed, ie. not just counters, including the ones above.

SharkSkin

I was kind of scratching my head until I came across the reason you want this, for a boost controller.

How many inputs will it accept? A quick google search indicated that one mode in addition to the one you mentioned is voltage based. Tricky to set up, but maybe you could make a DA-converter work.

If it will accept 4 switch inputs, you could set up some microswitches near the rear shift coupler, using the disc that the boot mounts on to activate fore/aft switches and adding a blade to the linkage to activate R/L switches. One each to detect lever full forward/full aft, another to detect 1-R plane and another to detect 4-5 plane. Both switches off in either F/R or L/R axis indicates neutral.

The big question is whether the boost controller will accept such a scheme. You could build a TTL circuit to convert the switch combinations to a binary number or to the scheme that you described above, bit of a hassle but doable. What types of inputs/number of switches beyond the one you mentioned are recognized?

ptuomov

Imo000, DR, cold_beer839:

I am no expert driver. However, since I grew up in Europe I have been mostly driving stick shift throughout my life. Once I am used to a car, I typically know what gear I have on without any indicator light. [That's not to say that I didn't scare a bunch of people in a stop-and-go traffic by pushing my 928 5-speed box forward to the first gear during my first drive... ;-)] But I have the basic handle of the 5 speed now. I can even tell without looking which one of my Bimmer's sequential seven is on, irrespective of whether it's me or the computer shifting. So the driver not knowing what gear is on is not the point.

The point is the traction limit. Many extreme-hp people on the list repeatedly make comments such as "my first and second gear are completely useless." I am planning to keep mine usable. Since I am building a turbo engine, this is in fact fairly simple ...IN THEORY.

I am planning to use the gear position sensor (or a set of switches as MBMB suggests) to select a boost program that delivers the appropriate maximum rear wheel torque irrespective of the gear. This I believe will maximize the acceleration by keeping the rear wheels slightly but only slightly slipping in all gears.

Here's my plan how to do his with eboost2 controller. This is all just at the brainstorming stage, so constructive criticism is greatly appreciated.

Run the car on a dyno at various boost levels. Do say five full throttle runs at different boost levels. Then, collect the triples [rpm, torque, boost] for all the runs and organize them as a single data set. This will give you a look up table that gives the boost required for a given engine torque at a given rpm. Guess the best rear wheel torque, and use the gear ratios to come up with the boost profile for each gear that delivers that rear wheel torque at all rpm points and gears.

I am sure that someone who's been involved in drag racing has a better method that uses actual 1/4-mile pulls with different boost settings. If that someone is reading this, please enlighten me how to do this better with actual acceleration data. In fact, if someone can point me to software to use the actual acceleration data to come up with the optimal boost profile for each gear, that would awesome.

PorKen

Simpler to use a different resistor for each switch through one A/D port.

ptuomov

That's the idea. This system is widely used in at least turbo drag bikes, because the bike would be pretty much uncontrollable if the max boost would be used on the first gear. The same idea is also used in staged nitrous kits, they have a separate juicing profile for each gear. I think BA with his 1000hp nitrous 928 has a system like that -- must have, because otherwise the 1000hp would not be usable.

ptuomov

Eboost2 can hold six profiles that can be switched sequentially. Four profiles can be accessed with random access by selectively grounding two wires (binary logic with four alternative states).

The eboost2 user manual is here: http://www.turbosmartonline.com/inde...ownload&id=175
The controller looks like a toy, but it's not. This is IMO best value out there. For the switch logic that I am planning to use, see section 6.2 on page 25.

If four profiles are enough, I'll use the simple random access switching and some microswithces. If it turns out that I need all six, then I need to program a splatco controller to read the position, remember the state of the boost control input, and sequentially switch up or down the right number of steps. But as you say, this is a hassle and one more thing to go wrong, so I am trying the simple 4-state random access selection logic first.

Now, this is all geared towards turbo cars originally. But I think this can be used with supercharged cars as well. Here's my idea. Fit a regular blow-off or by-pass valve to your supercharger system, unless you already have one. Connect the valve reference signal to the 928 vacuum reservoir. Then, use the eboost2 solenoid to control this reference vacuum line. Now, you can let the boost controller let off some "steam" from the manifold on low gears when the wheelspin is a problem. Sure, it's not thermodynamically efficient arrangement, but if you are running a mechanical supercharger that's obviously not your top concern -- throttle response and acceleration are.

Gtaebr

Tuomo
You could use either microswitches or linear potentiometers connected to an rpm controller. when first or second is selected with WOT it will ramp in the rpm at a preset rate or limit.

This is a simplified launch/traction control but should meet your needs.

SharkSkin

The problem with that approach(maybe) is that the gear selected will not necessarily be in sequence with the voltage range. As the voltage increases like so:

0v - 1v - 2v - 3v - 4v - 5v

The corresponding gears may be:

5 - 1 - 3 - 4 - 2

Just an "off the top of the head" example, not thinking too hard about the actual switch arrangements discussed.

Interesting problem. I only skimmed the documentation, but it seems that you have chosen the most appropriate scheme if you want the setup to be workable for a road course -- the sequential options don't seem to be very friendly to any application other than 1/4 mile, unless you translate switch state to a sequence of signals or a voltage.

If the eboost solenoid can be modulated rather than on/off your plan sounds workable.

ew928

I wanted a flashy LED display like the sequential shift rally cars.
Once in coming to a stop in heavy traffic, wondered why the car was so reluctant to move again.
Car had been in 2nd gear. Doh! (hate brain fade)

Maybe a thick polycarbonate plate can be cut like the Ferrari shift gate and mounted at the base of the shift boot. Mount the switches at each finger of the shift gate.

Gtaebr

Ernest
Come over to the Rogerbox side..gear indicator with lights and thrust pedal.
oh did i mention sequential shifting too!

Hilton

Nice idea Ernest - would give a more positive locator/detection rate too.

From memory, Pierre in South Africa has a shift-gate like this on his Fat Bully 928 racecar. Pictures on his "Fat Bully thread.

edit: Linky to shift-gate post

Can't remember if he ever sent a copy to anyone, but it might be worth contacting him about it?

ew928

Pierre had said he would look into production cost of his shift plate but never got back.


Last time I looked at the shift boot base, the darned thing was curved. As in it has a mild hump in the middle with the front and back sides lower.
So one would have to curve the shift plate blank before taking out the fingers if one is to fit it back into the stock location.

HowlerMonkey

Old post but the solution has been in use for a while.

We've been running a true boost by gear system for 2 years and it seems to have passed the durability test as it is still working now even as I am fitting an improved sensor system that also allows us varying stages of nitrous by gear as well.

Car speed has nothing to do with traction......unless you have wings capable of generating mucho downforce so boost by vehicle speed is not as accurate of a way to limit torque as true boost by gear.

Sure........you will have less possibility of wheelspin at higher speed but this is a given since your car will be in a higher gear by virtue of the higher vehicle speed.

Which gear you are in determines the torque available to the wheels for any given automobile configuration.

I'm currently finishing up work on weatherproofing and final fitment of a shift position sensor on GM six speeds that allows true boost by gear for many different cars including corvettes, lambos, ferrari.....anything with a manual transmission and a boost control system that is evolved enough to incorporate the "boost by gear" feature.



This is true boost by gear we use with the ams1000 boost controller meaning you can go through the gears and the display on the ams1000 will show what gear you are in and apply boost control based on that gear regardless of car speed.

For the last two years it's worked well enough for us to hit over 270 (unofficial) in the standing mile (252.9 official) even when someone set the boost levels far too high for the lower gears.

We did this on only 1400hp but we've got a 2500hp program nearing completion that will definately require boost by gear and careful programming.

http://www.youtube.com/watch?v=kMuPknhk67I

Look for this car to take the record over 270 sometime this year but don't expect that at texas as the venue is proving too short in length.

We've moved the shop and will be having a grand opening in march.

http://a8.sphotos.ak.fbcdn.net/hphot..._5610284_n.jpg


I have been released from my parts replacer jihad at Lexus and can now concentrate on something more interesting than throwing parts at cars and raping warranty to earn a living.

In the coming weeks, I will be working with other boost controller manufacturers to achieve true "boost by gear" with thier systems.