Shifting: Manual, Electric, or Pneumatic

[Chris Boyden]

We plan on doing an electromechanical shifter this year. In order to implement an ignition cut, I'm either gonna:
1. Put a strain gauge on the shift lever of the tranny, amplify, then ignition cut.
2. Cut the ignition when the button is pushed.
3. Try setting up a microswitch that is triggered by the shiftlever, then cutting the ignition.

Currently, I've specified the solenoid....but
I'm still interested in other methods.
Right now, we can buy the Tiptronics sytem for
~1000 USD (5 lb), the Klicktronics for ~600 USD(? lbs).
The solenoids I've spec'd out weigh 2lbs per and will give 50 lbs at .25" travel on a 1" lever arm. If we go with up/down shifts, I'm looking at 4 pounds plus ~.5 pounds for hardware.

" air system Tank (empty), regulator and gauge weigh 1.8 pounds" -Paradigm motorsports...

I imagine a bimba double acting cylinder with 2 solenoid valves weigh somewhere in the neighborhood of 2-3 lbs.

We're also considering a manual shifter w/ a push button system of some typebut that just adds weight. If we go with just a manual, then we'll definitely cut the ignition.

I've read through the forum and found quite a lot on this topic. Still, I'd like to see some more discussion.

[Chris Boyden]

We plan on doing an electromechanical shifter this year. In order to implement an ignition cut, I'm either gonna:
1. Put a strain gauge on the shift lever of the tranny, amplify, then ignition cut.
2. Cut the ignition when the button is pushed.
3. Try setting up a microswitch that is triggered by the shiftlever, then cutting the ignition.

Currently, I've specified the solenoid....but
I'm still interested in other methods.
Right now, we can buy the Tiptronics sytem for
~1000 USD (5 lb), the Klicktronics for ~600 USD(? lbs).
The solenoids I've spec'd out weigh 2lbs per and will give 50 lbs at .25" travel on a 1" lever arm. If we go with up/down shifts, I'm looking at 4 pounds plus ~.5 pounds for hardware.

" air system Tank (empty), regulator and gauge weigh 1.8 pounds" -Paradigm motorsports...

I imagine a bimba double acting cylinder with 2 solenoid valves weigh somewhere in the neighborhood of 2-3 lbs.

We're also considering a manual shifter w/ a push button system of some typebut that just adds weight. If we go with just a manual, then we'll definitely cut the ignition.

I've read through the forum and found quite a lot on this topic. Still, I'd like to see some more discussion.

[All Wheeler]

Try contacting our shifting guys. They might not be implementing your style of system, but chances are that they've looked into it.
timothy.donohue@studentmail.newcastle.edu.au
or
tivon.ostawari@studentmail.newcastle.edu.au http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Mechanicaldan]

Test, test, test. After going to competition early this year, that's what my comments are. How many teams didn't finish the endurance race because of air or electric shifters? I know of a few.

Also, make sure you take into account the additional electric draw from the solenoids if you go that route. Drawing more current than the stock alternator can supply is a bad deal.

Cyclone Racing
www.cyclone-racing.com/fhome.htm (http://www.cyclone-racing.com/fhome.htm)
Iowa State University
Project Director

[Charlie Monash OZ]

If you are using a Honda CBR600 be very careful choosing your solenoids, we have implemented both hydrolic and electromechanical shifting mechanisms on previous cars.

The problem we faced with electromechanical is the duration of the solenoid pressure, if the gear does not immediately engage then you are in trouble. Pressure needs to be held on the shifter for up to half a second at least, otherwise you wont engage the next gear. Unfortunately solenoids that are cabable of that usually draw too heavily on a fsae sized battery.

Good luck to you, if you want to know mre about what we tried I'm sure the Monash boys will dig up the information for you

There he goes -- one of God's own
prototypes -- a high powered mutant
of some kind never even considered
for mass production. Too weird to
live and too rare to die.

[Chris Boyden]

Thanks for posting some good info....and some very valid points.

1. The current draw cannot be higher than what the stock alternator can flow. I need to consider the entire pull on the electric system at a high duty cycle on the injectors and ignition and with the electric cooling fan on.
2. If we implement some pulse width control circuits, then we should be able to adjust the duration of the solenoid pressure.
Hopefully, we can use a peak and hold driver to reduce the solenoid current after the initial pull. That might help solve the current draw and size problem. Once the solenoid is seated or close to seated the pull increases which also leads me to think that a peak/hold driver would help. But like Dan said....test test test...all the thinking in the world won't help.

I have a solenoid on order. Should be here this week. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

The thought of a redundant mechanical shifter coupled to a solenoid shifter makes everybody feel better, but it still adds approximately a pound or two of weight by the time you install the brackets and push/pull cable.

Then there's the good ole neutral problem. A redundant mechanical shifter would make it easy to find neutral the old fashioned way, by hand. A separate N push button that would send a smaller pulse to the solenoid might do it as well.

There was another team that had electrical noise problems....and that the car would spontaneously shift itself. Sounds like they needed to filter the DC voltage better with a capacitor / regulator.

Any ideas about timing the ignition cut?
Most ECU's have an input that allows you to retard or cut the ignition/fuel. So that part is easy. A more generic shifter/ignition cut should probably control the signal that goes to the coil itself. But that's just more work and less integrated. A 555 timer w/ trimpot setup as a one-shot might allow you to tune the duration....I'm just not sure which order things should happen. Should the solenoid began engaging which triggers the ignition cut, or should the ignition cut trigger the solenoid?

I'd be glad to hear what you guys tried Charlie! Only a fool would learn from his own mistakes! .....I've been a fool way too often.... http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

[Banacek]

Make sure you know how much force you really need to pull to engage your shifter.

Having tested on an old F2 (I think that's what it was), it proved to be quite a bit and we discounted solenoids due to weight. (don't forget safety factors in your forces)

As for engine cutoff, check your electrical supplier for limit switches. They're cheap and they are designed to do exactly what you want.

For pneumatic, you only need 1 valve if you use a double acting cylinder: a 3 way, exhaust centre.

Banacek
"there's an old Polish saying..."

[Chris Boyden]

So far I've got a safety factor of 1.6 on the forces. I've used several different scales to measure the forces, and come up with about 30 to 35 pounds to engage the shifter on the R6.
The 2002 F4i might be a little different. But your point is well recieved. I'll need to do some more research on the air system. I think that the forces in an air system can be as high as 100lbs. rather easy. a 1 sq in piston will generate 100 lbs with 100psi. If the solenoid I purchased isn't enough, then I can try rewinding the coil with larger gauge wire to get more ampere-turns.

Do you remember what the forces necessary to push/pull the shift lever were? and at what lever arm length?
Chris

[Charlie]

I think Auburn's 1997 car was our first to implement electropneumatic shifting and it was among the first ever?

The system was largely unchanged until 2001. Our 2000 car (my first) lost all CO2 and was stuck in 3rd gear through the last 2/3 of the endurance event (which we did finish).

I remember designing the 2001 car we had a big push in the team for a mechanical system because many could not master the push-button downshifting and/or thought it was too unreliable. I fought for it and took up redesigning it, simplifying the system, using less parts and more reliable mounting and attachment. I did not change the basics like actuator piston size or system pressure.

It has worked well and through a couple years of tweaking the system is definitely at its best in the 2003 car. It is light, reliable (we have run practices weeks apart without refilling this year!). We could run a smaller tank easily. I believe it is faster and more predictable than a manual with enough driver training.

Well this all didn't say a damn thing about the systems specifications did it? http://fsae.com/groupee_common/emoticons/icon_wink.gif I will say that I think pnuematic is the best choice and our system is not outlandish, and does not utilize any custom or expensive parts.

-Charlie Ping
Auburn University FSAE (http://eng.auburn.edu/organizations/SAE/AUFSAE)
5th Overall Detroit 2003
? Overall Aussie 2003. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[karter]

In what order do ignition cut/shift need to occure? Seems like cut/shift/dwell/uncut??

[Chris Boyden]

Yea, i agree.

[Charlie Monash OZ]

Ok, after conferring with the boys apparently the battery is the only real issue in the end. Using peak and hold solenoids does work and with some super electronic trickery does so consistently. However we could never manage to maintain enough charge in the battery to keep it working for any lengthy period, so the system has been shelved. Although I'm sure a larger battery would do the job, we are not interested in the resulting extra weight.

So for the time being we have returned to manual shifing with a new edge!! I'm sure with enough R&D you could make it work, we also have issues with weight/advantage ratio as the solenoids are not exactly light either.

Once again good luck and let us know if it works

There he goes -- one of God's own
prototypes -- a high powered mutant
of some kind never even considered
for mass production. Too weird to
live and too rare to die.

[Banacek]

It just occured to me that I saw an all-electronic system in Detroit this year. I can't remember the school, I'll see if I can find it in my pictures. They had a klictronic for the gears and a solenoid the size of my arm for the clutch. They said they managed to do it without serious current problems.

You might want to track these guys down if that sort of thing interests you.

Banacek

[alfordda]

We were the ones with the noise problems, in 02. I was not super invovled with shifting, but what I gathered goes something like this. In 02 we used a 555 with some resistors and misc. other electrical stuff. When ever we reved the engine it would shift. We tried a couple of differnet filter combonations with no luck. So for competition, we removed the 555 circut and just used the stuff from Klictronic. Drivers had issues downshifting, probably because they were not use to the whole matching rpm thing. In 03 we got a new team member that had some knowledge of microcontrollers. So we dropped the 02 design and started over. In 03 we had no problems with noise. We still have problems with downshifting though. It doesn't seem to want to shift until the rpms are low. We didn't get the 03 car built in time to do much shifter testing, so it is probably just a delay is not long enoungh. I looked at the program, for the first time, and it is extremely simple. The process is cut->delay->shift->delay->uncut. Probably don't need the first delay. I think we are using a stock battery, with no current draw problems.

[Ali]

There's an SAE paper on an electric shifting system for FSAE cars by someone from Colorado State University. Uses a motor n cam to rotate the shift lever.

I'm planning to do a button shift system. I know it's ambitious for a 1st year team like ours, but my final year project grade depends on it!

Anyway, I think a pneumatic system is the way to go for the clutch. It's light n can provide the large force needed. For our 2002 (I think) GSXR600, the torque at the shift lever is about 5Nm, and it takes about 28kg to pull the clutch cable. All this was measured with a cheap spring gauge, so I can't vouch for the accuracy

[Chris Boyden]

well, the solenoid arrived and we've started testing the unit.
There can't be much sideload or the solenoid will start to bind. But the forces were the same as those advertised on the data sheet (within 5% using a cheap spring gauge). The initial results look promising. But 2lbs per solenoid kills me. Next, I need to mount it to the '03 car, implement the ignition cut, and start testing!

[Banacek]

In complaining about the weight of the solenoid, keep in mind that pneumatics aren't exactly light either. Those valves are a lot heavier than they look.

Consider whether you want to use your system to eliminate taking the hands off the steering wheel/removing the manual clutch or to gain performance. Getting heavy, fast, components can let you shift much faster than manually and the performance gain may be worth the weight.

I vaguely remember Waterloo having a very fast, well designed system this year that helped them do very well in acceleration. (on a 4s run, saving 1/4sec or more per shift is a big deal, note: numbers are more than approximate)

Banacek
Note: some or all of the information contained in this message may have been obtained from my rectal cavity. Use all information with discretion.

[Chris Boyden]

True...The weight of the tank alone is ~2lb.

I thought I smelled something stinky http://fsae.com/groupee_common/emoticons/icon_redface.gif http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Denny Trimble]

I'll have to take another look at our motec data, but I believe 1/8 second is the approximate shift time for our manual (no-clutch) upshifts. However, there are some consistency problems. I'm currently batting about .900 on upshifts the first try, we're hoping an ignition cut will make our shifts quicker and more consistent.

University of Washington Formula SAE ('98, '99, '03)

[Igor]

Heavy solenoids? I have a 5/3E valve here that weighs only 80grams. You must be on a really tight weight budget.

Igor
DUT Racing

[Garbo]

We've run a very simple (no logic or chips, just big wires that even an idiot like me can understand) pneumatic system. All the gear was from Festo (two solenoids, one double acting piston and two feet of hose) Without the bottle, I cannot see this system being over a pound (I might even weigh it). We also had an uber-sweet carbon fibre paintball tank from airAmerica paintball gear (~400USD new, pilfered from a team member for a case of beer) but ran into trouble getting it certified when some over-enthusiastic dive shop guy pressure tested it for us and somehow modified it to send all 3000psi of tank pressure into our poor little solenoid. Poof! no more shifter. I worked very well for two years until then, though.

Garbo

[Banacek]

What's the Cv on that 80g 5/3E valve out of curiosity? We're going on the assumption that bigger is better (that and our enine has an unusually heavy spring on the shift) and will probably use much heavier valves.

Also, when I mentioned heavy solenoids earlier, I meant solenoid actuators doing the shifting not solenoid valves. These are big things and the only team I ever saw using one said it was around 5lb.

Banacek

[Igor]

I'm not sure what you mean by Cv, but they flow about 3-5 liters per second depending on connector type.

Aha, mechanical solenoids are heavy indeed. We tried it and were rather displeased by its performance. One would need a high voltage power supply to get some decent pull out of them.

Igor

[boss451t]

I have not put an electronic system on our car yet, but from what I have heard the biggest problem is in the downshift. Upshifting with ignition cutting works out well, but on the downshift there are problems. I talked with a manufacturer for Kliktronic and he mentioned that clutching is still needed for downshifting. That is pretty much against the point of having electronic. Let me know what you think or what you have run into with electronic downshifting.

Mark
Virginia Tech

[woollymoof]

The whole reason I see for having electronic/pneumatic shifting is for fast up shifting. Does lighting fast downshifting really matter? Will it make the car faster (assuming that the downshifts are fast enough to get you in the right gear for corner exit)?

Cheers,

Kirk Veitch
Swinburne University of Technology

[racergirl55]

Right, Kirk ... as far as FSAE events are concerned, a fast upshift is necessary for the acceleration event. However you still need a reliable downshift that will actually shift when you want it to!! Why not have manual downshifts? You actually don't need a clutch if you have experience - just shift to neutral, blip the throttle to match revs, and downshift (I know ... easy to say, hard to do!). If you have electronic push button shifting there's no added weight from a mechanical shifter. The only question I have is whether you still need a clutch for "pit" work (changing drivers, etc).

Katherine
McGill Racing Team

[Charlie]

How do you shift into neutral?

-Charlie Ping
Auburn University FSAE (http://eng.auburn.edu/organizations/SAE/AUFSAE)
5th Overall Detroit 2003
? Overall Aussie 2003. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[racergirl55]

Well nevermind my thought, I was typing faster than my brain could think. I was thinking of using a hand clutch on the wheel. Pretty much you would need to have a well-programmed control system for your shifting so it could be simple downshifting via a button. Otherwise as Kirk said there would be no point in electronic shifts. There is always the option of having mechanical shifting as a "back-up," but then where is the point in that as well?

You teams who had electronic downshifting, what exactly were your problems with it? And where was your clutch?

Cheers!

[MikeWaggoner at UW]

I haven't done it pneumatically, but from driving I find downshifting in a corner while matching the revs hard to do. I could see how a system that just shifts would cause spinouts if the downshift wasn't smooth.

[racergirl55]

I know what you mean; it's never easy and especially not in competition (I've never driven at Detroit personally, just otherwise). That's where testing comes into play I suppose. But even with manual clutch downshifting you need to blip the throttle to the approximate rev range anyways to take care of your 'box and engine (can we say bye bye dog rings?!). An electronic system that cuts the engine long enough so the driver can blip the throttle would be ideal, then it downshifts at the end of the delay period. Again, drivers would need to practice being in sync with the program. All it takes is testing and iteration ... something most teams are short of!

Katherine Doyle
McGill Racing Team

[Denny Trimble]

Katherine,
It sounds like you have experience with "real" racecar gearboxes. Motorcycle trannies will take downshifting without rev matching, as long as you use the clutch. Of course the car will spin if you dump the clutch and don't match the revs. But it won't damage the tranny.

Our downshift sequence is:
clutch open
downshift
blip throttle to match revs
clutch closed

That's the way our clutch/shifter (manual, single lever controlling both) has worked for five years now. There's a delay mechanism for the clutch on downshifts, and no clutch for upshifts. Just lift for 1/8 of a second http://fsae.com/groupee_common/emoticons/icon_smile.gif

University of Washington Formula SAE ('98, '99, '03)

[GIJoeCam]

If the question is whether to use a manual lever type or an electric/pneumatic/electropneumatic system actuated by buttons, paddles, or some other method, there's no dabate. A wise man named Carroll Smith once told the 2002 LTU FSAE team that three pedals is one too many. He quoted the time it takes to shift your feet on the pedals at a particular speed and related it to the distance traveled on the racecourse, and, of course, he was right in every way.

In '02, we used a paddle-actuated electropneumatic system. The current draw for the sloenoid valves was miniscule (.3A if my memory serves) and a full tank and fully sealed system lasted us two weekends of racing easily. The keys are keeping the system sealed and sizing the components appropriately.

Another note which we capitalized on... remember that old greek dude Archimeded who said, Give me a lever and a place to stand and I can move the world? If the lever needs 50 lbs at 1 inch, it only needs 25 at 2 inches. The travel should still be less than an inch either way, and many companies (Pneumatics.com comes to mind) make double-acting pistons with roughly a half square inch of surface area.

The bottom line is, DO IT!!! There's no doubt that not only can you shift faster, more importantly you keep BOTH hands on the wheel at ALL times... critical for control!

As for driving technique, it takes a lot of practice, even shifting manually. I, for one, subscribe to the shift early school of downshifting, which is fine on a manual trans in an automobile, but in a sequential transmission, downshifting under hard braking even with the clutch disengaged STILL upsets the rear end very easily. I had to get in the habit of clutching, braking, and not downshifting until just before I was ready to accelerate! It's a learning curve.

Slipper clutches are a great help in downshifting...

Just some things I picked up along the way.

-Joe

2002 LTU FSAE Co-Leader
2002 LTU FSAE Brake System Engineer
Inaugural 2002 Road and Track FSAE Triathalon Champions!

[GIJoeCam]

Something I forgot to mention... during our testing, one afternoon, we ran out of CO2. So, in the effort to continue testing, we locked it in second gear and once we were rolling, 2-footed the gas and brakes alone. Around the course we had set up that day, (roughly 1/4 the distance of the endurance) we ran an average of 2 seconds per lap FASTER than when we shifted all the time. This underscores the importance of gearing... it's more critical than people believe.

-Joe

2002 LTU FSAE Co-Leader
2002 LTU FSAE Brake System Engineer
Inaugural 2002 Road and Track FSAE Triathalon Champions!

[Charlie]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Denny Trimble:
That's the way our clutch/shifter (manual, single lever controlling both) has worked for five years now. There's a delay mechanism for the clutch on downshifts, and no clutch for upshifts. Just lift for 1/8 of a second http://fsae.com/groupee_common/emoticons/icon_smile.gif
<HR></BLOCKQUOTE>

Denny-

Just curious, since I know you have some good logging equipment, what is a typical upshift time for your car, measured from longitudinal G peak-to-peak?

-Charlie Ping
Auburn University FSAE (http://eng.auburn.edu/organizations/SAE/AUFSAE)
5th Overall Detroit 2003
? Overall Aussie 2003. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Denny Trimble]

Charlie,
Unfortunately our accelerometer isn't mounted very well right now and it's a little noisy, as you'll see below. Here's the sequence of events from a 2nd to 3rd shift two weeks ago before we broke the car:

time event
0.00 peak g, throttle position
0.09 engine RPM peak
0.23 engine RPM minimum
0.28 peak g, throttle position

So, that's longer than I had commented, but you can see the engine RPM drop takes around .14 seconds. I'll let you know if ignition kill helps reduce the total shift time, we expect it will.

The plot below shows filtered g's at .100 second averaging.

http://students.washington.edu/dennyt/fsae/upshift.jpg

University of Washington Formula SAE ('98, '99, '03, '04)

[Charlie]

Interesting, thanks. We need to sample a bit finer before we can quantify ours, I was just curious.

-Charlie Ping
Auburn University FSAE (http://eng.auburn.edu/organizations/SAE/AUFSAE)
5th Overall Detroit 2003
? Overall Aussie 2003. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[awhittle]

http://www.holeshot.com/old/shifters/shifter3.html

http://www.sharpshifter.co.nz/

[MN]

Hi.
I have a idea about N system of Air shifter.
It is N button push,electric valve open just shift change to N.So I have three buttons,shift up, shift down, and N button.
But,I have not other idea.I want hear everyones ideas!!

[EgyptianMagician]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by awhittle:
http://www.holeshot.com/old/shifters/shifter3.html

http://www.sharpshifter.co.nz/ <HR></BLOCKQUOTE>

For ideas, yes, but for implementation ... I think a team SHOULD be able to build their own custom shifter without having to run off and buy something off the shelf.

I assume they would be expensive, but I didn't check.

[Agent4573]

I think I'm jumping in this a thread a little bit late, but we currently use a custom built electric shifter on our car. It weighs in at just under two pounds everything. Its controlled through a custom PCB board our electronics guru made up last year and activated by paddle shifters on the wheel. While the current draw on the motor is large, its only activated for such a short time we've never had a problem with the battery. Well ok, we have had a few problems with charging but thats only because we didn't have the alternative hooked up, and we still lastest over 6 minutes on battery alone, with normal shifting going on.

For upshifts we use the ignition cut in the ECU but have had a few little problems with the sampling rate of the ECU not being fast enough for extremely high RPM shifts. For downshits we use two paddles. Imagine on the left had side of your steering wheel two paddles, one for your clutch and one for you downshift activation. You can pull the clutch in by itself, but when you pull the downshift paddle, it pulls the clutch paddle in with it. You have to make sure though if you run a system like this, that your shifter motor only runs for a certain amount of time, and not just while the paddle is depressed.

[MN]

I appreciate your answer.
Now I have next question.
It is Control circuit of Ignition coil and ECU, Injector and ECU.
Do you need Control circuit of Injector and ECU??

[Nicholas]

<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR>Originally posted by Agent4573:
For downshits we use two paddles. Imagine on the left had side of your steering wheel two paddles, one for your clutch and one for you downshift activation. You can pull the clutch in by itself, but when you pull the downshift paddle, it pulls the clutch paddle in with it. <HR></BLOCKQUOTE>

Wow. Definitely a sweet setup.

[mysticv6]

How is everybody finding neutral on their electronic setups?

[gug]

use the clutch when stopping, kill the engine, then tell one of the pit monkeys (or pigeons) to reach down and find neutral by hand.

[Chris Boyden]

The push pull solenoid shifter that we partially developed was going to try to do one of two things to get into neutral.
1. use the neutral switch for feedback to the controller. i.e. stop the actuator when the switch is engaged and pray there was enough lead in the switch that you don't overshoot into 2nd.
2. use a linear pot as a feedback sensor with a voice coil and try to do position control with neutral as the setpoint when you push the green neutral button, using a pid controller.

So far these are just delusions of grandeur...but I think they would work....just takes hell of alot of work. Some of the aussie boys from UWA did some great work on this subject in a similar fashion. But for '04 US comp, it wasn't on the car because of driver transistor problems.

We got the shifter to upshift and downshift very quickly, but noise problems with the electronics, blown coils, and plain old lack of development kept it off the car. it was fun, but there's a lot to be said for KISS(keep it simple stupid) especially in FSAE.

If you ran a manual shifter along side the electric or pneumatic, that could be used to shift into neutral. The extra 2 pounds or less wouldn't kill you.

[Agent4573]

Our controller box is setup to only shift for about half the time when we hit a seperate neutral button. It also takes in a reading from the nuetral light output on the engine and keeps trying to hit neutral until it gets the light.

[gug]

hey Agent, what happens if the shift goes past neutral into second? wouldnt your box still think its in first and keep on giving the short upshift to get into neutral, and then eventually wind up in 6th?