dr bob

It's a crapshoot at best. How much force is transmitted with the "clutch partially engaged"? What exactly is "residual engagement"? ANY force transmitted by the clutch is force that the synchro need to manage. How big is the speed mismatch?

Think about it... Except for initial launching, an unloaded (not leading nor trailing throttle) disengage from any gear is painless. At the same time, a perfectly rev-matched engagement of any gear is painless without touching the clutch pedal. Yet you still use the clutch for shifts, to lessen the wear on synchro clutches and the gears themselves. Why would you bother, if you are that good at matching things up? There would be no grinding or crashing, the lever would almost move itself into the next gear. Meanwhile, when the speed match isn't perfect from initial gear contact all the way to full engagement, you end up with rounded and chipped gear teeth, worn and broken teeth on those little soft bronze syncro collars. You have whatever force you are transmitting through the clutch, plus the inertial load on the input train, times the square of the difference in RPM's, as the friction load on the syncros. Less load and less mismatch equals longer life and less reason for syncro wear and breakage. Easiest way to reduce the load on the synchro at engagement time? The left pedal. If you don't feel you need it, throw it and that heavy clutch out completely and save some weight. The engine might even rev quicker to make those shifts even faster!

littleball_s4

Easiest, yes. Best, no.

Best is to blip the throttle while totally or partially clutched, then use the left pedal to declutch totally just before engaging the gear. That's what most racers end up doing by pure instinct (the feedback through the lever is better the more you approach this).

At full speed, it's not a sequence of moves for your brain, it's a brisk stab with the two feet and the lever. Whatever adjustment it takes for it to work is justified, because it is not always possible to adjust timing in the brain.

If your blip arrives when the clutch is totally open, it will not help reduce the rpm difference for the syncro, and the shift will be slow and hard job for the syncro. Compared to that, given equal shift time, if the clutch drags a bit and you blip, it will have a relative (say 90%) syncro torque reduction due to reduced rpm difference and a constant (say 10Nm) syncro torque increase (the drag torque). Which one is greater depends on how good you are rpm matching and how big the drag is.

Downshifting with the engine connected is next to impossible in a syncro box at full speed, as you know. But helping the syncros is possible, even without double clutch.

littleball_s4

That being said, I would not recommend adjusting the clutch for slight drag to improve downshifts.

Which is not equal to say it's a crapshoot if someone encounter this effect accidentally.

Ultimately, the feedback through the lever is quite accurate. If you find that gears engage with less effort and/or you can donwshift faster, it has to be better for the syncro.

mark kibort

I agree. I think with a borg warner box, all things are so well within its capabilities, that there is no worries if you are shifting well. However, i guess it doesnt matter, about matching revs because the clutch disengages the engine rev from the synchro match to the new downshift driveline speed. however, i do think there is drag on the driveline with a partially engaged clutch. i do think the clutch wheeling around at 6000rpm, will have some residual force on the driveline via the INT plate making a very slight contact. this has to be it, why if i push the pedal to the floor and do a shift from 3rd to 2nd without a blip, it will grind . (or 4th to 3rd at moderate speeds, like 80mph) its probably a timing thing that most do when they shift and do this sequence of movements well.

dr bob

I love circular arguments/discussions where there is so much conflict within one side of the discussion. The "blip the throttle" part doesn't care about "totally" or "partially" uncoupled. If there's enough friction or internal drag within the clutch, the gearbox is going to see full engine speed. That's really the goal here, right? If not, why bother with the blip? And in your "while totally or partially clutched" we are referring to some driver action involving the left pedal? I think that's the best method that I described.


-----


Lots of opinions on the 'best way', and for sure those are different between street driving and race driving. Unless you happen to have a constant-mesh sequential gearbox, using the clutch to isolate engine thrust from the gearbox is mandatory as you engage the next gear, unless you are willing to sacrifice gear teeth and synchronizer teeth and hubs. The gearbox in the 928 was not designed with quick snick-snick shifting in mind, and the [lack of] availability of replacement bits will ultimately guide Mark in how he abuses the gearbox during races.

Modern technology in race cars has either a sequential box, and/or computer-controlled engine and gearbox management to make perfect shifts every time. I had a chance to review a Subaru factory rally car's early efforts, with some pneumatic actuators actually moving the forks to select gears, based on the driver's manipulation of a fore-and-aft "shift" handle. It could have been paddles on the wheel but that wasn't what they had available at the time. The little "controller" actuated a throttle bypass valve for rev matching on downshifts, a fuel and ignition cutoff for same on upshifts. There was a certain tolerance for driver participation on downshifts, but the preference was that the driver preselect the next gear and let the controller manage engine load, speed and the actual movement of the shift forks. This allowed foot-to-the-floor upshifts with no clutch actuation, and no- or part-throttle downshifts with no clutch actuation. Needless to say there was a learning curve for the drivers.

littleball_s4

I re-read post #24 and you and me are saying the same thing, only you say it more clear.

Just a little fun fact, Citroen WRC of 2003 used dog box with manual lever; but they had a "shifter" (essentially a extremely retarded timing triggered by a switch in the lever that was activated the millisecond you started to pull it) and a "kicker", (essentially a blip in the throttle by wire triggered by a switch in the lever that was activated the millisecond you started to push it).

I guess it took them some time to fine-tune both (the shifter was gear dependent, the timing was retarded inmediately but it came back gradually in some 100-200ms to normal value, while the kicker was independent of the gear). But once finished, gears went up and down, (without clutch, of course) almost effortless, and the wear of the dog rings was very good.

In fact, I heard the next evolution of the system was to add hydraulic actuation to the lever keeping the rest the same, to have paddleshift.

mark kibort

what i have found is that, my foot is never really fully depressed. you can see and i can feel this clearly. i suspect its one of the reasons i get really clean downshifts. If i EVER forget to blip or mis the blip that happens once every 1000 shifts if i was too guess) i will have a HUGE grind. what suspect i am doing is partially putting in the clutch, pulling it out of gear and blipping the engine as it passes through neutral. this spins up the driveline, but has so little force that it doesnt effect the force need to be added either higher or lower speed, to synchro nize the driveline with the internal trans shaft. and if i get it right, the syncrhos are not needed at all. if not, there is a little more of a clunk as the synchros fit together. they are just at a closer speed and they do less work this way.
its basicaly the same effect as double clutching. the real point is, you dont need but a hair of contact of the pressure plate touching the discs to spin up the driveline. we saw this clearaly in our int plate discussions.
that car would shift perfectly, yet there as a clear drag and coupling of the driveline to the engine, always, albiet very small. these days, with CF drivelines, or driveline in the gear box like a 911, this technique can be very useful and protect the synchros.

Fogey1

This.

Real double-clutching is a basic driving skill, imo. Definitely learnable for anyone with any feel - I'm proud to say my Mama could and did drive a crashbox.

I like your 911 story. I drove an M151 from Hue to Danang without a working clutch linkage.

ww

mark kibort

I did the same thing with my 84 with a broken clutch pedal. (snapped at work).. just had to anticipate all lights, drive 1st on approach... but it was easy for me to just match the RPM and fit it into the gear no grind if you know the RPM drop for gear and time it takes to make a smooth shift.

double clutching is really not needed due to the low mass of modern drivelines. driving the old trucks, absolutely

dan212

Double Clutching is a generation specific, coming of age technique. Those of a certain (ahem) age learned to do this by necessity. Synchros were something that might only last for a year or two.

My 72 Alpha Spider (bought with 3K miles, one year old) had synchro's that were a distant memory buried somewhere in the snowbound alps. All downshifts required a double clutch.

Jump 20 years (or more) later, it took me quite a while to get used to a modern gearboxes with durable synchros that actually worked. A small miracle in my eyes. Still does. If I skip the double going into 2nd on the track, I am filled with shame and remorse. In a race, I will cheat - knowing what it costs me.

There is a reason that Synchros are made of a softer material than the driving gears themselves.

mark kibort

with a modern or relatively modern gear box, even one with worn synchronous, its possible to drive them and engage the gears pretty smoothly, by using the technique im advocating. again, its the effect of double clutching, but doesn't take the time, skill, or effort and is much more precise.

GregBBRD

Up shifting or downshifting without using the clutch is absolute hell on syncros....unless the exact perfect rpm is found and maintained to accommodate the gear ratio change.

And since no one is good enough to match the exact rpms to the ratio change in every gear, syncro wear goes up at a parabolic rate.

Without the uncoupling of the clutch, there is no release of load as the shifting sleeve either travels over the syncro (early 928 transmissions) or pushes the tapered syncros onto the tapered hubs (Borg Warner transmissions.)

If the speeds are not exactly (and I mean exactly....not off one rpm) perfect, the syncro is trapped between two unmoveable forces as it tries to either accelerate or deaccelerate the "unyielding" gear.

Shifting without a clutch in a transmission with syncro is sometimes a necessity when things are broken and the car needs to get off the road, but any travel after that will be dramatically cheaper (over the long run) on a tow truck!

I sometimes have trouble getting my point across...so let me be clear...shifting a transmission with functional syncros, without a clutch, is stuff strictly done by fools. All of you readers, out there, thinking this would be a smart thing to do...forget it!

mark kibort

there is a reason my clutch packs still have the readable part numbers written on the surface over 10 years of racing Greg. I do a pretty good job of matching RPM. ii dont think i was that good in the 2001 to 2003 time period but I got pretty good after that.

also, its pretty easy to "feel" the right RPM when things are broken and you need to drive home with no clutch. it just slips into gear. ive done it without a grind and sounds like a few others have as welll. not recommended, but possible

again, you have to read this very carefully..... what i was talking about was shifting with a clutch partially engaged. , what i mean is im talking about pulling the shifter out of gear with the shifter partially engaged. maybe 1" from the floor. at the point, alll we have is a slight drag on the driveline...... so slight, its not noticealbe during a shift, but the dirveline still follows engine RPM pretty closely. remember the mis adjusted INT plates. just barely touching the flywheel, spins up the driveline to synchronous speed.

So then, as we pass through neutral, the skilled shifter, then blips the throttle, and the driveline is spun up to speed, the skillled shifter then pushes a little further on the clutch pedal to reduce even the slighest drag on the driveline, so that the synchros can now command the speed to match the driven gear set.

my synchros have been used for many many years. that holbert gear box was perfect for 10 years... it broke 5th gear set teeth, which probalby was just a fatique factor or as Bailey said, the expansion due to uncooled heat, finally took its toll on the gear edges.

Ive driven bad syncho gear boxes this way, with little or no grind on downshifts. but they sure work a lot better when the synchros are new or healthy. I do remember if no blip was used, there was NO way that the car could get a downshift from 4th to 3rd. however, with a healthy blip, there was a grindless match. this must mean blipping has drag on the driveline to spin it up, as i cant think of any other reason why that would change things during a shift.