Perfectly smooth upshifts-trick with disengaging the clutch

[zenfiz6]

Its not rev hang per say, the revs drop at a steady rate but its just slower. I think zenfiz is also experiencing the same thing in his A4 per his previous posts.

Anything other than the mechanical drop of revs is caused by rev hang. Do your revs fall at the same speed in every situation? Or can they vary?

Ok, go figure: my own technique was fooling me.

After a lot of playing on the trip to Williamsburg and back, pulling the car out of gear at speed and just letting the RPMs fall, I could tell no difference between that process and revving the engine in neutral.

It turns out that I was seeing my own habit of letting the clutch out at the bottom of the RPM trajectory while actively performing a shift. Clearly this will slow up the RPM drop rate. So, for my A4, I detect no rev hang.

My bad.

So is my penance for misinterpreting my tach having to drive a P.o.T Cruiser for a week?

[zenfiz6]

So, it's funny. Because I was sure I was seeing something.

After the trip up to Williamsburg, I had written this long post about how there was something like a 1 second difference between a real shift from 3k->2k (as well as from 4k->2k) as opposed to just rolling in neutral and revving the engine... but it wasn't really all that consistent. Then, as I was driving back, I just pulled the car out of gear and held the clutch in at 60MPH in 4th (3k).... no rev hang down to 2k. So I pulled over and quickly edited the post to what you see above on my smart-phone with a bad charger connection and 15% battery.

And yet, there were times during a shift on the trip back that it just seemed to take a bit longer to drop than it should have.

I have to take the car into the shop tomorrow to finish the oil consumption test (it seems like it will pass with flying colors: not a drop down in 800 miles), so maybe I'll try shifting into gear, but leave the clutch in and see if the ECU says "Hey, he's about to shift, hold up!"

I dunno... last ditch effort to find what I thought I was seeing.

[daleadbull]

Anything other than the mechanical drop of revs is caused by rev hang. Do your revs fall at the same speed in every situation? Or can they vary? BTW, flywheel weight has a big impact on the way the revs fall. Not all cars are the same, obviously. But don't make the mistake of thinking that rev hang automatically means that the revs vary in the rate that they fall as you watch the tach needle dropping. Sure, that can happen, but it can also remain steady, while falling slower than it would if the car had an old-school throttle cable.

Yep I'm aware that its the computer messing around, I just meant its not the "traditional" rev hang where the revs hang for a little bit then fall naturally.

So, it's funny. Because I was sure I was seeing something.

After the trip up to Williamsburg, I had written this long post about how there was something like a 1 second difference between a real shift from 3k->2k (as well as from 4k->2k) as opposed to just rolling in neutral and revving the engine... but it wasn't really all that consistent. Then, as I was driving back, I just pulled the car out of gear and held the clutch in at 60MPH in 4th (3k).... no rev hang down to 2k. So I pulled over and quickly edited the post to what you see above on my smart-phone with a bad charger connection and 15% battery.

And yet, there were times during a shift on the trip back that it just seemed to take a bit longer to drop than it should have.

I have to take the car into the shop tomorrow to finish the oil consumption test (it seems like it will pass with flying colors: not a drop down in 800 miles), so maybe I'll try shifting into gear, but leave the clutch in and see if the ECU says "Hey, he's about to shift, hold up!"

I dunno... last ditch effort to find what I thought I was seeing.

All I can say is don't worry its not in your head, it is very inconsistent. I'll have no rev hang sometimes even when shifting below 3k but other times it will feel like the revs are slower to drop. I've observed that the temperature has a lot to do with it, AC, ambient temp as well as the engine temp. But I've noticed that whenever I take it above 4k, all rev hang disappears. This is just something I've noticed with MY car. Its not that bad tho cus its makes sense that when you're shifting below 3k you are most likely trying to shift casually and the slower rate at which the revs fall isn't a huge deal. It does catch me by surprise sometimes so I just have to look at the tach to re-calibrate myself.

Shadow insists that Audi's have no rev hang whatsoever so I don't know. But the fact that you're even noticing some inconsistency makes me believe that you do have some computer interference.

[Soundbear]

What happens if you don't give it gas, under exactly those circumstances? Revs ought to drop the rest of the way, just the same.

When I don't give it gas, I get the head toward the windshield jerk. If I give it gas during this time, I'm either neutral or pressed slightly back into the seat, which to me is more comfortable and leads to better acceleration.

[daleadbull]

What happens if you don't give it gas, under exactly those circumstances? Revs ought to drop the rest of the way, just the same.

When I don't give it gas, I get the head toward the windshield jerk. If I give it gas during this time, I'm either neutral or pressed slightly back into the seat, which to me is more comfortable and leads to better acceleration.

That could also mean you're too slow releasing the clutch, are you holding the clutch at the friction point or easing through it?

[zenfiz6]

All I can say is don't worry its not in your head, it is very inconsistent.

Yah, but if it were in my head, it wouldn't be unexpected.

... I've noticed that whenever I take it above 4k, all rev hang disappears.

One thing I could verify last night is that, in all cases, the 4k->2k drop took longer than the 3k->2k drop. It didn't matter if I just revved to 3k or 4k and let it fall or if I was, say, at 45MPH in 3rd (3k) or 40 MPH in second (4k) and then pulled it out of gear. If there is some rev-hang, then it's not that strong.

It does catch me by surprise sometimes so I just have to look at the tach to re-calibrate myself.

This is exactly how I came to notice the issue. I watched the tach while counting so that I would know the timing for the shift. Almost always it was a 3-count. Then I noticed that, just revving the engine in neutral, it was a 2-count.

Shadow insists that Audi's have no rev hang whatsoever so I don't know. But the fact that you're even noticing some inconsistency makes me believe that you do have some computer interference.

Unless he has the software for the ECU, I'm guessing that he's basing his statement on empirical evidence. And---to be fair---the effect I'm seeing is small. So on the whole, I do agree with him.

One more niggly bit: I've noticed that, when I'm parked and just letting the engine idle, the RPMs will climb from 750 to about 1100 just by pushing the clutch in. Maybe that's just an unloading of the engine from the layshaft.... But it's interesting.

[daleadbull]

One thing I could verify last night is that, in all cases, the 4k->2k drop took longer than the 3k->2k drop. It didn't matter if I just revved to 3k or 4k and let it fall or if I was, say, at 45MPH in 3rd (3k) or 40 MPH in second (4k) and then pulled it out of gear. If there is some rev-hang, then it's not that strong.

See you keep saying 4k->2k and 3k->2k, but I'm confused here.

Assuming no rev hang (or consistent rev hang), wouldn't it always take longer for it to drop from 4k to 2k than it does to drop from 3k to 2k (about twice as long?)?

[theholycow]

One more niggly bit: I've noticed that, when I'm parked and just letting the engine idle, the RPMs will climb from 750 to about 1100 just by pushing the clutch in. Maybe that's just an unloading of the engine from the layshaft.... But it's interesting.

That sounds very much like a VAG-style feature. The idea would be to give it a little throttle for you when you launch.

If it was drag from the transmission...that'd be an awful lot of drag.

[zenfiz6]

One thing I could verify last night is that, in all cases, the 4k->2k drop took longer than the 3k->2k drop. It didn't matter if I just revved to 3k or 4k and let it fall or if I was, say, at 45MPH in 3rd (3k) or 40 MPH in second (4k) and then pulled it out of gear. If there is some rev-hang, then it's not that strong.

See you keep saying 4k->2k and 3k->2k, but I'm confused here.

Assuming no rev hang (or consistent rev hang), wouldn't it always take longer for it to drop from 4k to 2k than it does to drop from 3k to 2k (about twice as long?)?

Yes. It does. But not twice as long. Maybe 25% longer. Just revving the engine in neutral, it took something like 2.5s to fall from 4k->2k and 2s to fall from 3k->2k. Those fast 1k at the top drop faster than the lower ones. And people who know something more about engine suck (vacuum, whatever) could probably describe that better than the noob at this particular keyboard.

[zenfiz6]

One more niggly bit: I've noticed that, when I'm parked and just letting the engine idle, the RPMs will climb from 750 to about 1100 just by pushing the clutch in. Maybe that's just an unloading of the engine from the layshaft.... But it's interesting.

That sounds very much like a VAG-style feature. The idea would be to give it a little throttle for you when you launch.

Sounds laudable.

If it was drag from the transmission...that'd be an awful lot of drag.

I'm not too familiar with the moment of inertia of the transmission to know how much extra torque would be necessary to keep the engine at idle in neutral. But if you say that would be too high, I'll believe you.

[theholycow]

Those fast 1k at the top drop faster than the lower ones. And people who know something more about engine suck (vacuum, whatever) could probably describe that better than the noob at this particular keyboard.

Most likely the computer meddling. However, as I probably mentioned somewhere in this discussion, engine friction and reciprocating loss go up as a square (or some exponent) function of RPM rather than linearly...so at 4000 RPM there might be 2 to 4x as much loss as at 3000.

I'm not too familiar with the moment of inertia of the transmission to know how much extra torque would be necessary to keep the engine at idle in neutral. But if you say that would be too high, I'll believe you.

Well, look at it from the other end of things: that's an awful lot of energy to waste when fuel is $4/gallon, and an awful lot of energy for the transmission to dissipate (in the form of heat), when it's not even doing anything.

Just look at where the energy goes. This should come naturally to you of all people!

[zenfiz6]

I'm not too familiar with the moment of inertia of the transmission to know how much extra torque would be necessary to keep the engine at idle in neutral. But if you say that would be too high, I'll believe you.

Well, look at it from the other end of things: that's an awful lot of energy to waste when fuel is $4/gallon, and an awful lot of energy for the transmission to dissipate (in the form of heat), when it's not even doing anything.

Just look at where the energy goes. This should come naturally to you of all people!

So it makes sense that the engine should rev up a bit, but not because of the change in moment of inertia (gads, was I thinking like one of my students?), but because of some reduced friction in the system coming from the transmission. If the engine produces the same torque, but the frictional torque is lowered, then the engine RPMS increase until the two torques balance again.

Power is torque times RPMs (units aside). Even if the torque remains constant, to go from 750 to 1100 RPMs is an increase in power of the same magnitude (about 40%).

Ok, you definitely have a physics argument. But, still, 40% of a base consumption of 1 L/hr of gas is still not a lot. Even at $1/L.... Because I really don't know what to expect of frictional losses in a transmission, that could be reasonable in any car I'd build.

Now, had the RPMs gone up to 2k... then I might have questioned it a bit more deeply.

[theholycow]

It is my impression that in neutral the drag transmitted through the input shaft to the engine is so negligible that it's not even worth acknowledging. Without a computer meddling, as in my car that lacks any sort of computer at all (except the stereo), there is no measurable difference in RPM when in neutral with the clutch engaged or disengaged...and that's with my hackyard engineering that I'm sure has abnormal drag in various places. (Though, to be honest, I'm not equipped for particularly precise or accurate measurement.)

[Squint]

That sounds very much like a VAG-style feature. The idea would be to give it a little throttle for you when you launch.

[Shadow]

Shadow insists that Audi's have no rev hang whatsoever so I don't know. But the fact that you're even noticing some inconsistency makes me believe that you do have some computer interference.

No, I insist that my Audi has no rev hang whatsoever. I don't know about other models in the Audi line-up.