best shift points for most power????

[Veleno]

I'm wondering what's the best shift points on the NSX going from 1st-2nd-3rd-4th to get the most power???

I have an rx7 and I've found that the best shift points are right before redline, or when you hear the buzzer(at redline). Is it the same for the NSX? I have yet to run my NSX hard, but if/when I do, I'd like to know when is the best time to shift. So for all who run their cars hard, please elaborate.

Thanks for any help



------------------
Richard

 

[Lud]

Start the shift right before redline just like the RX.

 

[dswartz]

I have a 98 Coupe. I found the following:

Stock:
1-2 as close as possible to red line. Note, don't hit red line, the rev limiter will cause you to lose time. Therefore, if you power shift, shift at about 7800. If you speed shift, shift at 7950.
2-3 7700
3-4 7600
(never tried 4-5 or 5-6)
With Comptech intake, headers, exhaust, and nocats I can shift as close as possible to red line in all gears. While these mods do nothing for low end, they do give top end pull all the way to red line

 

[1BADNSX]

With both the 5 and 6 speeds you want to shift at redline to maximize either power or torque to the rear wheels.

With the 6-speed, if you shifted at 7700 RPM instead of redline going from 2nd to 3rd, the revs would drop to 5618, rather than 5837 - giving up power. If you shifted at 7600 RPM going from 3rd to 4th gear, the revs would drop to 5953, instead of 6266, again giving up power.

Bob Butler

 

[Veleno]

Thanks for the help guys. I guess it's pretty much the same for most sports cars.



------------------
Richard

 

[Number9]

Probably best to run the car on the dyno to get a look at the torque curve. For example, my car starts to lose torque at 7.5k RPM so shifting at redline isn't ideal (though my tach reads a bit high relative to the dyno, so I am shifting just before indicated redline). Also, as a previous poster noted, bumping up against the rev limiter is deleterious to good times. So on my car, making a mistake and shifting early has negligible consequences, but shifting late carries a penalty whether or not I hit the rev limiter (I'm boosted and run flat torque from 3.5 - 7.5, so YMMV.

Net takeaway is to dyno the car, see at what indicated RPM the rev limiter kicks in, normalize your tach readings to actuals, and then figure out your shift point(s) based upon the torque curve (i.e., if your torque curve falls off near redline, select a shift point such that the as-shifted RPM would give you the same torque as available just prior to that shift - you can fine tune for human/mechanical response latencies if you like...). If your car's torque curve is flat or rising to redline, then by all means, shift as near to redline as your reflexes reliably permit without bumping the rev limiter.

--twc

 

[nsxtasy]

Thanks for the help guys. I guess it's pretty much the same for most sports cars.

No, actually, it's not. See below.

Probably best to run the car on the dyno to get a look at the torque curve.

Or check out the manufacturer's published torque curve. For example, you can see the torque curve for the 3.2-liter NSX at http://www.nsxprime.com/FAQ/Technical/images/97/97NSXPowerCurve.gif It's very very flat, hardly changes at different RPM settings. This is why horsepower keeps increasing as revs go up (horsepower is directly proportional to torque).

For example, my car starts to lose torque at 7.5k RPM so shifting at redline isn't ideal

Actually, shifting at redline is probably still ideal. See below.

if your torque curve falls off near redline, select a shift point such that the as-shifted RPM would give you the same torque as available just prior to that shift

This statement is not correct because it doesn't take into account the gearing losses from going to a higher gear. It's the torque at the wheels that would be the same as available prior to the shift - and that means adjusting for the difference in gear ratios.

To maximize acceleration, you need to maximize torque at the rear wheels. Torque at the rear wheels is a multiple of torque at the crank, less drivetrain losses, multiplied by your gearing. This means that when you upshift, torque at the rear wheels drops by the amount that your gearing drops. For example, when you upshift from first to second in the stock six-speed, your gear ratio changes from 3.071 to 1.956. Thus it would make sense to shift at some point lower than redline only if your torque after the shift were 57 percent (i.e. 3.071/1.956-1) higher than before the shift. That's not true on the NSX, with its flat torque curve. But it could be true on some other cars, where the torque drops off rapidly before redline.

Bottom line for the NSX: Shift right before redline (so you can be sure not to exceed redline and/or bump into the rev limiter).

[This message has been edited by nsxtasy (edited 12 December 2000).]

 

[dswartz]

On the stock motor, the torque peak is 5500 RPM. The optimal shift point is when the (torque@RPM in current gear) x (current gear ratio) is less than the (torque@RPM in next gear x next gear ratio. If you use the factory torque curve and gear ratios of 3.07, 1.96, 1.43, 1.12, .91, .72 with a final drive of 4.06, the calculated shift points are (1-2) 8000, (2-3) 7880, (3-4)7750, (4-5) 7690, and (5-6) 7730. My shift points are based on real world testing at the drag strip (and 35 years of drag racing experience). The best results are as stated in my first post in this thread.

 

[nsxtasy]

A few questions - and please understand this is because I'm curious, not because I'm questioning the validity of what you say (which is the same as what I said, anyway

):

1. When you're shifting at a dragstrip, how precise can you be regarding shift points? For example, if I've got the pedal floored, I gotta believe that if I'm aiming to shift at say 7850 rpm, I could easily wind up shifting say +/- 150 rpm away from that figure - so for me, shifting "just as I'm approaching redline" is about as close as I can come to shifting at the calculated shift point, which for all gears is somewhere between 7690 and 8000. If you try shifting at say 7700 rpm, how close can you come? +/- 50 rpm? +/- 10 rpm?

2. Why do you recommend shifting a stock car at points in the rev band that are different from the calculated values for maximum acceleration?

3. What's "power shifting" and "speed shifting"?

Thanks Dave...

 

[Number9]

Thanks for the correction(s), Messrs. dswartz and nsxtacy. Agreed I was off (near optimal solutions are only good for computers and applied mathmeticians and not gearheads!) but I'm not quite sure I can reconcile both your formulations. In effect dswartz is optimizing shift point based on a where HP*gear ratio product equals as-shifted HP*gear ratio, whereas nsxstacy is optimizing based upon torque*gear ratio product - not the same. Sadly, I don't have a physics book handy (not sure I even own one anymore), and would appreciate some definitive clarification.

It is worth reiterating that the stock tachs aren't all that accurate, so it is important to calibrate it in conjunction with optimizing shift points. (I don't bump the rev limiter till I'm well into ferrari 355/360 peak RPM _indicated_ readings!)

--twc

 

[1BADNSX]

You want to maximize the torque at the rear wheels, which is like torque at the crank multiplied by the gearing. Because torque is related to power along with RPM, maximizing torque at the rear wheels is the same as maximizing power at the crank (In other words, it would prove the same shift points).

I agree that you need to look at actual NSX dyno data, but I have only found one marginal case where you would want to shift any NSX before redline to maximize acceleration. There are several dyno runs shown at:
http://southwest.nsxca.org/phoenix_dyno_day.htm

For all of the five speeds, this dyno data shows you want to shift at redline in every gear. There is one stock 6-speed 3.2L dyno run that shows you would want to shift at redline for all of the gear changes except 4th to 5th when you would want to shift at 7900 RPM.

Bob Butler

 

[Lud]

Originally posted by Lud:
Start the shift right before redline just like the RX.

Am I typing in invisible ink here??

 

[dswartz]

Dave, good questions!

1. How precise I can be is based on experimentation. Keep in mind that shifting is a process that begins at the stated shift points. Specifically, in first gear as the tach passes 7000 RPM I pay close attention, then at the shift point I press the clutch while simultaneously pulling the shifter towards the next gear. I then release the clutch. Practice (especially in first gear) tells me that if I wait to long, say to 7850, I hit the rev limiter. Therefore, I know that during the process, the engine will rev an additional 200 RPM during the shift. Someone else may perform the process faster and will be able to wait an additional 50 or 100 RPM.

2. If you notice, the calculated shift points are approximately 150 to 200 RPM above my shift points. After dozens of runs, I have achieved my best ¼ mile times at these points. However, since my shifts begin at my points, I’m probably shifting at exactly the calculated shift points.

3. Power shifting is shifting with the pedal on the floor. Speed shifting implies lifting off the throttle between shifts. In drag racing, I always power shift. In road racing or autocross, I speed shift because the lost time is insignificant. Furthermore, power shifting will most certainly unsettle the car around turns.

Bob, I’ve found with the Comptech Mods, (Intake, Headers, Exhaust, no cats) I must shift as close as possible to red line because the torque at lower RPM’s has decreased, while the torque at higher RPM’s has increased. However, after all is said and done, my ¼ mile times have consistently improved only one tenth of a second with the mods.

 

[Number9]

Lud, we're not ignoring you, just interested in figuring out the rationale! Seems there are some degenerate (or "marginal" per Bob's lexicon) cases that warrant more detailed analysis. Look at this URL for a Gruppe M modded car with a torque curve that falls off a cliff pretty early: http://www.daliracing.com/nsxcatalog/boy_racer/supercharger_graphs_03.cfm
You'd definitely want to recalculate/optimize your shift points if your torque curve started to fall off a cliff at 6k RPM. So in the majority of cases, shift just before bumping rev limiter is a "good thing(tm)" but depending on mods, it may not yield best 1/4-mile time... Anyway, I've learned a lot, thank you all.

--twc

 

[Lud]

That graph looks like Alex V's car before he fixed his fuel management with the Gruppe M blower. That dyno is obviously not a properly tuned NSX, even with the GMSC -- I have seen other GMSC dynos that were much flatter and stronger well into the 7k RPMs (similar to the stock torque curve). I would suggest that if that's what someone's NSX dyno looks like (all uneven with an early drop-off), they need to fix their car, not adjust their shift points. Just my opinion!

[This message has been edited by Lud (edited 13 December 2000).]

 

[Veleno]

wow! Lots of great info!! Thanks.

I'd love to try to dyno my car, unfortunately, there's no dyno around here. I was thinking about bringing one in...we'll see.

At least our FIRST EVER drag strip/race circuit is being built and the 1/4 mile should be completed sometime in February. I can't wait!!!



------------------
Richard

 

[nsxtasy]

Originally posted by Lud:
Originally posted by Lud:
Start the shift right before redline just like the RX.

Am I typing in invisible ink here??

I don't know; I thought I was echoing you when I said, "for me, shifting 'just as I'm approaching redline' is about as close as I can come to shifting at the calculated shift point".

 

[nsxtasy]

I'm not quite sure I can reconcile both your formulations. In effect dswartz is optimizing shift point based on a where HP*gear ratio product equals as-shifted HP*gear ratio, whereas nsxstacy is optimizing based upon torque*gear ratio product - not the same.

Re-read dswartz's post and you'll see he is optimizing where torque (not HP) times gear ratio is equal before and after the shift - and that is the same.

 

[Number9]

[Re-read dswartz's post and you'll see he is optimizing where torque (not HP) times gear ratio is equal before and after the shift - and that is the same.]

Not so - he mentioned comparing two numbers for equality that both include torque*RPM terms, i.e. Torque*PRM*gear-ratio. Note that if you multiply both sides of the equation by the appropriate constant, you have HP*gear ratio. As I noted, "in effect" it is an HP comparison. QED.

 

[Lud]

dswartz said torque @ RPM (torque at RPM), not * RPM (torque times RPM).

[This message has been edited by Lud (edited 13 December 2000).]

 

[Number9]

[dswartz said torque @ RPM (torque at RPM), not * RPM (torque times RPM).]

Mea culpa!

 

[ANDRUWNIAK]

postulate: acceleration is directly proportional to torque

formula: new rpm = old rpm x ((new gear ratio)/(old gear ratio))

application: if ((torque at old rpm)(old gear ratio)) > ((torque at new rpm)(new gear ratio)) then shift at a higher rpm.

the upshot: Only shift before redline, if you find there is more torque to be had at a higher gear and lower rpm than there is a a lower gear and higher rpm. For the NSX this means redline. Rather than trust the stock tach on my wrangler, I just shift at the point where i get the biggest chirps (themselves a byproduct of torque), indicating more torque (and therefore acceleration).

[This message has been edited by ANDRUWNIAK (edited 13 May 2002).]

 

[nsxtasy]

Originally posted by ANDRUWNIAK:
postulate: acceleration is directly proportional to acceleration

Directly proportional? I might go so far as to postulate that acceleration is exactly equal to acceleration.

 

[NSX4U2]

Originally posted by Lud:
Am I typing in invisible ink here??

Don't feel bad, I asked a question several days ago about cleaning the throttle body.

Actually I asked twice for feedback.

One response. Either that or the ink is invisible?