How to accelerate in P&G mode?

[sonyhome]

I'd like input from people with experience with optimal acceleration when in Pulse and Glide (keeping the engine ON in glide).

What's your real world experience on that topic? How hard to accelerate in P&G mode?


FE recommendations are to accelerate as slow as possible, in low RPMs. I believe the logic is you don't know what you're doing so we'll give you an idiot proof solution yet not optimal: low RPMs=less gas used. The recommendation also likely assumes you keep using a lot of gas once you've accelerated. HOWEVER, in P&G, the goal is to save gas on the total distance travelled, and in G mode gas used is little.

Let's set the rules here:

- Total distance is split in P-distance and G-distance.
T = P + G
- In G, car uses constant 700RPM, and burns a typical 0.3GPH.
- Total travel time is constant no matter what (else you could drive slower meaning higher FE in theory)
- City travel < 45MPH so aero impact negligible

My take is the best way to accelerate is a hard acceleration where the engine is in it's peak torque curve, so that distance P is minimized, in favor of G, and that doing so the engine runs at peak performance aka energy burned vs enrgy output to the wheels.

With the above, if we can plug in GPH for some accelerations and acceleration durations, we could tell what's optimal. However I don't have the data...

[theholycow]

I've been having great luck in my manual transmission car using WOT in the highest gear possible, though in cars that go to open loop mode under WOT it would be better to use 80% open throttle.

In my automatic, I use about as much pressure on the gas pedal as I can without causing it to downshift. For sure it is accelerating far less efficiently in 2nd gear under any amount of throttle than in 3rd gear.

Take a look at the fuel rate monitor thread linked in my sig, you can get a gauge that represents your GPH. A ScanGauge can't measure your GPH, it only guesses based on a calculation of MAF, O2, TPS, and other sensors.

[sonyhome]

Hmmm.. But in 2nd gear, you accelerate darn fast, so you're in acceleration mode less longer. For example you could be burning 50% more gas in P mode but need to do it for 1/2 the time before going into G mode.

You have to add time to the equation. That's why for me the equation is hard to figure out!

As for the "DIY fuel rate monitor" thread, I'd like to do that on my Del Sol, it's OBD-1 so no SG anyway. Someone posted a system with a wire connected to the PC audio port to collect duty cycles with a Java applet. I think it's a great idea, and wanted to do that but have yet to figure out what to tap where w/o messing up the car. I also have ideas on how to extend that to allow detailed experiments and conclusions. Pics of the tap setup would be great to get an idea of how it was done (but that's all off topic).

[PaleMelanesian]

http://www.autospeed.com/cms/A_110216/article.html

High throttle, low rpm is best. Get the most efficient power out of the engine.

But... I'm finding that from a complete stop up to maybe 20mph, super light throttle ends up better. Then, when I'm closer to the desired speed, I switch into a heavy-throttle P&G routine.

[Hateful]

I just go by sound;Automatic 4 speed BTW..... When accelerating I listen for the steady increase in RPMs and a nice smooth gear change, then the same repeated for the next gear.
When cruising at highway speed ( or my version there of) I listen for a certain up/down sound almost as if the transmission was slipping.Then shift to neutral when I feel I'm going fast enough for a good glide.

[theholycow]

You have to add time to the equation. That's why for me the equation is hard to figure out!

I figured for time. When the total acceleration is done, it's always been better for me to use a higher gear and lower RPM.

Pics of the tap setup would be great to get an idea of how it was done (but that's all off topic).

I could post pics, and I think monroe74 or fumesucker did (and I'm pretty sure it was fumesucker who hooked it up to his PC's sound card), but it's so incredibly simple that you really don't need pics. Your fuel injectors are probably kinda near your spark plugs. When you look at fuel injectors, you'll see that each has two wires going to it; one of the wires will be the same color for all, and the other wire will be a different color for each injector. Use the wire that's different.

But... I'm finding that from a complete stop up to maybe 20mph, super light throttle ends up better. Then, when I'm closer to the desired speed, I switch into a heavy-throttle P&G routine.

Accelerating from a stop, you're in very low gears with very fast acceleration and it's hard to control your RPMs and shifting fast enough. At least, that's what I've found with WOT in my manual tranny, and in my auto of course a quick stab of the gas pedal will put me almost immediately into high RPM before I get a chance to let off the pedal.

[R.I.D.E.]

Best mileage in my Del Sol was 47. Pale does much better in a heavier car with the same basic engine tranny combination (I think).

Aero drag in city EPA cycle is 29%, highway is 70%, so aero is always a factor.

70% throttle in low enough gear to have your revs peak at 2200, then shift to higher gear.

Draft if possible, it's easy here with fairly heavy traffic.

Instead of EOC (which you stated you don't use, like me) here is an option.

Whenever you need to slow down use the highest gear practical to keep engine revs above 1000 RPM. I used to never downshift to slow down, but now I think it is a big factor in my mileage improvement.

If you know your traffic light patterns, try to time it so you don't have to decelerate more than a coast, but if you find your self having to stop for a light, try to use downshifting to stop all fuel delivery, which is the same as shutting the engine off (no fuel delivery).

I find the ideal average for me is 1 to 3 pulse to glide, at about 40 MPH average speed. 2200 in your car should be about 52 MPH. Thats for a 1.5 engine (not SI) and manual tranny (which I assume you have).

At above 50 MPH average try to find a draft. Vehicles with larger frontal areas than yours are best. It may be better to just draft than to try to P&G at higher speeds, try both methods to see which works best for you in your situation.

The same applies to the CRV except drafting would be much more effective.

Drafting for me is 3 highway stripes between me and the host vehicle on the Interstate at 65 MPH (about 125 feet). Lower speeds would be closer. At 45 MPH around here if you give the car in front of you more than 6 lengths, someone will always pull over in front of you and then slam on the brakes and turn right. I try to keep my minimum stopping distance between the host vehicle and my car. Closer may be more effective but it also becomes a lot more dangerous.

Every suggestion should be considered and adjusted to your specific desires and the traffic you have to deal with. I have driven 64 east from Richmond at 77 MPH, in bumper to bumper traffic with less than 5 cars lengths between any vehicle, for 60 miles, without anyone moving more than 200 yards ahead or behind anyone else. Thats pretty much the saturation point for a 2 lane (each way) Interstate at 65 MPH.

regards
gary

[cugir321]

I'm still trying to figure out P&G. Are you guys saying very light throttle in say, first, until I get close to 1.5 rpm's, then floor it until 2k, then shift, light throttle to 1.5, then floor it to 2k...etc.

I've been using light throttle always to 2k, then shift.

http://www.autospeed.com/cms/A_110216/article.html

High throttle, low rpm is best. Get the most efficient power out of the engine.

But... I'm finding that from a complete stop up to maybe 20mph, super light throttle ends up better. Then, when I'm closer to the desired speed, I switch into a heavy-throttle P&G routine.

[sonyhome]

Hmmm... I'm not quite sure I follow.

From a stop, 1st quick to 2nd, up to 2000RPM, 3rd to 2000RPM, 4th to 2000RPM, coast.

That would be better than 1st/2nd/3rd then coast (higher RPMs), or 1st/2nd then coast (even higher RPMs)?

Would it help to take a counter videos while P&G'ding, with different scenarios to have experienced P&G'ers tell which looks and sounds liek the best P&G?


I usually P&G from 35 to 45, and I play the lights. If it turns orange and I'm close, accelerate, if I'm far, coast or even brake to get green before I have to stop.

I don't draft in city traffic because people's driving is erratic. I find P&G more effective, as I can coast 3-to-1 like said above, my commute is flat or slight inclines perfect for P&G with a low roll resistance car.

[sonyhome]

For the fuel rate monitor, photos would be good, not just the taping of the injector, but overall what was done, the tapped wire needs to go back to the cabin.

Also the tap gives you:
-duty cycle
- engine RPMs

But you're missing speed to get MPG.

Therefore it'd be nice to tap another signal to get speed or tire rotations.

With that, we could do a cool ultra precise hypermiling analyzer.

[theholycow]

I'm still trying to figure out P&G. Are you guys saying very light throttle in say, first, until I get close to 1.5 rpm's, then floor it until 2k, then shift, light throttle to 1.5, then floor it to 2k...etc.

I've been using light throttle always to 2k, then shift.

You seem to be under the impression that P&G is practiced when you might otherwise be normally accelerating from a stop. That is not correct. P&G is used when you'd otherwise be cruising at a steady speed.

Heavy throttle and lower RPMs (to result in the same amount of acceleration as light throttle and higher RPMs) is more efficient than the way you accelerate. Of course that only works if you've got a manual or a very obedient tiptronic/manumatic/etc transmission.

For the fuel rate monitor, photos would be good, not just the taping of the injector, but overall what was done, the tapped wire needs to go back to the cabin.

Okay, I'll post some photos, but it's really really incredibly simple. Just run the wire to wherever your meter will be, and connect it to the meter's positive lead. Ground the negative lead to any ground (I use a power outlet) and enjoy.

Also the tap gives you:
-duty cycle
- engine RPMs

The tap does not give you engine RPM, only fuel injector duty cycle.

But you're missing speed to get MPG.

Therefore it'd be nice to tap another signal to get speed or tire rotations.

Vehicle Speed Sensor (VSS) is what you need to tap for that, and I recently read (and surely posted) about a DIY MPG gauge that uses VSS and the fuel injector tap.

With that, we could do a cool ultra precise hypermiling analyzer.

If we could run this all (VSS, FI tap, and tach) into a computer or log it, you're damn right. That's a good idea. That's definitely a project that merely requires adding FI tap as an auxilliary input to a ELM327 interface. I have no idea how that would be done, it might have to be done through a separate serial port and combined in software.

[PaleMelanesian]

There are two different things going on here: accelerating, and cruising. For accelerating, use light throttle. For P&G where you would otherwise be cruising, use heavy throttle.

For accelerating, use first gear just to get moving. Go into 2nd gear at < 5mph, and from there use very light throttle up to a max of 2000 rpm before shifting, and keep shifting early as you gain speed. Use this to get up to, or nearly up to, your cruising speed.

Once you get up to speed, then start cycling between heavy throttle and neutral coasting - pulse and glide.

[theholycow]

Okay, I'll post some photos, but it's really really incredibly simple. Just run the wire to wherever your meter will be, and connect it to the meter's positive lead. Ground the negative lead to any ground (I use a power outlet) and enjoy.

I put up a couple photos. I thought I had more. Anyway, take a look at the DIY fuel flow meter thread again. It's linked in my sig. Let me know if I can take pictures of anything else for you.

[sonyhome]

[Off topic fuel meter]

Thanks! With that long a wire you still don't disrupt the ignition? Great, simple...

You do have RPMs with that tap, just not measured by your voltmeter. On a PC you'd see the frequency of the pulses, and the duration of on vs off (your duty cycle).

With VSS you have your speed.

With some software smarts, you have your knowledge of when you shift gears or coast...

That is definitively a project I'll want to do, including the software part.

I think using the audio input channel is the easiest. Can be done for injectors, but what about VSS? What kind of signal is it going to be? Gotta look around.

[sonyhome]

P&G and acceleration.

So if I understand right, experience shows it's most efficient to use the following two modes?

A P&G mode: Shift from N into 4th or 5th gear and floor the accelerator with a led foot (or close to that at 80%). That uses WOT with low engine RPMs to accelerate the car.

B Red light mode: Accelerating from a stop, use light acceleration (feather foot on pedal), and shift early (say around 2000rpm) until you reach coasting speed.

C alternative to B?: For starting from a stop light, we could also do this: 1st, 2nd with light acceleration and a 2000RPM shift. Then skip 3rd, and go directly to 4th gear (or 5th?), and use a hard acceleration but with low RPMs. Is that what you guys are doing?

I suspect the trick is we need the RPMs for 4th gear to be at least the minimum bearable for the engine, so for starters, the RPMs must be at least 1200 when getting into 4th gear.


Automatic: For A/T cars the above seems hard to achieve. as a kickdown will downshift, and a semi hard acceleration will lazilly shift up.
So that should allow you to do (A) and (B), but harder to do (C), unless you have a good pedal feel and can hear when you're in top gear.

The difference between (A) and (B) mode in terms of the driver is that in P&G, he would depress the pedal more than when he accelerates from a red light...

Did I get it right?

[R.I.D.E.]

Look at a Brake Specific Fuel Consumption Map.

There is an area on the map that represents the most horsepower for the least fuel. I like to call it the "sweet spot".

The most basic and essential component of effective P&G is to understand that you try to stay in the "sweet spot".

The Sweet Spot is easy to determine even without special equipment.

Put your car in a gear that is higher than you would normally use. The Sweet Spot is the point where when you apply increasing amounts of throttle, you get almost no difference in acceleration.

It takes a very long time for my VX to accelerate from 1200 RPM to 2000 RPM in 5th gear, compared to using second or third gear, or even fourth gear.

Given the option by the traffic around you, use the highest gear at the lowest RPM for the pulse.

In some cases I actually have to downshift in order to accelerate more rapidly. After driving for 42 years, my shifting is second nature, including any rev matching. I don't look at any instruments, it's pretty much instinctive.

I have to stay within a minimum and maximum range of average speed to keep in time with traffic lights. When it works I can go through 26 lights without idling for 1 minute total time in a 20 mile drive.

This requires me to accelerate faster than I would if I had no lights or traffic to deal with. MY success at "adaptive P&G" is reflected in my mileage.

If I was on Route 17 north of Gloucester VA, where the road is 2 lanes, divided, with a 55 MPH speed limit and virtually no traffic, I could probably get close to 70 MPG, without EOC.

I read here that turning off the engine on a VX, causes the lean burn system to not work for 30 seconds after a restart. I may test this one day when I have time, and I am working on a very simple system for topping off my fuel tank in increments of less than a gallon, so I can make precise comparisons under identical situations without haveing to drive 250+ miles between fillups. That range of distance for me only uses about 40% of my tank capacity.

regards
gary

[1993CivicVX]

Hypothesis: slow acceleration is always best (except in traffic light matching in Gary's case).

So I'm going back to old conventional wisdom that accelerating slowly is always best for the simple following reasoning: do you accelerate hard to get up to cruising speed when you don't want to P&G? Why, then, should accelerating hard for the purpose to glide afterward be any different?

[theholycow]

Should we take this over to the fuel meter thread?

[Off topic fuel meter]

Thanks! With that long a wire you still don't disrupt the ignition? Great, simple...

The wire isn't connected to the ignition, it's connected to the fuel injector. Anyway, I too was concerned about disrupting the fuel injector, but my FE keeps going up and I'm not throwing any codes, so I it seems to have no ill effect.

You do have RPMs with that tap, just not measured by your voltmeter. On a PC you'd see the frequency of the pulses, and the duration of on vs off (your duty cycle).

Careful with terminology. It's a dwell meter as used here. While you could get RPM data from it, why go to so much effort and do it the hard way when you can just get RPM the old fashioned way from the tachometer wire? Even easier if you're using a PC, use a $30 (on eBay) ELM327.

It probably won't work with your '93 Del Sol, though, unless that car is OBDII.

With VSS you have your speed.

You'll get that with the ELM327 too. Since it's so cheap, that probably ought to be your starting point. I'm pretty sure there's open source software for it so you can write your own piece without rewriting everything.

With some software smarts, you have your knowledge of when you shift gears or coast...

Again, th ELM327 will cover that too, with throttle position.

That is definitively a project I'll want to do, including the software part.

Whichever way you do it, I'd like to be involved a little. I can't really write code unless it's simple, and I can't help with hardware due to geographical differences (unless you happen to live in the RI/MA/CT area).

[1993CivicVX]

http://www.autospeed.com/cms/A_110216/article.html

High throttle, low rpm is best. Get the most efficient power out of the engine.

But... I'm finding that from a complete stop up to maybe 20mph, super light throttle ends up better. Then, when I'm closer to the desired speed, I switch into a heavy-throttle P&G routine.

But there's no power at low RPMS! But who am I to argue with a gas getter guru? But I am finding that I'm using less gas with light acceleration and slightly higher RPMs than with full throttle low as possible RPMs. We both have lean burn engines, too, so what's true for you should be true for me, right?

[PaleMelanesian]

Mine's not lean burn. You might actually have better results than I do with light acceleration. One more tool in your toolbox than in mine.

I'm not sure the exact dynamics of WHY slow acceleration works better from a stop, while hard acceleration works for P&G. But it does. For P&G, high throttle is unquestionably the best.

From a stop, for me in my car on my commute, harder acceleration will yield low 60's mpg, while feather-light acceleration will give me upper 60's and sometimes into the 70's.

(RIDE's advice differs from mine in part because he has LOTS of traffic lights to deal with, and I have relatively few. Also, he has lean burn and I don't.)

[theholycow]

P&G and acceleration.

So if I understand right, experience shows it's most efficient to use the following two modes?

Different people have different experience which leads them to different style (exacerbated by different vehicles). You may have to try a couple tanks one way, a couple the other way, until you choose between any two strategies.

A P&G mode: Shift from N into 4th or 5th gear and floor the accelerator with a led foot (or close to that at 80%). That uses WOT with low engine RPMs to accelerate the car.

This is pretty universal. If your car is equipped with a wideband O2 sensor then it probably does not go into open loop mode at WOT, so you can completely floor it. I found out that that's the case in my VW.

B Red light mode: Accelerating from a stop, use light acceleration (feather foot on pedal), and shift early (say around 2000rpm) until you reach coasting speed.

C alternative to B?: For starting from a stop light, we could also do this: 1st, 2nd with light acceleration and a 2000RPM shift. Then skip 3rd, and go directly to 4th gear (or 5th?), and use a hard acceleration but with low RPMs. Is that what you guys are doing?

I suspect the trick is we need the RPMs for 4th gear to be at least the minimum bearable for the engine, so for starters, the RPMs must be at least 1200 when getting into 4th gear.

I try to use WOT as much as possible, but my car has a lot of low-end torque and it can be tough to moderate my acceleration at WOT. When I do, I shift at 1200 or 1400 RPM, entering the next gear at about 1000rpm. My pattern is as follows:
1st: A light, quick application of gas gets the car rolling enough for 2nd.
2nd: A quick stab gets the torque-strong car fast enough for 3rd or 4th, which are uselessly close in my car.
Depending on my current pattern, I may skip 3rd.
3rd, if I use it: A quick stab and I go from 15 to 20mph, and am ready for 4th.
4th: Now I get a chance to drop it to the floor for a couple seconds, at which point I'm going 25 or 30mph and am ready for 5th.
5th: On the floor until I reach the top of my pulse, then I begin P&G cycling.

Automatic: For A/T cars the above seems hard to achieve. as a kickdown will downshift, and a semi hard acceleration will lazilly shift up.
So that should allow you to do (A) and (B), but harder to do (C), unless you have a good pedal feel and can hear when you're in top gear.

My GMC is nice about this, it shifts pretty decently for FE. I can briskly accelerate using a lot of the gas pedal and it shifts firmly at ~2500 rpm. It's nowhere near where I'd shift if I had control of shift points, and I hope soon to get a programmer that allows me to change shift points and such.

Other vehicles may have manumatic-style control, either tiptronic or paddle shifters or etc, and it may be feasible to use a lot of gas pedal and shift low.

The difference between (A) and (B) mode in terms of the driver is that in P&G, he would depress the pedal more than when he accelerates from a red light...

Did I get it right?

Yup.

[monroe74]

But there's no power at low RPMS!

This came up in another thread a few weeks ago. My VX pulls smoothly from 1000 rpm. Or even a bit less. Even in top gear, and even going up a moderate grade, and/or with one or more passengers. I think other VX owners have reported a similar experience. I think there might be something about your engine that is not providing proper power at low rpm. I realize you're still getting good results, but I think maybe you could do even better.

I think I suggested you check the timing. I don't recall if you mentioned doing that.

[theholycow]

Hypothesis: slow acceleration is always best (except in traffic light matching in Gary's case).

So I'm going back to old conventional wisdom that accelerating slowly is always best for the simple following reasoning: do you accelerate hard to get up to cruising speed when you don't want to P&G? Why, then, should accelerating hard for the purpose to glide afterward be any different?
But there's no power at low RPMS! (rest of quote finished later in this message)

I'm not sure the exact dynamics of WHY slow acceleration works better from a stop, while hard acceleration works for P&G. But it does. For P&G, high throttle is unquestionably the best.

You are both treating "throttle" and "acceleration" as synonyms. They are not. Slow acceleration and high throttle combined should provide the best results in most vehicles. Possible points of failure include cars equipped with lean burn (which may be able to accelerate more efficiently with light throttle) and cars that go into open loop at WOT (which is easy to remedy, still use high throttle but not quite WOT).

But who am I to argue with a gas getter guru? But I am finding that I'm using less gas with light acceleration and slightly higher RPMs than with full throttle low as possible RPMs. We both have lean burn engines, too, so what's true for you should be true for me, right?

When it's all said and done, you can't argue with results.

[PaleMelanesian]

You're right, I did word that confusingly. Using low rpm in all cases,

Heavy throttle, which still only gives moderate acceleration, is less efficient than light throttle, which gives "slightly less than moderate" acceleration.

This is comparing 80% throttle or so (no enrichment yet) vs. very light - just enough to be gaining speed.

[1993CivicVX]

most of my driving is short trips so instead of giving the motor a chance to warm up I EOC even if it's only for 5 seconds so it's not idling at 1500RPM. As a result, my battery is often too low to start the engine using the ignition, and on more than one occasion of late I've had to push it when I waited too long to bump start it and failed! The starter doesn't have enough juice to start the car, but enough to bump start. Good times, good times. Slightly embarrassing to have to push the car in the middle of town and then jump in and bump start it. My friend's think it is absurd/funny.

[1993CivicVX]

This came up in another thread a few weeks ago. My VX pulls smoothly from 1000 rpm. Or even a bit less. Even in top gear, and even going up a moderate grade, and/or with one or more passengers. I think other VX owners have reported a similar experience. I think there might be something about your engine that is not providing proper power at low rpm. I realize you're still getting good results, but I think maybe you could do even better.

I think I suggested you check the timing. I don't recall if you mentioned doing that.

Yeah, I remember this. Maybe there is a discrepancy between our definitions of "smooth" but my car will not pull at 1000 rpm in 5th gear smoothly--it will do it--but I wouldn't characterize the mild vibration and unwillingness as smooth. 1250RPM in 5th gear and the car is very happy. I don't think there is anything wrong with the car--just an old engine (200,000+ miles --dunno how many exactly since the odometer only works about 60% of the time)

[theholycow]

You might want to charge that battery from mains power to avoid having to get out of the car and push it to start. It really is absurd. This is 2008, not 1908. Plus, keeping it dead all the time will destroy it quickly so it will never hold a charge.

[1993CivicVX]

how do I charge it from mains power? I'm doing a highway trip tomorrow where I will not do too much EOC. That should charge it up good, right?

I'd like to hook up some solar panels to the battery, but too expensive.

[theholycow]

how do I charge it from mains power?

http://www.tayna.co.uk/getimage.php?image=tayna_chargers/2228g4138.jpg

[R.I.D.E.]

Here is my route one way.

4.5 miles, 55 MPH, 1 traffic light at the end, 4 lane non divided highway with Interstate on one side and railroad tracks on the other, perfect for hypermiling because everyone is on the parallel interstate. Plenty of room and time to DFCO and catch the light green every time. 4 different downhill slopes for coasting about 30% of the total distance. I usually stay about 50 MPH on this stretch.

3.5 miles, 45 MPH, 1 traffic light at the end, same 4 lane non divided highway, city park on left, Interstate on right, light traffic. More nice downslopes to help maintain speed when coasting.

3 miles, 45 MPH, 9 traffic lights mostly well timed but some will always catch you due to triggering of light by traffic entering road. Some construction, normally has a left turn lane, but not in construction area. Speed limit drops to 35 in construction part. Still 4 lane, but traffic can get heavy here especially if Interstate is not moving. Left turn at the end is most difficult to time properly, most idling at this point. Lots of DFCO here. Rabbit morons race to red light and mess up your timing.

1 mile, 4 traffic lights, 4 lane divided, very hard to time lights due to light timing controlled by entering traffic. Fortunately lights change fairly quickly.
Lots of DFCO on this short section. This is where the rabbit morons blast by you and crap up your light timing.

3 miles, 1 traffic light at end but right turn on red so waiting time is usually short, two lanes 45 MPH. Lots of elevation changes where coasting speed can remain alomost the same as average speed. Usually not much traffic. Lots of chances to coast down from 45-50 to 30-35, when no one is behind you. Pulse on inclines and glide on declines is the norm here, and that is typical of my strategy with the very gradual elevation changes that are typical on this whole route.

1 mile, 45 MPH, 2 traffic lights, 4 lane divided, traffic is usually fairly heavy. Hard to catch those 2 lights green, sometimes the traffic is 30 or more cars at the light when it changes and there is no way you can coast up on them and catch the light green. Nice chance to coast downhill then back uphill in this stretch.

6/10 mile, 1 traffic light at the end, exit ramp to overpass, good spot to use your speed to coast up the ramp to lower speed limit. DFCO to light at bottom of downslope to left turn at light. In most cases I can catch it green and turn immediately, but the light is switched by someone in the left turn lane, so on occasion you have some idling.

1.4 miles, 40 mph, two lane residential areas and two schools. Opportunities to coast on downhill areas and back uphill at almost constant speed due to elevation changes. Pulse begins halfway up the upslope to speed limit. Opportunities to coast to much lower speeds when no one is behind me.
Left turn at the end of thsi stretch with no traffic light. I make the turn in 3rd gear.

7/10 mile, 40 MPH, two lane, starts downhill so I can coast halfway then pulse and coast one more time on this stretch. Stop sign at the end for DFCO opportunity.

1/10 mile, 35MPH, to vehicle actuated left turn, always a crap shoot on this one. Sometimes I am the trigger, sometimes someone else catches it and I have to race to get through, sometimes I just have to trip it and wait 25 seconds. Ideally I can coast through when I have another vehicle to trip the light at the perfect time.

6/10 mile, 55 MPH, 2 lane, good chance to coast at the end, left turn trip actuated light, another crap shoot as far as timing, worst case is about 25 seconds wait.

19.4 miles 22 traffic lights, usually takes me about 40 minutes if I use this back road route.

In this area the notorious traffic jams are areas that can have 100,000 vehicles cross a single point in one day. Thats better than one vehicle per second (86,400 seconds per day). That means on a two lane choke point you have 50,000 cars daily per lane, every 86,400 seconds, or less than two seconds apart for 24 hours. Of course its always worse because there are times when it is really bad, and times when its not bad at all. 8 mile traffic backups happen almost daily in this area, called the Hampton Roads Bridge Tunnel.

These areas and traffic nightmares are usually not an issue for me. I stay off the beaten path and use roads I have used for 42 years. The same turn I now coast through at 35 MPH (the speed limit), I used to go through at 70 in my 59 Vette. It would do 70 on the average Interstate cloverleaf. I averaged 20-22 MPG in that car in 1973 the year the speed limits were reduced to 55 MPH nationwide. The Vette came with a 4.11 rear end which I ripped apart in two weeks. Changed it to a .308 ($25 part) and it tached 2100 at 62 MPH.

Current average 59.18 MPG

regards
gary

[monroe74]

my battery is often too low to start the engine using the ignition

I do a fair amount of EOC. I do tend to avoid it at night, because I'm a little concerned about the headlights draining the battery. Nevertheless, I think there might be something wrong with your charging system. Unless you're doing something extreme, I think you shouldn't often have this problem. I have a feeling maybe you need a new battery. Have you checked the level in it? It could also be something simple like a corroded connection.

Also, with a warm engine, the cranking required to start it should be very minimal. So this might be another clue that your engine isn't running quite right.

Slightly embarrassing to have to push the car in the middle of town and then jump in and bump start it.

Years ago I had a bunch of cars that worked this way. I would routinely look for parking spots that would let me roll downhill. But this was because they were old sports cars in poor condition. A few Fiat 124 Spyders, an Austin-Healey Sprite, an Austin-Healey 3000, and a TR-4. At one time or another, these cars had faulty charging/starting systems.

British car inside joke: Why do the British drink warm beer? Because Lucas makes their refrigerators. (Lucas is the company that makes electrical parts for British cars.)

Anyway, you have a Honda, not a Triumph, and I think you probably shouldn't be having this problem. Like someone said, you can solve the problem with an external charger, but I think that shouldn't be necessary, unless there is something extreme about the way you're using the car (e.g., I assume you don't have a 10,000 watt stereo).

I realize the VX has some special logic which restricts the way the alternator works, but I think you still shouldn't be having this charging/starting problem.

Maybe there is a discrepancy between our definitions of "smooth" but my car will not pull at 1000 rpm in 5th gear smoothly--it will do it--but I wouldn't characterize the mild vibration and unwillingness as smooth. 1250RPM in 5th gear and the car is very happy.

Yes, the word smooth is subjective. But when I say smooth, I mean really smooth. No vibration or unwillingness. Even down to around 800 or 900 rpm, although it's hard to tell just using the dash tach. This assumes fairly level ground, no passengers, warm engine.

Then again, 1250 is not radically greater than 1000, and the dash tach perhaps isn't a precision instrument, so maybe what you're observing and what I'm observing is not drastically different.

[1fixitman]

I pulse with about 1/3 throttle to 33 MPH then hit the EV(Electric Vehicle only) button I installed to force off the engine(less than 34 MPH) and feather the throttle to achieve a full glide, no arrow condition, but it is a hybrid so it is much easier to do this. The ScangageII is helpful too:)
Dwayne

[sonyhome]

(OFF TOPIC)

Should we take this over to the fuel meter thread?


Sorry, phased out of existence :)

Why not?

The OBD2 stuff is out the window, I want OBD1 or plain wire taps for the DelSol. For the CRV & OBD2, I have the ScanGauge.

Couple of notes:

I'm really slow at projects like that.
- This WE I stayed home, and got around to fix my CRV speakers
- ...and cut foam grill blockers for the CRV. Dunno if the foam worked, but got 30MPG @65MPH one way, and 26MPG @74MPH the other way (Short distances, flatish), without overheating apparently. Still warry to use that on long trips.

[sonyhome]

I try to use WOT as much as possible, but my car has a lot of low-end torque and it can be tough to moderate my acceleration at WOT. When I do, I shift at 1200 or 1400 RPM, entering the next gear at about 1000rpm. My pattern is as follows:
1st: A light, quick application of gas gets the car rolling enough for 2nd.
2nd: A quick stab gets the torque-strong car fast enough for 3rd or 4th, which are uselessly close in my car.
Depending on my current pattern, I may skip 3rd.
3rd, if I use it: A quick stab and I go from 15 to 20mph, and am ready for 4th.
4th: Now I get a chance to drop it to the floor for a couple seconds, at which point I'm going 25 or 30mph and am ready for 5th.
5th: On the floor until I reach the top of my pulse, then I begin P&G cycling.


Holycow, that's confirming my intuition is wrong and I need to review my red lights starts. Hondas don't really have much low end torque, they tend to have a longer power band and high-RPM limits with VTEC.

One more reason to hook up a PC to the car, to see where my gas is spent on my commute. i do get a lot of gliding, and I try to sync lights as much as possible.

I'll try to P&G in 5th gear (but I don't like -hum- slow accelerations in traffic).

BTW, I've found yet another way to hypermile: The AAA credit card gives 5% cash back on gas :-P

I not RPMs, flooring the accelerator, the injector duty cycle is close to 100%. So there's a max amount of gas in the piston. Is the explosion complete (gas fully used)? Is there always enough oxygen for the amount of gas pumped in the piston?

[R.I.D.E.]

My Del Sol was not VTEC, and I averaged about 42 MPG.

The worst you can do is hit the brakes.

When you use engine drag (car in gear) you are turning all fuel to the injectors off=DFCO=Decelerate Fuel Cut OFF.

Coasting is great, but if you have to stop or slow down a lot, it may be better to use DFCO. AS long as you can keep engine speed above 1000 RPM you have no fuel delivered to the injectors. Its the same as shutting the engine off without all the restart issues.

Try this. use a lower gear than you would use normally. Once you reach the point where more throttle really doesn't do much of anything, you have reached the best point for economy. Wot will use more fuel than this method. The ideal throttle percentage should be about 70%, could be less but not much more.

If you need greater acceleration use a lower gear. If you don't need greater acceleration, use the highest gear that you possibly can. When you want to maintain speed use the highest possible gear as soon as practical.

An example.

Light ahead is red and I know I will have to stop. I downshift through the gears as my engine speed approaches 1000 RPM.

5th gear down to 1000, then 4th down to 1000, etc.

To determine what RPM your enigne stops fuel shut off, just let off the gas in 5th gear and wait until you feel the surge when the fuel supply returns. In my car its about 860 RPM, but it is higher if you have the AC on. You can do the same in any gear, but it will tell you when the fuel shut off ends as far as engine RPM.

People do get frustrated with me when I spend a lot of time coasting to the light that is red. They fly by me and slam on the brakes, and sometimes they block my path through the light when I get there right at the point where it turns green. In that case I just keep using DFCO. I don't care what the other drivers do, I wont waste the gas idling or the gas it took to get to speed, when I throw it away braking for the light.

Remember DFCO is the same as shutting the engine off without the restart hassles, as long as you have to slow down more than you would coasting.
Coasting is alwats better as long as you do not have to use the brakes. Then DFCO is better.

regards
gary

[R.I.D.E.]

Also when using DFCO make sure your foot is not applying any pressure on the gas pedal or DFCO will not engage.

regards
gary

[cinnapatch]

What is P&G? No idea in North Carolina.

[dosco]

What is P&G? No idea in North Carolina.

Pulse and glide. A driving technique wherein you maintain an average speed - say 65 mph - by "pulsing" to 70 mph and then "gliding" (with the transmission in neutral) to 60 mph. Repeat.

There are variations on this theme. When gliding, the engine can be left to idle. The alternataive is to glide with the engine shut off, called "Engine Off Coasting" or "EOC."

I also have a 1999 Camry, mine has an automatic transmission. It is not possible to EOC because it will damage the transmission.

You should try P&G, I have a mixed commute from home to work, about 50% highway and 50% city, I'm getting about 33 mpg.

I noted your gaslog, if you have 90% highway commute, you could get much better mileage. On my last highway trip in the Camry, I achieved about 45 mpg using P&G driving.

[theholycow]

I not RPMs, flooring the accelerator, the injector duty cycle is close to 100%. So there's a max amount of gas in the piston. Is the explosion complete (gas fully used)? Is there always enough oxygen for the amount of gas pumped in the piston?

The computer won't let it run that rich. I have never managed to see more than 50% duty cycle. At low RPM, flooring it doesn't increase the duty cycle much vs. partial throttle. If I am light on the gas at a given RPM I might see 8% duty cycle, and if I floor it I might see 12%.

When you use engine drag (car in gear) you are turning all fuel to the injectors off=DFCO=Decelerate Fuel Cut OFF.

Coasting is great, but if you have to stop or slow down a lot, it may be better to use DFCO. AS long as you can keep engine speed above 1000 RPM you have no fuel delivered to the injectors.

If you need greater acceleration use a lower gear. If you don't need greater acceleration, use the highest gear that you possibly can. When you want to maintain speed use the highest possible gear as soon as practical.

Quoted just to reiterate. That was very well said. I would replace "may be" with "surely is".

Try this. use a lower gear than you would use normally. Once you reach the point where more throttle really doesn't do much of anything, you have reached the best point for economy. Wot will use more fuel than this method. The ideal throttle percentage should be about 70%, could be less but not much more.

Are you talking about your Honda in specific, or all cars? From actual measurements, I can tell you this is not universal. In my VW, injector duty cycle goes down when I go from 80% (maybe 70%, maybe 90%, my foot is just not that accurate) to WOT. I'm at a loss to explain it. At first I thought maybe the fuel rail pressure goes up at WOT, allowing a shorter duty cycle to feed in more fuel; but I can't argue with the gas pump, and using WOT (instead of that 80% target) has improved my FE.

To determine what RPM your enigne stops fuel shut off, just let off the gas in 5th gear and wait until you feel the surge when the fuel supply returns.

Also, note that the surge is barely perceptible, you really have to be looking for it, at least in my vehicles. It is gentler than, for example, the torque converter locking in my truck (which itself is almost impossible to feel).

[sonyhome]

My Del Sol was not VTEC, and I averaged about 42 MPG.

When you use engine drag (car in gear) you are turning all fuel to the injectors off=DFCO=Decelerate Fuel Cut OFF.


RIDE, thanks, I like your explanation!

However, for the CRV, it seems there is no DFCO, unless the Scangauge is wrong. Downhill in "D", I see 0.8GPH used, and up if the grade is steeper and the engine RPMs go up. Shifting to "2" or "1" increases consumption too (slower speed higher RPMs, duh! and no DFCO?).

What Del Sol model year were you driving? I have a '93 Si (D16z6 1.6L SOHC VTEC engine), there is also the S (1.5L non VTEC) and VTEC (160HP 1.6L DOHC VTEC). The S has the best fuel economy...

[sonyhome]

A driving technique wherein you maintain an average speed - say 65 mph - by "pulsing" to 70 mph and then "gliding" (with the transmission in neutral) to 60 mph.


For me P&G works better in the city with my DelSol car. My CR-V slows down too fast doing P&G on highway. Aero mods are better I thing for highway MPGs. (yet to be proven for my CRV)

[sonyhome]

I can tell you this is not universal. In my VW, injector duty cycle goes down when I go from 80% (maybe 70%, maybe 90%, my foot is just not that accurate) to WOT. I'm at a loss to explain it.

Also, note that the surge is barely perceptible, you really have to be looking for it, at least in my vehicles. It is gentler than, for example, the torque converter locking in my truck (which itself is almost impossible to feel).

Could it be that the computer switches mode in WOT, and computes the optimal duty cycle for acceleration, whereas at 80% it does what you tell it and runs too rich without returning any benefits in improved acceleration? (by too rich I means spend too much gas for no extra return, not spitting out unburned gas)

I'll try to look for that surge on my Del Sol. 875RPM is darn close to idle, no suprise there...

[theholycow]

However, for the CRV, it seems there is no DFCO, unless the Scangauge is wrong. Downhill in "D", I see 0.8GPH used, and up if the grade is steeper and the engine RPMs go up. Shifting to "2" or "1" increases consumption too (slower speed higher RPMs, duh! and no DFCO?).

The ScanGauge has to guess/calculate fuel usage. It can't measure GPH, there's nothing in the OBDII standard for that. I assume that it guesses based on O2 sensor reading combined with MAF, but if so then it's probably correct and your Honda has no DFCO. I have a hard time believing there's no DFCO on the CRV!

Could it be that the computer switches mode in WOT[...]

Could be. I don't know, but I hope to find out eventually.

[sonyhome]

The ScanGauge has to guess/calculate fuel usage. It can't measure GPH. I assume that it guesses based on O2 sensor reading combined with MAF, but if so then it's probably correct and your Honda has no DFCO. I have a hard time believing there's no DFCO on the CRV!

How can the SG compute MPG reliably and not be able to compute GPH? If what you say is true, then MPG results would be way off when coasting downhill (hard to prove, try downhill coasting for a full tank! LOL). People report that the SG should display "9999" in DFCO, right?

Well is there DFCO for other slush-box A/T vehicles in general?

[theholycow]

My slushbox GMC has DFCO, though it's not very useful -- it only comes on after ~10 seconds of non-DFCO engine braking (which uses lots of gas).

The inability to measure GPH is the main weakness that has held me back from getting a SG. That inability is confirmed, I'm 100% sure that OBDII and the SG does not support fuel rate measurement (though I'd still love to be proved wrong), and that the SG calculates it. I asked and researched a lot (here and elsewhere) until I was sure of it.

[dosco]

Aero mods are better I thing for highway MPGs. (yet to be proven for my CRV)

I don't think aero mods are *bad*, but I wouldn't put much stock in aero mods drastically reducing your fuel consumption. I recently did some calculations and found with a coefficient of drag of 0.28 the power consumed by drag at about 65 mph (assuming a standard atmosphere at 70 degrees F) is something like 16 hp.

Side note: assuming a coefficient of rolling resistance of 0.030, I calculated the power consumed by rolling resistance as something like 30 hp(!!).

Anyways, using P&G and some very mild aero mods (upper grill block) I have been able to hit up to 45 mpg. I'm not convinced that any single (or even a bunch of) aero mod would improve my mileage as drastically as implementing P&G.

I recently turned a buddy on to P&G and he has been hitting 50 and 60 mpg in his VW TDI.

[sonyhome]

Hollycow,

If DFCO activates after 10s, I would've seen it by now if it is either visible with the SG, or can be felt by the driver... Maybe rigging the DIY fuel rate monitor with the SG could help detect if the SG is lying...

Dosco,

I didn't mean P&G is not usefull on highway. I don't really know.

However, for my Gen'2 AT CR-V application I feel there's a lot of roll resistance and I suspect a lot of power lost thorugh aero at 75MPH: The drag coef. must be higher (like 0.30 maybe?) and the surface bigger too. I can't pump the tires much because of tire specs and SUV roll safety. All I know is P&G for an AT CRV on highway doesn't seem that feasible. Slows down too fast.

16HP on air friction sounds quite low. But then again power dissipation must be squared with speed or something, which would explain why a 120HP car can't go much past than 120MPH.

[PaleMelanesian]

According to this link (http://classicairshows.com/Education/Aerodynamics/SimpleAerodynamicsPartTwo.htm), the CR-V has a cd of .5! That's huge, combined with it's larger area.

P&G works on the highway, but still the slower the better, unless you drive an aerocivic. 45-60 P&G has given me 75mpg over 150 miles. Steady 65mph cruise gives me about 45mpg on that same segment.

and aerodynamic drag is proportional to the CUBE of velocity.

[sonyhome]

According to this link (http://classicairshows.com/Education/Aerodynamics/SimpleAerodynamicsPartTwo.htm), the CR-V has a cd of .5! That's huge, combined with it's larger area.


You might be quoting the coef for the Gen'1 CR-V (online mentions an unconfirmed 0.44). The Gen'2 can't be that high. Looking online, people quote 0.34... The Gen'3 CR-V Cx is quoted as 12% less than previously which would mean 0.30 or 0.29...

Gen'3 data:

One of the key elements behind the new Honda CRV’s excellent fuel economy is a 12 per cent lower drag coefficient. A significant part of that reduction is down to careful analysis of the air flow beneath the car. The use of strakes ahead of the front and rear wheels, cover panels beneath the cabin and a rear undercover decrease levels of turbulence.
The Honda CRV’s slippery body also keeps down wind noise, further helped by the curved shapes of the A-pillar cross section and door mirrors, and by using double seals around all the doors.


Gen'1: Cd 0.44, Up to 2001, boxy, smaller size
Gen'2: Cd 0.34, 2002-2006, more rounded, still boxy, bigger size
Gen'3: Cd 0.30, 2007 & up, very curvy body, probably less space inside now

If diesel was available in the US, the CR-V would sell with the 2.2L iCDTi engine and get 48.7MPG HW (Gen'2) and 49.6 HW (Gen 3)

[PaleMelanesian]

I dunno. That's just the first search result I found, and it doesn't look very reputable.

[monroe74]

The ScanGauge has to guess/calculate fuel usage. It can't measure GPH, there's nothing in the OBDII standard for that.

I'm pretty sure this is exactly true, but I want to add something to help people not get confused.

There are scantools that mechanics use, that plug into your OBDII port and report a lot of things. Some of the scantools report injector data. This seems to contradict what you said, but not really. I believe the folks who made the scantool paid a licensing fee to one or more car makers, and were given data about the proprietary method that maker uses to report injector data via the OBD port.

In other words, the data is not part of the OBDII standard, and therefore it's not easy to get at, even though it does exist, on some cars.

I think Scangauge decided to not to try to get at the data that way, presumably because it would have been difficult and added expense to the product. Maybe a lot of expense. I wouldn't be surprised if the car makers demand a licensing fee on a per-unit basis.

[monroe74]

I dunno. That's just the first search result I found, and it doesn't look very reputable.

A lot of cd data is here:

http://web.archive.org/web/20041204082513/www.teknett.com/pwp/drmayf/tbls.htm

I don't see the CRV, but maybe there's some similar vehicle, to make a comparison.

[theholycow]

I think Scangauge decided to not to try to get at the data that way, presumably because it would have been difficult and added expense to the product. Maybe a lot of expense. I wouldn't be surprised if the car makers demand a licensing fee on a per-unit basis.

You are most likely correct. It seems extremely expensive to get manufacturers' proprietary stuff like that.

And, here's proof that the SG definitely calculates/guesses fuel rate:
I EOC'd today on my commute and w/ engine off it was still reading 0.07 GPH and at 45mph it was only 477mpg.... this is with engine off!

[sonyhome]

[QUOTE=monroe74;106420]A lot of cd data is here:

http://web.archive.org/web/20041204082513/www.teknett.com/pwp/drmayf/tbls.htm
QUOTE]

Cool!
The Del Sol has a Cd of 0.42, but tiny frontal area. Could work on that!

The Ford Explorer has a Cd of 0.45, in line with what I found floating around for a Gen'1 CR-V, reported above.

[dosco]

and aerodynamic drag is proportional to the CUBE of velocity.

How do you figure that?

drag = 1/2 rho Cd S V^2

rho = air density

Cd = drag coefficient

S = area

V = velocity

[sonyhome]

In metric system:

rho: kg/m3
S: m2
V: m/s

Therefore drag = kg.m/s2 Are these the right units?

http://upload.wikimedia.org/math/7/2/d/72d686af4bdf6ffdff7928e3a20cb4b5.png
http://upload.wikimedia.org/math/1/d/6/1d6ff8f88450295e8a235a54d626ea78.png

The small v is just a unit vector that points the direction of movement in 3D space, doesn't change our computation for our application, rho=1.293 kg/m3. A means area, same as S (for surface) above.

from http://en.wikipedia.org/wiki/Drag_%28physics%29

Bottom line: 2X speed = 4X drag force, 3X = 9X force etc. Squared indeed. However, the power dissipated seems to be cubed... but you have to take into account the total time to travel a fixed distance, to figure out your net energy cost (you cube but then you actually go faster on that distance). Argh help.

BTW:

If you know the drag % of the underbody (or the wheels), and the drag of an air dam, you could tell if adding an air dam is likely to provide any benefits (increasing the frontal surface S, aka A, but reduces Cd by some amount)...

[dosco]

Bottom line: 2X speed = 4X drag force, 3X = 9X force etc. Squared indeed. However, the power dissipated seems to be cubed... but you have to take into account the total time to travel a fixed distance, to figure out your net energy cost (you cube but then you actually go faster on that distance). Argh help.


Yes, if I'd pulled my head out of my arse I would have taken the next step and calculated the work equation. Pale is correct. I'm a dullard. D'oh.


If you know the drag % of the underbody (or the wheels), and the drag of an air dam, you could tell if adding an air dam is likely to provide any benefits (increasing the frontal surface S, aka A, but reduces Cd by some amount)...

I think it's a bit more complicated than that, since the drag is a function of the entire body, and the interactions between the components is not linear. But I think your point has merit....

[sonyhome]

So I did a P&G experiment on my comute:

- Commute: 3 miles one way, with lights, 35PMH limit surface streets, return often with 1.5miles highway (next exit) and lights & stop sign. Night twice a week, drive to the gym, about 10 miles R/T surface street.

Experiment result

- Old method: Fast acceleration from light, 1st/2nd/3rd then coast down, 3rd or 4rth gear short pulse, then glide. Result is accelerations are short followed by long glides. Shifting usualy around 2500RPMs.

Result: 35MPG over 338 miles

- New method: Slow acceleration from light, 1st/2nd/3rd/4rth/5th, fast shifting to stay below 1500rpms, but anemic acceleration. I don't really do WOT acceleration, but my foot is pretty heavy after the 1st gear. End result is I spend more time accelerating than coasting.

Result: 35.22MPG over 191 miles

Bottom line:
Seems like both techniques work about the same in terms of fuel economy. However the new method supposedly with better MPGs is harder to drive I find because the car is less responsive. I am more likely to miss a light.

I wonder if I should try one run with normal driving, no P&G. I wonder if I'd fall back sub 30 or not, with somewhat controlled driving conditions. City EPA for my car is 25MPG (new rating) so I'm 40% overjust from P&G and some awareness of lights.