Computers don't lie - or do they?

HumberMart · 7th November 2007, 15:43

I drove to Leeds and back this morning from Scunthorpe.

At the moment, there are 50 mph roadworks on the M18, with an 'Average Speed Check'.

So I thought I'd be smart and set the A trip on entry to the speed restriction zone, and monitor my average speed.

Recognising the speedo reads about 4.5 mph over the gps readings, I generally run through such road works at an indicated 55 mph.

Surprisingly, despite maintaining 55 mph, for the whole zone, the average speed was shown by the display as 50 mph!! (Ironically very close to the gps speed, but obviously coincidental).

Later in the journey I set the cruise to 75, for 10 miles, reset the trip, and sure enough, it showed an average of only 70 mph!

Now I changed the wheels recently to Askenlads 18" ones, and I did have to raise the jack slightly to get the new ones on after removing the standard 17's (both sets of tyres were almost new, so no tread differences). So I did wonder if the TR wheels have a very slightly longer rolling circumfrance, which would I guess, lower the speed readings. But this can't be true, 'cos it would also affect the speedo as well as the average speed, and the speedo is just as optimistic as it was with the old wheels.

Another consideration is the accuracy of the milo. I've checked this against the gps, and it does read less, but only by 2%. Although I guess a gps reading may under read itself, due to the way it 'joins the dots' to read distance, thus will to some extent shorten all distances.

Has anyone else spotted such a difference between speedo and the average speed computer

Interestingly, due to my current concern regarding how poor the economy still seems on my 2.2. I monitored the mpg separately on the trip to and from Leeds. Going to Leeds it did only 40.6 mpg, but 49.5 on the return leg!!!

I drove at the same speeds, but went over with an empty tank, and came back with a full one, so if anything it should have been less economic on the return. However the wind was quite strong, and it was more with the wind coming back. Could the wind make such a difference?

Due to the low average speed reading, I'm now wondering if this could be also causing low mpg readings, so I'm going to do another fill-up to fill-up check, and compare that with the computer on this new tank full.

If you're still with me at this point, you need to get out more, and I need to take my anorack off

dylang652 · 7th November 2007, 16:00

Pottsy has done a thread which calibrates the speedo so that it doesn't underread by the 6 or 7% that they all seem to.

He reckons that does not impact the mileometer readings, so it may be that the average speed is calculated by the computer as "distance traveled" (from the mileometer) divided by "time taken" hence the more accurate reading

HumberMart · 7th November 2007, 16:17

Yes dylang, that makes perfect sense!

And it would even account for the fact that the calculated average is slightly less than the real gps average, if the milo is under reading by 2%.

Cheers

HumberMart · 7th November 2007, 16:21

............although unfortunately it wouldn't help convince me that I got more than 40.6 mpg on the way to Leeds, since I assume the mpg is calculated from the same 'distance travelled' variable

pjr · 7th November 2007, 18:43

Humbermart,
I've noticed exactly the same feature of the average speed shown on the trip computer - It seems to be the 'true' average speed - i.e. set cruise to 70, average speed will show 65ish.

TTDegs · 7th November 2007, 19:44

I guess it could be to do with how the average speed, and displayed speed are calculated?

Educated guess says that the speedo 'updates' every (say) half second.
And it works out the speed by taking the number of wheel revolutions in that half second.

I ought to sit down with a calculator and work out how many times the wheel goes round in half a second at 70mph, but I'm going out on a limb and saying that it will be something point 19562542983 (or some other equally non exact number

)

However, to get average speed over (say) 20 mins, the rounding from above will be much less pronounced.

Oh booger it - my OCD is reaching for the calculator...

TTDegs · 7th November 2007, 19:57

ok...

rolling circum of a Civic tyre is 195.65838cm
This is by taking an 18" wheel (times 2.46 cm per inch), and 2 times tyre wall (225mm times 40%)
and pi at 3.1415926

so we're already got a load of rounding going on

and that's without worrying about how much the tyre has compressed (not a true circle).

so.
at 80mph (128kmh to get everything metric!)
128kmh = 2.13333333333 km per minute = 0.03555555555 km per second

= 35.5555555 metres per second
= 3555.55555 cm per second.

meanwhile, our tyre is 195.65838 cm circum

which means that each seond, at 80mph, our tyre is making 18.172234 revolutions

or 9.086117 per half second.

so you can see how difficult it might be to get that exactly right!

TTDegs · 7th November 2007, 19:58

of course, all the above is a bit of rubbish, as I think the sensor comes from the engine instead of the tyres - but the theory is sound!

ie counts how many times something revolves per half second and translate that into speed...

TTDegs · 7th November 2007, 20:02

but back on the theory

in 20 mins, at the same 128kmh, you have done 42.6666666 km

or 42666.666666666 metres
or 4266666.6666666 cms

divide that by the 195 odd cm circum

= 21806.715 tyre rotations.

now, miscounting this as 21807 over the 20 mins isn't going to make much difference.

miscounting the 9.086117 as 9.1 in half a second, however, will make a much bigger difference.

phew!

anyone still there?

Hello?

HumberMart · 7th November 2007, 23:30

Gosh TT, and I thought I was crossing the attention span watershed of the average Civic owner

Sorry I've just been distracted by the thought that if this forum was reliating to the Civic of 10 years ago, it would be sponsored by Saga

HumberMart · 7th November 2007, 23:31

WHEEL NUTS ARE ACE

dylang652 · 8th November 2007, 13:44

TT - no wonder your "number of posts" figure is high if you take 4 posts to explain something as simple how we calculate speed!

HumberMart · 14th November 2007, 14:46

Well I've now refilled after an indicated 420 miles to check on the 'real' mpg, rather than the computer's suggestion. (Although I'm still reliant to some extent on the cars milo and Tesco's calibration of their diesel pump!).

Well took 43.6 litres to fill to the brim as before, so on the face of it, it's done 43.8 mpg, and the computer said "42.9 mpg"

Not a huge margin, but at least I can console myself to at least some of the lower than hoped for mpg (after 3700 miles) is down to computer error.

Of course if I take into account the apparent 2% under-reading of the milo against the gps, then that would suggest 44.7 mpg is nearer the mark, and that makes it at least on par with wot I got from the previous 520D.

Mystifying that some civic's mpg computer seems to be optimistic by quite a margin, yet mine (and a few others) are slightly pessimistic. Or does the 'space ship' have a 6th sense (like dogs), and assumes the owners character!!!!

Pottsy · 14th November 2007, 15:38

The fuel economy figure depends on the amount of fuel actually squirted being equal to the amount of fuel that the computer tells the injectors to squirt. Given the tiny quantities involved and the wild effects of temperature and mechanical inaccuracy, the mpg figure is going to wander.

The odometer (after the famous US court case) should be +/- 2.5%. But it tends to be nearer the +2% mark. The speedo seems to be +5%, so the overall effect is that the speedo over-reads by 7%, the odo by 2%, the mpg varying but on average +2% and the average speed (odo/time) +2%.

7th November 2007, 15:43	#1 (permalink)
HumberMart Supporter Locking Wheelnut Join Date: 31st August 2007 Location: North Lincolnshire Posts: 112 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	Computers don't lie - or do they? I drove to Leeds and back this morning from Scunthorpe. At the moment, there are 50 mph roadworks on the M18, with an 'Average Speed Check'. So I thought I'd be smart and set the A trip on entry to the speed restriction zone, and monitor my average speed. Recognising the speedo reads about 4.5 mph over the gps readings, I generally run through such road works at an indicated 55 mph. Surprisingly, despite maintaining 55 mph, for the whole zone, the average speed was shown by the display as 50 mph!! (Ironically very close to the gps speed, but obviously coincidental). Later in the journey I set the cruise to 75, for 10 miles, reset the trip, and sure enough, it showed an average of only 70 mph! Now I changed the wheels recently to Askenlads 18" ones, and I did have to raise the jack slightly to get the new ones on after removing the standard 17's (both sets of tyres were almost new, so no tread differences). So I did wonder if the TR wheels have a very slightly longer rolling circumfrance, which would I guess, lower the speed readings. But this can't be true, 'cos it would also affect the speedo as well as the average speed, and the speedo is just as optimistic as it was with the old wheels. Another consideration is the accuracy of the milo. I've checked this against the gps, and it does read less, but only by 2%. Although I guess a gps reading may under read itself, due to the way it 'joins the dots' to read distance, thus will to some extent shorten all distances. Has anyone else spotted such a difference between speedo and the average speed computer Interestingly, due to my current concern regarding how poor the economy still seems on my 2.2. I monitored the mpg separately on the trip to and from Leeds. Going to Leeds it did only 40.6 mpg, but 49.5 on the return leg!!! I drove at the same speeds, but went over with an empty tank, and came back with a full one, so if anything it should have been less economic on the return. However the wind was quite strong, and it was more with the wind coming back. Could the wind make such a difference? Due to the low average speed reading, I'm now wondering if this could be also causing low mpg readings, so I'm going to do another fill-up to fill-up check, and compare that with the computer on this new tank full. If you're still with me at this point, you need to get out more, and I need to take my anorack off Last edited by HumberMart; 7th November 2007 at 15:49.

14th November 2007, 14:46	#13 (permalink)
HumberMart Supporter Locking Wheelnut Join Date: 31st August 2007 Location: North Lincolnshire Posts: 112 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	MPG Reality Check Well I've now refilled after an indicated 420 miles to check on the 'real' mpg, rather than the computer's suggestion. (Although I'm still reliant to some extent on the cars milo and Tesco's calibration of their diesel pump!). Well took 43.6 litres to fill to the brim as before, so on the face of it, it's done 43.8 mpg, and the computer said "42.9 mpg" Not a huge margin, but at least I can console myself to at least some of the lower than hoped for mpg (after 3700 miles) is down to computer error. Of course if I take into account the apparent 2% under-reading of the milo against the gps, then that would suggest 44.7 mpg is nearer the mark, and that makes it at least on par with wot I got from the previous 520D. Mystifying that some civic's mpg computer seems to be optimistic by quite a margin, yet mine (and a few others) are slightly pessimistic. Or does the 'space ship' have a 6th sense (like dogs), and assumes the owners character!!!!

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads for: Computers don't lie - or do they?
Thread	Thread Starter	Forum	Replies	Last Post
Computers vs Cars	beepee	Any non Civic chat here please!	5	23rd October 2007 00:32


This thread is about: Computers don't lie - or do they?, it's in General Discussion at the Honda Civic forum Civinfo; I drove to Leeds and back this morning from Scunthorpe. At the moment, there are 50 mph roadworks on the M18, with an 'Average Speed ...

7th November 2007, 16:00	#2 (permalink)
dylang652 Supporter Triangular Exhaust Join Date: 6th September 2007 Location: Northampton Posts: 210 Thanks: 3 Thanked 0 Times in 0 Posts iTrader: (0)	Pottsy has done a thread which calibrates the speedo so that it doesn't underread by the 6 or 7% that they all seem to. He reckons that does not impact the mileometer readings, so it may be that the average speed is calculated by the computer as "distance traveled" (from the mileometer) divided by "time taken" hence the more accurate reading

7th November 2007, 16:17	#3 (permalink)
HumberMart Supporter Locking Wheelnut Join Date: 31st August 2007 Location: North Lincolnshire Posts: 112 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	Yes dylang, that makes perfect sense! And it would even account for the fact that the calculated average is slightly less than the real gps average, if the milo is under reading by 2%. Cheers

7th November 2007, 16:21	#4 (permalink)
HumberMart Supporter Locking Wheelnut Join Date: 31st August 2007 Location: North Lincolnshire Posts: 112 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	............although unfortunately it wouldn't help convince me that I got more than 40.6 mpg on the way to Leeds, since I assume the mpg is calculated from the same 'distance travelled' variable

7th November 2007, 18:43	#5 (permalink)
pjr Type S(cotland) Magic Rear Seat Join Date: 24th March 2007 Location: Perth, Scotland Posts: 556 Thanks: 9 Thanked 18 Times in 16 Posts iTrader: (0)	Humbermart, I've noticed exactly the same feature of the average speed shown on the trip computer - It seems to be the 'true' average speed - i.e. set cruise to 70, average speed will show 65ish.

7th November 2007, 19:44	#6 (permalink)
TTDegs Super Moderator Civinfo master Join Date: 4th August 2006 Location: Cardiff Posts: 5,634 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	I guess it could be to do with how the average speed, and displayed speed are calculated? Educated guess says that the speedo 'updates' every (say) half second. And it works out the speed by taking the number of wheel revolutions in that half second. I ought to sit down with a calculator and work out how many times the wheel goes round in half a second at 70mph, but I'm going out on a limb and saying that it will be something point 19562542983 (or some other equally non exact number ) However, to get average speed over (say) 20 mins, the rounding from above will be much less pronounced. Oh booger it - my OCD is reaching for the calculator...

7th November 2007, 19:57	#7 (permalink)
TTDegs Super Moderator Civinfo master Join Date: 4th August 2006 Location: Cardiff Posts: 5,634 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	ok... rolling circum of a Civic tyre is 195.65838cm This is by taking an 18" wheel (times 2.46 cm per inch), and 2 times tyre wall (225mm times 40%) and pi at 3.1415926 so we're already got a load of rounding going on and that's without worrying about how much the tyre has compressed (not a true circle). so. at 80mph (128kmh to get everything metric!) 128kmh = 2.13333333333 km per minute = 0.03555555555 km per second = 35.5555555 metres per second = 3555.55555 cm per second. meanwhile, our tyre is 195.65838 cm circum which means that each seond, at 80mph, our tyre is making 18.172234 revolutions or 9.086117 per half second. so you can see how difficult it might be to get that exactly right!

7th November 2007, 19:58	#8 (permalink)
TTDegs Super Moderator Civinfo master Join Date: 4th August 2006 Location: Cardiff Posts: 5,634 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	of course, all the above is a bit of rubbish, as I think the sensor comes from the engine instead of the tyres - but the theory is sound! ie counts how many times something revolves per half second and translate that into speed...

7th November 2007, 20:02	#9 (permalink)
TTDegs Super Moderator Civinfo master Join Date: 4th August 2006 Location: Cardiff Posts: 5,634 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	but back on the theory in 20 mins, at the same 128kmh, you have done 42.6666666 km or 42666.666666666 metres or 4266666.6666666 cms divide that by the 195 odd cm circum = 21806.715 tyre rotations. now, miscounting this as 21807 over the 20 mins isn't going to make much difference. miscounting the 9.086117 as 9.1 in half a second, however, will make a much bigger difference. phew! anyone still there? Hello?

7th November 2007, 23:30	#10 (permalink)
HumberMart Supporter Locking Wheelnut Join Date: 31st August 2007 Location: North Lincolnshire Posts: 112 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	Gosh TT, and I thought I was crossing the attention span watershed of the average Civic owner Sorry I've just been distracted by the thought that if this forum was reliating to the Civic of 10 years ago, it would be sponsored by Saga

7th November 2007, 23:31	#11 (permalink)
HumberMart Supporter Locking Wheelnut Join Date: 31st August 2007 Location: North Lincolnshire Posts: 112 Thanks: 0 Thanked 1 Time in 1 Post iTrader: (0)	WHEEL NUTS ARE ACE

8th November 2007, 13:44	#12 (permalink)
dylang652 Supporter Triangular Exhaust Join Date: 6th September 2007 Location: Northampton Posts: 210 Thanks: 3 Thanked 0 Times in 0 Posts iTrader: (0)	TT - no wonder your "number of posts" figure is high if you take 4 posts to explain something as simple how we calculate speed!

14th November 2007, 15:38	#14 (permalink)
Pottsy Administrator Civinfo master Join Date: 10th April 2006 Location: Leics Posts: 5,905 Thanks: 26 Thanked 234 Times in 126 Posts iTrader: (0)	The fuel economy figure depends on the amount of fuel actually squirted being equal to the amount of fuel that the computer tells the injectors to squirt. Given the tiny quantities involved and the wild effects of temperature and mechanical inaccuracy, the mpg figure is going to wander. The odometer (after the famous US court case) should be +/- 2.5%. But it tends to be nearer the +2% mark. The speedo seems to be +5%, so the overall effect is that the speedo over-reads by 7%, the odo by 2%, the mpg varying but on average +2% and the average speed (odo/time) +2%.