Prius T4 Speedo

[B&J]

I decided to check the speedo readings on my new Prius T4 so I went on the motorway and set cruise control to 70mph on a level stretch. I checked this speed against my SatNav. This showed 64mph! The following day I checked again against a different, hand held GPS devicewhich also showed 64mph against an indicated 70mph on the dash. This is a 10% inaccuracy give or take a decimal point.

I think this is disgraceful; with a completely electronic display I don't believe there should be any inaccuracy.

I have some questions which I hope others can answer.

1. Which should I believe - the speedo or the GPS?

2. If it's the GPS, is it possible to get the speedo recalibrated so it's accurate?

3. Most important for me - if the speedo is 10% out, doesn't this mean that my mileage is being incorrectly recorded i.e. when it shows I've done 110 miles, have I really only done 100?

If this is the case, then over the life of the car there will be literally thousands of miles erroneously recorded, with a severe impact on trade in value.

Thanks

B&J

[Glyyde]

I think it is well known that Toyota (or most other make) overestimate the speedometer by about 5-10%, so this is completely normal. The mileage, however, will be accurate as far as I know.

[Jan&tone]

I think it is well known that Toyota (or most other make) overestimate the speedometer by about 5-10%, so this is completely normal. The mileage, however, will be accurate as far as I know.

+1. all manufacturers speedo's are between 5 and 10pct fast. Not sure whether it's compulsory, but better to be over than under!

[mandon75]

It was the same in my mondeo and my wife's toyota yaris, I thought it was the Tom Tom that was out of spec, I didn't know the manufacturers had a variance too. Never been flashed by a camera doing 70mph in my Prius. Just go by what your speedo says and not your sat nav :-)

[Ancient Nerd]

Based on my phone's GPS, mine is almost exactly 10% over at all speeds from 30 mph to 70mph.

I checked by setting cruise control to a speed that gave a steady 30 mph on the phone GPS, and then reading the speedometer. Then repeated for 40,50,60,70.

This is very different from previous cars, where the percentage has varied significantly with speed.

[timberwolf]

It's an EU regulation (and there was probably a UK one before) that the speedometers can over-read by 10% + a few km but not to under-read.

I can understand why the legislation may have originally been drafted and that was most likely because of limitations in the mechanical technology although you can't excluding lobbying by manufacturers on cost grounds, but as the technology moved to electronics it is puzzling why the speedometer accuracy wasn't improved? The only reasons I can think of for purposely making the speedometer over-read by the full amount permitted by the regulation is to deceive us.

[Grumpy Cabbie]

Check this link out - esp the section 'international agreements' as this will explain all

http://en.wikipedia.org/wiki/Speedometer

[Sagitar]

It's an EU regulation (and there was probably a UK one before) that the speedometers can over-read by 10% + a few km but not to under-read.

I can understand why the legislation may have originally been drafted and that was most likely because of limitations in the mechanical technology although you can't excluding lobbying by manufacturers on cost grounds, but as the technology moved to electronics it is puzzling why the speedometer accuracy wasn't improved? The only reasons I can think of for purposely making the speedometer over-read by the full amount permitted by the regulation is to deceive us.

As long as you have tyres that wear and flex and which vary in diameter because of things like the combination of wheel diameter and tyre profile. And while you have differential rotation of inner and outer wheels in order to go round corners, and tyres that scrub and hop for various reasons, there will always be error in a device that depends upon relating the rotation of the gear box output shaft to the rotation of the wheels.

The point of setting up a deliberate error in the speedo is to ensure that all the other inadvertent errors never result in a reading that is lower than your actual speed.

I think the time may come when speed is measured in some other way, using GPS for example, or the Doppler effect, but it hasn't arrived in the motoring legislation yet.

[ian rooke]

Hi I have had 2 Gen2s and 1 Gen3 and all have been 10% over on the speed so it seems the norm.

[timberwolf]

It's an EU regulation (and there was probably a UK one before) that the speedometers can over-read by 10% + a few km but not to under-read.

I can understand why the legislation may have originally been drafted and that was most likely because of limitations in the mechanical technology although you can't excluding lobbying by manufacturers on cost grounds, but as the technology moved to electronics it is puzzling why the speedometer accuracy wasn't improved? The only reasons I can think of for purposely making the speedometer over-read by the full amount permitted by the regulation is to deceive us.

As long as you have tyres that wear and flex and which vary in diameter because of things like the combination of wheel diameter and tyre profile. And while you have differential rotation of inner and outer wheels in order to go round corners, and tyres that scrub and hop for various reasons, there will always be error in a device that depends upon relating the rotation of the gear box output shaft to the rotation of the wheels.

The point of setting up a deliberate error in the speedo is to ensure that all the other inadvertent errors never result in a reading that is lower than your actual speed.

I think the time may come when speed is measured in some other way, using GPS for example, or the Doppler effect, but it hasn't arrived in the motoring legislation yet.

Mmmm, I don't think that explains why as the technology has apparently improved that the speedometer accuracy has decreased i.e. it has moved closer to the upper limit allowed for over-read.

[Sagitar]

it has moved closer to the upper limit allowed for over-read.

I would be interested in a reference to the evidence that you have for that.

[B&J]

Interesting debate, people. I knew about the allowed 10% error - I just don't think that manufacturers need to take 100% advantage of it.

However, apart from Glydde, there hasn't yet been any other comment about whether recorded mileage is correct with such an inaccurate speedo. Is Glydde right that

The mileage, however, will be accurate as far as I know.

?

Thanks

B&J

[Grumpy Cabbie]

I would say not. On my taxi there is a 100% accurate speedo built into the taxi meter. This matches the satnav whereas the car speedo is out. As the taxi meter is connected to the same pulses as the car speedo I imagine the car is actually adding the error later but would imagine the mileage to be correct.

Easy way to check - reset trip meter and go 50 miles with sat nav. If they both car and sat nav tally then you've answered your question.

[timberwolf]

it has moved closer to the upper limit allowed for over-read.

I would be interested in a reference to the evidence that you have for that.

No sorry I don't have any hard evidence.

It's only my own observation over time, it is not recorded, so quite rightly you'd be able to refute my claim as me getting muddled up.

I can still recall that my earliest hand-held GPS was a Garmin II+. I've know about the Speedometer and GPS show different speeds from back then, which doesn't prove anything.

I also sometimes used to connect the GPS with a serial cable to a Psion S3 PDA. I don't think there was turn-by-turn navigation in those days, at least not within my budget, and I used to program in the limited number of way-points manually and it used to take ages. Happy tinkering

[Sagitar]

No sorry I don't have any hard evidence.

It's plausible that manufacturers might be using more of the tolerance on a consistent basis and I was hoping that you might be able to quote some research.

I think there are good engineering and safety reasons for using as much as possible of the tolerance and it need not be because there is some kind of conspiracy. As I said before, it is simply not possible to set up a speedo in such a way that it will remain accurate over a prolonged period. Vehicles that have accurate speed measuring devices, tachometers for example, depend upon regular calibration to account of the changes that occur over time.

In a set up that is not recalibrated regularly, it makes sense to use as much as possible of the tolerance initially so that the period before the device goes out of tolerance is as long as possible.

Older speedos, depending on mechanical drives and eddy current effects, were relatively inconsistent and so part of the tolerance was taken up in accommodating this inconsistency. If the inconsistency is diminished by using e.g. a pulse generator feeding an electronic counter, then more of the tolerance can be used to accommodate wear and other factors.

So there could be good reason why some speedos of the past would have less error initially than speedos of today. I had hoped that you might be about to tell me of information that proved the case. It would form a good basis for arguing that it might be time to reduce the tolerance.

[Opifex]

I would say not. On my taxi there is a 100% accurate speedo built into the taxi meter. This matches the satnav whereas the car speedo is out. As the taxi meter is connected to the same pulses as the car speedo I imagine the car is actually adding the error later but would imagine the mileage to be correct.

Easy way to check - reset trip meter and go 50 miles with sat nav. If they both car and sat nav tally then you've answered your question.

Interesting that the taxi meter reads the exact speed if it's using the same pulse source as the speedometer. I wonder how it is compensating for the tyre variations that have been mentioned.

An alternative to measuring distance with the satnav is to use the distance markers down the centre of motorways. Either way you have to choose a straight road. Going clockwise round the M25 is further than going the other way.

[johalareewi]

... would imagine the mileage to be correct.

Easy way to check - reset trip meter and go 50 miles with sat nav. If they both car and sat nav tally then you've answered your question.

Have done that many times and can confirm my prius's gen3 trip meter and satnav are within 0.1 miles of each other after test drives (even long ones > 150 miles).

[acetip]

I dont own a Prius,though i would like one,but my pensions dont strech to do the change from the Auris, so i dont know if any of the Prius has a rev counter.

i managed to find out by checking several different road test the road speed per thousand revs in each gear of my Auris.In top gear (5th) it should read 20 mph per thousand revs.so at 60 mph the revs should be 3000.

i cant remember the exact fiquire but the speedo reading is several mph above above that 60 mph.

Maybe the rev counter is more accurate.I have a TomTom so i shall have to compare readigs on that.

On motorways i tend use the rev couner to set up the cruiise control.other times i use the speedo,thinking that i am well within any speed limit and not going to activate any speed cameras accidently.

[B&J]

. . .i dont know if any of the Prius has a rev counter.

There is no rev counter on a Prius or an Auris Hybrid.

Cheers

B&J

[Ancient Nerd]

There is no rev counter on a Prius or an Auris Hybrid.

Anyway, it would be meaningless in this context, because of the continuously variable gearing.

But you are right with a conventional car, automatic or manual. I think it was my second car, a 1956 Riley Pathfinder, that had a very iffy speedometer. So I used the rev counter all the time. It helped that it did exactly 20 mph per 1000 rpm.

That was an interesting car, but I wouldn't want to go back to it.:

• A sports saloon with a bench front seat. The obvious column change wasn't an option, as that was distinctly non-sporty.So it had a cutout in the seat by the drivers door to house a short gearstick,. With a column change linkage under the driver's seat! It had the side effect that you could get a gear lever up your trouse leg as you got out of the car, making for a less than graceful exit..
• No power steering, so a big steering wheel, with about 2 inches of free play (well within spec & MOT requirements).
• No servo for the drum brakes.
• 110 bhp, which was a lot at the time.
• Umbrella handbrake - Truly awful.
• You think the standard Prius headlights are bad!
• A starting handle. Emergencies only, but you really don't want to know!
• Optional radio. Big with valves, so most of it sat under the bonnet.

[acetip]

You cant beat a bench seat when you have women beside you,very cnvenient.nots so good on he corners.my old Victor FB had one and a 3 speed column change,but soon converted that to 4 on the floor,pretty easy to do as well.

Drum brakes as the Victor had, are no worse as braking power goes compared to disc brakes of the same friction area.but dont run as cool.

But it had no servo.both Girling and Lockheed made servo kits you could fit to almost any car.i got a Lockheed one and used pipes etc off a i think was called a Victor XR which came with a servo,what a differce.you could lean down and brake to a shuddering halt just using one finger.things we do when we are (was) young.

Like most cars then bodywork rust was a problem.the Victor FA was terrible,it had a warraty that it would not rust for a 100000 miles or to the edge of the dealers forecourt, which ever occured first.

But my old victor was one of those cars you knew would get you there come hell or high water.it never missed a beat.

It went through several floods in its time,that more expensive cars would not dare go.never had a drop of water get into it.Not a bad engine cosidering it had only 3 main bearing crankshaft,and bypass oil filter.

[DrCez]

Interesting debate, people. I knew about the allowed 10% error - I just don't think that manufacturers need to take 100% advantage of it.

However, apart from Glydde, there hasn't yet been any other comment about whether recorded mileage is correct with such an inaccurate speedo. Is Glydde right that

The mileage, however, will be accurate as far as I know.

?

Thanks

B&J

If I reset Trip A or Trip B and maintain a constant speed (based on the main speedo) for a short period, the average trip speed (AVS) is always lower than that indicated by the speedo (e.g. travelling at 50 mph indicated the AVS might settle down at around 45 mph). This is another experiment that supports the idead that the 10% error is just added in (assuming the car uses the same signal for both measurements).

As others have said I find that the mileage is accurate when compared with sat nav or tools like Google Maps. Fuel consumption, however, is not (out by about 6%) and Range is literally miles out.

10% seems overly generous for a modern vehicle - my back of an envelope calculations suggest you'd need the tyre diameter to change pretty dramatically to get anywhere near misreading by that amount; I would have thought a 2-3% over-read would easily be enough tolerance to ensure the speedo never under-reads.

[johalareewi]

If I reset Trip A or Trip B and maintain a constant speed (based on the main speedo) for a short period, the average trip speed (AVS) is always lower than that indicated by the speedo (e.g. travelling at 50 mph indicated the AVS might settle down at around 45 mph). This is another experiment that supports the idead that the 10% error is just added in (assuming the car uses the same signal for both measurements).

Nice one. Hadn't thought of doing that.

I feel an experiment coming on with the prius average speed meter and the satnav...

[B&J]

Easy way to check - reset trip meter and go 50 miles with sat nav. If they both car and sat nav tally then you've answered your question.

I've finally been able to do this test and the result was that the trip meter showed a 1% difference to the GPS. I can definitely live with this now I know that the car is not going to record thousands of virtual miles travelled. But as pointed out elsewhere on this thread, if the odometer can be within 1% of actual, there is no necessity for the speedo to be so far out, even to save huge numbers of drivers from getting speeding tickets. 2% - 3% overread would be quite adequate.

Thanks for all the input and interest

B&J

[johalareewi]

If I reset Trip A or Trip B and maintain a constant speed (based on the main speedo) for a short period, the average trip speed (AVS) is always lower than that indicated by the speedo (e.g. travelling at 50 mph indicated the AVS might settle down at around 45 mph). This is another experiment that supports the idead that the 10% error is just added in (assuming the car uses the same signal for both measurements).

Nice one. Hadn't thought of doing that.

I feel an experiment coming on with the prius average speed meter and the satnav...

Compared the satnav average speed against the prius trip average speed over 80 miles.

Satnav and Prius were within 2mph of each other on average speed and within 1 mile on distance travelled (new satnav doesn't have tenths of mile on trip meter). The average speed readings were identical until I had to stop for traffic lights (after 40 miles) :( I guess the satnav and prius handle idling slightly differently.

So it does indeed look like the Prius knows pretty accurately how fast it is going but inflates the displayed speedo figure by a good %age.