Speedometer appears to be under-reading, who can fix it?

TouchingVirus

This is going to sound like a pile of shíte but hear me out...I've been done for driving at 59km/h in a 50km/h zone last Saturday. It's Cork Street in Dublin and those that know it as well as I do know it has a Garda speed van on it at least twice a week. Since this was only 7 days ago and I drive my car maybe twice a week I've a full memory of the event, the van's location etc and am full sure that like pretty much every time I drive down that road I had the cruise control on at 50km/h to avoid a heavy foot getting me some points. I made no attempt to slow down, after all I was doing 50

So I'm gutted I got the notice yesterday but figured something was amiss. I took the car out today and with my dad using a GPS app on his phone I had some interesting results: When I'm doing 40kph on the speedo it's really 48, when doing 50 it's 58/59, 60 => 68, 80 => 88, 100 => 109; I didn't do 120, this was just an N road :pac: We repeated the test several times to be sure and the difference was fairly constant. I'm running stock wheels so there's no trickery on that front. Clearly something's wrong with it under-reading so much but does anybody know somebody who can fix it?

Before anybody thinks I'm working some sort of angle here I don't think I'll bother appealing the points - it doesn't change the fact I was speeding and while 3 points is sickening risking 5 for an appeal isn't worth it IMO. I'm just looking not to get collared again for speeding when I think I'm fine; eventually I'll forget to shave 10kph off my current reading in my head and this'll happen again.

Special Circumstances

What car.
What size wheels.
What size tyres.

TouchingVirus

E46 M3. 19" OEM CSLs. F:225/40/19 R: 255/35/19

challengemaster

http://forum.e46fanatics.com/showthread.php?t=999740

ironclaw

If you had a bit of cash, I'd be interested in seeing how this could be challenged. If the car was absolutely stock, and you were willing to swear this legally, and have it independently assessed and again, sworn legally that the fault was unknown to you (e.g. ECU Error) Would you have a case? There is no onus on the State for their equipment to be proved calibrated, accurate or in working order. Therefore, can it be expected that the accused has all their equipment in the right order? Its quite possible to buy a car, and unknown to you, have a dodgy speedo. Granted we'd all have a fair idea of 30km/h versus 80km/h, but 30km/h vs 50km/h would be harder to know. 35km/h on the Quays would be enough to have you nabbed.

I'm not read in law but there is probably a case if you had enough moola to pursue it. I know for a fact cases like this in the States are heard widely.

TouchingVirus

I hear what you're saying but spending a lot of cash and risking 5 penalty points in front of a judge isn't something I'm up for. The way I see it is it was 59km/h so I can't argue on the grounds that I wasn't speeding and arguing on the grounds that I didn't know I was speeding seems doomed to failure. I found out last night that some cars have their needle slip and the E46 has a cluster test mode that will show your speed according to the ECU. I can use that to compare to my needle reading to ECU and see if there's a discrepancy or if this is an actual internal fault that I need to take out a hexeditor or go to BMW to fix.

Skatedude

just download a gps speedometer for your phone, it will confirm your speedo readings. best done on a straight road.

visual

Car speedo was set for stock size wheels changing the diameter effects both gear ratio and speed reading.

Bit like wearing high heel shoes makes you taller

visual

Car speedo was set for stock size wheels changing the diameter effects both gear ratio and speed reading.

Bit like wearing high heel shoes makes you taller

TitianGerm

Could you look to put a car on a rolling road ? This would be able to give more accurate speed measure than a phone GPS

CiniO

visual wrote: »

Car speedo was set for stock size wheels changing the diameter effects both gear ratio and speed reading.

Bit like wearing high heel shoes makes you taller

How can changing wheel size affect gear ratio?

Gear ratio depends on amount of teeth on cogwheels inside gearbox.

TheChizler

emmetkenny wrote: »

Could you look to put a car on a rolling road ? This would be able to give more accurate speed measure than a phone GPS

As long as you're keeping a constant speed on the flat the GPS derived speed will be correct.

TitianGerm

TheChizler wrote: »

As long as you're keeping a constant speed on the flat the GPS derived speed will be correct.

I always found the GPS on phones to be a little off or inconsistent between different phones. I tried it before where we were using an iPhone 5 and a Samsung S4 at the same time and both were 3/4 KPH different in there speeds.

ironclaw

emmetkenny wrote: »

I always found the GPS on phones to be a little off or inconsistent between different phones. I tried it before where we were using an iPhone 5 and a Samsung S4 at the same time and both were 3/4 KPH different in there speeds.

Actually depends how the programmer worked with the data being fed from the GPS itself, and how accurate they were willing to accept. This, and rounding, can greatly affect accuracy and hence the numbers you see.

Either way, on average, it will be fairly spot on.

TouchingVirus

Thanks to the powers of the internet it turns out that the E46 has a cluster test that displays the speed in KPH as read by the ECU. So naturally enough I went out again today for more testing.

We put a GPS app against the kph readout from test mode and found they were constantly within 1 of each other. I've got reasonable confidence in the accuracy of the GPS speed and the cluster's test mode. No idea why my needle is off by 8/9. I've no idea about that 8.9% its supposed to read over the ECU speed according to challengemaster's link either.

I'll call a few of those odometer places tomorrow to see about getting it tested/fixed

visual

CiniO wrote: »

How can changing wheel size affect gear ratio?

Gear ratio depends on amount of teeth on cogwheels inside gearbox.

The final cogwheel is the wheel it's self.

TouchingVirus

visual wrote: »

The final cogwheel is the wheel it's self.

I hear what you're saying but this is stock wheels & stock tyres though. This isn't wheel or tyre variance, we are talking almost 10km under real speed and 20km under what the speedo is supposed to show (8.9% over real speed).

challengemaster

TouchingVirus wrote: »

I hear what you're saying but this is stock wheels & stock tyres though.

It's not really though. Stock wheels for an E46 M3 are 18's. Sticking 19" CSL wheels on it (despite being OEM) means there'll be a difference. The speedo will have been set up for the original alloys.

Same as if you stuck CSLs on a E46 that came from factory with 16's. The speedo would be way out.

TouchingVirus

challengemaster wrote: »

It's not really though. Stock wheels for an E46 M3 are 18's. Sticking 19" CSL wheels on it (despite being OEM) means there'll be a difference. The speedo will have been set up for the original alloys.

Same as if you stuck CSLs on a E46 that came from factory with 16's. The speedo would be way out.

Oh? Jaysus, I bought it with 19" non-CSLs Would the difference of 1" cause it to go from overeading by 8.9 percent to underreading by 9km?

pippip

TouchingVirus wrote: »

Oh? Jaysus, I bought it with 19" non-CSLs Would the difference of 1" cause it to go from overeading by 8.9 percent to underreading by 9km?

Ask here - http://www.boards.ie/vbulletin/forumdisplay.php?f=380

TouchingVirus

So I thought about it a bit more, it can't be the wheels. The engine speed estimation shown by the cluster test mode was calibrated to 18" wheels if you're right challengemaster. If it says I'm doing 60, the GPS says 60/61. So the wheels might account for that 1km difference, but the needle is way back at 52kph when it should be at 64/65 given it is supposed to overestimate. Correct me if I'm wrong

Edit: Using 2*Pi*r as the circumference of the wheel says there's just over a 5% difference. So yeah

challengemaster

TouchingVirus wrote: »

Oh? Jaysus, I bought it with 19" non-CSLs Would the difference of 1" cause it to go from overeading by 8.9 percent to underreading by 9km?

Actually it seems like the 19's were an option, don't think I've ever actually seen them before though. 18's are certainly more common and would be the base option for M3s.

I wonder if they were calibrated from factory for the base. I doubt it would cause that much of a change though.

Special Circumstances

I'd say some offset fault on speedo or someone was dicking around with it previously.

ongarite

Based on this site and the standard 18" size of 225/40 your speedo will be 4% out with those 19" CSL wheels.
http://tire-size-conversion.com/tyre-size-calculator/

CSL alloys, if they are OEM not replicas should be running 235/35/19 front and 265/30/19 rear.

Bongalongherb

What's it now, €30 to hook up the machine to the engine control unit ? if the results received show an error with the speedo then that will probably quash your penalty points if you present this information in court or maybe before court. It seems reasonable enough to pay €30 for the information regarding any errors found as it will remove the 3 penalty points.

The above comment is just an idea as I am not sure if an ECU check will actually show an error from the speedo.

TouchingVirus

ongarite wrote: »

Based on this site and the standard 18" size of 225/40 your speedo will be 4% out with those 19" CSL wheels.
http://tire-size-conversion.com/tyre-size-calculator/

CSL alloys, if they are OEM not replicas should be running 235/35/19 front and 265/30/19 rear.

So I just double checked, looks like i was wrong last night. I'm running 235/35 on the front, 255/35 on the rear. The CSLs are OEM.

I'll bring it to the garage and get the entire setup looked at. Could be a sensor, bad cluster or a heap of other things so I'll let the pros sort it out and check for codes etc. Cheers for all your help

OldMrBrennan83

This post has been deleted.

9935452

TouchingVirus wrote: »

This is going to sound like a pile of shíte but hear me out...I've been done for driving at 59km/h in a 50km/h zone last Saturday. It's Cork Street in Dublin and those that know it as well as I do know it has a Garda speed van on it at least twice a week. Since this was only 7 days ago and I drive my car maybe twice a week I've a full memory of the event, the van's location etc and am full sure that like pretty much every time I drive down that road I had the cruise control on at 50km/h to avoid a heavy foot getting me some points. I made no attempt to slow down, after all I was doing 50

So I'm gutted I got the notice yesterday but figured something was amiss. I took the car out today and with my dad using a GPS app on his phone I had some interesting results: When I'm doing 40kph on the speedo it's really 48, when doing 50 it's 58/59, 60 => 68, 80 => 88, 100 => 109; I didn't do 120, this was just an N road :pac: We repeated the test several times to be sure and the difference was fairly constant. I'm running stock wheels so there's no trickery on that front. Clearly something's wrong with it under-reading so much but does anybody know somebody who can fix it?

Before anybody thinks I'm working some sort of angle here I don't think I'll bother appealing the points - it doesn't change the fact I was speeding and while 3 points is sickening risking 5 for an appeal isn't worth it IMO. I'm just looking not to get collared again for speeding when I think I'm fine; eventually I'll forget to shave 10kph off my current reading in my head and this'll happen again.

Personally i dont think its the tyres/wheels that is causing this.
Logic says if the outside diameter of the wheel/tyres is 10% bigger than standard, you should do 10% more at 50k and and 10% more at 100k, which is not the case.
The speedo is reading 8/9k out at all speeds.
Most likely there is an offset (of sorts ) that is out.
Its something i would talk to a BMW specialist about , they would have more info.
As regards using this as a excuse in court i wouldnt say it would fly. It would open an aweful loophole . Everyone would start using it as an excuse in court.
but that said if it could be proved that the car was supplied day one from bmw like that you would have a case i court

TouchingVirus

9935452 wrote: »

As regards using this as a excuse in court i wouldnt say it would fly. It would open an aweful loophole . Everyone would start using it as an excuse in court.
but that said if it could be proved that the car was supplied day one from bmw like that you would have a case i court

I've no interest in taking it anywhere near a court but thanks, I'm agreeing with you; it's not like the cluster has a warranty sticker on it that can prove I wasn't near it. And that becomes relevant when you read the following...

So I just did a cluster test lights & gauges test, I think the needle is off. During the test all the gauges go right the way around - petrol, oil, engine-temp, rpm and speed. The clock ends at 185mph but the speedo needle never makes it there, stopping short around 179 or so. On closer inspection it does appear to be off - the last mark on the gauge is 5mph but if you take the distance from 10mph to 5mph and infer where the needle should be for zero it's not there but a bit further back. The more I think about it as the morning passes the more I'm thinking this is the problem - it certainly explains the reason for the speedo being constantly out by 8/9kph rather than a percentage. I'll see if I can remove the needle and refit it so it's just shy of the 5mph and do some GPS tests to make sure I'm back over-estimating.

VeVeX

TouchingVirus wrote: »

So I just double checked, looks like i was wrong last night. I'm running 235/35 on the front, 255/35 on the rear. The CSLs are OEM.

I'll bring it to the garage and get the entire setup looked at. Could be a sensor, bad cluster or a heap of other things so I'll let the pros sort it out and check for codes etc. Cheers for all your help

E46 M3 OE sizes are -

235/35R19 87Y & 265/30R19 89Y
or
225/40R19 93Y & 255/35R19 96Y

TouchingVirus

VeVeX wrote: »

E46 M3 OE sizes are -

235/35R19 87Y & 265/30R19 89Y
or
225/40R19 93Y & 255/35R19 96Y

It's not the wheels or tyres, but thanks

Chemical Byrne

If there's a question mark over the accuracy of the GPS app, why not just go simple old school on it, ie, V=D/T , calculate your true speed based on travel time between two points which are a known distance apart while keeping at a constant speedo reading . Like go between 2 motorway junctions that are a known distance apart. The longer the distance you travel, the smaller the error. Do it at nighttime so that you can keep going a constant speed. You can also use this to validate the accuracy of the GPS

It could be the speedo gauge itself is faulty - reset your trip meter to 0. try travelling at say, a constant speedo reading of 100kph for 10 minutes straight on the motorway. Your trip meter should read 16.7km. If the distance clocked is greater, the gauge itself is probably faulty.

Also get the likes of a bluetooth ELM 327 and Torque Pro app and bring up the speedometer on it to see what the ECU is registering and if this differs from the car speedo.

If the ECU registered speed is different from that known to be correct from the GPS or calculating using V = D/T then you've narrowed it down to the sensor being faulty or something in the ECU being set up wrong or something.

challengemaster

OSI wrote: »

What drives the speedo in an e46, is it the ABS sensors, or a cable or what? I wonder, if it's sensor based if the sensors are on their way out. I'd give the ABS a quick test on the next outing if I was you TV.

Left rear ABS/Wheel speed sensor - I'd mentioned this (via PM) to TV lastnight alright.

The fact the cluster test reads the right speed is confusing the matter - but there's a chance it's not using the same sensor to read the speed from. I'd expect some warning lights to pop up with a 10km/h wheel speed difference though.

listermint

Have you considered doing the test in another M3? See if it's consistent.

Would do it mine but I sold it

JohnH1099

TV,
I had the clock from our e30 down with Motor Clocks in Gorey. They used to be based in Blessington St where I had used them years ago but now are a bit further away. I found them pretty good on both occasions.

They might be able to help.
Cheers
John

TouchingVirus

Cheers John. I paid my fine yesterday and took the needle off, resetting it at 4mph when stationary. It now reads a constant 7/8kph under so I'll not be getting caught due to a faulty speedo again.

9935452

TouchingVirus wrote: »

Cheers John. I paid my fine yesterday and took the needle off, resetting it at 4mph when stationary. It now reads a constant 7/8kph under so I'll not be getting caught due to a faulty speedo again.

just in case it decides to move again, take note of what RPM the car is at when it is doing 100kmph. My audi at 120kmph does 2000rpm.
If it changes in the future you will know the needle has skipped