Speedo vs GPS

[slofox]

For what it's worth - just in case anyone is wondering...

I ran a GPS on The Vixen this morning to check speedo reading against GPS speed readout...

Essentially, the speedo is about 9% optimistic against the GPS - i.e. at 100km/hr on the GPS, the speedo reads 109km/hr.

This is an average, since there was some variation involved - a bit of lag on the GPS and a tendency for it to read the same speed over a speedo range of 2-3 km/hr. Probably due to lag but also to GPS accuracy as well - maybe..

Sufficient to give me a better estimation of how fast I can go on the speedo without straying into ticketing territory.

[avgas]

Don't trust the GPS speed - the algorithm for its calculation is a 'best guess' method rather than raw data.
It can't use raw data due to update times.

What brand gps is it?

[slofox]

Don't trust the GPS speed - the algorithm for its calculation is a 'best guess' method rather than raw data.
It can't use raw data due to update times.

What brand gps is it?

Yep. Like I said - it is an estimate to give me some idea...

GPS was a tomtom start...

[neels]

I did the same the other day.

On the bike the speedo is about 8k's fast at 100, which is about the same error I get from those roadside speed signs. The car was about the same error as well.

I had suspected the old scooter was a bit out, given that a chinese scooter passed me a while ago when I was doing 30mph on the speedo. Actually it's reading about 1.5 times faster than what it's doing, so at 30 on the speedo it's only doing 20. So it's all good, I'll just pretend it's marked in kph instead :laugh:

[schrodingers cat]

Accuracy of a GSP depends on its Frequency. Really top end stuff is 20Hz but you can get really good results with 5Hz (sample every 0.2 sec)

I've done a quick search to see what the GPS frequency of a Tom Tom is and can't find it (easily) but I suspect it is 1 Hz or worse. Under hard acceleration or braking it will lag.

Quote from a GPS data logger site:
Speed Measurement. While speed is probably the most important parameter that anyone wants to measure using the data logging system, it is also the most inaccurate in a "conventional" system. The normal way to measure speed is to simply attach a pickup to a wheel to detect how fast it is rotating - but the rolling circumference of a tyre changes by 4% just with wear and temperature. Even worse, the error increases significantly under race conditions where the tyre is under load - typically the tyre slips by upto 20% under hard braking going into a corner. Measuring speed using GPS is now common practice in high-end systems - under typical conditions speed error is well under 1%!

This is true of cars but moreso with a motorcycle as the rolling radius of the tyre alters as you lean

[p.dath]

Don't trust the GPS speed - the algorithm for its calculation is a 'best guess' method rather than raw data.
It can't use raw data due to update times.

If you are travelling at a constant speed then the GPS speed should be accurate.

[sinned]

If you are travelling at a constant speed then the GPS speed should be accurate. The GPS check of a speedo is as accurate as your ability to maintain a steady speed and the inherent accuracy depending on how many satellites the receiver is locked on to. The top end GPS units display the accuracy as +-distance. If you can be bothered the maths can be done.

[p.dath]

Don't trust the GPS speed - the algorithm for its calculation is a 'best guess' method rather than raw data.

I'd trust the GPS reading more than the speedo reading ... the speedo has no way of self-calibrating.

[slofox]

The GPS check of a speedo is as accurate as your ability to maintain a steady speed and the inherent accuracy depending on how many satellites the receiver is locked on to. The top end GPS units display the accuracy as +-distance. If you can be bothered the maths can be done.

Ok. We had six satellites. I took the 100km/hr reading on a looooong straight on Kaipaki Road for those who know it. I kept as steady a speed as I could. (Not perfect, but over time etc etc...)

All I wanted was an indication. Which I got. 109 on the clock was around 100 on the GPS. Better than guessing which is what I have been doing up 'til now.

Incidentally, I somehow got the idea that the GPS was +/- 2km/hr...according to the joker who owned it I think...

As p.dath says, I'd trust the GPS more than the speedo...

[cowpoos]

If you are travelling at a constant speed then the GPS speed should be accurate.

Totally agree!!

[sinned]

Speedo reading 9kph fast is about right for a Suzuki; my check on 2 Zooks indicated 8kph. Your measurement technique is probably good enough to be relied on.

[vifferman]

The GPS (TomTom) consistently indicates the car's speedo is 2% optimistic at 100km/h. Good enough for me. Oh - that's when the car's speedo's actually working. About 60% of the time, it consistently reads 0. :blink:
I checked the bike with the TomTom back when I was running lowered gearing and a speedo corrector. My guesstimate for the speedo corrector's settings of -12% involved bad maths, so it turned out the speedo was under-reading by a few percent. Apparently 6% lowered gearing and 6% inaccuracy (based on polling other VFR owners) doesn't add up to 12%. Well... duh!

[schrodingers cat]

Incidentally, I somehow got the idea that the GPS was +/- 2km/hr...according to the joker who owned it I think...


With a 5Hz reciever the accuracy is 1% so +/-2km/hr @ 100kph ish is believable for a 1 hz reciever.
At (near enough to) constant speed I'd trust it.

[Filterer]

With a 5Hz reciever the accuracy is 1% so +/-2km/hr @ 100kph ish is believable for a 1 hz reciever.
At (near enough to) constant speed I'd trust it. How is update time in anyway shape or form related to accuracy at a constant speed?

[Padmei]

What I've always wanted to do is to find a friendly policeman at the end of a long straight with a speed gun & do a couple of tests. I would like to ride consistently at 100k on the GPS, then with 100k on the speedo & see what it reads on his speed gun. No matter how good your GPS is, ultimately if there is a descrepency between his reading & your GPS then you'll be gettting a ticket.
I don't think an argument that your GPS is top of the line will go down too well.

Anyone tried it?

[MarkH]

Measuring speed using GPS is now common practice in high-end systems - under typical conditions speed error is well under 1%!

This would be true. Even if a GPS can only give your location within 10 metres it can still give an accurate speed reading, it only requires consistency from one reading to the next, not accuracy on each location reading. A GPS is crap at showing the changes in speed in real time, but travel at a pretty constant speed and they are probably reading correctly to the nearest 1 kph.

[schrodingers cat]

How is update time in anyway shape or form related to accuracy at a constant speed?

Because speed = distance over time

The unit has a built in clock (Time)
The GPS co-ordinates provide distance as a differential between 'plots'

The refresh rate is critical if there is acceleration/decellaration. 5 samples per second ensures accuracy of around 1% A lower rate obviously gives a different 'average' (mean)

You are correct in saying that if the speed is constant then the refresh rate is irrelevant. Personally I'd like a speed calculation from 20 samples not 4 (5hz vs 1Hz)

[schrodingers cat]

This would be true. Even if a GPS can only give your location within 10 metres it can still give an accurate speed reading, it only requires consistency from one reading to the next, not accuracy on each location reading. A GPS is crap at showing the changes in speed in real time, but travel at a pretty constant speed and they are probably reading correctly to the nearest 1 kph.

I find that some systems (designed in the northern hemisphere) have 'issues' from time to time down under. I would suspect that satillite flight paths are more interested in good coverage for the bulk of the world land mass.

GPS co-odinates can 'jump' when the unit has to 'drop' one of the satillites it is triangulating off and pick up another. It is funny to see the jump on track maps during a session...

[MarkH]

I find that some systems (designed in the northern hemisphere) have 'issues' from time to time down under. I would suspect that satillite flight paths are more interested in good coverage for the bulk of the world land mass.

GPS co-odinates can 'jump' when the unit has to 'drop' one of the satillites it is triangulating off and pick up another. It is funny to see the jump on track maps during a session...

Sure, a 'jump' will throw out the speed readings, but unless the unit is constantly 'jumping' then the speed value that it is mostly showing when you are travelling at a constant speed WILL be correct. If generally you get a reading of 100kph on the GPS when travelling at a steady 108kph on the speedo then you can be fairly sure that 100kph is the real speed when you are going 108kph on the speedo. If occasionally you get a reading of something else then that could be a 'jump'.

My experience: readings on the GPS are VERY consistent when travelling at a steady speed in a straight line and not changing elevation. From this I have a good idea of where the speedo needle should be if I don't want a speeding ticket. While accelerating or braking I wouldn't bother using the GPS reading of speed for any idea of my actual speed.

[miloking]

For what it's worth - just in case anyone is wondering...

I ran a GPS on The Vixen this morning to check speedo reading against GPS speed readout...

Essentially, the speedo is about 9% optimistic against the GPS - i.e. at 100km/hr on the GPS, the speedo reads 109km/hr.

This is an average, since there was some variation involved - a bit of lag on the GPS and a tendency for it to read the same speed over a speedo range of 2-3 km/hr. Probably due to lag but also to GPS accuracy as well - maybe..

Sufficient to give me a better estimation of how fast I can go on the speedo without straying into ticketing territory.

Iam getting speedo healer for mine because i changed the sprocket so iam sick of guessing, especialy if they finaly decide to get rid of the tolerances for good (few more public holidays with low road toll i reckon)....maybe worth the idea for yours too (they are about $114USD from egay)

Edit: actualy there is another alternative and thats something called speedo tuner, seems to be way cheaper then speedo healer ($84) and from reviews it seems to be easier to setup....proably worth the money if you dont like doing the math everytime you go for ride.

[onearmedbandit]

When I've compared GPS vs speedo I've noticed something other than a discrepancy between the two, and that is that the difference between the two is always the same when held at a steady speed, so I'm quite confident that my GPS is tracking accurately.

[Max Preload]

...a tendency for it to read the same speed over a speedo range of 2-3 km/hr. Probably due to lag but also to GPS accuracy as well - maybe..

It's called hysteresis. They intentionally do it so the retards don't stare at it watching the numbers changing and drive off the road.

[LBD]

What I've always wanted to do is to find a friendly policeman at the end of a long straight with a speed gun & do a couple of tests. I would like to ride consistently at 100k on the GPS, then with 100k on the speedo & see what it reads on his speed gun. No matter how good your GPS is, ultimately if there is a descrepency between his reading & your GPS then you'll be gettting a ticket.
I don't think an argument that your GPS is top of the line will go down too well.

Anyone tried it?

Yup twice.......and I have the infringment notices pinned to my camp wall s a souvenir....Radar said 115....lights came on and I looked down and gps was 114...on a long straight south of levin....buggar...$80

Kumara straight one week later119 on GPS and 119 on the notice.....$120....Double buggar

To make matters worse....my radar detector mount broke and was lost the week before the first booking....tripple buggar

From experience and talking to the dealer, Ducati speedos read about 8 % optimistic....108 on bike speedo is a neat 100....Which makes the 200 on the race track a dismal 180 ish

[NZsarge]

Essentially, the speedo is about 9% optimistic against the GPS - i.e. at 100km/hr on the GPS, the speedo reads 109km/hr.

That's pretty much the same as my ZX14, reading 109 as opposed the gps at 100. Using a Garmin Nuvi 500.

[avgas]

If you are travelling at a constant speed then the GPS speed should be accurate.
And direction, and altitude.........

I can get GPS time sync down to Sub-uS on structures that aren't moving.
What most cheap GPS units do is window sampling when calculating speed. Depending on what data sits in the window buffer will determine the speed.
But then again its a case of what you are using if for and GPS does what it was designed to do perfectly. The Mil-spec crystals are good for about 1m. I believe the commercial ones are good for 2m.

But yes if you travel at constant speed, with constant direction at a constant altitude - you accuracy CAN be a good as the freq used. Exponential decay outside those parameters.

[Max Preload]

And direction, and altitude...

Only direction. Your altitude is a function of direction. You need constant velocity, the vector quantity, as opposed to constant speed, the scalar quantity.

[p.dath]

... Ducati speedos read about 8 % optimistic....108 on bike speedo is a neat 100....Which makes the 200 on the race track a dismal 180 ish

Assuming the error in your speedo is constant with speed ... maybe the error gets worse the faster you go ...

[onearmedbandit]

Assuming the error in your speedo is constant with speed ... maybe the error gets worse the faster you go ...

On my gixxer the error reads at a constant 13% error at any speed. (Down 1 on the front, std gearing is 8% error).

[Jantar]

I now have 2 GPS units on my bike displaying the speed. My radar detector has a GPS built into it, and I have a Garmin for recording. Both GPS displays are always within 1 km/h of each other when accellerating/decellerating, exactly the same at constant speed and around 8% slower than indicated on the speedo.

[Reckless]

Seems like everyone is searchng for a more accurate speed reading, GPS or speedo!
I imagine the thoughts are if the cops are going to be ticketing down to 104k then we don't want to loose another 8% in speedo error!

To take another angle
The person your following thats doing 80k and is so lacking in confidence they won't go faster because they've been told Speed KILLS is actually doing 73.6k. LOL!!!

You just gotta laugh its all quite funny what the boys in blue are creating! This scenario must be safer surely we are all going slower now?? LOL!!

[slofox]

What I've always wanted to do is to find a friendly policeman at the end of a long straight with a speed gun & do a couple of tests. I would like to ride consistently at 100k on the GPS, then with 100k on the speedo & see what it reads on his speed gun. No matter how good your GPS is, ultimately if there is a descrepency between his reading & your GPS then you'll be gettting a ticket.
I don't think an argument that your GPS is top of the line will go down too well.

Anyone tried it?

Yeah. I found a bunch of coppers playing with their laser guns one morning. I stopped and asked them to ping me at 50k on the clock to check accuracy (it was in 50k zone). They did. Reading was about 6% out. That was on the SV.


It's called hysteresis. They intentionally do it so the retards don't stare at it watching the numbers changing and drive off the road.

At which point it's called hysterics...


And direction, and altitude.........

But yes if you travel at constant speed, with constant direction at a constant altitude - you accuracy CAN be a good as the freq used. Exponential decay outside those parameters.

It was a long flat straight road...


Assuming the error in your speedo is constant with speed ... maybe the error gets worse the faster you go ...

The percentage error was very close to the same at 50km/hr and at 100km/hr...

[wysper]

Seems like everyone is searchng for a more accurate speed reading, GPS or speedo!
I imagine the thoughts are if the cops are going to be ticketing down to 104k then we don't want to loose another 8% in speedo error!

To take another angle
The person your following thats doing 80k and is so lacking in confidence they won't go faster because they've been told Speed KILLS is actually doing 73.6k. LOL!!!

You just gotta laugh its all quite funny what the boys in blue are creating! This scenario must be safer surely we are all going slower now?? LOL!!

and from yet another angle... when you are stopped for a reading of 104kph your speedo might have been indicating 113kph, if it was 110kph then your indicated is pretty much 120kph etc.

So if you travel at 100kph indicated on your speedo, you should be sweet. However if your speedo is actually accurate then your margin of error is significantly reduced.

[sinned]

Assuming the error in your speedo is constant with speed ... maybe the error gets worse the faster you go ...
Some cars have an xKph offset built in. This provides a set Kph error at all speeds rather than a percentage error. The Mazda 3 appears to have a 5Kph offset so reading of 50Kph= 45Kph and 100Kph=95Kph. This contributes to slow driver frustration.

[miloking]

Seems like everyone is searchng for a more accurate speed reading, GPS or speedo!
I imagine the thoughts are if the cops are going to be ticketing down to 104k then we don't want to loose another 8% in speedo error!

To take another angle
The person your following thats doing 80k and is so lacking in confidence they won't go faster because they've been told Speed KILLS is actually doing 73.6k. LOL!!!

You just gotta laugh its all quite funny what the boys in blue are creating! This scenario must be safer surely we are all going slower now?? LOL!!

You know whats even funnier, people driving at 45km/h indicated around town...just because speed limit says "50" so they think "i better be few ks under to be safe that way" all caused by mass hysteria created by police which makes these people think they are in mortal danger on every corner...(sinister man with spinning wheel comes to mind) ...what is that like actual 40km/h hour??? I think olympic sprinters can run at that speed, lol

[Flip]

The electronic spedo on my harley is 3 or 4 kph under at 100 and the old mechanical unit on the old boxer is almost exactly spot on.

[avgas]

Only direction. Your altitude is a function of direction. You need constant velocity, the vector quantity, as opposed to constant speed, the scalar quantity.
Haha yes of course - I was merely explaining it in a way for the poor fellas who haven't done Dynamics.
I still don't like the fact that is does a window average before it does the derivative of the distance though.
But I guess at the end of the day that what you get if you use extraterrestrial telemetry to do such a task.

[avgas]

Assuming the error in your speedo is constant with speed ... maybe the error gets worse the faster you go ...
That would explain "the faster you go the bigger the mess"

[eldog]

The electronic spedo on my harley is 3 or 4 kph under at 100 and the old mechanical unit on the old boxer is almost exactly spot on.

Electronic speedos - are they more accurate than mechanical units?
Or do they just show the same error with figures instead of a needle and a dial?

When travelling on the mway at the signposted speed, I am passed like I am standing still.

Any preferred way of correcting this or do I use a GPS and make an allowance on my Speedo reading

I prefer to be on or below the Hway limit, when travelling in unfamiliar locations (where I might just miss a posted speed limit sign).:headbang:

[Jeeper]

If you are in Auckland, on SH1 Southern Motorway near the Bombay Hills going south there are marked measured 5 kilometers. Yellow round markers on the side starting from 0 to 5. Maintain a constant speed and measure your time it takes to cover those 5 kilometers. That will give you the most accurate actual speed. There are similar marked kilometers on the SH20 going from Onehunga to Manukau travelling south.

[MarkH]

Electronic speedos - are they more accurate than mechanical units?
Or do they just show the same error with figures instead of a needle and a dial?

When travelling on the mway at the signposted speed, I am passed like I am standing still.

Any preferred way of correcting this or do I use a GPS and make an allowance on my Speedo reading

I prefer to be on or below the Hway limit, when travelling in unfamiliar locations (where I might just miss a posted speed limit sign).:headbang:

Both my car and my bike have digital speedo readouts, they are both as inaccurate as any other speedo.
I ALWAYS use a GPS to work out what the error factor is, so I can set a speed that is really 100kph instead of just an indicated 100kph. In the car I travel to & from work with, I set the cruise control keeping me at an indicated 110kph, at this speed I don't need to worry about the police, because I'm travelling at a real 100kph.

Since the modern smart phone can provide an accurate speed reading there is no excuse for anyone not knowing what their speed reading really means. Just use the GPS to find out what speed you need on the speedo to travel to maintain a steady 100kph. If it is 110kph then your speedo reads 10% over, 108kph means 8% over, etc. My experience is that the error is quite consistent, so 110kph at a real 100kph means that 55 on the speedo is going to be a real 50kph - the same percentage out throughout the range.

[SaferRides]

Electronic speedos - are they more accurate than mechanical units?
Or do they just show the same error with figures instead of a needle and a dial?

When travelling on the mway at the signposted speed, I am passed like I am standing still.

Any preferred way of correcting this or do I use a GPS and make an allowance on my Speedo reading

I prefer to be on or below the Hway limit, when travelling in unfamiliar locations (where I might just miss a posted speed limit sign).:headbang:I've checked the car speedo against the phone and it's high by 4 km/h from 50 to 110 km/h. So I usually keep to the limit +10 and haven't had a ticket.

Dunno about the R1 as I don't have a secure way of mounting the phone yet. But I've had a few cop cars flash their lights when the speedo has been about 115!

[slofox]

If you are in Auckland, on SH1 Southern Motorway near the Bombay Hills going south there are marked measured 5 kilometers. Yellow round markers on the side starting from 0 to 5. Maintain a constant speed and measure your time it takes to cover those 5 kilometers. That will give you the most accurate actual speed. There are similar marked kilometers on the SH20 going from Onehunga to Manukau travelling south.

This is exactly how I calibrated both my GPS unit and also various speedos. After installing a speedo healer on the bike (years ago now) everything measures up perfectly now.

[Jeeper]

The problem with GPS measurements of speed with a lot of units is their inability to adjust for altitude changes (i.e. going up and down hill versus on a flat piece of road).

[F5 Dave]

So you calibrate your Speedo over a fair straight flat bit of ground.

Tank bag or duct tape it.

12 o'clock labs do a cheaper Speedo healer which is small.

TuneECU allows me to adjust my pre13 Triumph for free directly. Always set to accurate that way you don't have to do fast arithmetic when you meet a police person on your trip.

[Hoonicorn]

Your speedo should never under-rate your speed, GPS may not have that same requirement so it could under-read. Speedo + 5% should keep you within the speed limit when the cops decide to go Zero Tolerance.

[eldog]

Checked the car speedo tonight

road was busy so had to do it at 90 kph as the fastest speed I could do constantly with the heavy traffic.

car 90
gps average 86 fluctuated 84 to 88 most of the time at 86
calc 86.1

that same difference was noted when I did 100 the other day with the gps

now only to try it on the bike:eek5:

done. Bike says 94/95, calculation from 5k signs and stopwatch 88
suggest don’t stop on side of motorway at night, truck drivers tend to follow red tail lights. Even though I was well off to one side one truck driver tried to make me into another lane:no:
will try gps next time on bike.

[FJRider]

What I've always wanted to do is to find a friendly policeman at the end of a long straight ...

Anyone tried it?

One found me ... and although he wasn't that friendly ... the number on the ticket matched the max speed on my hand held (jacket pocket actually) GPS.

The GPS was right on the money. And I'd trust it over the speedo anytime ...