Hi (speedometer accuracy)

decrep · Dec 25, 2024

Some GPS chips report accuracy, so it knows if it has a good satellite signal. In that case, with a good signal, the calibration could be instantaneous. In this day and age it would be fairly obvious for the on board GPS to be involved somehow. Even lesser chips can be accurate over longer times and distances, so if not instantaneous, over a few minutes would be possible.

decrep · Dec 25, 2024

Just remembered, I saw a GPS file in Carscanner, haven't seen one since, don't know what I did different.
Anyway, this is the same chip we use, a UBLOX chip set to 10HZ.

Date Time Speed +/- Sats m/s^2 Dist

So this is showing an accuracy of +/- 0.19 kts for every 0.1sec reading.
Over 10s the accuracy really improves.

The speed is knots. The tracks are in a 70km/hr zone.

decrep · Dec 25, 2024

here's the accuracy over 10s, Talking miles/hr, that's about +/-0.5.

So speed correction could be virtually instantaneous.

10 sec	1	41.428	0.042	10:52:40.900
10 sec	2	38.931	0.051	09:59:48.300
10 sec	3	38.726	0.047	10:44:17.100
10 sec	4	38.363	0.045	09:49:28.900
10 sec	5	38.182	0.041	09:57:20.700

ReintjeWA · Dec 25, 2024

Do you think that means you could theoretically put any size wheel on and the car will recalibrate the speedo accordingly? Or do your think it checks different data point against each other to look for potential malfunctions?

decrep · Dec 25, 2024

So if you see a strange .TXT file i n your car scanner, it's probably a .UBX GPS file. You need to ask me how to open it.

decrep · Dec 25, 2024

ReintjeWA said:
Do you think that means you could theoretically put any size wheel on and the car will recalibrate the speedo accordingly? Or do your think it checks different data point against each other to look for potential malfunctions?

I've no idea I'm only surmising possibilities.
We need somebody to do it to find out.

Coulomb · Dec 25, 2024

ReintjeWA said:
Do you think that means you could theoretically put any size wheel on and the car will recalibrate the speedo accordingly?

Within reasonable limits, I'd say yes, it's theoretically possible.

Whether MG actually do this, as above, we don't know yet.

siteguru · Dec 25, 2024

ReintjeWA said:
That’s an interesting point you raise. I would hope MG had the foresight to leave enough margin for tyre wear, but perhaps there is less margin in the speedometer to play with when getting different rims/tyres. Normally you want to stay within 2-3% of your original overall tyre diameter but perhaps for the MG4 this margin should be much less.

Per my previous post, as tyres wear the circumference of the wheels+tyres reduces. That means that for each rotation the distance covered reduces, therefore actual speed reduces vs displayed speed (assuming speed is based on wheel revolutions per second/minute rather than GPS measurement).

SteveMSJ · Dec 25, 2024

siteguru said:
Per my previous post, as tyres wear the circumference of the wheels+tyres reduces. That means that for each rotation the distance covered reduces, therefore actual speed reduces vs displayed speed (assuming speed is based on wheel revolutions per second/minute rather than GPS measurement).

Surely the smaller the wheel gets the more times it has to rotate for the same horizontal distance travelled. Therefore if speed reading is based on revolutions it will show faster as the tyres wear?
Small wheels spin much faster at the same vehicle speed.

DixieTGS · Dec 25, 2024

Both are correct. @siteguru worded it as actual speed reduces vs displayed speed. You worded it as displayed speed shows faster (than actual speed).

SteveMSJ · Dec 25, 2024

DixieTGS said:
Both are correct. @siteguru worded it as actual speed reduces vs displayed speed. You worded it as displayed speed shows faster (than actual speed).

Good point

siteguru · Dec 26, 2024

I worded it specifically that way to address @ReintjeWA's earlier comment that actual speed may be higher than displayed speed when tyres wear, to show that in fact it is the opposite. (Speed is distance over time ... if the distance travelled is less due to reduced circumference, then actual speed is lower*).

* This all assumes that the speedo is calibrated based on wheel rotation speed, and not using GPS tracking.

ReintjeWA · Dec 26, 2024

To be fair I never implied the speedo would show higher, I just spoke of discrepancies in actual speed and shown speed. I know the issue is increased diameter when fitting after market rims and tyres, and am mostly interested in the car potentially correcting this automatically.

Anyway seems we’re all on the same page and in agreement

siteguru · Dec 26, 2024

ReintjeWA said:
That’s an interesting point you raise. I would hope MG had the foresight to leave enough margin for tyre wear

That's what you said ... to me that infers that tyre wear = actual speed increases. I'm glad we're all aligned.

decrep · Dec 27, 2024

So who's going to do the test?

2 space saver wheels would be good if anybody can lay their hands on them.
I'm also wondering what would happen if you jacked a drive wheel up and just ran it slowly, or put the car on a dynamometer. But I suspect that would be too big a jump, from any speed to 0.
If I get bored, I may just jack up one wheel, fully chock the rest and run it at a crawl, will it adjust 2mph to 0?

ReintjeWA · Dec 27, 2024

decrep said:
So who's going to do the test?

2 space saver wheels would be good if anybody can lay their hands on them.
I'm also wondering what would happen if you jacked a drive wheel up and just ran it slowly, or put the car on a dynamometer. But I suspect that would be too big a jump, from any speed to 0.
If I get bored, I may just jack up one wheel, fully chock the rest and run it at a crawl, will it adjust 2mph to 0?

Before you jack up the car, I’ve read @nexia201 is running a smaller than OE fitment tyre, so they might be able to provide us with an answer

Sorry to @ you twice nexia, just really hoping you can come help us out with this one haha

kc1966 · Dec 28, 2024

dawson2099 said:
Me too - the MG4 has the most accurate speedo that I've ever had - it's bang on with detectors and my GPS speedo

me too

TonyF · Dec 28, 2024

According to Waze, my XPower never reads more than 2km/h above my actual speed. Occasionally they match, but usually the car claims a mere 1km/h faster than Waze.

In mumblety years of driving/riding numerous vehicles I've never had such an accurate speedo. My motos have been the least accurate, consistently out by 5-10%. Knobbier tyres (same nominal size as OEM) made a huge difference when I had a DR650. I assume the knobbier knobs increased the overall circumference.

Corindikev · Dec 28, 2024

I don't care how the accuracy is achieved but a check with a speedo app on my phone was enough for me when it aligned almost perfectly with the cars digital readout. So when on the highway I set the cruise control to the actual limit. My last car was terrible & displayed about 118 kmh at 110kmh actual speed.

Joningy · Dec 30, 2024

Thebigone10 said:
Has anyone had the car tell you you're going faster than you are? I have been past the flashing lights that tell you your speed when passing them. Three of them all said I'm going 23 mph, but whinny car is telling me I'm doing 28. Has anyone got anything to say? What I can do about this?

This is odd as I've found my Oct 2023 MG4 Trophy is the most accurate speedo reading I've ever had, the street speed signs have consistently read exactly the same as my speedo reading, so too has the GPS and Waze speed readings.

Hi (speedometer accuracy)

Established Member

Established Member

Established Member

Established Member

Established Member

Established Member

Moderator

Distinguished Member

Standard Member

Established Member

Standard Member

Distinguished Member

Established Member

Distinguished Member

Established Member

Established Member

Standard Member

Established Member

Established Member

Established Member

Similar threads

Yes

I'm in the middle

No

Latest MG EVs video