THSCC Forums

Since I don't have lots of HP ad TQ I don't *know* that. Maybe next year I'll get an opportunity to experince driving a 300HP first gen RX-7 though. I'll let you know how that feels.

_________________
Vincent Keene
'06 Ford Mustang GT (track rat)
'15 Dodge Charger R/T (yeah, it's got a HEMI!)
'07 Ford Fusion SE (205,000 miles and counting)
'98 Chevy Z-24 (retired)
'93 Acura Integra (Team SWB 24HOL Car)

Well, I know of a stock WRX that was reportedly exceeding 130 on Full Course, so high 140's sounded very reasonable to me for an STi. The person who told me the speed for a stock STi was Jon Krolewicz, and I would trust what he said unlessl I could personally prove him wrong... I guess I'll find out firsthand in a few months, unless I sell mine before then or something bad happens to it.

Any of you guys who are quoting top speeds on the straights based on speedometer readings, do me a favor. Start looking at the tach instead -- calculate the actual speed based on RPM, tire size, gear, and FD, and then post the number. As we all know most speedos are optimistic, but I believe most tachs are better calibrated. That typical 5% error is only 3 MPH at 60 but more like 6+ at 120+MPH. Plus a lot of us run smaller tire sizes which give high-speed speedo readings even more of a "holy shit" factor.

_________________
Mike Whitney
whit32@gmail.com, 919-454-5445
V10, V8, V8t, I6, I6, V6, F4t, I4, I4, I4, I4, I2, 1, 1

Z06; 4th gear = 1.0; Diff=3.42; Tire=26.1" dia {295/35-18}. Typical RPMs = 6000-6300. Rev limiter = 6600rpm.

On VIR Full depending on how I come out of Oak Tree and Hog Pen depends on whether I'm at 6000 or 6300 at the end of the straight. I have hit 6400rpm a bunch of times. I bounce between 136-143mph, sometimes 146mph, according to the speedo. So I don't think it is that far off.

Mid season I went to Nitto R2 - 305/35-18, dia 26.3". It only saves me about 50rpm at 140mph. So it wasn't a big difference on top speed. Grip was better though

_________________
http://www.greywinds.com

I've noticed that a lot of newer cars no longer have a percentage discrepancy on the speedo. They seem to get to a 3mph error by 45mph or so, and then it *stays* a 3mph error (or thereabouts). Observed using a GPS and an occasional lead foot.

I think back when they were basically mechanically driven and/or had a simple circuit driving them, the percentage made sense. Now that the gauges are pretty much completely digital driven by the ECU, they can be "smarter" about how they do it. Thus the algorithmic change.


--Donnie

So my miata says 172 on the front straight. That's gotta be close to right...right?

_________________
Marty Howard
2011 NASA SE Factory Five Challenge Champion
Track Events Logistics Coordinator - TZC/THSCC
2007 Factory Five Challenge Car.
http://www.mh-motorsports.com

Interesting but I gotta say, back it up with some tech.

All of my "modern" cars that I have inspected still have a sensor on the diff counting teeth as they fly by.

_________________
Mike Whitney
whit32@gmail.com, 919-454-5445
V10, V8, V8t, I6, I6, V6, F4t, I4, I4, I4, I4, I2, 1, 1

Which are just pulses going to the ECU. That's why you can now flash your ECU to adjust for changing a tire size on many cars. I don't believe I claimed the ECU got its data any different. It just does some more advanced math than "add 3%".


--Donnie

Well since you seem to be in the making of creating an urban legend, let me just explain my skepticism:

I have a hard time believing that someone, who works for an hourly wage at an electronics supplier for the automotive industry, would come up with an idea to build an algorithm into a modern ECU to provide speed-sensitive speedometer mis-calibration. There is no commercial value to that, and in the absense of market value there will be no solution. Just don't see it happening. Even if it did once at one electronics supplier, I can't see it as a standard, let alone a regular practice.

Maybe your GPS is out of whack? Or maybe there are some tinfoil hat-wearers at the GPS company? I'm not saying I disbelieve, I'm just saying I need some fact or well-correlated data or even some informed knowledge to buy into it!

_________________
Mike Whitney
whit32@gmail.com, 919-454-5445
V10, V8, V8t, I6, I6, V6, F4t, I4, I4, I4, I4, I2, 1, 1

The cars I have owned and checked the speedos on have tended to have an error that rose with speed. However the also were mechanical speedos. In fact two of the three vehicles I own still have speedometer cables. The Focus seems to be fairly close according to the police radar signs I have passed. Of course I have not checked it towards the upper limit yet.

_________________
George Bright Member # 141
One of the youngest of the old farts.
Yes I voted against you joining the club.

On the GM cars the sensor is called a VSS and it is connected to the transmission, not the diff. It is calibrated against the factory gearing and tire sizes in the ECU. I'm sure the VSS has some margin of error since it is an electromechanical device. All OBD2 cars can be reprogrammed for changes in gearing and tire sizes. I doubt too many people would bother for tire sizes but for a gear change it would be a big deal.

On my last road trip I had a GPS in the car and watch the speed. The speedometer and the GPS were within 2-3mph easily. That was up to about 80mph. There will always be some amount of error. 3% @ 140mph is 4.2mph. So it is reasonable that people see a 140mph reading on the straight and it is really 136mph. But it certainly shouldn't be off by 10mph+ unless the car is messed up or the tires sizes were swapped as you mentioned before.

_________________
http://www.greywinds.com

I have merely reported what I have observed. I'm sorry you don't agree with my observations and/or the accuracy of my equipment. I'm quite certain of what I've seen, however, and quite certain my GPS was operating just fine. But I hardly feel compelled to go out and contract and independent test just to appease you.

I will point out that the "work" involved in what you're talking about is at most a change of approximately 1/3 of a LINE of code. I will also point out that the electronics suppliers are usually *obligated* to do NOTHING other than what is asked, so I can only assume that the auto manufacturers asked for this "change." Why? Beats me. Why do they do a lot of crap that they do? It does seem to make sense to me that a speedo need not be off by more than 3mph once you hit 60mph or so, but it should obviously ramp up to that 3mph "error". Was it for magazine tests? Consumers asking questions about speedos being "way" off? *shrug* I honestly don't really care.

What I do know is I've seen vehicles that were off right about 3mph at 60mph and were still *only* off right about 3mph at 120mph. If it were a straight percentage, as it was in the old days with mechanical stuff, that would have been 6mph. No way my powers of observation or my GPS were off that much.


--Donnie

Ah, but the only "error" capable here is in tire wear, really. I contend that the industry (or at least some players in it) have decided the algorithm should basically be 5% up to 60mph and then 3mph from there (or pretty close to those numbers) whereas it used to be straight 5% (and the indication on the speedo is *high* by that amount). The "error" in this case is always going to be sub-1% or so and is fairly negligible for the purposes of the discussion.


--Donnie

very good point...i pulled my numbers from my run data and my wheel speed numbers match up with the engine speed numbers...the only time i ever look at the speed display while i'm driving is when i have to worry about a pit lane speed limit...

_________________
David White
95 M3 - #275 BMWCCA JP, NASA GTS-3
02 M3 - street/track
05 Suburban 8.1L

That's to be expected. As Graham said, the VSS isn't lying...it's just the speedo output from the ECU that's lying. We all know it's lying, the question is by what algorithm is it lying.

On a related note, I wonder if VSS is the number from the actual sensor or not. If so, my Jeep is doing a divide-by-three *at* the sensor because the tone ring on the Jeep has three "teeth" that a hall effect sensor seems to be seeing. So either it's sending three times the real pulses or it's dividing by three as it goes. If it's sending three times the pulses, then the ECU must have a VSS output that's truly the sensor's output divided by three since all in-dash NAV systems (aftermarket and OEM) want a VSS signal, and based on the ones I've installed they all expect a "true" VSS, not something they have to do math on (if they needed to do math then you'd have to configure it for what kind of math, and I've never had to do that).


--Donnie