 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Car Forum / BMW Cars / September 2005Speedo Accuracy - '94 E36
I set the Cruise to 80 and timed two miles by the mile markers; it took 92 seconds to travel the distance. This works out to 78.26 mph, so the speedo is fast by 1.74 mph at 80. The SPEED reported by the OBC was 78.2, so it seems that all is well in my universe. The speedo is accurate to within 2 mph at 80, what more can a guy ask? By the way, I run 225/45 x 17s on my 325i Convertable. > I set the Cruise to 80 and timed two miles by the mile markers; it took 92 > seconds to travel the distance. This works out to 78.26 mph, so the speedo [quoted text clipped - 4 lines] > > By the way, I run 225/45 x 17s on my 325i Convertable. I did a similar test just after buying my e32, although over five miles - it was very early in the morning - and checked against my GPS. The results were also similar with about a 2mph error at 60, 70 and 80, increasing to 3mph at 90. Didn't test above that. My V8 Discovery gives almost identical results too. The GPS was +/- 1mph at all speeds.  SignatureRegards Steve G > I set the Cruise to 80 and timed two miles by the mile markers; it took > 92 seconds to travel the distance. This works out to 78.26 mph, so the > speedo is fast by 1.74 mph at 80. The SPEED reported by the OBC was > 78.2, so it seems that all is well in my universe. > The speedo is accurate to within 2 mph at 80, what more can a guy ask? I'd ask for a lot more. A pulse counting device can be made accurate to within 1% for pennies. Check out any DVM. Ancient eddy current speedos could better this too - let alone chronometric types. BMW - and others - deliberately make their speedos over-read. Which is disgraceful on a car of this cost. They should be accurate to within the width of the needle. Or so that with worn tyres so they don't under-read with new. Signature*Few women admit their age; fewer men act it. Dave Plowman dave@davenoise.co.uk London SW To e-mail, change noise into sound. >> I set the Cruise to 80 and timed two miles by the mile markers; it took >> 92 seconds to travel the distance. This works out to 78.26 mph, so the [quoted text clipped - 11 lines] >width of the needle. Or so that with worn tyres so they don't under-read >with new. The trouble is all tyres are slightly different. The diameter of a new tyre from manufacturer A will be slightly different to one from manufacturer B. During the life of a tyre you loose approx 2% of the diameter too. Then there's the tyre pressure and temperature. Your tyre expands when it get up to temperature, so there's yet another variation. And how many of you out there obey the tyre pressure guide to the letter? Do you drive to the store and then put the extra 2psi in when you've dumped your shopping in the back? No, of course you don't. Oh, and you'd have to let the tyre cool down completely before checking it too, and when you get home, put the shopping away, let the tyres cool down, let the pressure back out... Yeah right! lol! So I'd much rather the manufacturers read a little over, than read under. Mr policeman doesn't like people saying "Oh really officer, the speedo said I was doing less than that". With all that, I think 2.5% is damn impressive. Dodgy. SignatureMUSHROOMS ARE THE OPIATE OF THE MOOSES >> I set the Cruise to 80 and timed two miles by the mile markers; it took >> 92 seconds to travel the distance. This works out to 78.26 mph, so the [quoted text clipped - 11 lines] > width of the needle. Or so that with worn tyres so they don't under-read > with new. In the UK speedo's can over-read by up to 10% but is not allowed to under read (or you fail your MOT / type approval). Since a new car has new tyres, one has to make an allowance of a further ~2% for that compared with the reading when the tyres are worn (and so smaller). 5% is typical, 7% on older caars and more for Fords. You can do better than that with doppler radar or GPS. >>> I set the Cruise to 80 and timed two miles by the mile markers; it took >>> 92 seconds to travel the distance. This works out to 78.26 mph, so the [quoted text clipped - 19 lines] > >You can do better than that with doppler radar or GPS. And don't we just know about the doppler radar! Dodgy SignatureMUSHROOMS ARE THE OPIATE OF THE MOOSES "Dodgy" <Dodgy@earth.planet.universe> wrote in message >>You can do better than that with doppler radar or GPS. > > And don't we just know about the doppler radar! > > Dodgy Oh yes. TC's used to quote XX.123mph, although the last figure could vary depending on the officer's pulse! > > BMW - and others - deliberately make their speedos over-read. Which is > > disgraceful on a car of this cost. They should be accurate to within > > the width of the needle. Or so that with worn tyres so they don't > > under-read with new. > In the UK speedo's can over-read by up to 10% but is not allowed to > under read (or you fail your MOT / type approval). I've had many many MOTs, but never had the speedo tested. > Since a new car has new tyres, one has to make an allowance of a further > ~2% for that compared with the reading when the tyres are worn (and so > smaller). Smaller tyres (wear) would make the speedo over-read, so is fine with the concept of never under-reading. > 5% is typical, 7% on older caars and more for Fords. If you bought a one dollar digital watch that gained 5%, would you be happy? Mind you, the digital clock in my E39 is poor too. Signature*Make it idiot-proof and someone will make a better idiot. Dave Plowman dave@davenoise.co.uk London SW To e-mail, change noise into sound. > > > BMW - and others - deliberately make their speedos over-read. Which is > > > disgraceful on a car of this cost. They should be accurate to within [quoted text clipped - 17 lines] > If you bought a one dollar digital watch that gained 5%, would you be > happy? Mind you, the digital clock in my E39 is poor too. My speedo, '94 E34 seems to be reasonably accurate. Within about 1-2 mph of the reading given by a sat nav at 70 mph. The analogue clock keeps excellent time. I've only had to reset it once in 18 months, and that was about 8 months ago after fitting a new battery. It's still correct within about half a minute. Mike. That means you don't reset it twice a year? DAS SignatureFor direct contact replace nospam with schmetterling --- [...] > The analogue clock keeps excellent time. I've only had to reset it once in > 18 months, and that was about 8 months ago after fitting a new battery. > It's > still correct within about half a minute. > Mike. > That means you don't reset it twice a year? Not in the sense of correcting it's timekeeping. I've simply added or subtracted an hour. I haven't needed to adjust the minute reading. Mike. > DAS > [...] [quoted text clipped - 3 lines] > > still correct within about half a minute. > > Mike. Ok. Yes, I did mean the seasonal changes spring & autumn. You called it "analogue". Does that not mean a display with hours and minutes hands? Whatever, your clock's time-keeping seems pretty good. DAS For direct contact replace nospam with schmetterling --- [...] > Not in the sense of correcting it's timekeeping. I've simply added or > subtracted an hour. I haven't needed to adjust the minute reading. > Mike. [...] > Ok. Yes, I did mean the seasonal changes spring & autumn. > > You called it "analogue". Does that not mean a display with hours and > minutes hands? It does. On my E34 it's fitted instead of computer. I'm still debating whether to change it for an E34 computer I've picked up. Apparently it would give all the std functions, apart from outside temperature. > Whatever, your clock's time-keeping seems pretty good. Just luck I suppose. I can't imagine they're as accurate in general. Mike. > > You called it "analogue". Does that not mean a display with hours and > > minutes hands? > It does. On my E34 it's fitted instead of computer. I'm still debating > whether to change it for an E34 computer I've picked up. Apparently it > would give all the std functions, apart from outside temperature. > > Whatever, your clock's time-keeping seems pretty good. > Just luck I suppose. I can't imagine they're as accurate in general. It's not rocket science to design a clock or watch that keeps time to within 30 seconds every half year. I've got a 10 year old plus quartz analogue wrist watch that does just that. Of course a wrist watch tends to have the its temperature kept near constant. But my E39 clock looses some 2 minutes per week. Strange that they didn't drive it off the RDS part of the radio since it's in the same panel. Signature*Microsoft broke Volkswagen's record: They only made 21.4 million bugs. Dave Plowman dave@davenoise.co.uk London SW To e-mail, change noise into sound. > > > You called it "analogue". Does that not mean a display with hours and > > > minutes hands? [quoted text clipped - 11 lines] > analogue wrist watch that does just that. Of course a wrist watch tends to > have the its temperature kept near constant. Still relatively new then. :-) My regular wristwatch is a quartz Seiko analogue baught in 1984. Is only on it's 3rd or 4th battery, and keeps similarly good time. Expensive when I baught it, but cheap when it's still running perfectly after 21 years. > But my E39 clock looses some 2 minutes per week. Strange that they didn't > drive it off the RDS part of the radio since it's in the same panel. Even if they didn't, I'd have thaught it would be quartz regulated, therefore having similar accuracy as that of most quartz timekeepers. 2 minutes a week is disgraceful in a car purporting to be luxury drivers car etc. Come to think of it, why not radio controlled. Accurate to within a second in umpteen years. They're cheap enough these days. Mike. >> I set the Cruise to 80 and timed two miles by the mile markers; it took >> 92 seconds to travel the distance. This works out to 78.26 mph, so the [quoted text clipped - 6 lines] > within 1% for pennies. Check out any DVM. Ancient eddy current speedos > could better this too - let alone chronometric types. Sure it can, but the variables are numerous. Tire size, timing of my pressing the start and stop buttons, to name but two. > BMW - and others - deliberately make their speedos over-read. Which is > disgraceful on a car of this cost. They should be accurate to within the > width of the needle. Or so that with worn tyres so they don't under-read > with new. I don't agree that it's disgraceful at all. >>I set the Cruise to 80 and timed two miles by the mile markers; it took >>92 seconds to travel the distance. This works out to 78.26 mph, so the [quoted text clipped - 6 lines] > within 1% for pennies. Check out any DVM. Ancient eddy current speedos > could better this too - let alone chronometric types. SG: Had to do it. Just looked up the specs on several of my DVM/DMM's. Depending on the quantity being measured (e.g. volts, amps, ohms, etc) a 1% error is quite typical. Then you have to go on to read the conditions under which this can be achieved - ambient temperature having a big influence as does wave shape for ac measurements - and the over-riding caveat to all displayed values +/-1 digit. The accumulative errors can easily be more than 2.5% > BMW - and others - deliberately make their speedos over-read. Which is > disgraceful on a car of this cost. They should be accurate to within the > width of the needle. Or so that with worn tyres so they don't under-read > with new. SG: You should actually thank BMW (and all other European manufacturers for that matter) for making their speedometers read high - they may well save you from a speeding ticket one day. No matter how good the pulse counting system is, the final display that you, the driver, sees is electro-mechanical and there's probably more than a 1% error simply in the conversion. If you really want needle width accuracy then go buy a car with a fat needle :-)) SignatureRegards Steve G > > I'd ask for a lot more. A pulse counting device can be made accurate > > to within 1% for pennies. Check out any DVM. Ancient eddy current [quoted text clipped - 3 lines] > Depending on the quantity being measured (e.g. volts, amps, ohms, etc) a > 1% error is quite typical. Err, the likely error will be different depending on what it's measuring. > Then you have to go on to read the conditions under which this can be > achieved - ambient temperature having a big influence as does wave shape > for ac measurements - and the over-riding caveat to all displayed values > +/-1 digit. BMW uses an analogue reader. And anyone designing a pulse generator - and reader - would be well aware of the pitfalls. > The accumulative errors can easily be more than 2.5% Not on a custom designed system. In any case, BMW speedos don't have an 'accumulative error' of 2.5% - they all over-read. If it were a tollerance, some would be near accurate and some over-read by 2.5% above this. But this isn't the case. > > BMW - and others - deliberately make their speedos over-read. Which is > > disgraceful on a car of this cost. They should be accurate to within [quoted text clipped - 4 lines] > for that matter) for making their speedometers read high - they may well > save you from a speeding ticket one day. Heh heh. Since I know my E39 speedo over-reads, I just put this into the equation. So happily pass speed cameras at 5 mph over what it says. > No matter how good the pulse counting system is, the final display that > you, the driver, sees is electro-mechanical and there's probably more > than a 1% error simply in the conversion. Don't be silly. The analogue display can be easily made to be accurate to within the width of the needle at any spot speed. In the UK, this would be 70 mph. And if BMW etc use such crappy meters that it isn't still within tolerance at 30 or whatever they need to get some decent engineers. > If you really want needle width accuracy then go buy a car with a fat > needle :-)) I've got a 20 year old car with a Smith's pulse counting speedo. Remember them? Similar to Lucas. 0.5 mph fast at 70, and spot on at 30. As checked by Autocar at the time using a fifth wheel driving a chronometric speedo - same as the police used to use for checking speed. And exactly the same on the five variates of this model they tested, from basic to top of the range. With 5 different axle ratios and four different tyre sizes. BMW *deliberately* make their speedos read high. Draw your own conclusions. Signature*If you don't like the news, go out and make some. Dave Plowman dave@davenoise.co.uk London SW To e-mail, change noise into sound. <top post> Sorry to beat this dead horse to death, but my only point was that people that claim they have a speedo that is off by 5 ~ 7 mph are making an error. Using the formula, <width> * <aspect ratio> / 25.4 * 2 + <rim> = <overall diameter>, I find that all available tire sizes are within 0.5 inches of having the same diameter. Surely if all of the tires are within a half inch of being the same overall diameter, then the speedo readings should be pretty close no matter which tire is installed, and how far it's worn down. Assuming my beat up old car is anywhere near approaching normal, then everybody should have similar error rates as I have, and my error rate is below 2%, which is pretty damn close considering the application. My car is the E36, '94 325i Convertable (which replaced a '94 325i Sedan), carrying the 225/45 x17 tires. I captured the tires from the sedan and put them onto the convertable, and the sedan came with the Sports Package that included 225/55 x 15 tires that are only made by Michelin. The sedan always had the 225s on it while I owned it - except that it had 205/60 for a few weeks that the PO had installed, and were near end-life when he sold the car to me, and I did not check speedo accuracy with them -- but I digress. Sure, it would be nice if the speedo was dead nuts accurate, but is that _really_ a requirement? Not to me. Width Ratio Rim Diameter 225 55 15 24.744 205 60 15 24.685 225 45 17 24.972 235 40 17 24.402 245 35 18 24.752 185 65 15 24.469 </top post >> > I'd ask for a lot more. A pulse counting device can be made accurate >> > to within 1% for pennies. Check out any DVM. Ancient eddy current [quoted text clipped - 54 lines] > BMW *deliberately* make their speedos read high. Draw your own > conclusions. >>>I'd ask for a lot more. A pulse counting device can be made accurate >>>to within 1% for pennies. Check out any DVM. Ancient eddy current [quoted text clipped - 5 lines] > > Err, the likely error will be different depending on what it's measuring. SG: That's why I said the error was "typical". Do you have a problem with English comprehension? >>Then you have to go on to read the conditions under which this can be >>achieved - ambient temperature having a big influence as does wave shape [quoted text clipped - 10 lines] > tollerance, some would be near accurate and some over-read by 2.5% above > this. But this isn't the case. SG: What makes you think BMW speedo's are custom designed. To my certain knowledge they use instruments designed and manufactured by Bosch and Siemens and probably others too. The speedo display tolerance is -0/+6% - as defined in the current version of construction and use regulations. There is, in any case, an accumulated error as with any system that goes from an electro-mechanical origin through digitisation, computation and back to electro-mechanical display. >>>BMW - and others - deliberately make their speedos over-read. Which is >>>disgraceful on a car of this cost. They should be accurate to within [quoted text clipped - 16 lines] > 70 mph. And if BMW etc use such crappy meters that it isn't still within > tolerance at 30 or whatever they need to get some decent engineers. SG: Yes, an analogue meter can be made to be spot-on accurate but at what cost? And we're not just talking about the meter are we? Errors are introduced at every point in the chain - right from the car tyres to the actual instrument. You have to compromise between absolute precision and lowest cost (as the two extremes) and come up with something that the masses will accept. >>If you really want needle width accuracy then go buy a car with a fat >>needle :-)) [quoted text clipped - 5 lines] > the five variates of this model they tested, from basic to top of the > range. With 5 different axle ratios and four different tyre sizes. SG: I've got a 30+ year old motorbike (1972 Royal Enfield Interceptor) with one of the Smith chronographs on. I'd hate to think what it would cost to replace if it got broken :-( > BMW *deliberately* make their speedos read high. Draw your own conclusions. SignatureRegards Steve G >SG: You should actually thank BMW (and all other European manufacturers >for that matter) for making their speedometers read high - they may well >save you from a speeding ticket one day. Sorry, but that's ridiculous. The thing's job is to report your speed. I no more want a high-reading speedometer than I want a fast-reading watch that may "save me" from being late to an appointment. I don't want to start anything new and don't feel like bringing myself into this but I often set my watch a minute or two fast just for that reason, although I do prefer a more precise watch. I guess you are right, it is better to have the precision and then decide what you want to do with it. Maybe bimmers read the speed high just in case there is a change to the car that would make the speedo read a little low and get you closer to a ticket? To be honest with you guys, I really don't care that much about what my speedo reeds, I don't think 2 mph in any direction will do any harm or good. > I don't want to start anything new and don't feel like bringing myself > into this but I often set my watch a minute or two fast just for that [quoted text clipped - 5 lines] > car that would make the speedo read a little low and get you closer to > a ticket? SG: It's not just BMW - ALL European manufactured cars do the same thing. In fact any car sold within the EU irrespective of where it is manufactured should also not read low - it's a legal requirement on the manufacturer (not the owner/driver I hasten to add). > To be honest with you guys, I really don't care that much about what my > speedo reeds, I don't think 2 mph in any direction will do any harm or > good. SG: Hooray, some sense at last :-) SignatureRegards Steve G >> I don't want to start anything new and don't feel like bringing myself >> into this but I often set my watch a minute or two fast just for that [quoted text clipped - 10 lines] > manufactured should also not read low - it's a legal requirement on the > manufacturer (not the owner/driver I hasten to add). It's not even an EU Thing. > > SG: It's not just BMW - ALL European manufactured cars do the same > > thing. In fact any car sold within the EU irrespective of where it is > > manufactured should also not read low - it's a legal requirement on > > the manufacturer (not the owner/driver I hasten to add). > It's not even an EU Thing. But there's no requirement for the speedo *not* to be accurate. Merely that it must not under-read. And assuming it's spot on with brand new tyres, wear on those will cause it to over-read. Signature*Why is it considered necessary to screw down the lid of a coffin? Dave Plowman dave@davenoise.co.uk London SW To e-mail, change noise into sound. >> > SG: It's not just BMW - ALL European manufactured cars do the same >> > thing. In fact any car sold within the EU irrespective of where it is [quoted text clipped - 8 lines] > And assuming it's spot on with brand new tyres, wear on those will cause > it to over-read. My guess, and my expereince by the way, is that the speedo is pretty close to dead nuts accurate with the largest factory tire available. As the tire size moves down the available fitment table, the speedomer becomes increasingly fast. I see no requirement for the manufacturer to precisely calibrate the speedometer for each and every tire that might be fitted on the car beczuse this would require the owner to recalibrate the speedo -- or the dealership -- when a factory optional tire package is installed. If they make the speedo work well for the largest tire they recommend/install, then all is well with the universe. >>SG: You should actually thank BMW (and all other European manufacturers >>for that matter) for making their speedometers read high - they may well [quoted text clipped - 4 lines] > fast-reading watch that may "save me" from being late to an > appointment. And I'm sorry that you didn't recognise my tongue in cheek comment. Whether you want a high reading speedometer, or not, is irrelevant. Current EU construction regulations make it compulsory for designers and manufacturers to ensure that the instrument doesn't read low under normal working conditions (i.e. taking into account tyre wear, differences in tyre circumferences, mechanical tolerances in the instrument itself and transmission, etc.) so they err on the safe side and make them read high. The fact that your BMW costs umpteen tens of thousands of Pounds/Euros/Dollars/whatever is also irrelevant. BMW don't actually make the speedo; they buy them in from whichever manufacturer offers them the best price and, therefore, potential profit margin. The fact that you don't like it bothers neither them nor I one little bit. SignatureRegards Steve G >>>SG: You should actually thank BMW (and all other European manufacturers >>>for that matter) for making their speedometers read high - they may well [quoted text clipped - 14 lines] >instrument itself and transmission, etc.) so they err on the safe side >and make them read high. That's quite a different statement than the one I responded to, which said something about thanking them for their high-reading speedometer. That's the statement that I objected to, understand? >The fact that your BMW costs umpteen tens of thousands of >Pounds/Euros/Dollars/whatever is also irrelevant. Please point-out where I claimed that my car's cost was relevant. Oh, you can't, because I didn't. >BMW don't actually make the speedo; they buy them in from whichever manufacturer offers >them the best price and, therefore, potential profit margin. Please point-out where I claimed that they did make the speedo. >The fact that you don't like it bothers neither them nor I one little bit. Please point-out where I asked whether twits like you are bothered by my opinion. The problem withthat analogy is that if YOU set youir own watch, then you know that it's fast, and you compensate accordingly. If _I_ set your watch fast, you will not know until you have had a chance to check it against another watch or clock. Assuming you do not get the chance, you will end up being early to wherever you are going until you figure out that the watch you are using is set ahead. Your car is much the same, if the speed limit sign says 45, or whatever, then you tend to plant the speedometer needle on what the sign says is the limit. The sign says 45, you go 45. You don't want the cops nailing you for doing 48, and they are not going to bother nailing you for going 42, so you prefer the speedometer to error on reading your speed faster than actual, not slower. The discussion was that some people claim that their speedo reads from 5 ~ 7 mph too high, I suggest that the error is really more like 2 mph too high, or maybe even a bit less. Mine is under 2 mph fast at 80, and I assume the error is pretty much linear, so it should be 2 mph fast at other speeds as well. I haven't measured this though, and I have no idea that the error is linear or proportional to the actual speed, and don't much care. >>SG: You should actually thank BMW (and all other European manufacturers >>for that matter) for making their speedometers read high - they may well [quoted text clipped - 4 lines] > fast-reading watch that may "save me" from being late to an > appointment. > Mine is under 2 mph fast at 80, and I assume the error is pretty much > linear, so it should be 2 mph fast at other speeds as well. I'd be most surprised if it were. It's likely to be a percentage error. Signature*And don't start a sentence with a conjunction * Dave Plowman dave@davenoise.co.uk London SW To e-mail, change noise into sound. > I set the Cruise to 80 and timed two miles by the mile markers; it took 92 > seconds to travel the distance. This works out to 78.26 mph, so the speedo [quoted text clipped - 4 lines] > > By the way, I run 225/45 x 17s on my 325i Convertable. You started a good one here, mate :-)) SignatureRegards Steve G |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.