jeh@cmkrnl.uucp (05/31/91)
In article <1970@ole.UUCP>, ssave@ole.UUCP (Shailendra Save) writes: > [...] > Sometime ago, I asked the net about a weekend project to > make a tachometer for my car. I got numerous replies, for > which I am thankful. Hopefully I will soon have a complete > digital dash. I will post a summary for those interested. Projects for such have been appearing in the hobby mags every now and then for almost forever. There was even one in one of the trade magazines, _EDN_ I believe, circa 1974 (!) or 75 or so. That one was for a motorcycle. Personally I would steer clear (ha ha) of a digital dash. Saab, who builds fighter planes as well as cars, has done extensive research on what kinds of gauges can be read the most easily. The answer? Old-fashioned analog, with a moving pointer against a stationary background. The pointer and scale can be electronically generated (ie on a computer-generated display) or mechanical, circular or bar-graph style, but whatever the variant, it beats a digital display six ways from zero. You see, it is usually far more important to know that "you're nearly at redline" than that "engine speed is 6,253.8 RPM", and it's MUCH more important to know that "your oil pressure is low" than that "your oil pressure is 21.3 psi". For rapidly-changing things like a tach, the pointer and scale have another big advantage: It is easy to see how fast the reading is changing. This is almost impossible to derive from a digital readout. Color-coding the scale for good, warning, and bad ranges -- or adding color-coded idiot lights -- obviously helps to draw attention to those gauges that need reading at the moment. But color-coding gets tricky even for something as simple as an automobile engine. For instance, my water temp. gauge swings almost up to the red part of the scale as the car warms up, and then the thermostat opens and it goes down to the middle of its range where it's supposed to be. What would be nifty would be something that would draw my attention to that gauge if the reading stayed up near the redline for more than a few minutes. How about a heads-up display? And ONLY display those instruments that are indicating out of "nominal" range. Mount a backlit LCD atop your dash, wire it so the display is mirror-imaged, and read its reflection off the windshield. Of course since both the inner and outer surfaces of the windshield will reflect you will get some ghosting in the vertical dimension. This is, I think, why the fighters with heads-up displays have a special piece of optic surface just for this purpose (with an antireflection coating on the far surface). --- Jamie Hanrahan, Kernel Mode Consulting, San Diego CA Internet: jeh@dcs.simpact.com, hanrahan@eisner.decus.org, or jeh@crash.cts.com Uucp: ...{crash,scubed,decwrl}!simpact!cmkrnl!jeh
tindle@ms.uky.edu (Ken Tindle) (05/31/91)
In article <12974@mentor.cc.purdue.edu>, wilker@gauss.math.purdue.edu (Clarence Wilkerson) writes: }I used to have an add-on cruise control for a Plymouth. }The pickup was activated by 2 small magnets taped and glued }to the drive shaft, fairly near the transmission. The weight }was small, and fairly balanced anyway. In article <54642@rphroy.UUCP>, vbreault@gmr.com (Val Breault) writes: }There's a lot of stuff happening under your car. Stones, water, vibration, }dust. Then too, the driveshaft will wobble several thousandths on it's spline. }Rather than attempt to make something that will work in that rather hostile }environment, why not try this idea? }In the days before automotive computerization, a cruise control setup }consisted of a mechanical device that was attached via flexible cable }between the transmission and speedometer. You should be able to get one }of these pretty cheap at a nearby junk yard (oops, 'Automotive Recycler'). An interesting idea- how did this mechanical setup work? It seems to me that an encoder wheel attached to the speedometer cable, with an interrupter IR diode/ phototransistor pair is the way to go. You have to rig the cable drive to pass through, since it is illegal under Federal law to disable the odometer on purpose. This has the benefit of being more universal as well, since one can find the speedometer cable on older cars without too much hassle (fiddling with it in such a way that it doesn't break in two later is the hard part). This will get you away from the driveshaft, and provide pulses you can count in a digital circuit. With these pulses, you can use the circuit provided by Dakota Digital in Radio-Electronics. That speedometer circuit also has a twenty-step LED bar driven from two LM3914N's for an "analog" display, acknow- ledging the need for such. You have to change their input conditioning stage, since they use an inductive driveshaft sensor. It's interesting to note that most aftermarket cruise controls now use a connection to the ignition system instead of the driveshaft, which works IF you don't have an "overdrive" transmission and this transmission is in its highest gear (1:1) all the time the cruise is on. So, can one of these older in-line mechanical devices be made to output a pulse train? --------------------------\ /----------------------------------------------- INTERNET:tindle@ms.uky.edu | "Could you please continue the petty bickering? BITNET: tindle@ukma.bitnet | I find it most intriguing." --- Data, Ken Tindle - Lexington, KY | Star Trek, The Next Generation, "Haven" --------------------------/ \-----------------------------------------------
vbreault@rinhp825.gmr.com (Val Breault) (05/31/91)
...<discussion of using early cruise control as part of digital dash>...
> An interesting idea- how did this mechanical setup work?
In the olden days, when the only electronics on a car was the radio, the
cruise control worked this way:
The speed signal was carried by the speedometer cable to a mechanical
device that was positioned under the hood. Another speedometer cable
carried the speed signal from that device to the speedometer itself.
The device compared the desired speed against the actual speed and
created an error signal using manifold vacuum. The vacuum was supplied
to a diaphragm that actuated the throttle.
If you get one of those buggers you can gut it and salvage some nifty stuff.
The housing with it's bearings and speedometer cable connectors will get
the speed signal to the existing speedometer. With the mechanical comparitor
out of the way, you'll have plenty of room to mount your magnetic pickup.
-------------------------------------------
Val Breault - vbreault@gmr.com
The opinions expressed by the author do not necessarily reflect the position of the General Motors Research Laboratories or the General Motors Corporation.
tonya@hpldsla.sid.hp.com (Tony Arnerich) (06/01/91)
I added on an after-market cruise control that used this setup. A magnet and a weight (or was it 2 magnets?) was tie-wrapped onto the driveshaft (the half-shaft, actually - the car was a SAAB with FWD, and ~1" dia shafts). The inductive pickup was mounted on a sheetmetal bracket that I custom-bent to get about 1/4" clearance. It worked very well for at least 6 years (until I traded the car in) without any need to readjust, replace, or repair. Maybe you could find the relevant parts as a replacement kit? tonya@sid.hp.com
john@newave.UUCP (John A. Weeks III) (06/03/91)
In article <1991May30.151040.50@cmkrnl.uucp> jeh@cmkrnl.uucp writes: > Personally I would steer clear (ha ha) of a digital dash. I would like to see both types of displays in my car. I like the analog needle to see rate of change, but I would also like to have a digital display to see the precise speed my car is moving at. For example, I know the State Patrol in Wisconsin (a very picky state about traffic speed) normally does not give out speeding tickets unless you are going 5MPH or more over the speed limit. I would like to set my cruise control at exactly 69MPH on the interstate. With the analog display, I never really know exactly how fast I am going. Of course this assumes that a digital display is actually accurate to the resolution that it displays at--which is not a good assumption unless it has been calibrated. -john- -- ============================================================================= John A. Weeks III (612) 942-6969 john@newave.mn.org NeWave Communications, Ltd. ...uunet!tcnet!newave!john
lester@infonode.ingr.com (Lester Bartel) (06/03/91)
john@newave.UUCP (John A. Weeks III) writes: >In article <1991May30.151040.50@cmkrnl.uucp> jeh@cmkrnl.uucp writes: >For example, I know the State Patrol in Wisconsin (a very picky state >about traffic speed) normally does not give out speeding tickets unless >you are going 5MPH or more over the speed limit. I would like to set >my cruise control at exactly 69MPH on the interstate. With the analog >display, I never really know exactly how fast I am going. >Of course this assumes that a digital display is actually accurate to >the resolution that it displays at--which is not a good assumption unless >it has been calibrated. The speed of your readout can vary by over 2% depending on if you have new or wornout tires on your car. Figure about 3/16 inch usable tread on a tire. Subtract 2 times this from the diameter of a new tire and recompute the circumference. At 60 MPH thats over 1 MPH error. So, even if the display is accurately calibrated, tire wear will cause some error. Also, there is some tire slipage depending on road material and conditions like water or sand (or ice :-) ). -- Lester Bartel b23b!naomi!lester@ingr.com Dazix, An Intergraph Company uunet!ingr!b23b!naomi!lester
ssave@ole.UUCP (Shailendra Save) (06/03/91)
In article <1991May30.151040.50@cmkrnl.uucp>, Mr. Haha Jamie writes: > > Projects for such have been appearing in the hobby mags every now and then for > almost forever. There was even one in one of the trade magazines, _EDN_ I > believe, circa 1974 (!) or 75 or so. That one was for a motorcycle. I am aware of this. The idea behind this venture is to make more than a just the run of the mill "digital dash" Most articles (projects) in the mags do not come close. Even if they did, I would not use them. Takes the fun out of experimenting. > Personally I would steer clear (ha ha) of a digital dash. Saab, who builds > fighter planes as well as cars, has done extensive research on what kinds of > gauges can be read the most easily. The answer? Old-fashioned analog, with a This is perfectly true. But in the case of the speedometer, I would certainly like to have a completely digital readout. It is just easier to read. The tach remains "analog" though I am still using comparators to make comparisons to light LEDs. The advantage: I can use different colored leds. I have a green section, an orange, and a red. When I am in the orange, I know I am approaching red. As far as reading battery voltage, I do need to know when I am charging and when I am discharging. (think of multicolored digital 7-seg displays). > moving pointer against a stationary background. The pointer and scale can be > electronically generated (ie on a computer-generated display) or mechanical, > circular or bar-graph style, but whatever the variant, it beats a digital > display six ways from zero. You see, it is usually far more important to know > that "you're nearly at redline" than that "engine speed is 6,253.8 RPM", and > it's MUCH more important to know that "your oil pressure is low" than that > "your oil pressure is 21.3 psi". > BTW, the resolution of the tach is 50 rpm. (borderline too much) and oil will be green-orange-red. > For rapidly-changing things like a tach, the pointer and scale have another > big advantage: It is easy to see how fast the reading is changing. This is > almost impossible to derive from a digital readout. > > Color-coding the scale for good, warning, and bad ranges -- or adding > color-coded idiot lights -- obviously helps to draw attention to those gauges > that need reading at the moment. But color-coding gets tricky even for Not so idiotic. The way I designed all of the stuff, is that as long as all the lights are green, I am fine. At a glance, I can find out if anything is wrong without actually taking the complete reading. For example, if I look down, and see everything green, I know everything is normal. If I look down, and see rpm approaching red (ie orange) I do a double take and see how close I am to it in terms of actual rpms.. > something as simple as an automobile engine. For instance, my water temp. > gauge swings almost up to the red part of the scale as the car warms up, and > then the thermostat opens and it goes down to the middle of its range where > it's supposed to be. What would be nifty would be something that would draw my > attention to that gauge if the reading stayed up near the redline for more than > a few minutes. I think you need to have your thermostat looked at. The temp should approach red. Never be *in* it. Especially with summer coming up. > > How about a heads-up display? And ONLY display those instruments that are > indicating out of "nominal" range. Mount a backlit LCD atop your dash, wire it > so the display is mirror-imaged, and read its reflection off the windshield. Of > course since both the inner and outer surfaces of the windshield will reflect > you will get some ghosting in the vertical dimension. This is, I think, why > the fighters with heads-up displays have a special piece of optic surface just > for this purpose (with an antireflection coating on the far surface). Thought of this one. Learjets have this. The only problem with this is that at noon or high sun, the reflection is almost non-existent. Would require high powered laser leds to light this. One final note. Digital becomes better when there is something intellegent you can do with the reading. I intend to use a 6811 microcontroller which will tell me things like trip distance, miles per gallon, total distance, shift up (rpm too high), shift down (angle of car upwards or downwards, rpm too low) and other things like that. Why all this? Because it is fun. Because it comes free. I have lots of free ports on the I/O. Because it is a challenge. Cars are some of the noisiest environments you can work with. Because I like to do things myself. And finally, because I don't have a tach or battery guage on my dash ^_^ > > --- Jamie (Haha)Hanrahan, Kernel Mode Consulting, San Diego CA Shailendra ssave@caen.engin.umich.edu
agodwin@acorn.co.uk (Adrian Godwin) (06/03/91)
In article <847@newave.UUCP> john@newave.mn.org (John A. Weeks III) writes: >Of course this assumes that a digital display is actually accurate to >the resolution that it displays at--which is not a good assumption unless >it has been calibrated. Don't just calibrate the digital speedometer - the pulse train it gets from the speedo is dependent on wheel diameter, which is dependent on temperature as well as tyre wear. If you want accuracy, consider using a Gunn-diode microwave unit to measure ground speed by doppler shift. -adrian -- -------------------------------------------------------------------------- Adrian Godwin (agodwin@acorn.co.uk)
jeh@cmkrnl.uucp (06/04/91)
In article <1973@ole.UUCP>, ssave@ole.UUCP (Shailendra Save) writes: > In article <1991May30.151040.50@cmkrnl.uucp>, Jamie Hanrahan writes: >> Projects for such have been appearing in the hobby mags every now and then for >> almost forever. There was even one in one of the trade magazines, _EDN_ I >> believe, circa 1974 (!) or 75 or so. That one was for a motorcycle. > > I am aware of this. The idea behind this venture is to make more > than a just the run of the mill "digital dash" Most articles > (projects) in the mags do not come close. Even if they did, I would > not use them. Takes the fun out of experimenting. Then, taking your position to an extreme, you shouldn't be asking the net for hints on driveshaft speed pickups -- experiment and figure it out for yourself! Seriously, stumbling down blind alleys where others have gone before takes some of the fun out of experimenting too. It's useful to look at others' work if only to get an idea of what not to do -- not necessarily for making a direct copy. That _EDN_ project, for instance, used only SSI and MSI TTL!!! And it provided only "dumb displays". No way would anyone copy that today. Still, there might be some good ideas on signal pickup techniques, if nothing else. >> . For instance, my water temp. >> gauge swings almost up to the red part of the scale as the car warms up, and >> then the thermostat opens and it goes down to the middle of its range where >> it's supposed to be. What would be nifty would be something that would draw my >> attention to that gauge if the reading stayed up near the redline for more than >> a few minutes. > > I think you need to have your thermostat looked at. The temp should > approach red. Never be *in* it. Especially with summer coming up. I think you need to read the articles you respond to a bit more carefully. I said it gets "almost up to the red part", not *in* it. The car is running fine, thank you, and can take either side of the Grapevine at prevailing traffic speeds in the summer with the air conditioner *on*, with not even a hint of overheating. But my underlying point is that a simple good-warning-bad indicator, based solely on statically-defined ranges, is not really adequate. And as long as you're using a micro, you *can* do a better job. In my example, I'd want to be warned if my temp gauge was up *near* the red area for more than fifteen seconds or so, even if it isn't *in* the red area, because that would be a change from the car's typical behavior. Similarly I'd want to be warned if the gauge did *not* get near the red area during warm-up -- either the thermostat is staying open all the time, or the temperature sensor has gone bad. > One final note. Digital becomes better when there is something > intellegent you can do with the reading. No argument here! But that doesn't mean that the readouts have to be digital, only the processing. > I intend to use a 6811 > microcontroller which will tell me things like trip distance, miles > per gallon, total distance, shift up (rpm too high), shift down > (angle of car upwards or downwards, rpm too low) and other things > like that. Why all this? Because it is fun. Because it comes free. > I have lots of free ports on the I/O. Because it is a challenge. > Cars are some of the noisiest environments you can work with. > Because I like to do things myself. And finally, because I don't > have a tach or battery guage on my dash ^_^ Hey, no one said "don't do it"! All I said was "*I* wouldn't use digital readouts on the dash". Go for it. Experiment. Try lots of options. Interface to a G-Meter while you're at it!!! Then write it up and sell it to one of the hobby mags, both to recoup some of your parts costs, and so that the rest of us can learn from YOUR experience! --- Jamie Hanrahan, Kernel Mode Consulting, San Diego CA Internet: jeh@dcs.simpact.com, hanrahan@eisner.decus.org, or jeh@crash.cts.com Uucp: ...{crash,scubed,decwrl}!simpact!cmkrnl!jeh
humphrey@hpcvca.CV.HP.COM (Tom Humphrey) (06/08/91)
For most Ford, GMC, VW and Chrysler applications, There is a company that I am aware of that sells a pulse signal generator for these applications. The unit attaches to the transmission where the speedo cable goes in and puts out a given number of pulses per revolution. This is a company that makes and sells custom gauge sets. Classic instruments at (503)- 655-7900