Page 4 of 6 FirstFirst 123456 LastLast
Results 31 to 40 of 60

Thread: How do I...

  1. #31
    Variable Bitrate Grrrmachine's Avatar
    Join Date
    Sep 2005
    Location
    Warsaw, Poland
    Posts
    346
    I've bitten the bullet and ordered my brain, so as soon as its done I'll make a photo-write-up so that people can see what measurements they have to take, what wires to run and how to calulate what resistors to use.

    Im thinking about making a board that runs from a 12V molex from the PC, so that you get a stable supply voltage, with four voltage dividers on it (with 1k potentiometers), four inputs and four outputs. That way there's a simple plug-in-and-go solution for most old-style automotive sensors to the Fusion Brain. If I can put it on the same board I've built my tachometer on, then there's a real all-in-one car sensor solution there...

  2. #32
    Constant Bitrate
    Join Date
    Jun 2008
    Location
    Georgia, USA
    Posts
    188
    I can't wait to see how it goes...
    Thanks,

    Dan

  3. #33
    Variable Bitrate Grrrmachine's Avatar
    Join Date
    Sep 2005
    Location
    Warsaw, Poland
    Posts
    346
    Success! Using a 3k resistor I had lying around and my increasingly-abused multimeter, I just managed to get voltage readings from my car's VDO gauges. Nothing like accurate, but it's a start.

    The gauges have 12V and earth wires, plus the sensor wire, and those are the three I needed for this test. The 3k resistor was connected to the 12V at one end and the sensor wire at the other, and then I put my voltmeter between the sensor wire and the ground. Bingo! Three different readings at IGN ON, engine idle and engine hard rev.

    It would be very easy to replace the 12V wire I used with either a regulated 12V from the PC power supply or the Fusion's 5V wire. The earth wire is unimportant as long as the FB and power supply are connected to the same earth as the sensor you're measuring (ie, the car body) because, presumably, the FB references the analogue input voltage to earth in the same way my multimeter does.

    Make sense now?

  4. #34
    Constant Bitrate
    Join Date
    Jun 2008
    Location
    Georgia, USA
    Posts
    188
    That is the same thing described earlier by greenman. As he said though, you have to know the max resistance of the sensor to size the other resistor correctly.

    Using a regulated input voltage will stop the three different readings you saw. By using the +5vdc on the FB you don't have to use a voltage divider for the 12vdc.
    Thanks,

    Dan

  5. #35
    Variable Bitrate Grrrmachine's Avatar
    Join Date
    Sep 2005
    Location
    Warsaw, Poland
    Posts
    346
    I was expecting three different voltages; these were engine related sensors (oil temp and oil pressure) so that's just what I wanted to see.
    And you'll still need a voltage divider if you use the 5V from the Fusion; just not so big a resistor. Hence my idea of using variable resistors; you can tweak them until the voltages you are getting are within your range.

    What h3rk was saying I think bears some relevance to thermistors (I temperature-related resistors); they don't change their resistance in a linear way, but rather exponentially, which means that if your resistor is too big then you won't register a big enough voltage change for the FB to pick up.

  6. #36
    FLAC
    Join Date
    Oct 2006
    Location
    Las Vegas
    Posts
    1,286
    Quote Originally Posted by Dan2008 View Post
    I don't know??? I was just going by the solution greenman provided in an ealier reply.

    I'm SO confused...
    Quote Originally Posted by Dan2008 View Post
    Okay, I found the ranges for all of the senders.

    The temp and oil pressure are both 240-33 ohms (low-high) and the fuel level is 0-90 ohms (empty-full).

    Thanks,

    Dan
    Assuming that the parameters that are being measured are mid-range (i.e. when the engine is running and heated up and everything is normal, the resistance of the sensor is about 1/2 of the range), and that's a big if, then for your temp and oil, the mid value would be 240-33/2 =~ 100 ohms

    {R1 is paired resistor, R2 is sensor}
    Using Vout=Vin(R2)/(R2+R1) volts should be about 2.5 out with 5 in. for a given change in resistance of R2, say 10 ohms, the difference in voltage will be larger closer to R2= minimum. A larger voltage change for a given change in resistance means 'more sensitive'

    If you want the voltage out to be more sensitive at high temps (resistance is low for NTC sensors) then R2 should be below R1. If the sensor's voltage should be more sensative at higher resistances, then R2 should be above R1 (using your diagram as an example).

    See in the beginning you wanted the top resistor to "knock down" what would have been a 0-12 signal to something that wouldn't go above 5V. So the main concern was finding a proportion of R1 to R2 that would accomplish that. But now, fortunately, you discovered a way around that, and no longer have to worry about that. Now you just have to find an R1 that gives a good range, in this case, smaller would be better for giving you the 5 V range (or close) but because the resistance of your sensors are so low... unusually low in my book, you should give some thought to how much current each of these sensors draw. Unfortunately, limiting current with a bigger R1 will limit the range of voltage available.

    If you used a 10 ohm R1, the max voltage would be 4.8V which is a nice spread, but the sensor current would be 1/8 amp. The brain can only source 1 amp of current, total, so this would overload the brain if all 10 channels were done this way. With a 330 ohm resitor your max voltage would only be 2.1, but your current would be 13.8 mA maximum. That's probably too much. So it's a balancing act.


    Here's a little spreadsheet I made with 10 channels and all of them were set up with the range of your temp/press sensors. It tracks individual and total max amps, and highlights in yellow what the max voltage will be. It is a simplistic way to try out different paired resitors. It assumes you use FB's 5V as Vin for all of the sensors.

    The initial point I made about more sensativity near the midpoint escapes me now, it just doesn't make sense with this example. I'll have to find where I saw that and clarify the context. Now I think it may apply more toward the initial way you were going to do this with 12V input. Now I'm just not sure, sorry. Anyway, I hope some of what I said helps.

  7. #37
    Variable Bitrate Grrrmachine's Avatar
    Join Date
    Sep 2005
    Location
    Warsaw, Poland
    Posts
    346
    unfortunately when the car is up and running the resistors won't be midway: the main function of the oil temp is to tell you when the engine is warmed up enough for hard revving, so the usual resistor value will actually be pretty close to the end of the desired measured range (ie usual running temp is maybe 90 degrees, and if it goes over 120 then that's the worry point.) Similarly with oil pressure, which actually has its highest resistance at engine start and drops its value as the oil warms up.

    Water temp will be same as oil temp - in fact, the only "regular" resistor would be the fuel gauge, which is basically a rheostat controlled by a float. That's the only one that would sweep in a linear way between its highest and lowest resistance.

  8. #38
    Variable Bitrate Grrrmachine's Avatar
    Join Date
    Sep 2005
    Location
    Warsaw, Poland
    Posts
    346
    Well, I've made my input board:



    Labels added in Photoshop. Im no electrical engineer, which is why there are lots of little wire tracks all over the place which make it look a little clumsy, but its functional and I get readings from all parts of it. It's basically two frequency-to-voltage converters and four voltage boards, where each terminal block just needs the relevant input from the car's sensor and then passes back out to the Fusion Brain. Tweaking comes from the pots, which are 47k for the F-V converters, then two 470R, a 10k and a 3k. The whole thing can be powered from the CarPC's PSU or any other regulated source.

    Just waiting for my brain to be delivered now

  9. #39
    Newbie
    Join Date
    Aug 2008
    Posts
    3
    I don't know a thing about the FB, but I do know a bit about auto gauges! The sensor itself is nothing but a variable resistance between the power source and ground. The FB and sensor share a ground, so that's not a problem. The gauge is nothing but a volt meter. 5V from the regulated power supply runs through the gauge then through the sensor to ground.

    This is where it all comes to a stand still for me since I don't know exactly how the FB works. If the FB supplies 5V on the inputs then just hook it up. If it needs 5V supplied then you need an adjustable 0-5V PS tied into the line. Adjustable PS regulator chips use a resistor to control voltage output. Tie the sensor into that line as the resistance, the variable voltage output to the FB. You'd need a separate 0-5V PS for every input. If the FB supplies 5V on the input lines (I doubt that!), then you just connect the sensor and you're done. All the instruments in a vehicle use the same 5V regulated power source, but if the FB doesn't have 5V on the input I can only see using a variable 5V PS for each unit or the FB would see the combined resistance of all. Variable PS chips don't cost much, luckily! You could build a small board with all the variable PS chips on it and mount that close to the FB.

    The oil pressure sending unit works the same way. AMC typically used Ford type instruments. I'm an AMC nut and have a set of factory manuals from 55-82, so I have all the specs!

    On early models (prior to 1978) 0 psi should be 69-77 ohms, 10 psi 35-38 ohms, 60 13-15, 80 9.5-10.5. You're not looking for voltage, you're looking for resistance! Stewart-Warner gauges may use the same range. Since you indicated the car uses a 30-240 range for the fuel sending unit, it probably uses a similar range oil pressure sending unit (1978 and later, though a few vehicles, such as Jeep, continued using the older style). In that case 0 psi is 240 ohms, 60 psi is 67 ohms. My manuals don't show any other test points. You'll have to measure resistance of the unit to determine which it is. Just connect the ohmmeter between sensor case and output terminal with no pressure. If the reading isn't within the 5% range of 240 ohms or isn't in the 69-77 ohm range, it's either defective or not one of those type units. Some aftermarket gauges use different ranges, and GM and Chrysler may use different ranges.

    All the mechanical gauges are called "5% gauges". In other words, they have a 5% range -- +/-2.5% -- at every value.

  10. #40
    Fusion Brain Creator 2k1Toaster's Avatar
    Join Date
    Mar 2006
    Location
    Colorado, but Canadian!
    Posts
    10,049
    The FB has an onboard regulator. You can hook it up directly to the car.

    Each input has a +5v supply, ground pin, and the actual input pin.

    And the FB reads voltage, not resistance. Since most of the sensors are resistive like you said, you need another resistor to make a voltage divider.
    Fusion Brain Version 6 Released!
    1.9in x 2.9in -- 47mm x 73mm
    30 Digital Outputs -- Directly drive a relay
    15 Analogue Inputs -- Read sensors like temperature, light, distance, acceleration, and more
    Buy now in the MP3Car.com Store

Page 4 of 6 FirstFirst 123456 LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •