sensors connection question

#1

Connecting TPS string pot to a Traqmate. What’s happening is I see a voltage change off of the pot with the car off or if I manually actuate the pot with the car running (i.e. 1.35v-2.57v idle to max throttle), but if I monitor the signal voltage while revving the motor I get hardly any change at all. I’ve tried both 5v and 12v input off the TM, and 12v off the car with similar results, and I’m getting steady 5v or 12v input voltage even while revving. All monitoring is done with a voltmeter at the pot, but I have verified its consistent with what I see at the Traqdata as well as on my Traqmate display. I’ve tried introducing a 1k/1w resistor in line with 5v and with ground with no change in results. I’m thinking current must be increasing from the alternator which is limiting the voltage change but I’m baffled.

The variable is the motor running. I’m not good enough at the electrical stuff to figure out what current is doing or where it’s coming from. Been trying different things and online reading for the better part of two days now and I’m out of ideas.

I figure I must have something wrong like missing a resistor, but the TM manual says 470 ohm minimum, and this is a 10k ohm resistance sensor. Do I still need a resistor on the 5v wire, and if so what ohm/watts is correct? I’ve tried with a 1K ohm / 1 watt. The other thing I might try is Ranger’s suggesting in another post to use an external power source to see if I can isolate

Many thanks for any guidance!

#2

I fought this sort of thing a lot. What is the “string pot”? You’re not going to get good throttle position info off of the TPS. If you were really determined to get that info, you’d have to install a secondary rotary or linear resister.

There’s 2 ways to get good analog data on to the Traqmate. The first way is stepper gauges with analog outs. These are pricey but it’s a good solution.

The 2nd solution is to get a better reference voltage. Get a 12V power supply chip from Mouser. It will take a wide range of Vin but always provide 12V out. Then figure out how much series resistance you need to keep the current demand down on your chip. At endstate you’ll probably end up with 5v of analog signal delta, instead of 1-2v of signal delta, and that will make your analog data accurate enough to be useful.

#3

I saw that in your other post, im going to do a test tmrw to see if i can get a clean 12v signal off a separate battery or something. A string pot is a variable transducer, except instead of rotary it works off a very thin cable and based on the changing length it varies the resistance and provides an outpit voltage.

I believe my issue is either the ref voltage as youve discovered and i will test, or its the amout of resistance. The odd thing is that i only fail to get a variable voltage if the motor is turning an rpm over about 2000. What resistance values didyou use and where did you locate in series? I did some reading today that led me to believe that anywhere between the source voltage and the ground would provide the same result but since all of the locations for my 1k oh,/1w resister provided no change to my results im unsure.

Sorry for my weak ass spelling. Im typing on my ipad and the soft keyboard blows and i dont want to go back and fix em all.

#4

I tried to write out a couple scenarios, but in the absence of a diagram it got too complicated.

You can’t use vehicle voltage because it’s no where near stable enough.

I considered using a battery as a source voltage too, but it just seemed like too much of a cludge. The voltage regulator from Mouser was easy and elegant.

Is there a voltage and/or current rating on the sensor?

There has to be two resisters in this, otherwise you are just measuring source voltage. This is called a “bridge” circuit, look that up so you can see a diagram.

Consider this…

Vs----o1----R1----o2----Rs----G
Vs,12V voltage source
o1, o2 places to measure voltage
R1, a current limiting resister, prob 10W
Rs, senser resistance
G, ground

V=I/R

Now play around with some values of R1 and Rs and figure out what the current thru the circuit is and what voltage you will see at o2. It sounds to me like you are measuring the voltage at o1 and that is always going to be the same as the reference voltage.

You need to pick at R1 such that the amount of current thru the circuit isn’t too high. If the voltage supply is rated at 1.5A for example, you’d probably want to pick a resistance such that peak current was limited to 750mA. Maybe even 500mA. This will keep the voltage supply cool and it’s V steady.

Lets say the sensor’s min resistance is at idle and it’s 10 Ohms. If R1 is 0, that means 1.2A would go thru the circuit and that’s too much current. But if R1 was 10 Ohms, now the total resistance is 20 Ohms and current is limited to 600mA, so now we’re getting in the ballpark.

So if the sensor’s resistance (Rs) is 10 Ohms…
Total R = 20 Ohms so I=600mA
V across R1 is 6V
V across Rs is 6V. This is also V as measured by the TM at locaton O2.

Now we move the throttle to WOT and note that the sensor reads 110 Ohms (Rs). Total resistance is now 120 Ohms so current is 100 mA.
Rs=110 Ohms
Total R = 120 Ohms so I=100mA
V across R1 is 1V
V across R2 is 11V. This is also V as measured by the TM at O2

So now you have to voltage #'s with which to tell the Traqmate about. Idle is 6V at O2 because there’s 600mA across the 10 Ohm resistance provided by the sensor (Rs). WOT is 11V at O2 because there’s 100mA going across the 110 Ohm resistance provided by the sensor (Rs)

Power checking. P=IV. The peak power scenario is WOT with 600mA across Rs@10 Ohms. That’s 6W which almost surely too much heat for the sensor to dissipate. So now you have to go back and make R1 bigger to reduce peak current to something the sensor can live with. Find a spec for the sensor. More than a couple hundred mA is probably too much.

It takes some trial and error to calibrate this with Traqmate. A surprising amount of trial and error. I’m not quite sure why that is.

#5

As Scott states, you will have to have to have a source of regulated power. The value of the current limiting resistor is a function of the sensor resistance, reference voltage and input range of the TracMate. In Scott’s diagram you read the sense voltage at o2.

For a 0-180 ohm sensor and a 5v reference supply a 180 ohm 1/4 watt resistor is a good combination. And will result in a voltage input over the range of travel of the sensor of approximately 0-2.5v

#6

I had good success with a string potentiometer that I got from gormans. But for the most part, I’m 100% full throttle. With our cars, I’m thinking time can be had in the brake zones, so if you want to get nerdy a pressure sensor to measure brake pressure would be a good tool. I bought one, but never got around to installing it. I’m debating on selling it.

#7

I probably study data more than most folks. Braking and accel info is already available to you via g data. One of the charms of TM is being able to compare your data to folks that are better than you. And it will be braking and accel g’s that you’ll be comparing. IMO trying to log throttle pos and braking psi isn’t worth the effort.

If you want to log something interesting, look into steering angle and yaw. Getting data on yaw would be really cool. Imagine being able to detect that you’ve got 3deg of yaw (slip angle) late in the turns, but never early. That would tell you that you’re not carrying enough speed into the first half of the turn. That would be primo data that no one else has that I know of.

#8

Hey guys, lots of testing today I think I’m on to something. I do have it working in certain configurations, so I’m getting close.

Rob, do you know what the size of the series resistor you have on your string pot?

My sensor is not like the VDO 0-180 ohm type, it is actually a 3-wire sensor with variable resistance 0-10K ohms, and it supports up to 30v source voltage. What I don’t know is the expected input current it’s not in the data sheet. Through my testing I’m pretty sure it’s expecting between 500-1000 ma, and when translated to Ohms Law using 5v I think my resister will need to be 50-100 ohms. But I want to confirm that with Kenny or from a working setup like yours Rob.

Oddly, I think the Traqmate supplies too little amperage (100 ma) and the car’s fusebox is delivering too much when the car is running (because the alt is delivering 60-70 amps or so). But when the car is off the amperage off the battery is low enough to allow the resistance/voltage to vary.

#9

[quote=“Age” post=54751]Hey guys, lots of testing today I think I’m on to something. I do have it working in certain configurations, so I’m getting close.

Rob, do you know what the size of the series resistor you have on your string pot?

My sensor is not like the VDO 0-180 ohm type, it is actually a 3-wire sensor with variable resistance 0-10K ohms, and it supports up to 30v source voltage. What I don’t know is the expected input current it’s not in the data sheet. Through my testing I’m pretty sure it’s expecting between 500-1000 ma, and when translated to Ohms Law using 5v I think my resister will need to be 50-100 ohms. But I want to confirm that with Kenny or from a working setup like yours Rob.

Oddly, I think the Traqmate supplies too little amperage (100 ma) and the car’s fusebox is delivering too much when the car is running (because the alt is delivering 60-70 amps or so). But when the car is off the amperage off the battery is low enough to allow the resistance/voltage to vary.[/quote]

The low current capability of the TM 5V reference is an issue, but the real problem is simply the low voltage. Once you set up your bridge, you’ll find that min to max sensor values range only across 2V or so and you’ll never get accurate data out of that. By going to a 12V reference voltage you’ll end up with sensor V values of 5-6V and that works.

Using vehicle voltage as a reference flat-ass doesn’t work. It’s not stable enough. The current that the alternator is putting out is not part of the problem tho.

#10

That makes sense - when I use a 5v reference signal, the transducer shows a very granular voltage change (in hundredths of a volt), but the TM shows it in “steps”. Over my 1.29v-2.57v change (a total range of only 1.28 volts), I was seeing 8 steps or increments. Now this is only on the “Check Inputs” function since I can’t get a test strip without GPS which would require taking my car out into the snow, but I am picturing my Traqmate strip looking pretty stair-stepped rather than a smooth line up and down.

A 12v voltage would provide a range roughly 250% the size of of the 5v, and assuming the same TM increment value exists the values would have much more granularity, although would still show some “stepping”.

Do you guys on Traqmates see the sensor voltage as a very smooth curve?

#11

The “stair stepping” in the data is because the TracMate has a low resolution A/D converter (probably only 8 bits). That along with the small change in sense voltage (1.28v) causes the “stair stepping”.

Since your string pot is a 0-10k ohm device you can use a larger current limiting resistor to increase the sense voltage swing and decrease the load on the regulated 5v reference in the TracMate. A 2k current limiting resistor will result in a sense voltage swing of 0-4v.

With the string pot having three terminals you have to make sure you hook it up right. A three terminal potentiometer has one terminal at each end of the resistance element and the third goes to the wiper. In this application you want to use the wiper and one end of the resistance element. Which end you use will determine whether the since voltage rises or falls with extension of the string. To get the best resolution, you need to rig up the string pot so that the string is almost retracted at idle and fully extended at WOT. This may require adding a sector wheel to the throttle to pull more string as the throttle opens.

#12

Age, here’s what my TPS graph looks like for a lap @ VIR. I am using the 5v reference from the traqmate, and my pot at idle is not fully retracted. At WOT it isn’t fully extracted either, so I had to play with the voltage ranges to get it where I liked it.

[attachment=1370]untitled.jpg[/attachment]

#13

I think you made an error.

P = IV = I^2*R = (0.600 A)^2(10 ohms) = 3.6 W

Still quite high but not as bad.

#14

[quote=“MrDomino” post=64545][quote=“Ranger” post=54734]
Power checking. P=IV. The peak power scenario is WOT with 600mA across Rs@10 Ohms. That’s 6W which almost surely too much heat for the sensor to dissipate. So now you have to go back and make R1 bigger to reduce peak current to something the sensor can live with. Find a spec for the sensor. More than a couple hundred mA is probably too much.
[/quote]

I think you made an error.

P = IV = I^2*R = (0.600 A)^2(10 ohms) = 3.6 W

Still quite high but not as bad.[/quote]
I think you’re right. Good correction.

We’re worrying here too much about the sensor’s ability do dump heat. Worry about current limiting so that power supply and power resister can survive.