Cheap Sensors

[Rodiman]

Nothing good comes cheap, blah blah blah. I have to imagine that there are very cheap alternatives to buying sensors from the main motorsports sensor companies or big DAQ companies. Typically the guts of any sensor is based on individual components that are dirt cheap. Granted, that purpose built sensor will be more accurate, have a faster time constant, or more resolution, but could a team on a budget not get usable data for a handful of change?

Our team could only afford 2 linear pots from the suspension, and ran them diagonally opposed to estimate 4 wheel data. A few years ago, Claude asked, "how many wheels do you have? Then why do you only have 2 poh-tench-ee-oh meters?" He then proceeded to talk about a team that made a usable system with cheap radio shack rotary pots and told us that budget should never be an excuse, because you can almost always come up with a clever solution on the cheap.

For example: Cheap chinese eBay IR guns. Can they not be taken apart, repackaged and become IR tire temp sensors? They have to have some voltage output that could be read.

Wheel speed sensors are relatively expensive for what they are. Could those not be replaced by simple hall effect PCB's?

[Warpspeed]

Oh absolutely Stephen!

Cheap, homemade or improvised sensors and transducers can work really well if they are carefully calibrated and installed with tender loving care.

Now maybe they won't last 300,000 road miles, or work over aerospace temperature ranges, or survive 50G shock and vibration, or submersion to 200 feet.

But they can work wonderfully well over a short intensive do it yourself on a shoe string R&D effort.

In the end, it is what you do with the data that really counts.

[MarkSchaumburg]

You can get Melexis IR temp sensors for ~$12 on digikey. These can be configured to output a PWM signal. Run this through a low-pass filter (1 resistor and 1 capacitor) and you have a voltage output with relation to temp. They also have an I2C interface, so if you have a competent EE/CE or 2, it's really simple to make an IR to CAN device using a microcontroller. We're running 3 of these on each wheel and logging the data via CAN to a DL1. Overall it'll cost you about $50 in parts per wheel compared to ~$200 per individual IR sensor from somewhere like Texense.

Mark Schaumburg
Head of Electronics and Data Acquisition (2011-2013)
Mizzou Racing
University of Missouri - Columbia

[Drew Price]

Exactly. If you have people who can read (and make sense of) the datasheets of sensors on Digikey / Mouser / Etc., then you can find cheap industrial alternatives.

Just keep in mind that a lot of industrial sensors are meant to be powered by 24vdc, check the power requirements extra carefully and you will be fine.

[Pennyman]

Sparkfun electronics all day long.

[Zac]

I'm still waiting for a team to show up with the junkyard sourced Cadillac night vision cameras Bill Cobb has mentioned a couple times.

[MCoach]

Originally posted by Zac:
I'm still waiting for a team to show up with the junkyard sourced Cadillac night vision cameras Bill Cobb has mentioned a couple times.

...In all seriousness, Baja teams have talked about it. They want it for a rear view camera.

[Racer-X]

Home made daq is going save you a ton of money and would be a neat project. What you save in $$$$ you will be paying for with time and headaches though. Totally possible and cool in scope of fsae none the less.

[Claude Rouelle]

This what I have learnt in terms of cheap sensors

1. What is cheapo is not good. What is good is no cheap.

2. Three goals: Quick (delivery), good (quality), cheap. Pick two.

3. Worth to pay a bit more for the possibility to be in contact with the sensors customer support service and ask any question and tips and trick you want.

4. Beware of manufacturing copyright infringement (Did I say Asia?). If you do not want your intellectual property to be stolen, do not stole others rights. What goes around comes around. Do not do to others what you do not want others to do to you.

[Menisk]

In terms of cheap position and pressure sensors we're about to have a crack at using Honeywell RTY's for suspension position and steered angle and SSI P51 series pressure sensors for fuel pressure and both brake circuits. When I've actually had some experience with them I'll post here and let people know how they went, but on paper both look good.

Honeywell's should stand up real well. Weather sealed sensor with an amp termination and they're hall effect rather than resistive so no friction on a resistive strip to fuck shit up. Hard to justify couple hundred dollar linear pots when we can get these at 70AUD(ish) each.

[Warpspeed]

Claude.

Often we just need to data log something to track down and nail some vexing problem.

It may be fluctuating oil pressure, fuel pressure, induction air temperature, suspension movement, steering angle or whatever....

Now suppose for example, we have an oil surge problem turning in one direction, and we wish to experiment with a variety of oil pickups and baffles.
So we fit an oil pressure transducer and an accelerometer to our logger, go for a thrash around our test track.

Now we don't need to know oil pressure to +/- 0.01 psi with a fifteen hundred dollar fully certified pressure transducer.
A ten dollar Mickey Mouse transducer that is +/-1 psi would tell us everything we need to know.

Either the oil pressure is holding up quite well, or it is frighteningly unstable.
That's all we need to know, one or the other, no splitting of fine hairs necessary.

Likewise many other data logging situations, such as mysterious engine overheating problems.

We don't need to know temperatures with exact fractional degree precision, just a whole bunch of air and water temperatures monitored around the whole system to get a handle on what may possibly be going wrong.

[B Lewis @ PE Engine Management]

For 'inexpensive' sensors we always first try to find an example of a sensor used on a production vehicle. Good examples are TPS sensors to measure suspension travel and GM oil pressure sensors for fuel pressure, oil pressure, etc. These sensors are already designed for the automotive environment and can be modified to work in many other instances.

[Mbirt]

Speaking of cheap sensors, GM has a new small, lightweight flex fuel sensor that costs only $50 new. The old bulky $400 sensor had factors working against it for FSAE use, but this one might be something worthwhile for teams running the mystery e85 at comp.

[jlangholzj]

Originally posted by Mbirt:
Speaking of cheap sensors, GM has a new small, lightweight flex fuel sensor that costs only $50 new. The old bulky $400 sensor had factors working against it for FSAE use, but this one might be something worthwhile for teams running the mystery e85 at comp.

birt you have a link to that guy? I did a quick google at work and didn't come up with anything.

[Mbirt]

Originally posted by jlangholzj:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Mbirt:
Speaking of cheap sensors, GM has a new small, lightweight flex fuel sensor that costs only $50 new. The old bulky $400 sensor had factors working against it for FSAE use, but this one might be something worthwhile for teams running the mystery e85 at comp.

birt you have a link to that guy? I did a quick google at work and didn't come up with anything. </div></BLOCKQUOTE>Here's the one I used on our snowmobile, John: http://paceperformance.com/i-1...fuel-sensor-e85.html (http://paceperformance.com/i-10801493-13577429-fuel-composition-flex-fuel-sensor-e85.html)

[TMichaels]

As Brian mentioned, you should always look for sensors from production vehicles.
I can assure you that everything that you need is mounted to current mid-class or premium vehicles including exhaust gas temperature and pressure sensors, acceleration sensors, potentiometers, wheel speed sensors etc.. They are cheap, reliable and usually provide sufficient accuracy. Additionally you may buy/order them at the next official brand dealer or get them from the junk yard. Getting data sheets / documentation is sometime hard though and some are quite bulky.

[murpia]

Originally posted by TMichaels:
As Brian mentioned, you should always look for sensors from production vehicles.
I can assure you that everything that you need is mounted to current mid-class or premium vehicles including exhaust gas temperature and pressure sensors, acceleration sensors, potentiometers, wheel speed sensors etc.. They are cheap, reliable and usually provide sufficient accuracy. Additionally you may buy/order them at the next official brand dealer or get them from the junk yard. Getting data sheets / documentation is sometime hard though and some are quite bulky.
I agree with Tobias. If you can find a suitable production sensor it will be usually be preferable to a 'motorsports' sensor, especially in cost, but also durability and repeatability.

I see an opportunity here for a 'wiki' or other collaborative online resource. If your team finds a good sensor, let others know about it!

Regards, Ian

[B Lewis @ PE Engine Management]

One more thought..... Even with production vehicle sensors, a very high level or accuracy is attainable if you calibrate the sensor yourself using known inputs.