Various newby questions.

All sensor specific discussions! Temperature, pressure, steering angle, brake and throttle, etc. Post adaptations of OEM-style sensors, and also your clever DIY hacks and custom designs here too!

Moderators: JeffC, rdoherty, stieg

Post Reply
davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Various newby questions.

Post by davef_dci »

Howdy,

Chumps endurance race team running a 1996 Dodge Neon. We've ordered Racecapture Pro and are excited to hear it's on its way. First race in is a month so we've got some work to do and will certainly need some help from this forum. A few introductory questions.

The Neon is pre-OBDII so that's not an option. The car is currently configured with the following analog after-market sensors/gages. All of these are redundant the inputs to the ECU - they were purchased as secondary sensors/analog gauges.

RPM - Tach
Wideband O2
Oil Pressure
Engine Oil Temp
Transmission oil temp
Engine Coolant Temp

First question might be a dumb one. Is it worthwhile to try to run both our analog gauges and an on-dash display via RaceCapture? With the input to the analog gauges being 12v I assume we'd need to work up some voltage divider to get to the 5V max signal RaceCapture is expecting. The only reasons to continue to run the old analog gauges AND a tablet in the car are a) as a fall back in case we can't get this system up and running before our first race b) the analog tach has a bright shift light and I noticed the Racecapture shift light is on back order and c) we are slightly concerned about visibility of the tablet in bright sunlight. All this being said, if running 12V into the gages and then somehow getting that down to the 5V needed by Racecapture is difficult or possibly dangerous to our new hardware we might want to go all-in and simply count on getting racecapture up and running. Anyone have any advice here?

Second question - for this style of car running conventional factory OEM ECU, is the Coilx isolation board likely required or do you think we can run directly from the available signal?

Thanks in advance and we really appreciate the advice and help!

Dave

davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Post by davef_dci »

Hmmm... a bit more information and another question. We just noticed the statement on the web site that

Low resistance gauge type sensors are not recommended. Which is the type of sensors we have.

Cany anyone expand on what's meant by "not recommended"? Is that "not recommended" as in not accurate or is it not recommended as in "will damage your hardware"?

brentp
Site Admin
Posts: 6274
Joined: Wed Jan 24, 2007 6:36 am

Post by brentp »

Hi,

The newer versions of RaceCapture/Pro (MK2, MK3) as well as RaceCapture/Apex all have very high impedance inputs and can generally safely be used to piggy back on existing sensors.

Each sensor should be considered separately, and some can be more easily tapped into and piggy-backed than others.

The overall theme you'll see below is that RaceCapture/Pro can map *any* 0-5v sensor, providing you have the documentation for the sensor calibration, or you spend the work to reverse engineer the sensor.

We have a general guide for mapping analog sensors that should be helpful: https://wiki.autosportlabs.com/RaceCapt ... ate_Sensor

And our sensor guide here: https://wiki.autosportlabs.com/RaceCapturePro_Sensors

Overall, it is really helpful to know the response curve of the sensor. If you don't know the response curve and insist on using it, you'll need to reverse engineer the sensor output range and map the voltages.



RPM:

If there is a reasonably clean tach output from the ECU, you can use that. If you have to tap into the primary trigger wire of the coil, you'll need the CoilX adapter.

Two points of interest:
https://wiki.autosportlabs.com/RaceCapturePro2_RPM
https://wiki.autosportlabs.com/CoilX

Wideband O2:

Presumably you're using a wideband that has a calibrated output. You can map that output using the general guide here:
https://wiki.autosportlabs.com/RaceCapt ... ate_Sensor

Oil Pressure:

Recommended to use a high quality dedicated sensor with a 0.5 - 4.5v linear output:
https://www.autosportlabs.com/product/1 ... re-sensor/

Alternatively you can tap into an existing sensor; as before, you'd need to know the output voltage of the sensor across the pressure range.

Transmission oil temperature / Engine coolant temperature:

Same theme - get the output curve across the voltage range and map that to an analog input. Or, use a known dedicated sensor with a known calibration: https://www.autosportlabs.com/product/e ... re-sensor/

Notes about your existing gauges:

* Higher quality gauges have a regulated voltage reference for the sensors; lower quality gauges just use the fluctuating voltage from your alternator.
* Even though gauges may be supplied by 12v, the sensor range may be different, such as 0-5v range, so you may not need a voltage divider.
* Definitely look at the technical information for your gauges, it may have all the information you need.
* Your RaceCapture/Pro has a high accuracy 5v voltage reference that is ideal for powering sensors such as oil pressure sensors and providing a voltage reference for two-wire temperature sensors.

Hope this helps!
Brent Picasso
CEO and Founder, Autosport Labs
Facebook | Twitter

brentp
Site Admin
Posts: 6274
Joined: Wed Jan 24, 2007 6:36 am

Post by brentp »

Hi,

we also have a freshly minted MK3 hardware install guide as well - you can find it here:
https://wiki.autosportlabs.com/Racecapt ... re_install
Brent Picasso
CEO and Founder, Autosport Labs
Facebook | Twitter

davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Post by davef_dci »

Brent,

Thanks for your note. Very helpful. I have just finished characterizing our temp sensors using an inductive hotplate and a hot oil bath. I have a thermcouple I used to measure temp. I did a curve-fit and have equations for resistance as a function of temperature.

The (good) news is that I checked the output of our existing analog gauges and it's regulated down - I'm getting 5.09 Volts with no sensor hooked up.

Do i still need to add a pull-up resistor (between the analog gauge and the sensor), and if I do, won't that mess up the analogue gauge reading Sorry if my electrical engineering is a bit rough.

Thanks for the advice!
Attachments
1.JPG
1.JPG (71.74 KiB) Viewed 7007 times
2.JPG
2.JPG (57.97 KiB) Viewed 7007 times

brentp
Site Admin
Posts: 6274
Joined: Wed Jan 24, 2007 6:36 am

Post by brentp »

That's impressive. Should make this part of the wiki as a how-to!

The gauge will have it's own pullup, so all you have to do is tap into the existing signal.
Brent Picasso
CEO and Founder, Autosport Labs
Facebook | Twitter

davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Post by davef_dci »

Sorry.. A bit more information. I feel like I'm pulling the thread on a sweater here.

I checked the output sensor terminal on the analogue gauge without a sensor attached and its 5V.

However when I hook up the sensor, the voltage jumps up. With the sensor at 220 degrees the voltage is is at 8.14 volts. Clearly I can't run that into the MK3.

I'm not certain I understand how to add a voltage divider without altering and screwing up the reading on the analog gauge. Can this be done?

Thanks!

davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Post by davef_dci »

Brent, Sorry to be needy. My weak electrical engineering skilz are showing through.

I'm wondering if the work I did characterizing the sensor resistance as a function of temp isn't going to help because i need it to continue to function inside the circuit with the existing analog gauge. I was assuming the existing analog gauge would work like a resistive element (i can do V=ir) but clearly that's not the case.

The only thing i can think to do is to create a voltage divided using some very large resistor values. In this way, the new resistive path would not (significantly) change the resistance being seen by the analog gauge. However I'd have to re-do the existing mapping and instead of mapping resistance as a function of temp i'd need to map voltage at the divider as a function of temp.

Am I looking at this wrong? I don't have much confidence i know what I'm doing here.

to make matters all the more delightful, the analog gauge is proving to be off by about 15 degrees. Not sure what I expected but I was hoping for better than that. So I need to decide if I want to map RaceCapture to the analog gauge readings or to the real temperature. Chinese saying - man with one watch knows what time it is but man with two watches is no longer sure. Wonderful.

Thanks for any advice.

Dave

davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Post by davef_dci »

Brent,

I've got it figured out I think. It's actually pretty cool and i am pretty sure I know how to piggyback my gauges. If I get organized I'll write it up.

Thanks!

Dave

brentp
Site Admin
Posts: 6274
Joined: Wed Jan 24, 2007 6:36 am

Post by brentp »

Hi, Sorry I missed this earlier.

Did you end up making a voltage divider? The inputs to RCP are very sensitive, so you could make a voltage divider that has very high resistance so your gauge's readings aren't altered.

Curious how you figured it out!
Brent Picasso
CEO and Founder, Autosport Labs
Facebook | Twitter

davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Post by davef_dci »

Brent,

That's exactly what I did. It took me a bit of time to figure out how the analog gauges work but I found a schematic someplace that made it clear. For the analog temperature gauges there are two wound coils inside the gauge that act as electromagnets to deflect the needle but basically appear as resistors. One of the gauge coil/resistors leads to 12v and the other leads to ground and the sensor input is tied between these. Basically the sensor input resistance and ground resistor are in parallel which in turn are in series with the ignition resistor so as the sensor resistance changes the magnetic attraction to the needle changes. Based on the gauges I've measured, these internal resistances are in the 100-300 ohm range. So I build a voltage divider out of much larger resistive values (300K or so) that don't impact readings. A schematic is attached.

It all works as expected and I characterized all my sensors as functions of resistance.

To be candid though we're still on the fence on whether to run piggyback gauges or go all-in on race capture. It's a PITA to wire this into the existing gauge cluster which is one issue. The other is that the analog gauges are pretty stunningly inaccurate. For example, in the attached picture you can see i rigged up a simple pressure vessel out of some pipe to allow me to characterize the oild pressure gauge. Unless I've screwed something up, the analog gauge is showing 50psi and my calibrated pressure gauge shows 80+. Yikes. The temperature gauges have been closer but are still off by as much as 15 degrees which could make a difference around boiling. Of course we really use our gauges as a relative reference - not really absolute anyway.

So part of us wants to can all our analog gauges and go all-in on race capture. What's dissuading us is a) if we have some sort of a failure of race capture or our tablet we'd have no gauges in the car (the OEM dash and all the OEM gauges are long gone) and would either have to shut down or run blind. In addition, we are sort of struggling to get all the information we'd like to show on one race capture screen - we obviously would need all our gauges but really want predictive and segment timing as well and I think we run out of channels and real estate.

Oh well. We'll sort it out!

If anyone else is curious I can share how I characterized all the gauges - I did polynomial curve fits and reverse engineered resistance equations from measurements. I suspect that's probably a waste of time and it's far easier to just map the resistance and voltage to mirror the readings on the analog gauges - even if they suck.
Attachments
Capture.JPG
Capture.JPG (25.47 KiB) Viewed 6987 times

davef_dci
Posts: 39
Joined: Wed Mar 22, 2017 4:57 am

Post by davef_dci »

pics
Attachments
f.JPG
f.JPG (119.12 KiB) Viewed 6986 times
g.JPG
g.JPG (86.3 KiB) Viewed 6986 times

brentp
Site Admin
Posts: 6274
Joined: Wed Jan 24, 2007 6:36 am

Post by brentp »

Those types of oil pressure sensors are flat out terrible. Highly recommend a powered sensor that has a linear output between 0.5 - 4.5 volts

https://www.autosportlabs.com/product/1 ... re-sensor/

https://wiki.autosportlabs.com/RaceCapt ... re_Sensors
Brent Picasso
CEO and Founder, Autosport Labs
Facebook | Twitter

Post Reply