Custom Sensor Installations

[RV8Driver]

OK, first of all, I love learning about this electrical stuff because I am an engineer at heart; therefore I have taken on the challenge of setting up my own sensor files per the WIKI directions and I have a few questions. Please don't tell me just to buy the dynon supported sensors, becuase then I wouldn't learn anything

1. What is the purpose of the pull-up resister settings ("low_resistance=" & "high_resistance=") in each sensor file? Is this an internal thing or do I need to install an in-line (in-series) resister for each "standard" (config A) general purpose pin sensor that would correspond to the pull-up number in the sensor file. As the wiki direction state, every config A General Purpose pin must have a pull-up resister setting (not "NONE") but the installation manual doesn't mention installing resisters into your sensor signal wires. If it is strictly internal (no installaton of a resister), what is the philoshophy or effect of different values from high to low (sensitivity, noise, etc)? Example; My Van's resistive fuel senders go into pins 20 and 21 and the sensor file indicates a 200 ohm pull-up on the low_resistance line. Do I need to install a 200 ohm resister in the line coming from the fuel sender?

2. I have an Air/Fuel ratio gauge that outputs 0 to 5 volts for it's signal that I would like to display on the Dynon. I have hooked it up to pin 31 (config C). This sensor has no units so I assigned it the "function=position". I plan on calibrating it so it wll display the correct value without any unit. Is this the best way to have a unitless sensor? Can I calibrate it using the Dynon wizard instead of my proposed process below?

3. I plan on calibrating all my non-Dynon supported sensors by initially setting the sensor file with C_x1=1, and all others zero. I will then record the readings (which I am guessng will be the raw voltage of the sensor) and plot them against the calibrated value gained from other accurate instrumentation. I can take that data and create a polynomial equation to put into the finished and calibrated sensor file coeficients. Is this the best course of action or is there a better way?

 

[mmarien]

I'd be interested in the air/fuel mixture gauge/sensor when you have it working. I'm looking at an electronic fuel injection system and don't really want to add a separate gauge for the air/fuel mixture sensor.

 

[RV8Driver]

MMarien,

I'll let you know what I find. I had it working great on the Blue Mountain (or rather Black Mountain) EFIS so I am confident it will work on the Dynon.

 

[mmarien]

Sounds good. I was looking at this system:

http://www.flyefii.com/

They suggested this sensor and gauge.

http://www.plxdevices.com/wideband.html

I probably won't start anything until winter. What sensor are you using?

 

[dynonsupport]

The pull up resistors are inside the SkyView EMS module. When you choose an option here, this turns on the pull up to the selected value. That's why the wiki says "This is the internal pull up resistance used when this polynomial is active"

The install manual is correct and not missing anything. The user does not install resistors.

There really isn't any device in the world that reads resistance. Pretty much everything reads voltage. Pull up resistors of a known value allow you to hook an unknown resistance up and the voltage that exists tells you the unknown resistance.

We support different pull-ups in order to optimize against the resistance you will be running in your sensor. If you have a sensor that goes from 9-200 ohms, you want a 200 ohm pull up. If you have a sensor that goes from 8K to 20K, you want a 20K pull up. If you have a sensor that goes from 300 ohms to 100K ohms like some temperature sensors, then things get complex.

You cannot use position, as these are special in the system. You need to use volts. AFR sensors are far from unitless- they are in AFR, Lambda, or % oxygen. So what you do is use volts in and give it a formula to convert to the ratio that you prefer. The AFR sensors I've used let you program a conversion ratio, and this should make it easy to put in the same formula in SkyView.

Your method for calibration will sort of work, but you'll get low resolution since we only display voltage to 0.1 volts, which is not enough to do a good calibration. What you need to do is use our EMS status page, which shows voltages to 0.001V, and will show the voltage even if the conversion factor is not just 1. This voltage is at the pull-up you have chosen, so you do need to start with a sane pull-up.

The normal way to calibrate a sensor is to know it's conversion beforehand. All the sensors we use have a table of resistance vs. pressure (or temp), and we use that table to build the formula. We of course check them afterwards, but you kind of need to know something about the sensor before you start.

 

[RV8Driver]

Mmarien,

Email me and I'll let you know what I have done. Thanks for the links. I did not know someone was as frustrated as I was about the low tech aviation engine technology of today. I built my system in 2001 and it has been flying trouble free ever since. It is very similar to that of what you are looking at with a few key exceptions. Namely, how the AFR gauge is incorporated into the whole system. It is a very key piece that lets the EFII run in a closed loop mode and schedules the fuel based on the AFR and not a pre-programmed open loop fuel schedule. I also run dual computers with dual pairs of direct fire units. I'm curious if the flyefii can do the same with the AFR gauge. I did not notice any need for an AFR gauge in their package so I assume it is not incorporated into the system but just a gauge for the curious and well informed pilot. My email is RV8Driver@yahoo.com.

 

[dynonsupport]

Just to mention, the reason we don't support an AFR sensor like that directly is because they die quickly in the presence of lead, generally within 25-50 hours, and get inaccurate long before that. They work in planes if you are using MoGas, but expect them to be pretty useless if you run AvGas. Thus we get basically no demand for them.

This is a common issue in race cars that use O2 sensors and run leaded fuels. They just replace them often, which gets expensive.

One interesting thing also is that the main goal or O2 is for emissions, not performance. Even the highest performance cars don't use O2 for real time feedback at high throttle positions, they use airflow sensors and lookup tables, since airflow is so much quicker and smoother than O2 readings. Trying to run closed loop when at WOT is a large challenge that as far as I know no system has implemented in production. Given the horsepower wars that the car companies are in, I'm sure they would have done it if they made more horsepower or reduced engine failures.

Since that airplanes are pretty much always at WOT, and your mixture isn't supposed to be stoichometric at that point, I'm not 100% sure what your AFR target is going to be. Generally EGT's give you a really good sense of AFR, which is why we support EGT on every cylinder.

 

[RV8Driver]

Wow. You guys are tireless. Answering forum write-ups on a Saturday night. Thank you very much!

I did reread the wiki doc again just after my original post and picked up on that "internal" verbage. Thanks for the answer and sorry for the question.

I did however calibrate my pressure sensors today very well by initially setting the sensor file resolution to 2 and the round to 0.01 and the display showed me the voltages very consistently and what seemed very accurately at x.xx. I used my newly calibrated compression tester (which has two gauges for redundancy) with a pressurized air source to run the pressures up on the sensors as I recorded the voltage readings on the Dynon gauge. It seemed to work like a champ and it was steady with consistent results to a .01 resolution. I plugged the data into Microsoft excell to come up with the polynomial coefficients and viola! I think I may have the most accurate pressure readings around because it is now an installed calibration. With the new sensor files, I re-ran the test and the Dynon now reads exactly as the calibrated compression tester. Thanks for the heads up on the EMS status page. I'll take a look at that the next time.
As far as the AFR goes, I know about the avgas vs mogas problem and I replace my O2 sensor a lot when I run avgas. I just wish more airports sold mogas, but we'll see when avgas goes away. I like your volts idea better, but will the display place a "V" or the word volts next to the readout? The position function I found out today has a very limited display style too which I didn't really like for the AFR gauge. I just want the readout to be the number ( i use the ratio like 12.3 or 14.7) with the title AFR and no units. It's a wideband sensor that really can keep up with the aircraft engines at 2700 rpm and below. I've been flying with it for 6 years (500 hrs) and it really works great.

Thanks again for the great support. I'll let you know how it all comes out in the end.

 

[rgcrothers]

Regarding the oxygen sensor, my experience has been positive also. Installing the sensor in a special bung that keeps the tip back out of the main exhaust flow made a big difference in the life of the sensor as do the cooling fins on the bung. Also, running a good lead scavenger like the Decilin product helps.

I am trying to find some help in getting a sensor definition put together for my oxygen sensor. I think they all work on 0 to 5 volts. In my case 0 volts = 7.35 and 5 volts = 22.39.

I have also posted a request for help up in the "Create your own sensor definition" area up at the top of the opening pages in the forum.

Has anyone been able to put an oxygen sensor into the Skyview EMS? Can you share the definition here?

FYI, I use the air fuel ratio gauge on a Subaru installation in my RV7A. I ate up plenty of O2 sensors until I learned about the special mounting piece available from Innovate Motorsports. I do try to avoid running too much straight 100LL, mainly to keep the rings from getting gummed up from lead deposits...

Randall Crothers
rgcrothers@gmail.com

 

[RV8Driver]

Randall,

I have mine working great in my Skyview. It is a wideband Bosch O2 sensor for an AEM gauge. The gauge installation manual came with a volt vs AFR calibration curve for the sensor which I used in the sensor definition file for the Skyview. It is exactly linear. I did however adjust it for any installation variations by unplugging the sensor and reading the voltage. It should correspond to a 14.7 AFR reading.

 

[rgcrothers]

RV8Driver,

Glad to hear someone has been succesfull at this. I use the Bosch sensor also. One of the things I have learned that helps it live longer in a 100LL environment is the use of a bung extension. Innovate motor sports sells one with cooling fins, getting it cooler helps too. The bung extendor gets the sensor tip back out of the main flow of the exhaust and away from the lead contamination, but does not seem to interfere with it's operation at all. I think the heat is also an issue with an aircraft as our exhaust systems are short and we run at high power settings continuously, very different from even a racing car application. Anyway, the extension seems to have enabled me to run for much longer periods of time on a sensor. It is still probably a good idea to keep a spare in the bag.

Can you elaborate a little more about how you configured the sensor definition with information from the voltage vs AFR curve? Did you calculate polynomials etc. and enter them into the definition file?

Randall

 

[shenweas]

The pull up resistors are inside the SkyView EMS module. When you choose an option here, this turns on the pull up to the selected value. That's why the wiki says "This is the internal pull up resistance used when this polynomial is active"

The install manual is correct and not missing anything. The user does not install resistors.

There really isn't any device in the world that reads resistance. Pretty much everything reads voltage. Pull up resistors of a known value allow you to hook an unknown resistance up and the voltage that exists tells you the unknown resistance.

We support different pull-ups in order to optimize against the resistance you will be running in your sensor. If you have a sensor that goes from 9-200 ohms, you want a 200 ohm pull up. If you have a sensor that goes from 8K to 20K, you want a 20K pull up. If you have a sensor that goes from 300 ohms to 100K ohms like some temperature sensors, then things get complex.

You cannot use position, as these are special in the system. You need to use volts. AFR sensors are far from unitless- they are in AFR, Lambda, or % oxygen. So what you do is use volts in and give it a formula to convert to the ratio that you prefer. The AFR sensors I've used let you program a conversion ratio, and this should make it easy to put in the same formula in SkyView.

Your method for calibration will sort of work, but you'll get low resolution since we only display voltage to 0.1 volts, which is not enough to do a good calibration. What you need to do is use our EMS status page, which shows voltages to 0.001V, and will show the voltage even if the conversion factor is not just 1. This voltage is at the pull-up you have chosen, so you do need to start with a sane pull-up.

The normal way to calibrate a sensor is to know it's conversion beforehand. All the sensors we use have a table of resistance vs. pressure (or temp), and we use that table to build the formula. We of course check them afterwards, but you kind of need to know something about the sensor before you start.

You stated here that volts is the thing to use for this "type" of sensor.  I am doing something a little different.  I have an input that is 0.01 volt/deg of timing from the ignition.  I have used both P1 and P2 for the electrical buss.

How do i configure one of the general purpose inputs for voltage measurement and hopefully have the widget display in deg/timing?

I tried setting  it up using a level widget and modifying the calibrations to match the inputs and have not tried it yet but the widget shows "timing gal" :-/ Also plays havoc with the computed fuel level.

 

[dynonsupport]

The simple answer is that you can't do exactly what you want.

We only support certain "functions." These are Pressure, Temperature, Voltage, and Amps. We don't let you call something "angle".

We do support "position" but these have specialized, fixed names of Elevator, Ailerons, Rudder, and Flaps.

We do have level as well, and as you say, this is in Volume (gallons) always.

So, you aren't going to be able to get it to say "Timing Degrees."

You can get it to say "timing volts," and have it read exactly the number you want it to, such as "24.8." But "24.8 timing volts" not "24.8 timing degrees"

In order to do this, you need to make a custom sensor file, which calls out the fact that this is a voltage input (which is only allowed on 3 of the GP inputs). Then you just need the simple conversion from volts to degrees, which in your case is 100 (there are 100 degrees per volt of timing).

Now, be aware, that you won't get a lot of resolution with this because the output is so small. You likely run 25-ish degrees of timing in flight, and that's only .25V. We have about .01V accuracy on voltage, so we'll likely only be within 1 degree at any time, and you need to make sure your grounds are all perfect.

So, you can get a display on the screen which is informative, albeit labeled a bit weird.

I do have to ask: How is knowing timing angle in flight helpful?

 

[mmarien]

It has the same function as temperature, pressure or volts. If it's not in the green there is a problem that needs to be rectified. Not a big deal with mags as it should always be 26 degrees, but more so with variable timing that's available with electronic ignition. It that "experimental" thing.

 

[shenweas]

The simple answer is that you can't do exactly what you want.

We only support certain "functions." These are Pressure, Temperature, Voltage, and Amps. We don't let you call something "angle".

We do support "position" but these have specialized, fixed names of Elevator, Ailerons, Rudder, and Flaps.

We do have level as well, and as you say, this is in Volume (gallons) always.

So, you aren't going to be able to get it to say "Timing Degrees."

You can get it to say "timing volts," and have it read exactly the number you want it to, such as "24.8." But "24.8 timing volts" not "24.8 timing degrees"

In order to do this, you need to make a custom sensor file, which calls out the fact that this is a voltage input (which is only allowed on 3 of the GP inputs). Then you just need the simple conversion from volts to degrees, which in your case is 100 (there are 100 degrees per volt of timing).

Now, be aware, that you won't get a lot of resolution with this because the output is so small. You likely run 25-ish degrees of timing in flight, and that's only .25V. We have about .01V accuracy on voltage, so we'll likely only be within 1 degree at any time, and you need to make sure your grounds are all perfect.

So, you can get a display on the screen which is informative, albeit labeled a bit weird.

I do have to ask: How is knowing timing angle in flight helpful?

It is electronic ignition and the timing changes with MAP and rpm. The curve is tunable but I want to be able to monitor it during adjustments. Thanks for the user file set up. I will see if I can make it work.

 

[shenweas]

In order to do this, you need to make a custom sensor file, which calls out the fact that this is a voltage input (which is only allowed on 3 of the GP inputs).

Witch 3 gp inputs? Or can it be any unused as long as its no more than 3?

 

[mmarien]

The three enhanced general inputs.

See table 14 and 15 on page 7-5 and table 57 on page 17-9

 

[shenweas]

Then you just need the simple conversion from volts to degrees, which in your case is 100 (there are 100 degrees per volt of timing).

I have successfuly made a custom user file with the "Timing Volts" and have it working however I do not seem to be able to get the conversion right.

I tried making the low coeff c_x1=100 that made no differance as c_x1=1   I would have thought that would do it but I am no code writer so got lost here.

I also tried making a user cal in the DFG file with timing as the lable and having 2 points with 0=0 and .4=40 but that didnt make any differance.

any thoughts?

Edit: I got it working now. had the lables incorect in the Calibration file