Rotax 915is and VPX installation discussion

[VictorV]

Hi Everyone,
I've been having a side discussion via email with Fred B. (Doggitz) who's also building a Kitfox with a 915 and VPX and
there have been some interesting topics so we thought we'd bring the discussion into the forums so that
we can share what we've learned and how we have approached the design/implementation. This discussion could
also be helpful to those installing a 912is since the electrical designs are almost identical.


The majority of our discussion has been regarding the best approach to start-power switching. For those unaware of
this concept, the 915 is self-contained from an electrical standpoint and supplies it's own redundant power via 2
internal generators. In order to start the engine you need to apply momentary power and ground from the battery to run
the ECU, electronic ignition, and fuel pumps. Once the engine hits 1500 rpms the internal generator takes over.
The external "start-power" must then be removed otherwise the engine controller will not be able to determine
if the primary generator has failed.


As I mentioned, we're both using the VPX as a power control center. Initially I was considering using the PPS as well
but the engine has so much functionality built-in that the PPS does not add any value. We've also heard that
running engine circuits through the VPX will void the Rotax warranty. I'm still trying to get a copy of the actual
warranty to understand what types of things will void the warranty.


Fred has chosen the GRT avionics suite and I am going with the G3X Touch. I don't believe this will have much, if any,
impact on the design of the systems supporting the engine.


Now, back to start-power....


So far I'm aware of 2 designs to deal with the start power issue:


1) A 3-position switch -- on, off (center), on-momentary. The simplified start procedure is as follows:
- Master on
- Press and hold start-power switch to momentary
- Lanes A and B on
- Wait for Lanes A and B red status lights to go out
- Fuel pump A on
- Hit the start button
- Once the engine starts, release the momentary start-power switch.


2) Use this device: https://www.stockflightsystems.com/emu-912is.html ($3900)
It has a built-in relay that turns on when power is applied and stays on until the engine hits 1500 rpm. This low-current relay is used
to trigger a higher-current relay that supports the 8 amps needed by the engine electronics during start. This is what is used in the
Kitfox SLSA as well as Trent Palmer in his Freedom Fox.


The primary flaw in #1 is that you need one hand to operate the switch and the other to operate the momentary switch which leaves
0 hands available to control the throttle. #2 is a nice solution but it's pricey and takes up panel space.


The solution we've been working on is shown below. It consists of a interval timer that's triggered by turning on either Lane A or Lane B
ignition switches. It drives a higher-current relay which controls the start-power. The "on" time is programmable from 10 to 1000 seconds.




The diode "OR" gate allows either ignition switch to enable the start power. There's also a bypass switch that will manually enable start power.


If anyone has any thoughts/ideas on this or any other topics related to 915 and VPX integration please chime in!


Victor

[Eric Page]

1) A 3-position switch -- on, off (center), on-momentary. The simplified start procedure is as follows:
- Master on
- Press and hold start-power switch to momentary
- Lanes A and B on
- Wait for Lanes A and B red status lights to go out
- Fuel pump A on
- Hit the start button
- Once the engine starts, release the momentary start-power switch.

The primary flaw in #1 is that you need one hand to operate the switch and the other to operate the momentary switch which leaves 0 hands available to control the throttle.

Two options occur to me, both similar:

1. Use a momentary OFF-(ON) toggle for the Starter and put it next to the Start Power toggle. With this arrangement you push up Start Power with, say, your middle finger, then push up Starter with your index finger. Once the engine starts, release Starter, and once it passes 1,500 rpm, release Start Power.

2. Put the Start Power toggle and a Starter push-button next to each other and manipulate them in a similar way.

The first option is probably easier to work from the standpoint of finger mechanics than using two different types of switches as in option two.

The solution we've been working on is shown below. It consists of a interval timer that's triggered by turning on either Lane A or Lane B ignition switches. It drives a higher-current relay which controls the start-power. The "on" time is programmable from 10 to 1000 seconds.

The diode "OR" gate allows either ignition switch to enable the start power. There's also a bypass switch that will manually enable start power.

Unfortunately, it won't work as shown in your drawing. The 10k resistor is pulling the timer input to ground, and your Lane switches are switching to ground as well. Closing a Lane switch will not change the input to the timer.

I would be leery of connecting anything to the Lane switch wires without knowing what's driving them inside the Fusebox (although it's possibly safe if Rotax wants them switched directly to ground).

The pull-up inside the Fusebox may activate your timer as soon as the Master switch is turned on, depending on its value and what voltage it pulls up to (don't assume that it's 12V).

It's possible that your 10k pull-down is strong enough to signal "Lane On" to the Fusebox, making it impossible to turn off the Lanes.

Your 10k pull-down could create a voltage divider with a pull-up resistor inside the Fusebox, or the diodes could raise the line slightly above ground. Either of these might create an indeterminant state and cause unpredictable results.

It could probably be made to work with a sufficiently high impedance tap from the Lane wires and an inversion of the Lane switch logic to the timer, but then a couple of other questions come to mind, both related to finding the correct relay timing:

1. Will the engine's ECU or Fusebox be unhappy if Start Power is maintained too long after start? What is "too long?"

2. What happens if the pilot is distracted briefly between turning on the first Lane switch and engaging the starter, or the engine takes a tad too long to start, and the Start Power relay drops out before the engine passes 1,500 rpm?

If you could identify a signal elsewhere in the Rotax loom that indicates switch-over to the internal generator, or create an output from your EFIS that signals >1,500 rpm, then you could reliably automate Start Power release with a bit of logic.

All of this is a lot of work to eliminate a switch...

[Eric Page]

One more question occurred to me last night:

3. What happens if the timer/relay fails closed during start and Start Power is never released? Is there a way to know that this has happened?

[Geek]

.......The primary flaw in #1 is that you need one hand to operate the switch and the other to operate the momentary switch which leaves
0 hands available to control the throttle. Victor

How about locating the momentary switch just to the left of the throttle. Wrap your hand around the throttle and use your thumb to pull in the momentary switch. When it starts, let go of the momentary but you still have the throttle controlled. I think that's actually how Kitfox does their panels.

This however is a great post that I am certain is going to help a lot of people out. Thank you

Gary

[alexM]

I have very limited time around Rotax 4 cylinder engines but my recollection is that they come to life abruptly (unlike the cantankerous clattering stuff I'm used to) and pass through 1500 rpm before you let go of the starter button/switch.

[VictorV]

Two options occur to me, both similar:

1. Use a momentary OFF-(ON) toggle for the Starter and put it next to the Start Power toggle. With this arrangement you push up Start Power with, say, your middle finger, then push up Starter with your index finger. Once the engine starts, release Starter, and once it passes 1,500 rpm, release Start Power.

2. Put the Start Power toggle and a Starter push-button next to each other and manipulate them in a similar way.

The first option is probably easier to work from the standpoint of finger mechanics than using two different types of switches as in option two.


Unfortunately, it won't work as shown in your drawing. The 10k resistor is pulling the timer input to ground, and your Lane switches are switching to ground as well. Closing a Lane switch will not change the input to the timer.

I would be leery of connecting anything to the Lane switch wires without knowing what's driving them inside the Fusebox (although it's possibly safe if Rotax wants them switched directly to ground).

The pull-up inside the Fusebox may activate your timer as soon as the Master switch is turned on, depending on its value and what voltage it pulls up to (don't assume that it's 12V).

It's possible that your 10k pull-down is strong enough to signal "Lane On" to the Fusebox, making it impossible to turn off the Lanes.

Your 10k pull-down could create a voltage divider with a pull-up resistor inside the Fusebox, or the diodes could raise the line slightly above ground. Either of these might create an indeterminant state and cause unpredictable results.

One other solution that I have seen is to use a different momentary switch: off, on, momentary. The off and on states control the start
power and the momentary controls the starter. The disadvantage is that you have to remember to turn off the start power after the engine
starts though. Having your hand already on the switch though might make that easier.

There are definitely some things to work out with this design. I was not sure what happening on the Lane switches. IIRC the wiring diagram in the
915 installation manual just says to place a switch between 2 pins on a connector but I was not sure exactly what signal was on
those lines. If one is ground and the switch just connects to ground then you're right that the circuit will not work as designed. I was assuming
that one side of the switch had 12v and the other was floating near ground. Do you know for sure that the Lane switches just connect one
pin to ground?

I also read this in the installation manual:

"When the engine is started the engine internal power supply system Generator B is supplying the EMS. At a certain engine speed , and after a certain amount of stable operating time, the ECU commands power supply system A to take over. power supply system Generator A remains engaged except in the case of failures. This procedure ensures that both power supply systems are checked prior to flight. It is therefore important that a start power switching system releases power prior to commanding power supply system A. This is important so as not to compromise the check of power supply system Generator B."

If you're using a manual switch it's unlikely that you'll get the timing correct each time but maybe that's not a huge deal. You're correct, though,
in that we don't know what the exact timing should be. When the engine is running on Generator B it does not output a charge current to the
airframe until it switches over to Generator A. One possibility would be to use the power going to the airframe to turn off the start power. I could
also switch the logic so it engages when either Lane switch connects to ground.

It could probably be made to work with a sufficiently high impedance tap from the Lane wires and an inversion of the Lane switch logic to the timer, but then a couple of other questions come to mind, both related to finding the correct relay timing:

1. Will the engine's ECU or Fusebox be unhappy if Start Power is maintained too long after start? What is "too long?"

2. What happens if the pilot is distracted briefly between turning on the first Lane switch and engaging the starter, or the engine takes a tad too long to start, and the Start Power relay drops out before the engine passes 1,500 rpm?

If you could identify a signal elsewhere in the Rotax loom that indicates switch-over to the internal generator, or create an output from your EFIS that signals >1,500 rpm, then you could reliably automate Start Power release with a bit of logic.

All of this is a lot of work to eliminate a switch...

1. Agreed. The timing is unknown. I was going to have a long timeout like 1 minute.

2. This could happen. I was assuming the engine would not start and you'd have to turn the Lane switches off and on again to restart the sequence.

With my original solution #2 using the Stock Flight Systems engine monitor we still don't know exactly when the relay is disabled other than
when the engine reaches 1500 rpm. This may or may not be sufficient from a timing perspective.

[VictorV]

One more question occurred to me last night:

3. What happens if the timer/relay fails closed during start and Start Power is never released? Is there a way to know that this has happened?

I was planning to have an indicator light on the panel when start-power was active. This begs the question....what action does the pilot take
if the light stays on?

[VictorV]

How about locating the momentary switch just to the left of the throttle. Wrap your hand around the throttle and use your thumb to pull in the momentary switch. When it starts, let go of the momentary but you still have the throttle controlled. I think that's actually how Kitfox does their panels.

This however is a great post that I am certain is going to help a lot of people out. Thank you

Gary

The start-power switch is indeed close to the throttle on the Kitfox 912is panels but I was not able to get my hand on both simultaneously.
Maybe my hand is just too small

[efwd]

I Have this setup. I used a three way switch for the start power. On-off-momentary. I placed it behind a trigger guard. My switch is just right of the throttle. I can lift the guard, it parks and stays up. I can use my throttle hand (while holding the throttle ) to push down on the momentary switch if I choose but I throw that switch up to the on position and wrest my hand on the throttle. Once I have controlled the RPM to 2000rpm I throw the guard down over the switch which automatically returns that switch to the off position. I have forgotten to throw the guard down but seem to catch it quickly when I look at the throttle and see that the guard is still up.

Page 12 of "show us your panel" you should find my pic.

[VictorV]

Does anyone know how I can determine exactly what type of signal the 915is Lane A and B switches generate? It's just
connected between 2 pins on an engine connectors HIC-A and HIC-B. Is 1 pin ground and the switch just connects the other pin to ground
or is there power on one side and switch connects power to the other pin?

Eric, you implied that it's the former but is there a way to verify this? Maybe someone that has a 912is or 15is can measure
what's happening on their system.

Victor