A/P Kill switch?

[claycookiemonster]

Installing Dynon Pitch and Roll A/P's in an RV8. I will also have Ray Allen Trim systems operated via the Dynon A/P's.

I will have an A/P disconnect switch on the stick, but I also wanted a switch to shut down the entire A/P system. As I wire this up, I was assuming I'd place the red guarded A/P OFF switch on the power lead to the A/P servos, but each servo has their own lead. Wiring them together complicates the VPX and prevents shutting down each axis independently. On the other hand, I could put the red guarded switch on the single power lead to the A/P Panel, but that would leave the servos powered even after shutting down the control panel. It would also kill the trim systems, as they are operated by the Dynon A/P.

How are you guys wiring things so a single switch can kill the entire A/P system?

 

[N890GF]

Dynon doesn't recommend having an "Autopilot" main switch. See this post from Dynon Support: https://forum.flydynon.com/threads/servo-power-switch-requirement.271/post-1409

Also the entire autopilot system can operate with or without the control panel. So shutting off power to that doesn't affect the Autopilot system at all, the only thing it does it shutoff power to the Trim system (auto-trim or otherwise).

I wired up my VPX to A/P servos on a single pin (since I don't plan on using either axis independently), and can disable that via the VPX screen on the Dynon (if I need to). So - No Switch for A/P power in my system.

However, IF you wanted an A/P Off switch, you could use one of the switch pins on the VPX assigned to both power pins for each servo. For example - J2 pin 2 could be Autopilot, when flipped enables Pitch Servo on J8 pin 1 and Roll servo on J8 pin 2. You can control each pin independently via the VPX screen on the Dynon, or use the switch to control both.

If you do kill power to the servos the Dynon system will show the A/P servos are missing, and throw warnings on the screen.

 

[dmn056]

I'm planning my system with Dynon SkyView and Ray Allen trims, but without the VPX. I've worked out some wiring diagrams, but have not yet built them, so no guarantees that this is correct.

In the Dynon system, pin 9 on SV-AP-PANEL is the trim servo power input, so a switch in the wire from this to an electrical bus will provide the trim disconnect function, but will not affect the A/P function. It's very low power, so can be run through an APEM 8636 push-button toggle such as used by the Infinity stick grip.

The autopilot servos are controlled through the SV-NET system, with power supplied from an electrical bus to pin 7 of the servo inputs. Servos can be toggled on/off by grounding pin 3 of their inputs. So, joining pin 3 of the roll and pitch servos to one side of a normally-open switch, with the other side grounded, will provide an autopilot servo disconnect/reconnect function. This is also a low power connection.

If you want an option to turn off all power to the autopilot servos, a switch is needed in the pin 7 inputs from the electrical bus. May not be necessary, but not likely to do any harm either. It's not very low power, so you probably wouldn't want it on the stick.

 

[climberrn]

Not using the VPX, but the power to both my Dynon servos are run through a dpst switch, so they can be fused independently.

 

[Draker]

I'm in the Garmin world and chose to use a single DPST switch attached to the power pins of each servo. I do not anticipate I would need to turn off one servo by itself.

 

[PhatRV]

I connected the power wires for the AP roll servo, pitch servo, and trim servo to a switch on the panel. If there is a runway condition or the CWS failed to disconnect the AP, then flipping the switch kills the power to all the servos which will prevent further runaway condition.

 

[Jpm757]

AFS installs an autopilot master on their Quick Panels. AFS uses a mostly Dynon based system. I don't have the prints handy, but I believe the are available on their website.

 

[dmattmul]

Garmin A/P system

I know trim runaway could be an issue but what am I missing? Touching the trim cooley hat in either direction disconnects the A/P and then there is the A/P disconnect on the stick and lastly the A/P switch on the panel. How much redundancy is needed to keep the A/P from being the source of a NTSB investigation?

LASTLY THE A/P SWITCH ON THE PANEL??

 

[newt]

I know trim runaway could be an issue but what am I missing? Touching the trim cooley hat in either direction disconnects the A/P and then there is the A/P disconnect on the stick and lastly the A/P switch on panel. How much redundancy is needed to keep the A/P from being the source of a NTSB investigation?

Depending on your autopilot (and I'm not familiar with the Dynon one, so take this on general principles):

The various "disconnect" signals are interpreted by software, which reacts by taking the servos out of circuit (de-energizing the clutch, whatever)

The software is a single point of failure: No matter how many coolie hats and disconnect buttons you have, you're relying on code running on a single device to disconnect the servos from your controls.

If you don't trust the software, some other kind of "hard" disconnect should be used to bypass it.

Perhaps those of us who do software for a living trust it a lot less than people who don't

(servo circuit breakers are adequate for my installation. Once they're open, it doesn't matter what the software does)

- mark

 

[PhatRV]

I was definitely following Dynon guidance of wiring the servos. Here is the screenshot of the Dynon installation page

 

[claycookiemonster]

As far as I can tell, the stick A/P disconnect switch is a momentary ground which should stop whatever mode or operation is happening at that moment. Yes, there is also simply reaching up to the A/P panel and de-selecting the A/P. My thought was to have a single Red Guarded switch which would remove power from the A/P completely. That is turning out to be harder than I thought.
Wiring all of this into my VPX, the power sources are; A/P panel is on pin J8-1, Roll A/P is on J8-6 and Pitch is on J8-5. So, a single simple switch to de-power it all is not simple. The Ray Allen trim lives in the A/P panel as I'm doing it.

Hmmm....?

 

[PaulvS]

As far as I can tell, the stick A/P disconnect switch is a momentary ground which should stop whatever mode or operation is happening at that moment. Yes, there is also simply reaching up to the A/P panel and de-selecting the A/P. My thought was to have a single Red Guarded switch which would remove power from the A/P completely. That is turning out to be harder than I thought.
Wiring all of this into my VPX, the power sources are; A/P panel is on pin J8-1, Roll A/P is on J8-6 and Pitch is on J8-5. So, a single simple switch to de-power it all is not simple. The Ray Allen trim lives in the A/P panel as I'm doing it.

Hmmm....?

A DPST switch can control the power from the VPX to the two AP servos. If the servos don't have power then the AP is effectively taken offline. The system software will also detect that the servos are not connected.

 

[N890GF]

As far as I can tell, the stick A/P disconnect switch is a momentary ground which should stop whatever mode or operation is happening at that moment. Yes, there is also simply reaching up to the A/P panel and de-selecting the A/P. My thought was to have a single Red Guarded switch which would remove power from the A/P completely. That is turning out to be harder than I thought.
Wiring all of this into my VPX, the power sources are; A/P panel is on pin J8-1, Roll A/P is on J8-6 and Pitch is on J8-5. So, a single simple switch to de-power it all is not simple. The Ray Allen trim lives in the A/P panel as I'm doing it.

Hmmm....?

You can program a single switch on the VPX to enable multiple power pins. So that shouldn’t be a problem

 

[mburch]

The various "disconnect" signals are interpreted by software, which reacts by taking the servos out of circuit (de-energizing the clutch, whatever)

Depending on which type of autopilot you have, this may not be entirely the case. With the one I'm most familiar with, the autopilot disconnect input electrically interrupts power to the servo motor and trim outputs, regardless of what the software might be doing. Maybe not a surprise since the manufacturer has a lot of experience with certification requirements.

 

[upperdeck]

I had a Garmin AP malfunction. I have the G3X and associated AP servos with the GMC 507 control head. During a practice approach into KOSH, at night, the AP refused to disconnect. I did not have any switches installed to shut the servos off. The ONLY way to disconnect the servos was to press and hold the AP disconnect switch all the way to the ground. Even after landing, when releasing the disconnect switch, you could feel the servos reengaging. Garmin made me pay to swap out the control head after their equipment failed so miserably. I sent in data logs. I was never told what went wrong with their hardware.

I'm currently recutting my panel to add the switches needed to disconnect power.

 

[wirejock]

AFS installs an autopilot master on their Quick Panels. AFS uses a mostly Dynon based system. I don't have the prints handy, but I believe the are available on their website.

My AFS ACM has an autopilot master. It was on the schematic design so I installed it. All Dynon.

 

[rv8ch]

I have a power switch for my GRT AP servos, just in case. Also have one for the electric trim servos.

I too am in software.

(First post on the new forum - exciting!)

 

[rmartingt]

I have a power switch for my GRT AP servos, just in case. Also have one for the electric trim servos.

I too am in software.

I took the same approach, one switch on the AP power and one on the trim power. No mucking about with which axis to kill; just shut it down.

 

[2theskyy]

I had a Garmin AP malfunction. I have the G3X and associated AP servos with the GMC 507 control head. During a practice approach into KOSH, at night, the AP refused to disconnect. I did not have any switches installed to shut the servos off. The ONLY way to disconnect the servos was to press and hold the AP disconnect switch all the way to the ground. Even after landing, when releasing the disconnect switch, you could feel the servos reengaging. Garmin made me pay to swap out the control head after their equipment failed so miserably. I sent in data logs. I was never told what went wrong with their hardware.

I'm currently recutting my panel to add the switches needed to disconnect

I had a Garmin AP malfunction. I have the G3X and associated AP servos with the GMC 507 control head. During a practice approach into KOSH, at night, the AP refused to disconnect. I did not have any switches installed to shut the servos off. The ONLY way to disconnect the servos was to press and hold the AP disconnect switch all the way to the ground. Even after landing, when releasing the disconnect switch, you could feel the servos reengaging. Garmin made me pay to swap out the control head after their equipment failed so miserably. I sent in data logs. I was never told what went wrong with their hardware.

I'm currently recutting my panel to add the switches needed to disconnect power.

Do you have a VPX or AFM? Did you try pulling C/B’s (electronic or mechanical)?

First post on new format, so apologize for some oddities

 

[upperdeck]

First post on new format, so apologize for some oddities

I have a combination of fuses and circuit breakers. Circuit breakers are for anything critical to keep the engine running everything else, including avionics are on fuses. Fuses are not reachable and thus the installation of an auto pilot switch.

 

[Carl Froehlich]

I’m not sure why this is such a debate. Is this because using a VPX puts you in a corner forcing yet another power switch?

I have a pull breaker labeled “TRIM”, another pull breaker labeled “AutoPilot” and a pull breaker labled “Flaps” to use in abnormal situations. No switches other than the stick for Trim, Flaps and AP disconnect and the Dynon Autopilot module.

Carl

 

[2theskyy]

I’d make C/B’s accessible via old school or thru electronic C/B. Pull those in non normal before letting AP drive me into the ground. But that’s why having memory items and non normals built into POH based on individual systems is important. IMHO
I just like to hear what others have and do so I can learn.