What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

G3X/VPX trim - first flight eye opener

ctennis

Well Known Member
I had my RV-10's first flight yesterday, and thought I'd share something interesting that occurred, mostly as a discussion point: My trim did not work in flight.

System: G3X controlled AP, fronted by a VPX Pro, wiring setup as per them and Garmin.

I tested the system on the ground multiple times, and also had another person sanity check my results. This is my first "electric trim" aircraft, having flown a 182 for the past decade and having built an RV-8 with manual trim. A few days ago I realized I had an error: I had the up and down trim buttons on the control sticks reversed - the more forward button should be "down" trim, I had it as "up" trim.

Not a problem, the VPX has a control for both "normal" and "inverse". As well, the G3x configuration mode system also lets you change between a normal and reverse setting. Whew, I can just change this setting and I don't have to fix my wiring, great.

Swapping the VPX value, I checked the system again, and it functioned as expected.

First flight is here, and one of my sage pilot friends suggests I have a little nose-up trim for takeoff since I'm fairly forward CG on the flight, it'll help a bit with getting off the ground. Ok, on my checklist I go ahead and take his advice and do that - trim system move and responds as expected. Trim is now slightly nose up.

Takeoff, everything is great, but man that climb is a bit aggressive - need a lot of forward stick to overcome that trim. No problem, attempt to trim nose down - and nothing happens. Multiple attempts, no movement. I can see the g3x screen indicator for trim turn from white to blue every time I push the trim button, but it is most definitely not moving. This is becoming a problem, I'm pushing really hard to keep relatively level if not modest climb. Nothing is working though. Luckily I had a chase plane to help with traffic and tower while I debugging. Tried the copilot stick - same results. Went into the VPX screen on my MFD and attempted to the push trim up/down from there, same results. It was getting the signal, but not moving the servo. Meanwhile I'm pushing extremely hard with my left arm.

Ended up circling back and coming in - luckily with less power and after applying flaps, the nose up trim was in a pretty good spot to make a nice landing. Luckily I didn't have any other major issues on that flight to deal with, as my attention was 100% focused on this problem.

Taxied back, and while doing so exercised the trim - and it worked fine!

First thought was maybe we had it on a really slow movement setting at airspeed, but checked the g3x config settings against my chase plane's equivalent, and they were basically the same. Also thought maybe something was just binding it up and airspeed pressure was just too strong to overcome, but putting pressure on it manually with our hands, it would still move fine.

I ended up going into the VPX configuration and changing normal back to inverse as it was before. Flew again, and this time it worked fine (albeit, the buttons on the stick were backwards). Used that as an opportunity to just rewire the sticks the correct way, and on a later flight verified it's working as expected.

Why this was an issue: I dunno! Why it worked on the ground but not in the air: I dunno! I suspect there is some kind of vpx/g3x logic that kicked in with airspeed at which they were fighting with each other and neutraling out the command, but that's just a pure guess.

Food for thought.
 
Memories fade over time, but I had somewhat the same issue a while ago after getting the plane painted and re-connecting the pitch trim servo backwards. I tried inverting the control but that didn't work right either. The only thing I could do was re-wire and it worked fine after that. I have a Dynon set up so there may not be a 1 to 1 correspondence.
 
Why does trim need to go thru the VPX in the first place?
Seems like you've added another level of logic/complexity/failure mode to what is normally a pretty simple system.
 
Why does trim need to go thru the VPX in the first place?
Seems like you've added another level of logic/complexity/failure mode to what is normally a pretty simple system.

I wondered how many responses it would be before someone questioned "why did you do it that way?". In the handful of posts I've put on here in the past 7 years, I don't think a single one has gone without someone making a comment like this.

Because it's my experimental aircraft, that's why.
 
I wondered how many responses it would be before someone questioned "why did you do it that way?". In the handful of posts I've put on here in the past 7 years, I don't think a single one has gone without someone making a comment like this.

Because it's my experimental aircraft, that's why.

OK to put it another way, what is the benefit of integrating the VPX into the trim system? You should know this because you built it...

PS: you're the one that stated "I dunno why", I would think you would want to know the technical reason for the trim behavior rather than taking a defensive attitude to my question. Your I dunno reminds me of my favorite pilot saying when using the autopilot: "what's it doing now?"
 
Last edited:
I wondered how many responses it would be before someone questioned "why did you do it that way?". In the handful of posts I've put on here in the past 7 years, I don't think a single one has gone without someone making a comment like this.

Because it's my experimental aircraft, that's why.

You designed a system for which you clearly don't understand fully how it operates under ALL conditions. That is your right and no one is suggesting it is not. However, it would seem that those questioning that approach and suggesting a move to a simpler system where you CAN fully understand what it is doing in ALL circumstances is good advice, even if it seems unreasonable to you. This does beg the question of why did you post this? If you weren't looking for advice, it might have been better to state that your intention was only to warn others and that you didn't want guidance.

Safety often comes from fully understand how your system will operate in ALL conditions. Consider the fact that you still don't know why the trim didn't work when it should have. Are there other things that the system is going to do under other unique circumstances that will create problems? Probably a good thing to know.

Larry
 
Last edited:
Trim

Well, integrating the vpx with the trim allows for speed trim, much like every other trim controller out there for the Rv-10.

...and I agree with the OPs post about “why did you do it that way” comment.

Anyway, I had a problem with the VPX trim on my -10, too. Check to see if you have the speed trim set up in the VPX config page. You can set it to slow the speed of the servo above a specified speed. The problem comes when you set the % servo value. I found that anything below about 80% for the servo would cause the servo to not move while in flight. The VPX reduces the voltage to the servo to slow it down and by doing so, reduces its ability to move the trim tab. With flight loads and less than 80% power my tab did not move but the indicator would change from white to blue with every trim input. Slowing the aircraft to below the specified speed would immediately restore trim function... it was easily repeatable and I even tried it at multiple specified speeds and power %s.

So check your speed trim settings in the vpx.

Also, knowing what I do now about it, I would likely use the garmin trim functions if I were to do it again. I do not think I will rewire the harness to change it, though...
 
I would also surmise that if you had to push so hard with your left hand you could have easily reduced that push by pulling back with your right hand.
 
If you have decided to use a VPX, as I did, why wouldn't you use the built in trim controller? Seems pretty moronic to ignore the built in runaway trim detection, and electronic CB protection. Of course, this just tees up the never ending argument about VPX vs CBs vs fuses, etc, etc. Personally speaking I'm happy with my VPX and would definitely do it again. Signing off, argue amongst yourselves! :rolleyes:
 
VPX trim versus older technology

I think the actual question here is why would you use old technology with relays that can weld shut causing runaway trim when you have solid state control available through the VPX? How many of our group have replaced their tried and true magnetos for electronic ignition? This could go on and on!
 
Well, integrating the vpx with the trim allows for speed trim, much like every other trim controller out there for the Rv-10.

...and I agree with the OPs post about “why did you do it that way” comment.

Anyway, I had a problem with the VPX trim on my -10, too. Check to see if you have the speed trim set up in the VPX config page. You can set it to slow the speed of the servo above a specified speed. The problem comes when you set the % servo value. I found that anything below about 80% for the servo would cause the servo to not move while in flight. The VPX reduces the voltage to the servo to slow it down and by doing so, reduces its ability to move the trim tab. With flight loads and less than 80% power my tab did not move but the indicator would change from white to blue with every trim input. Slowing the aircraft to below the specified speed would immediately restore trim function... it was easily repeatable and I even tried it at multiple specified speeds and power %s.

So check your speed trim settings in the vpx.

Also, knowing what I do now about it, I would likely use the garmin trim functions if I were to do it again. I do not think I will rewire the harness to change it, though...

I think this nails the cause of this problem that the OP had. I have the same setup (G3X touch, Garmin autopilot and VPX pro) and never had any issues, but I think my % servo values are all greater than 80%.

Thea great thing is, we as a group have another RV-10 flying, so Congratulations on that point:)........
 
I suspect this is the real answer to the OPs question. Let us know.

Carl

Anyway, I had a problem with the VPX trim on my -10, too. Check to see if you have the speed trim set up in the VPX config page. You can set it to slow the speed of the servo above a specified speed. The problem comes when you set the % servo value. I found that anything below about 80% for the servo would cause the servo to not move while in flight. The VPX reduces the voltage to the servo to slow it down and by doing so, reduces its ability to move the trim tab. With flight loads and less than 80% power my tab did not move but the indicator would change from white to blue with every trim input. Slowing the aircraft to below the specified speed would immediately restore trim function... it was easily repeatable and I even tried it at multiple specified speeds and power %s.

So check your speed trim settings in the vpx.

Also, knowing what I do now about it, I would likely use the garmin trim functions if I were to do it again. I do not think I will rewire the harness to change it, though...
 
You designed a system for which you clearly don't understand fully how it operates under ALL conditions. That is your right and no one is suggesting it is not. However, it would seem that those questioning that approach and suggesting a move to a simpler system where you CAN fully understand what it is doing in ALL circumstances is good advice, even if it seems unreasonable to you. This does beg the question of why did you post this? If you weren't looking for advice, it might have been better to state that your intention was only to warn others and that you didn't want guidance.

Larry

Your first sentence is incredibly rich.

I posted it as I hadn't seen anyone else with this equipment have a similar experience, and wanted to try and share some information. If a net result is also putting a nail in the coffin for someone considering this setup as well, then that is seemingly useful.

"Just another data point in why I don't like this particular setup" seems like a much more reasonable response.

Anyone here is welcome to question my approach all that they want. But I literally said this to my wife before posting it: "I wonder how long until the first person questions why I went with this setup". I feel like it's a much more systemic issue with the attitude of a few people in these forums than it is with the people who are just trying to share either ask for help, or share something they learned.
 
I suspect this is the real answer to the OPs question. Let us know.

Carl

I think this is likely it, thanks Carl! This value was below 80 on that flight, and i subsequently changed it for the following one while troubleshooting. Unfortunately at the time I didn't think to try the pitch trim again while really slowed down for the landing, but I suspect it may have worked once slow enough.
 
I wish

I wish the % power adjustment was PWM instead of voltage decrease...well, maybe in the next version!

Chad, are you listening?...😎
 
I wish the % power adjustment was PWM instead of voltage decrease...well, maybe in the next version!

Chad, are you listening?...😎

Garmin uses PWM so you can slow it way down with no issue.
Using both VPX and Garmin at the same time may not work great.
 
As a registered electrical engineer who worked doing instrumentation and control systems for several years, I would recommend using one or the other (Garmin or VPX) but not both. It is generally not a good design to place 2 different controllers in series connected to a single control device (servo, etc.). Each of the controllers has typically been designed to work by itself, and with them in series each one's ability to control is uncertain.

In this case, if you don't want to re-wire, perhaps you could consider setting one to just act in a pass-thru mode, and do the actual control with the second unit. Since the G3X trim speed control uses PWM, it might be best to set the VPX to always output full voltage (speed) and normal (not reversed) direction, and then adjust speeds and direction with just the G3x.

Just a suggestion. As stated by the OP, it is his build and so his decision as to how he wants to set it up.

Regards,
 
As a registered electrical engineer who worked doing instrumentation and control systems for several years, I would recommend using one or the other (Garmin or VPX) but not both. It is generally not a good design to place 2 different controllers in series connected to a single control device (servo, etc.). Each of the controllers has typically been designed to work by itself, and with them in series each one's ability to control is uncertain.

In this case, if you don't want to re-wire, perhaps you could consider setting one to just act in a pass-thru mode, and do the actual control with the second unit. Since the G3X trim speed control uses PWM, it might be best to set the VPX to always output full voltage (speed) and normal (not reversed) direction, and then adjust speeds and direction with just the G3x.

Just a suggestion. As stated by the OP, it is his build and so his decision as to how he wants to set it up.

Regards,

Good advice. No one installing a VP-X with a G3X Touch autopilot (with trim motors connected directly to the GSA 28 servos) should have the variable trim speed function enabled on the VP-X.

Vertical Power is quite clear about this in their installation manual.

VPX_Trim.png

Steve
 
What Steve posted, plus it sounds likely that the Garmin autopilot is either wired on configured incorrectly. Above you mention that the EFIS trim indicator changes color when you press the trim switch. This is the VP-X telling the EFIS that the trim circuit is powered, so the VP-X IS sending current to move the trim motor.

However, the Garmin autopilot is wired "downstream" of the VP-X (between the VP-X and the trim servo) and therefore is likely opening the circuit for some reason.

My 2 cents... hope that helps.
 
As a registered electrical engineer who worked doing instrumentation and control systems for several years, I would recommend using one or the other (Garmin or VPX) but not both. It is generally not a good design to place 2 different controllers in series connected to a single control device (servo, etc.). Each of the controllers has typically been designed to work by itself, and with them in series each one's ability to control is uncertain.

But what are the alternatives here?

If I wire trim from the VPX directly to the trim servos and bypass the Garmin GSA28s, then my Garmin autopilot will not have auto trim capabilities.

If I didn't use the VPX's trim feature, then I would need to add Garmin's GAD27 module.

Or I could just have switches and relays.

It seems like every combination is just a different set of tradeoffs. I recognize different people have preference for different approaches, but I don't see where removing the VPX doesn't add some other system back into the mix, without also losing some functionality.
 
Last edited:
If I didn't use the VPX's trim feature, then I would need to add Garmin's GAD27 module.

Or I could just have switches and relays.
.

Autotrim and trim control in general does not require the GAD 27 in a G3X system. Just relays and switches wired like you would normally in a non-efis plane, Just take the output of them into the 28 and back out to the trim motor, following the input and output ports. The 27 has the relays built in, so avoids that step and makes that part of the wiring easier. However, you now have to wire the 27 into power and the can bus.

larry
 
Last edited:
We've had some really good responses here and also had a couple of PMs, so wanted to circle back with some analysis.

The VPX's pitch control has two configuration items of note: "Speed" and "% Power". In my case, the default factory settings for those were "90" and "70%" respectively.

The VPX manual (revd) section 5.19g states in a GSA28 system to make sure you set "Speed" to 0 to disable the VPX's variable trim system.

Section 6.2 table later defines the speed (though they call it "Reduce Power above", but in the UI it's actually just labeled "Speed"), and also a definition of %Power.

Obviously speed should have been set to 0, and then %power wouldn't even matter. But, I missed that step - and it sounds like others may have as well.

As a result, above 90 knots, power was reduced to 70%. This was clearly not enough to overcome the forces on the elevator trim tabs.

On a subsequent flight I had upped these values to 110, and 90%. 90% apparently enough to overcome the external forces.

I would posit that *maybe* one of the issues is that because of the larger surface area of the RV-10s trim tabs, the forces are much higher on the servo than the other models. As a result, if you miss this "step" in a smaller aircraft - the out of the box config items may still work because 70% power is enough to still move the servo.
 
We've had some really good responses here and also had a couple of PMs, so wanted to circle back with some analysis.

The VPX's pitch control has two configuration items of note: "Speed" and "% Power". In my case, the default factory settings for those were "90" and "70%" respectively.

The VPX manual (revd) section 5.19g states in a GSA28 system to make sure you set "Speed" to 0 to disable the VPX's variable trim system.

Section 6.2 table later defines the speed (though they call it "Reduce Power above", but in the UI it's actually just labeled "Speed"), and also a definition of %Power.

Obviously speed should have been set to 0, and then %power wouldn't even matter. But, I missed that step - and it sounds like others may have as well.

As a result, above 90 knots, power was reduced to 70%. This was clearly not enough to overcome the forces on the elevator trim tabs.

On a subsequent flight I had upped these values to 110, and 90%. 90% apparently enough to overcome the external forces.

I would posit that *maybe* one of the issues is that because of the larger surface area of the RV-10s trim tabs, the forces are much higher on the servo than the other models. As a result, if you miss this "step" in a smaller aircraft - the out of the box config items may still work because 70% power is enough to still move the servo.

This is excellent info to be on here for other builders who may be troubleshooting this same problem in the future !!
 
I would posit that *maybe* one of the issues is that because of the larger surface area of the RV-10s trim tabs, the forces are much higher on the servo than the other models. As a result, if you miss this "step" in a smaller aircraft - the out of the box config items may still work because 70% power is enough to still move the servo.

This makes a lot of sense. The same trim motor is used on the two seat models and the 10, yet on the 10 that same motor must move two tabs that are almost twice the size of the one tab on the 2 seat models (excluding the 14). It would stand to reason that reducing voltage very much would put the torque output under the threshhold needed to move the tabs at moderate speeds.

The 6 has too much trim authority at high speed and I made a voltage reducer to slow down the motor at cruise speeds. at 160 KTAS, 9 volts is plenty to move the tab. I doubt that would work on the 10.
 
Last edited:
But what are the alternatives here?

If I wire trim from the VPX directly to the trim servos and bypass the Garmin GSA28s, then my Garmin autopilot will not have auto trim capabilities.

If I didn't use the VPX's trim feature, then I would need to add Garmin's GAD27 module.

Or I could just have switches and relays.

It seems like every combination is just a different set of tradeoffs. I recognize different people have preference for different approaches, but I don't see where removing the VPX doesn't add some other system back into the mix, without also losing some functionality.
Ctennis,

The choices for trim switch mixing in a G3X Touch system are GAD 27, VP-X, or a relay deck.

There is nothing inherently wrong with your choice, but you were trying to use more functionality in the VP-X than needed since your GSA 28 servo based autopilot is extremely capable and has PWM trim speed control built-in.

This diagram from page 27-12 of the Rev. AN G3X/G3X Touch Installation Manual shows the system architecture when using a 3rd party mixing device like the VP-X. You will note that it also states that any trim speed control capability of the 3rd party trim mixer should be disabled.

ACtC-3fOe3OO7qDWuL1G0XIZEt0uhFbe_7o14trKvVcszJSfoJeHjnwG9jT1rH_qC5azbZCUzmvVjsnrFN7xYorBdnfy6duOwXPxTQnqiCJVRpTqltdnA3TMG6a9P0v1FY3UA-dpPob2yBjf1fh2EwkC824M=w840


Your low power stick mounted trim switches are wired into the VP-X, which then simply mixes the two sets of inputs, and provides a single set of power/ground outputs which are connected to the GSA 28 servo. This works the same as a trim mixing relay deck, but without the relays.

When the GSA 28 servos are powered on, the GSA 28 does all of the trim motor driving and adds speed scheduling to move the trim motors slowly when the aircraft is flying fast and faster when the aircraft is flying slow like in the pattern. This speed scheduling occurs both when the autopilot is engaged (auto-trim function is active), and when manually trimming when the autopilot is disengaged.

When the GSA 28 servos are powered off, the power/ground inputs from the 3rd party trim mixing devices (e.g. VP-X) drive the trim motors directly since the powered off GSA 28 just passes these signals straight through.

Steve
 
By setting the VPX trim to Speed = 0, I believe you are taking the VPX out of the control system, other than as a method of mixing the switch inputs from both sticks. In this way you avoid the 2 different systems from "fighting each other". This is basically what I was suggesting and is a reasonable way to achieve the desired results.

Sorry if I caused confusion, but I don't have a VPX system in my RV-10, and I do have the GAD 27 (no relays). I didn't check my wiring diagram before posting to see that the trim switches are run through the GAD 27 instead of the 24.

Regards,
 
I have an identical avionics architecture in Glasair 3rg to fly in the fall. So I understand value set >80 or zero? It’s run through the vpx for wiring.

Edit: Steve’s lady post clarifies, set it to 0.
Thx
 
Last edited:
Because it's my experimental aircraft, that's why.

In that case you can further experiment by having the G3X GAD, and the VPX, and the auto pilot, and the buttons, and the buttons on the other side control it. Heck, add in a few fuses and breakers, get a few more grounds, maybe add an Arduino in there. Lots of ways to experiment!

Why did you do it that way is a perfectly reasonable question to ask when it's not working right, especially when run away trim could possibly convert the experiment into a smoking hole at the end of the runway.
 
Airspeed Thresholds

The Garmin set-up screen I’m not positive what their intent is. You need to input the airspeed thresholds (in knots) for the fastest movement and slowest movement. The manual doesn’t document what the meaning is. For me I used 80 knots as my fastest movement and set the speed and 120 knots knots as the airspeed threshold I wanted the slowest movement. It’s logical that implies from 0 to 80 knots you get the fastest movement and then 80 to 120 knots it’s a linear step down to 120 knots where you get the slowest movement. This is not documented anywhere in the 496 page user manual I’m pretty sure. By the way when you set the VPX parameters with speed set to zero (which disables the speed function in VPX) according to VPX help line the % power remains active. Can anyone confirm this ? If this is true probably better off not reducing the voltage and let the Garmin PWM control the speed and not have voltage drop risk no movement of the servo in flight like other readers have had issues. This does get a little confusing which system specifically is handling the control. By the way Stein did my panel and it’s sounds like it’s wired exactly as the op did his panel.
 

Attachments

  • IMG_6931.jpg
    IMG_6931.jpg
    190.3 KB · Views: 89
  • IMG_6930.jpg
    IMG_6930.jpg
    167.4 KB · Views: 84
What revision of the G3X installation manual are you looking at? The current version is Revision AQ. Trim speed configuration is documented in section 35.4.13.4, and Figure 27-1.10 tells you what to do with your Vertical Power box.
 
% Power

With the speed set to 0 the % Power adjustment (40 - 90%) has no impact on actual speed of the pitch trim motor. Makes since but obviously the VP help desk didn’t state this correctly. Would make more since to write the program input screen as NA to let people know this has no effect on trim speed. I’m guessing thread authors who said they could dial there % power down to 50% and the servo still worked did not realize if they have the speed at 0 this disables letting VP control this function and the Garmin system takes over this function. I’m going to test the theory that the Garmin speed control is linear by timing full up elevator trim to full down and back while I’m flying between my upper and lower airspeed thresholds in the next few days.
 
As a quick background, let me say I design 3-phase 450V electric vehicle-scale (100-300kW) inverters and their control software.

It sounds like Garmin/VPX's approach to speed control for these trim servos using % PWM is inadequate because with flight loads the lower %PWM settings are insufficient to exceed the stall torque, and the problem gets worse as airspeed increases. Are there any advanced settings, perhaps using closed-loop velocity+position control using the servo's position feedback? I think the simplest answer would be to apply 100% PWM for 50-100ms to get the motor moving and then drop it to the reduced setting for the current airspeed.

As a side note, the (85 lb) motor next to me requires 45-50A just to break free of its cogging torque, but then continues to spin readily with 4A at a few hundred RPM.
 
GSA 28 Trim Output Questions

Good Morning,

The Garmin set-up screen I’m not positive what their intent is. You need to input the airspeed thresholds (in knots) for the fastest movement and slowest movement. The manual doesn’t document what the meaning is. For me I used 80 knots as my fastest movement and set the speed and 120 knots knots as the airspeed threshold I wanted the slowest movement. It’s logical that implies from 0 to 80 knots you get the fastest movement and then 80 to 120 knots it’s a linear step down to 120 knots where you get the slowest movement. This is not documented anywhere in the 496 page user manual I’m pretty sure. By the way when you set the VPX parameters with speed set to zero (which disables the speed function in VPX) according to VPX help line the % power remains active. Can anyone confirm this ? If this is true probably better off not reducing the voltage and let the Garmin PWM control the speed and not have voltage drop risk no movement of the servo in flight like other readers have had issues. This does get a little confusing which system specifically is handling the control. By the way Stein did my panel and it’s sounds like it’s wired exactly as the op did his panel.

If you are using a VP-X, you need to disable trim speed control in the VP-X LRU. As Matt mentioned, this is stated on the GSA 28 Trim Switch Interconnect drawing, for reference.

The value you enter for slowest movement is the slowest trim will move under manual trim command. The value you enter for fastest movement is the fastest trim will move under manual trim command. Between your low and high airspeed threshold, trim speed will increase linearly from your programmed lowest trim speed to your programmed highest trim speed.



Thanks,

Justin
 
Back
Top