What's new
Van's Air Force

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

A Red Cube, a G3X touch, and thou...

bjdecker

Well Known Member
Ambassador
The riddle of the day --

What shows 18.2 GPH at take off, 18.1 GPH at top of climb (5500ft), 12.5 GPH in cruise (2500RPM, 24" MAP)?

My RV-7 :(

---

Back story: I've been trying, with varied degrees of success, to get my G3X Touch to display correct fuel flow #'s for the past 200+ hours.

I have a Red Cube (FT-60) with straight steel fittings (AN816-6) and fuel lines without any appreciable bends within 6" of the cube. The cube is mounted with Adel clamps and AN4 bolts to the #2 & #4 intake runners (identical to RV-14 suggestion, page 43-09), and a heat shield mounted on the #4 exhaust stack.

Connections from cube to GEA 24 is a 3 position TE Mate-N-Lok connector, which in turn is strain relieved and secured to adjacent structures (read: not wiggling or vibrating in the breeze...)

I started with a K-factor of 68000, then adjusted to 72000, then 84000, then 77400, then 75009, then 73201, then 83024, and so on and so on. Adjusting the parameters became a lot easier with the more recent software revisions of the G3X touch (v8.6 and later), but each/every flight I adjust the gallons burned/filled as reported by the G3X to what the pump says I took on board, a gallon or two over, or under, ad-infinitum...

For what its worth, I had this exact same setup (engine, propeller, cube, cube location, ignitions, airframe, etc...) except I was using Rob Hickman's AF-3400EM -- the fuel flow readings were D.N.O., each time, every time...

So, I'm begging for the wisdom of the hive-mind; Is this just the way it is? Is there something else I should look at?

B
 
Temperature change of the fuel when filling up? I recall that a 20-30 degree change in fuel temp can be, what, a half gallon or more difference on tanks our size? (Don't have the formula or constants handy, but if one time you refuel the fuel is cold, and the next it's significantly warmer, you'd get different readings on the fuel pump).

Just a thought.
 
How close is the cube to the exhaust pipes? Any heat shields on the exh? How many hours on the red cube? Occasional boiling of the fuel in or near the cube could account for the variances. On my 10, the cube is installed in a cooler area in front of the servo. After a few adjustments, it is very stable. Quite unlikely that this is an issue with the Garmin s/w. I believe that bubbles in the fuel will create inconsistent readings with the red cube, as can intermittent connections in the wiring.

A picture would help. Sorry, don't know what the 14 layout is.

Larry
 
Last edited:
Brian,
I had a similar issue and replaced the Red Cube. Same issues with it mounted per the RV14 suggestions. I decided to mount it on the firewall similar to the way my RV7A was set up. It is downstream of the high pressure pump before the mechanical pump. I know it's not the recommended location but I can tell you all my issues went away, that was 700 hours ago. I check the fuel quantity added vs what the totalizer indicates at every fill up. It is always within .3 - .5 gallons. Some of that difference is who tops the tanks off.

My setup is all Dynon but the Red Cube issues were very similar to what you described.

Gary
RV10
 
Generally, the recommended location is between the servo and fuel distribution manifold (spider). Not sure where yours is located from your post.
 
Temperature change of the fuel when filling up? I recall that a 20-30 degree change in fuel temp can be, what, a half gallon or more difference on tanks our size? (Don't have the formula or constants handy, but if one time you refuel the fuel is cold, and the next it's significantly warmer, you'd get different readings on the fuel pump).

Just a thought.

Good thought -- The error is in the 2~3 gallon range after burning 25 or so. It
s been hotter than blazes (this is Texas after all), but it doesn't seem to track temperatures. It's just off, and adjusting it only makes it still off...
 
Brian,
I had a similar issue and replaced the Red Cube. Same issues with it mounted per the RV14 suggestions. I decided to mount it on the firewall similar to the way my RV7A was set up. It is downstream of the high pressure pump before the mechanical pump. I know it's not the recommended location but I can tell you all my issues went away, that was 700 hours ago. I check the fuel quantity added vs what the totalizer indicates at every fill up. It is always within .3 - .5 gallons. Some of that difference is who tops the tanks off.

My setup is all Dynon but the Red Cube issues were very similar to what you described.

Gary
RV10

Gary,

I'm beginning to wonder if some of the variability is the last gallon or so needs to be put in to get it full (touching the bottom of the flange). The line folks always forget to do that, or spill it all over the wing :/

B
 
Generally, the recommended location is between the servo and fuel distribution manifold (spider). Not sure where yours is located from your post.

Hmmm, following the recommendations from EI tends to preclude mounting it between the spider and the servo (vertical orientation of the probe wires.).

I may pursue this if other experiments and getting the data set squared away doesn't solve the issue.

B
 
How close is the cube to the exhaust pipes? Any heat shields on the exh? How many hours on the red cube? Occasional boiling of the fuel in or near the cube could account for the variances. On my 10, the cube is installed in a cooler area in front of the servo. After a few adjustments, it is very stable. Quite unlikely that this is an issue with the Garmin s/w. I believe that bubbles in the fuel will create inconsistent readings with the red cube, as can intermittent connections in the wiring.

A picture would help. Sorry, don't know what the 14 layout is.

Larry

Hey Larry -

Attached pic is the RV-14 guidance. I opted for an AN816-6 on the inlet to the cube instead of the AN823-6D that Van's shows.

There is a heat shield installed around the #4 exhaust stack.

This cube has about 100 hours, I replaced the previous one due to the same issue -- waste of money since the problem persisted.



B
 

Attachments

  • Screen Shot 2022-07-11 at 2.58.34 PM.png
    Screen Shot 2022-07-11 at 2.58.34 PM.png
    1.5 MB · Views: 119
  • Screen Shot 2022-07-11 at 2.59.12 PM.png
    Screen Shot 2022-07-11 at 2.59.12 PM.png
    1.3 MB · Views: 124
Hmmm, following the recommendations from EI tends to preclude mounting it between the spider and the servo (vertical orientation of the probe wires.).

I may pursue this if other experiments and getting the data set squared away doesn't solve the issue.

B

Check the CGR30 install manual for many options for installation of the FT60.

"D. The flow transducer must be mounted so the wires exiting the transducer are pointing up, or the cap with five bolts is pointing up, or the output port is pointing up, or any combination thereof."
 
Check the CGR30 install manual for many options for installation of the FT60.

"D. The flow transducer must be mounted so the wires exiting the transducer are pointing up, or the cap with five bolts is pointing up, or the output port is pointing up, or any combination thereof."

Totally missed that note -- thanks.
 
Hey Larry -

Attached pic is the RV-14 guidance. I opted for an AN816-6 on the inlet to the cube instead of the AN823-6D that Van's shows.

There is a heat shield installed around the #4 exhaust stack.

This cube has about 100 hours, I replaced the previous one due to the same issue -- waste of money since the problem persisted.

B

In the second pic, it looks like the 45* AN fitting is about an inch from the exhaust pipe; Much closer than I would like it. How close are your fittings or the cube getting to the exhaust pipe?
 
In the second pic, it looks like the 45* AN fitting is about an inch from the exhaust pipe; Much closer than I would like it. How close are your fittings or the cube getting to the exhaust pipe?

Just checked a friends 14A, it's a bit more than an inch, my install it's about 2"...
 
Last edited:
Based on the system working well with the AFS 3400 before the GX3 change, it may be possible the Red Cube wire is introducing some EMI into the EIS module causing the erratic indication. Can you reroute the wire or slide a braid over the existing wire and ground the braid on one end?

Not sure how the AFS 3400 sensor inputs compare with the Garmin EIS module inputs, but with a few generations of chip differences between the two units it could be more sensitive to EMI over the previous generations.
 
Just checked a friends 14A, it's a bit more than an inch, my install it's about 2"...

2" is probably enough. However, an interesting experiment would be to heavily wrap the cube and fittings, anywhere near the exhaust, with fiberglass and see if the behavior changes. That would rule heat in or out as a cause.

Larry
 
Based on the system working well with the AFS 3400 before the GX3 change, it may be possible the Red Cube wire is introducing some EMI into the EIS module causing the erratic indication. Can you reroute the wire or slide a braid over the existing wire and ground the braid on one end?

Not sure how the AFS 3400 sensor inputs compare with the Garmin EIS module inputs, but with a few generations of chip differences between the two units it could be more sensitive to EMI over the previous generations.

It is a pretty simple system that should tolerate EMI relatively well. The cube sends a 5 or 12 volt (memory is failing me) square wave pulse every time that a paddle on the wheel passes the optical sensor. The K factor is what the EMS uses as a multipler. For example, a K factor of 68000 means that 68000 pulses equals one gallon. Pretty simple stuff.
 
Based on the system working well with the AFS 3400 before the GX3 change, it may be possible the Red Cube wire is introducing some EMI into the EIS module causing the erratic indication. Can you reroute the wire or slide a braid over the existing wire and ground the braid on one end?

Not sure how the AFS 3400 sensor inputs compare with the Garmin EIS module inputs, but with a few generations of chip differences between the two units it could be more sensitive to EMI over the previous generations.

Thanks Don as always -- just a point of clarification, the airplane in question has >always< been a G3X platform; The AF-3400EM system was in the prior airplane that I built.

The similarities are the engine (IO-360-A1B6), ignitions (Plasma IIIs), Prop (Hartz BA), Red-cube, Wiring to/from red-cube (22ga 3 conductor+shield; shield dis-contiguous - open at the Mate-N-Lok - case grounded at the EIS/GEA 24.
 
Update -- It wasn't obvious...

After a few folks dropped by and poked around the airplane, the observation was made that the Red Cube wasn't as secure as it could/should be.

I attempted to change the clamps from WDG-26's to WDG-25's but they were just not having any of that. I then cut a bit of baffle material (Aerosilk) and worked it into the gap between the intake runners and the WDG-26 clamps. Using this method, I was able to eliminate the majority (read: all) of the wiggle that had been observed (Thanks again Check6).

Today's flight and refueling showed flows, and ultimately fuel burned totals, were much closer in tolerances -- Adjusted K factor is now 74012.

We'll see how it behaves subsequent flights.

B
 
Back
Top