What's new
Van's Air Force

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

Red Cubes - Junk?

FasGlas

Well Known Member
This is the third Red Cube I've installed in my Lancair.. And this is the third Red Cube that's junk! The wiring is solid, the connections are solid, the cube is behind the baffling mounted level with the wires up. I tried fire shield to see if there was any heat getting to it. I removed them and checked for any debris.. I even changed the RDAC to make sure it's not that. Nothing... Just intermittent - All three. Next flight it might work, next flight it might not. These are junk. I've seen many complaints here about them and I've had it.
 
This is the third Red Cube I've installed in my Lancair.. And this is the third Red Cube that's junk! The wiring is solid, the connections are solid, the cube is behind the baffling mounted level with the wires up. I tried fire shield to see if there was any heat getting to it. I removed them and checked for any debris.. I even changed the RDAC to make sure it's not that. Nothing... Just intermittent - All three. Next flight it might work, next flight it might not. These are junk. I've seen many complaints here about them and I've had it.

Are you connecting with the Faston terminals or soldering - there have been several reports about issues with the terminals. I soldered mine. Also using shielded wire properly grounded to the receiving equipment?

Did you send one back to the manufacturer for confirmation of failure?
 
Alternative

Mine has never inspired confidence, but today it went completely erratic.
What is the logical alternative to this sender?
 
This is the third Red Cube I've installed in my Lancair.. And this is the third Red Cube that's junk! The wiring is solid, the connections are solid, the cube is behind the baffling mounted level with the wires up. I tried fire shield to see if there was any heat getting to it. I removed them and checked for any debris.. I even changed the RDAC to make sure it's not that. Nothing... Just intermittent - All three. Next flight it might work, next flight it might not. These are junk. I've seen many complaints here about them and I've had it.

I've installed plenty and had the same unit on my own aircraft for over 10 years, never had any issues with any of them.
 
There is also the Floscan 201 sensor. I have one. 9+ years. I confess it gets a little erratic once in a while. But it works right ALMOST all the time. It seems to be a little G-sensitive, and of course it does not like being a foot downstream of the boost pump.
 
I'd replace the wiring to the red cube entirely. It not likely that you would have three go bad. I would also make sure the supply voltage is coming from your engine monitor.
 
Last edited:
I'd replace the wiring to the red cube entirely. It not likely that you would have three go bad. I would also make sure the supply voltage is coming from your engine monitor.

I've changed the wiring. I use shielded Teflon cable 3 connector. I use aviation spade connectors from the RDAC cable to the Red Cube. I've metered the wires and shook them every which way, solid connections, solid voltage from the RDAC. I've replaced the RDAC. I've fire shielded the Red Cube. The connections to the RDAC are screw down solid. The clamps have rubber on them for some vibration cushion. I've installed many Floscans and a few Red Cubes. It's not rocket science. It's connected between the servo and spider. I'm so sick of these things. I've returned the last two to MGL now I'm on my third one in the last two years.
 
Last edited:
Almost 9 years on mine. Works fine. Might die tomorrow but for now it works.

The one time I did have an issue, turned out it was a poor crimp on the factory connector.
 
Last edited:
Over 9 years with two of them installed in my plane (have return fuel line with one to measure return fuel). They have worked flawlessly. I do not have them on the hot side of the firewall. I do not think it wise to install them in that environment.

If you have had three of them go out with all of them mounted with the same results, perhaps you should explore the notion there is something else killing them besides bad manufacturing.
 
MGL is either a ending you the same cube back each time or there is something off in your wiring....or panel. My red cube is mounted inside my plenum, is not rubber mounted, shielded or wrapped. Ive put enough miles on it to lap the equator three times. The red cube is not junk. Look into your wiring or look very carefully at what you?re wiring it to.
 
My son Peter had trouble with fuel flow using a MGL RDAC-XF.
MGL tried to help him many times. The customer relations were good. He tried all the recommendations and we reviewed the circuit, locations, 'noise' potential, everything we could think of...
He tried the Red Cube and the Floscan with the same result....erratic and useless.
Then one day I suggested he connect the leads to the second flow monitor input (FF2). Now it works fine. I think its the input for a fuel return flow meter? Anyway, the RDAC may have damage to the first fuel flow circuit?
Try the second fuel flow input channel?
 
Like the others, I have zero problems with my red cube (over 720 hours and counting). One other thing to check: I recall that the cube must be oriented a certain way, one end low and one end high (I don?t remember which end goes low, but it?s in the manual).
 
I think the next thing I'm going to do is connect a pocket scope to the pulse wire and fly. There is nothing left I haven't done to troubleshoot this intermittent fuel flow problem. I can monitor the pulses from the EFIS and see they are slowed down. Sometimes it works and sometimes it doesn't. It never goes to 0, it just slows down to maybe ~1 GPH then after a while will come up to ~10 GPH. Hot out, cold out makes no difference, just slows down when it feels like it. I run 2 RDAC's, the other RDAC has 2 Red Cubes that read the fuel flow from the main tanks to the header tank. They seem to work but are used only when moving fuel to the header, only ~5 minutes, not constantly like the engine fuel flow.
 
My Lancair uses the FTX-201 Floscan transducer that came with my JPI EDM-700 and it has worked well, even though it's not in an ideal location, plumbingwise. On a carbureted Lycoming, it's not really possible to mount the transducer lower or even with the carburetor inlet port. Mine is located in a horizontal run between the Facet boost pump and mechanical fuel pump. The only "problem" I've had is that it reads incorrectly when the boost pump is running. Since I rarely use the boost pump, that hasn't been a problem.
 
How is your engine grounded?

As Rocketbob stated how is your engine grounded? Just thinking out loud hear. If the engine is not grounded well when the starter is activated anything attached to the engine, or engine baffling could be damaged. Make sure wherever on the engine you have the engine ground connected that the paint is also sanded off in that area.

Bill
 
Must be a Lancair thing.. I've been through 2 of them and the third is starting to come and go randomly. The last one I sent back to EI to test and the response was "Yes, it's bad, we'll sell you one at a discount." Not wanting to change all my plumbing I bought it again. It is installed between the mechanical pump and the fuel servo. I is under fire sleeve. Regular spade connections. I've redone the wiring to the UGB-16 once before, no change. I'm about to scrap the whole panel so I'm looking forward to making some kind of change.
 
Randy, with literally 1000's of them flying, I would suspect that there is something in your system that the little electron thingys dont like. Doubt that its a mounting issue because they are mounted all over the place in different attitudes with no issues. Even in high G aerobatic planes.

Tom
 
One other thing to check: I recall that the cube must be oriented a certain way, one end low and one end high (I don?t remember which end goes low, but it?s in the manual).

The red cube seems to work fine in many "non-approved" orientations, and with many "not-recommended" inlet and outlet fitting arrangements. I have 45 degree connectors on both sides, it's mounted downstream of the fuel pump, and it works perfectly with the only exception being it reads a little high when the pump is on. Connected to an MGL Velocity fuel flow/level meter, FWIW.
 
Me too

Less than 80 hours on my Red Cube and it went erratic. New one installed, works perfectly so far.

Haven't sent the old one back yet to find the failure mode, but would like to know what happened.
 
I installed my "Red Cube" back in 2012, it now has more than 500 hours and is still running great. Sounds like there may be something in your particular installation.

:cool:
 
I installed an EI CGR30P on my Aviat Husky last year. Installed the red cube between the mechanical fuel pump and Berndix FI servo, per EI instructions (my Husky is fuel injected). The red cube itself is attached to the firewall. The inlet and outlets are both 90 degree elbows. EI said that was okay but I might have to do more K-factor tuning.

I flew about 80 hours and the flow never registered below 1200 rpm. EI had me send in the cube and said it was bad and replaced it under warranty. The warranty replacement (installed in October) worked great but just stopped working after 20 hours. I am sitting here looking at it on my workbench. I am going to call EI on Wednesday.

I have had great luck with Flow-Scan transducers on JPI monitors but this first foray into red cube-ville hasn't been pleasant. My red cubes have been FAA PMA TSO'd units since the Husky is type certified.

Jim
 
Last edited:
I've seen lots of complaints here about Red Cubes failing so I know it's not just mine. I know it's mounted right and the wiring is right and solid. It's out of the heat and the cable is away from plug wires.... The power and ground that runs the Red Cube comes from the RDAC, no other sensors are failing. The only thing left is to scope the pulses to isolate the cube from the RDAC. Always something............
 
Install

Have you installed the unit in line after all fuel pumps according to the instructions? Mine is mounted on the fire wall between the carb intake and
Engine driven fuel pump. It also has straight fittings in and out. Works well.
 
Just curious, what do the spam cans use for a fuel flow sensor?

They mostly use Floscan sensors. Mounted between the servo and divider (FI). I've installed many in certified and they seem to work very well if installed properly.
 
Can i use one

They mostly use Floscan sensors. Mounted between the servo and divider (FI). I've installed many in certified and they seem to work very well if installed properly.

Can I use one of these with a g3x? How would i wire it up?
 
Almost 9 years on mine. Works fine. Might die tomorrow but for now it works.

The one time I did have an issue, turned out it was a poor crimp on the factory connector.

Almost 3 years and 370 hours on mine, and the only issue I have had was due to my poor wiring crimp.
 
Over 10yrs and 800ish hours mounted on the hot side of the firewall on a shelf and still going strong. Accurate to within .2 gallons most top offs. No issues at all. Hope I haven't jinxed myself :)
 
Just out of curiousity...what do people think is happening to cause the FF transducer to read "a little high" when the boost pump is on? All the thing does is measure, via a little impeller sticking into the flow, the amount of fuel *flowing past it*. It doesn't "know" that the boost pump is on.

I guess, in some weird way, depending on where in the system the thing is mounted, it might see some odd pulsing somehow. But, e.g., if it's mounted between the servo and the spider, I don't see any way that it could register "a little high" because of an increase in *pressure* going to the servo from the boost pump.

BTW, FWIW, 600 hours on my red cube and not a hint of trouble. Mounted between the servo and the spider, via a bracket to sump bolts, firesleeved.
 
I'm not qualified to render an opinion on electronics quality, but the red cubes's circuit board is potted in such a way that vibration should not hurt it.

Potted%20Electronics.jpg


The only moving part is the rotor. It spins on a jewel in the main body, and another in the rotor cover.

Rotor.jpg


Here's the rotor in the main body. The arrow points to what I assume to be either a light source or a receiver. The little tabs on the rotor interrupt a light beam between it and...

Rotor%20and%20Eye.jpg


...its counterpart located in the cover.

Trigger.jpg


The device cannot be dismantled without damaging it, which appear to be by design, so cleaning the rotor and its chamber is only possible by flushing. I can see how some loose fuel lube might gum up the rotor, but can't think of much else an owner might do beyond hit it with a voltage spike.
 
Last edited:
Fuel lube???

Seems like a good theory. A few months back I had an issue with my RED CUBE where it fell to Zero intermittently . Then after a few hours it came back.

Maybe it was a little fuel lube I placed on the o-ring of the fuel caps...???

I didn’t like the stickyness amd switched to vasiline. It works much better and “fingers crossed” I don’t have any future RED CUBE issues.
 
My take

Just out of curiousity...what do people think is happening to cause the FF transducer to read "a little high" when the boost pump is on? All the thing does is measure, via a little impeller sticking into the flow, the amount of fuel *flowing past it*. It doesn't "know" that the boost pump is on.

I guess, in some weird way, depending on where in the system the thing is mounted, it might see some odd pulsing somehow. But, e.g., if it's mounted between the servo and the spider, I don't see any way that it could register "a little high" because of an increase in *pressure* going to the servo from the boost pump.

BTW, FWIW, 600 hours on my red cube and not a hint of trouble. Mounted between the servo and the spider, via a bracket to sump bolts, firesleeved.
I suspect the pulse from the fuel pump causes the rotor to jump back and forth sending extra pulses to be counted. I also suspect the fuel servo will smooth these out making those installations more accurate with the fuel pump on. Jmho ymmv
 
FasGlas;1311620 The only thing left is to scope the pulses to isolate the cube from the RDAC. Always something............[/QUOTE said:
When mine was acting up I started thinking about a wiring issue between the Red Cube and the Dynon EMS. On the phone with tech support they mentioned a quick way to test the wiring is to ground the sensor wire running from the Dynon system. It seems the Red Cube is grounding the sensor wire and Dynon system was measuring how long it was grounding. So if I grounded the sensor wire and got a reading on fuel flow the Red Cube was the problem, if no reading the problem was the wiring. I do not know, but it seems your system must be measuring the Red Cube data in the same manner.

As your problem is intermittent it would seem to indicate a problem with that aspect of the wiring...or another bad fuel flow transducer. Hope that helps in some way.
 
When mine was acting up I started thinking about a wiring issue between the Red Cube and the Dynon EMS. On the phone with tech support they mentioned a quick way to test the wiring is to ground the sensor wire running from the Dynon system. It seems the Red Cube is grounding the sensor wire and Dynon system was measuring how long it was grounding. So if I grounded the sensor wire and got a reading on fuel flow the Red Cube was the problem, if no reading the problem was the wiring. I do not know, but it seems your system must be measuring the Red Cube data in the same manner.

As your problem is intermittent it would seem to indicate a problem with that aspect of the wiring...or another bad fuel flow transducer. Hope that helps in some way.

The way the Red Cube works, and probably the Floscan, it's an open collector output (switch). The LED light source, in the cube, is on one side of the slotted wheel and the phototransistor is on the other side. As the wheel spins it shines light on the transistor in a "Pulse" fashion. When the transistor gets that light it opens the circuit. Connected to the signal wire, on the RDAC end, is "Bias" voltage. This bias voltage is what creates the pulses from 0 ~ 5 volts. The pulse is open bias voltage and shorted closed transistor voltage. In the case of the MGL RDAC the bias voltage comes from a 5.6K resistor that comes with the Red Cube and is mounted between the 5 volt connector and the Fuel Flow sensor connector (FF1 or FF2). EI, JPI and others apply the bias voltage internally but they all work the same.
Connecting a scope to the sensor signal wire should show a square wave like pulse as the fuel flows through the sensor. In my case 18000 pulses = 1 liter of fuel or 68,000 per gallon. The math is different per manufacture. This is called the "K" factor.
Any random pulses from stray interference like spark plug wires, etc, will also show on the scope. Missing pulses as well. Depending on the electronics the pulses could be canceled out by interference, depends on the filtering designed into the converter circuits. Other factors like multiplex timing, etc could also effect the RDAC output to the EFIS.
Anyway...... Reading the sensor output to the RDAC is where I'm going to start. It's either pulsing steady and changes with the flow of fuel or it's not. If I'm reading bias voltage and no or intermittent pulses it's a bad Red Cube.
I was PM's by a VAF reader that had to increase his bias voltage from 5 volts to 12 volts to get a more consistent reading from his Red Cube. If this is the case then it could be the RDAC is not getting a strong enough swing in voltage or the Red Cube is leaking light to the phototransistor or the transistor is not fully open and closed during a pulse. This is a next step after scoping the signal. One way or another I will get to the bottom of this.
 
Last edited:
Maybe that’s your problem (and solution). I found an (older) EI manual that specified the voltage to the red wire as 9 -15 volts, e.g., 5 volts may be too low for reliable operation.
Edit. I see someone else also suggested this.
 
The image that Dan posted, presumably from MGL, shows a +12 volt supply to the red cube. Same as what Dynon, GRT and perhaps others supply from their engine monitors as a regulated output for driving fuel flow sensors. So that doesn't appear to be the issue here. It's the bias voltage being applied to the output that's in question. Dynon and GRT both have fuel flow inputs on their engine monitors that are simple one wire hookups (no external bias resistor) so the bias voltage they presumably apply to the signal input isn't obvious.

I too have had a red cube fail, though in my case I think I cooked it due to mounting location. Even so it went to almost 500 hours. Because I had a spare on the shelf I never bothered to determine if it was botched wiring as opposed to something else. Replacement is now up to about 250 hours. I've installed several on other RVs now with a combined several thousand hours of flight time and no failures, all with GRT or Dynon electronics to read the outputs.

Would be interesting to know what bias voltage the two above-mentioned manufacturers supply, and why. Not that it explains the failures...
 
The image that Dan posted, presumably from MGL, shows a +12 volt supply to the red cube. Same as what Dynon, GRT and perhaps others supply from their engine monitors as a regulated output for driving fuel flow sensors. So that doesn't appear to be the issue here. It's the bias voltage being applied to the output that's in question. Dynon and GRT both have fuel flow inputs on their engine monitors that are simple one wire hookups (no external bias resistor) so the bias voltage they presumably apply to the signal input isn't obvious.

I too have had a red cube fail, though in my case I think I cooked it due to mounting location. Even so it went to almost 500 hours. Because I had a spare on the shelf I never bothered to determine if it was botched wiring as opposed to something else. Replacement is now up to about 250 hours. I've installed several on other RVs now with a combined several thousand hours of flight time and no failures, all with GRT or Dynon electronics to read the outputs.

Would be interesting to know what bias voltage the two above-mentioned manufacturers supply, and why. Not that it explains the failures...

That's right. The 12 volts (8~30v) (red wire) is used to power the electronics in the Red Cube (Regulator, LED, phototransistor, etc). The bias voltage is external and separate to the input voltage. Since MGL was designed to use many different sensors they didn't preset the bias voltage internally, they allow the user to have different setups. This is the same with all the sensor inputs on the MGL RDAC. EI, JPI, Dynon, etc all require the setup to be Red Cubes or Floscans. Anyone curious about they're bias voltage can just meter the white wire from the Red Cube and there ya have it.
 
Last edited:
That's right. The 12 volts (8~30v) (red wire) is used to power the electronics in the Red Cube (Regulator, LED, phototransistor, etc). The bias voltage is external and separate to the input voltage. Since MGL was designed to use many different sensors they didn't preset the bias voltage internally, they allow the user to have different setups. This is the same with all the sensor inputs on the MGL RDAC. EI, JPI, Dynon, etc all require the setup to be Red Cubes or Floscans. Anyone curious about they're bias voltage can just meter the white wire from the Red Cube and there ya have it.

Completely as an aside, I am installing a red cube in the latest project, which has a 28 volt system. I contacted EI, and they told me that he cube works perfectly well at 5 - 30 volts - its just not documented. I?ll be using it with MGL.

I now return you to your ongoing debate.... (I have four red cubes in operation that have had no issues BTW).
 
Must be a Lancair thing.. I've been through 2 of them and the third is starting to come and go randomly. The last one I sent back to EI to test and the response was "Yes, it's bad, we'll sell you one at a discount." Not wanting to change all my plumbing I bought it again. It is installed between the mechanical pump and the fuel servo. I is under fire sleeve. Regular spade connections. I've redone the wiring to the UGB-16 once before, no change. I'm about to scrap the whole panel so I'm looking forward to making some kind of change.

Working great on my Lancair. Mine is mounted to the firewall. Amazingly accurate.
 
Well, I went over everything, as I have so many times before. I connected my scope to the RDAC input from the Red Cube sensor. As I said prior there will be a square wave coming from the sensor. Here is a pic of the pulsing. You can see the swing is a bit over 10 volts (2 volts per division), this is the bias voltage I've set the sensor to. This is about 4.5 GPH. Next is to record this in flight to see what's going on.

CIMG5899.jpg
 
I think the next thing I'm going to do is connect a pocket scope to the pulse wire and fly. There is nothing left I haven't done to troubleshoot this intermittent fuel flow problem. I can monitor the pulses from the EFIS and see they are slowed down. Sometimes it works and sometimes it doesn't. It never goes to 0, it just slows down to maybe ~1 GPH then after a while will come up to ~10 GPH. Hot out, cold out makes no difference, just slows down when it feels like it. I run 2 RDAC's, the other RDAC has 2 Red Cubes that read the fuel flow from the main tanks to the header tank. They seem to work but are used only when moving fuel to the header, only ~5 minutes, not constantly like the engine fuel flow.

Potentially a lot of lead swapping, and re programming K factor #'s HOWEVER
Just curious if you swap the suspect red brick leads to your other RDAC if the problem will follow the red brick or stays with the RDAC?

"The wiring is solid, the connections are solid, the cube is behind the baffling mounted level with the wires up."

Any chance it is physically vibrating in a manner that could alter the rotation of the turbine? I suspect it would easy to alter or damp the vibes if on the cool side and see if the problem is altered.
 
Last edited:
Just noticed this thread so let me add what I know.

Here at MGL (in SA) we have never had any reports of problems with the red cube so this comes as a surprise. We don't sell them here (I know the guys in the U.S. do but they operate independent from us).
The red cubes are fairly popular though and most here import them direct.

This thread has good info and some really high quality suggestions and comments so there is not terribly much I can add.

The RDAC input for fuel flow is very simple - resistor, cap (as filter), zener as overvoltage protector and all that is feeding into a schmidt trigger input and from there it goes to a processor. We count the pulses and integrate over a 4 second period and then use the k-factor to work out the flow. The 4 second period was decided on after a lot of experiments with various engine types with the aim of providing a good compromise between speed of updates and unnecessary fluctuations (flow is not really all that constant).

The input can easily handle voltages in excess of 50V - no issue there. The critical part is the voltage with the sender output at the low level - this should be less than one volt - preferably a lot less.
The 5K6 resistor to 5V was chosen based on experiments with a sample sensor. However, neither the voltage (5V) nor the resistor are particular critical.
However - optoelectronics as used in many flow senders can be iffy. The amount of light arriving at the transistor in the sender affects how well the transistor can "pull" down the voltage at its output. Not enough light and it does not go properly to ground. To make matters worse - this has a very high degree of tolerance and is also affected by temperature and aging. The actual light source (usually an infrared LED) also ages and slowly gets dimmer with age. The optical surfaces can also be tainted over time by fuel residue reducing the amount of light that makes it through. Composition of fuels may also have a small effect.
A simple fix for that would be to increase the resistance of the pull up resistor. Perhaps 10K, 15K or 22K could make all the difference with a sender that is not quite up to snuff. Te RDAC does not mind at all (and the reason we did not put a fixed pull up inside is to keep things flexible).

Using a scope is the best way to diagnose issues - a portable scope like the many cheapies you can now get that connect to your phone is a great way to check what happens in flight. Take particular care of noting the "zero" volt level...

As far as flow senders over reading with fuel pumps "on" - yes that is pretty normal and happens with any flow sender I have ever played with. This happens with "pulsed" pumps. What happens is that during the pump stroke fuel is forced under pressure through the sender - this can expand some of the softer bits of your fuel system like perhaps a piece of rubber fuel hose on the other side of the flow sender. When the pump is in its "off" stroke that fuel simply flows back through the sender - maybe creating additional pulses - even if not - this creates an apparent flow that is not real as far as consumption is concerned.
On our own MGL senders you can just make out the rotor - you can see the rotor shunting back and forth with the pump strokes. The MGL sender uses magnets and a hall effect sensor rather than light. However it is not available in the U.S. due to legal concerns.

Rainier
CEO MGL Avionics
 
Ranier, since this is your field of expertise, what if the cube is being suspended between ground contact points using the hose as flow material. And lets assume the hose material does not have the carbon grounding internally. We know that fast flowing fuel will create a static charge (thus the carbon lining for aviation hose) so is it possible that this installation has highly charged fuel going through it and is failing the electronics? Should the case of all fuel transducers be grounded?
 
Ranier, since this is your field of expertise, what if the cube is being suspended between ground contact points using the hose as flow material. And lets assume the hose material does not have the carbon grounding internally. We know that fast flowing fuel will create a static charge (thus the carbon lining for aviation hose) so is it possible that this installation has highly charged fuel going through it and is failing the electronics? Should the case of all fuel transducers be grounded?

This is an interesting question. Static is one of my favorite topics as it happens to be perhaps one of the most neglected issues in many constructions.

In my opinion the chance of a static discharge damaging the senders opto electronics is not going to happen - the metal body of the sender, grounded, will shunt any change to ground easily. Should the housing however not be grounded it would form a ready interface for any charge to cause a discharge internally into any of the three wires connected to the sender or to any part of the internal electronics. Just depends on the path of least resistance. In a case like this (as unlikely as it may be) you could very well end up with damaged electronics.
Of course the charge has to be high enough to overcome any insulation - gut feeling tells me this is not going to happen.

Rainier
CEO MGL Avionics
 
Back
Top