What's new
Van's Air Force

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

Hooking up dual resistor float senders in a single fuel tank

R

RVFan671

Well Known Member
I'm installing the RV-10 ER tanks from Sky Designs and at the moment I'm planning to use the Garmin G3X avionics. I am using the Stewart Warner 385 series floats that Vans sells.

Since I am about to proseal the floats to the baffle I want to know if there is any surgery required on the internals of the sender so I don't have to find out later and cut them out. My preferred approach is to just hook the one output wire from the outboard sender to the output wire location of the inboard sender and then run the single wire from that inboard sender output to the avionics. It'll make wiring simple and removing the tank simple, if required.

Is my preferred approach the best (most accurate, or not far off the most accurate option)? Will it work if I change my mind and go with Dynon or something else?
 
What you are proposing is to run the senders in parrallel. As both senders are locally grounded there is no option to run them in series.

So here:
- Both tanks full, each sender is at ~30 ohms so the resistance as seen by your EMS is ~15 ohms.
- Both tanks empty, each sender is at ~240 ohms so the resistance as seen by your EMS is ~120 ohms.
- Any level in between will be somewhere from 15 to 120 ohms.

I suggest you contact your EMS provider and ask if the EMS will calibrate using a sending resistance of 15 to 120 ohms.

Carl
 
The thing to note here is that only the inboard sensor moves when near empty, and only the outboard sensor moves when near full. e.g., the response is not linear, and would be noticeable on a meter. But most efis units allow you to provide multiple calibration points (GRT does, I presume they all do). Just make sure you use 3 or 4 points for the first 10 gal, 3 or 4 for the last 10 gal, 3 or 4 (or more) for the levels in between.
 
It seems like this approach would result in a U-shaped calibration curve, which your avionics won't like.
 
Piper Dakota PA-28-236 uses two fuel senders in each tank in a summing configuration. As I recall the outboard sender is isolated from the airplane with non-conductive washers and bushings. Different part # but basically the same. Only one terminal on each sender.


2i5qdwE.png
 
Glad

RVFan671 said:
I'm installing the RV-10 ER tanks from Sky Designs and at the moment I'm planning to use the Garmin G3X avionics. I am using the Stewart Warner 385 series floats that Vans sells.

Since I am about to proseal the floats to the baffle I want to know if there is any surgery required on the internals of the sender so I don't have to find out later and cut them out. My preferred approach is to just hook the one output wire from the outboard sender to the output wire location of the inboard sender and then run the single wire from that inboard sender output to the avionics. It'll make wiring simple and removing the tank simple, if required.

Is my preferred approach the best (most accurate, or not far off the most accurate option)? Will it work if I change my mind and go with Dynon or something else?
Click to expand...

I had this idea 6 years ago when I had just finished my tanks and the rear baffle was pro-sealed in. If i ever have to rework my tanks, this is exactly what i will be doing. Good luck and let us know how it works out.
 
RVFan671 said:
I'm installing the RV-10 ER tanks from Sky Designs and at the moment I'm planning to use the Garmin G3X avionics. I am using the Stewart Warner 385 series floats that Vans sells.

Since I am about to proseal the floats to the baffle I want to know if there is any surgery required on the internals of the sender so I don't have to find out later and cut them out. My preferred approach is to just hook the one output wire from the outboard sender to the output wire location of the inboard sender and then run the single wire from that inboard sender output to the avionics. It'll make wiring simple and removing the tank simple, if required.

Is my preferred approach the best (most accurate, or not far off the most accurate option)? Will it work if I change my mind and go with Dynon or something else?
Click to expand...

I’ve completed my tanks and have almost completed the wing modifications.
I also have a G3X and what you propose is exactly what I am planning to do. I briefly considered passing the ground through the tank to series rather than parallel but I don’t think it’s necessary.
I’ll likely calibrate in 10L increments. Hopefully I’ll get that done in the next week or so and report back.
Perhaps I’ll start a new thread with some photos about the wing butchering. It’s not too hard and the instructions and tooling is first rate.
Cheers
 
Last edited:
Total fuel capacity readings on gauges isn’t necessary

I don’t really see the necessity since the float gauges aren’t too accurate. Let’s take the RV7’s. The float senders read full at around 15 gals for a 21 gal tank. The importance of your accuracy in fuel is not when you are mostly full but when you are half empty and less when you still have plenty of time to make fuel decisions, and that’s if you don’t have a fuel flow meter. Of course for most of your RV’s with 38-42 gal capacities you’ll start seeing movement when you’re at 30 gals or less.

Most of us by now have fuel flow meters which are a heck of a lot more accurate than gauges. The gauges are there to verify large differences between the gauges and fuel remaining calculated by the fuel flow meter. IMO total fuel capacity accuracy (whether I read 42 gals) for the gauges isn’t necessary. Seeing changes after 30 gals remaining is enough
 
FireMedic_2009 said:
I don’t really see the necessity since the float gauges aren’t too accurate. Let’s take the RV7’s. The float senders read full at around 15 gals for a 21 gal tank. The importance of your accuracy in fuel is not when you are mostly full but when you are half empty and less when you still have plenty of time to make fuel decisions, and that’s if you don’t have a fuel flow meter. Of course for most of your RV’s with 38-42 gal capacities you’ll start seeing movement when you’re at 30 gals or less.

Most of us by now have fuel flow meters which are a heck of a lot more accurate than gauges. The gauges are there to verify large differences between the gauges and fuel remaining calculated by the fuel flow meter. IMO total fuel capacity accuracy (whether I read 42 gals) for the gauges isn’t necessary. Seeing changes after 30 gals remaining is enough
Click to expand...

I do agree to a point.
But for the ER tanks the delta is far more significant than that
If you’ve seen what I wrote previously about the 7 vs the 10 you’ll see the rationale.
I’ve had all types of senders over time.
IMHO staring at a fuel gague for HOURS on end with it grossly inaccurate is asking for trouble eventually. It’s trivial to make it accurate in all phases of flight. So why not do it?
ER tanks are much more like a real airplane. Some days you’ll fill it to full. Other days you won’t.

Totalizers based on flow only are a super big hole in the Swiss cheese in these scenarios. With the ER tanks the inboard sender is pegged before you’ll see fuel at the outboard end. Are you sure you know how much fuel you have?

Whereas in my RV7 I don’t think I ever partially filled it.
 
Richard Connell said:
IMHO staring at a fuel gague for HOURS on end with it grossly inaccurate……

With the ER tanks the inboard sender is pegged before you’ll see fuel at the outboard end. Are you sure you know how much fuel you have?
Click to expand...

If your second statement, above, is true (hard for me to believe) then you have exactly the scenario you fear - there is a region where you don’t know how much fuel you have in the tank.

I disagree with your first statement (‘grossly inaccurate’). My single float in a 30 gal tank won’t read above 25 gal, but below that is good to a gallon or two (data calibrated and massaged thru the (grt) efis.)
 
BobTurner said:
If your second statement, above, is true (hard for me to believe) then you have exactly the scenario you fear - there is a region where you don’t know how much fuel you have in the tank.

I disagree with your first statement (‘grossly inaccurate’). My single float in a 30 gal tank won’t read above 25 gal, but below that is good to a gallon or two (data calibrated and massaged thru the (grt) efis.)
Click to expand...

I’ll verify exactly this week when I hopefully finish install and cal, but given where the tab is and what I know from flying around my stock 10 then I suspect it’s true. If not it’ll be close.

When I said grossly inaccurate im referring to the hours on end where it’s reading full but you have burnt some number between 0 and 30gallons.
 
This concept is very common with Cies Fuel senders. There are several examples for ER tanks with dual Cies senders feeding G3X’s flying today.
 
The issue with hooking the sending units up in parallel is that you will have less resolution over the range. It appears that there will be 1/4 the resolution in parallel as in series. I have attached a spreadsheet showing the difference between parallel and series resistance. I made some gross assumptions on the fuel level verses resistance for each sending unit. In series the range is 40-480 ohms. In a parallel configuration it is 10-120 ohms. If I were doing this I would modify the inboard unit to bring the ground out through a sealed feedthru and hook them up in series.
 

Attachments

  • fuel level.jpg
    fuel level.jpg
    72.2 KB · Views: 77
mburch said:
I suppose that's correct... not U-shaped, but definitely not linear. Careful calibration would be a must.
Click to expand...

Correct. It will be non linear, in part due to the outter sender moving faster as well as the overlap area. However, by about 1/3 full, it will be exactly the same as having one sender (just a different ohm range, but same linearity) and this is the safety critical area. Also, The one sender approach is quite far from linear, as it doesn't even record 25% of tank volume and once past that is still not linear, due to the dihedral angle. The non-linear nature is absolutely a non issue with an EMS, as many calibration points can be registered to deal with the non linear nature.
 
Last edited:
Another poster

Bill.Peyton said:
The issue with hooking the sending units up in parallel is that you will have less resolution over the range. It appears that there will be 1/4 the resolution in parallel as in series. I have attached a spreadsheet showing the difference between parallel and series resistance. I made some gross assumptions on the fuel level verses resistance for each sending unit. In series the range is 40-480 ohms. In a parallel configuration it is 10-120 ohms. If I were doing this I would modify the inboard unit to bring the ground out through a sealed feedthru and hook them up in series.
Click to expand...

Another poster mentioned adding additional fixed resistors in line to help balance the readings and bring the resistance up of each leg. This might work for audio signals but I think will exasperate the problem with the fuel quantity system. I think it would be better to run a low resistance and get as much change as possible, then to add resistance and reduce the range of change. JMHO
 
Bill.Peyton said:
If I were doing this I would modify the inboard unit to bring the ground out through a sealed feedthru and hook them up in series.
Click to expand...

Please elaborate on how to do this. Maybe pictures would help, too. I'm not sure how one would break the existing ground setup. Stuffing it through the tank in a newly drilled hole is self explanatory, but then how do you propose hooking it into the other senders ground?
 
I think all that is required is to take the signal wire from the outboard sender, pass it into the tank at the inboard sender end using a wire gland/mini bulkhead fitting - ideally small enough to fit in the sender plate itself.
Desolder the ground at the sender and solder the signal wire in its place.
Now they are in series.
 
Last edited:
Richard Connell said:
I think all that is required is to take the signal wire from the outboard sender, pass it into the tank at the inboard sender end using a wire gland/mini bulkhead fitting - ideally small enough to fit in the sender plate itself.
Desolder the ground at the sender and solder the signal wire in its place.
Now they are in series.
Click to expand...

Seems like something like this would be a good option. Slight modification would be required since the tank baffle is thinner than the 1.5MM called out in their tech sheets.
 
RVFan671 said:
Seems like somethinglike this would be a good option. Slight modification would be required since the tank baffle is thinner than the 1.5MM called out in their tech sheets.
Click to expand...

As it happens I have procured something very similar and I’m trying to install it into the sender plate itself. Has an o-ring on each half too. I’ll still goop plenty of proseal on the inside as well.
It’s tight but it seems to be working out so far. It’ll be nice to have it all self contained.

F42194E4-FE9E-4ADB-B944-83BE6A87EC08.jpg
 
Have you done a bench test - series vs parallel.

A few minutes with an ohm meter might be time well spent.
 
Webb said:
Have you done a bench test - series vs parallel.

A few minutes with an ohm meter might be time well spent.
Click to expand...



I’ve bench tested my continuities and grounds.
I’ve got no reason to believe the senders and efis won’t meet specifications.

Ultimately this series v parallel discussion comes down to whether the particular efis resistance measurement is accurate enough to have say 20 calibration points spread over only 100Ω (cal point every 5Ω) If that’s the case then parallel is fine.
20 points spread over 400Ω (cal point every 20Ω) on the face of it is better.

My DMM spec is +/- 1% plus one digit. So in the 200Ω range it’s really 197-203
If EFIS is the similar then we really want to use the biggest range possible.

Thanks Bill for making me think about this a bit more.
 
Last edited:
Garmin and Dynon resistance ranges

Confirmed with Dynon via email that the Skyview setup reads 0 to 500 ohm. Garmin G3x system also does (found it in their documentation. Not sure how many or how small of ohm changes can be set for each though.
 
You must log in or register to reply here.
Back
Top