JPI 700 K-factor calibration question

[Stryk77]

Hello everyone!

I have a quick question for you all.  I recently had a JPI 700 installed and now need to calibrate the K-Factor. I downloaded the instructions on how to calibrate it over the course of 3 flights. 

My question is, that in the manual it does not state what mixture setting you should use for the 3 "test" flights. Because you need to add fuel after your test flight, (value added is part of the formula) running full rich, 65%, 55% etc, in my opinion will fluctuate the readings.  My thoughts, run all 3 test flights at full rich>?  

Any comments? 

Thanks!

[jetdriven]

Normal cruise  mixture.  Burn off 30 gallons then plug the actual fuel used, the JPI fuel used into the formula 

Edited May 28, 2021 by jetdriven

[larryb]

The mixture or power setting does not matter. You are comparing total gallons used between the JPI and the fuel pump. The more gallons used the better. I kept records for a few fill ups but found the default K value was within 1%. The hard part is fueling to the exact same level.

[jaylw314]

The k-factor is for your fuel flow meter, which does not care what happens after it leaves the meter.  Full power, low power, full rich, full lean it won't matter.

[Stryk77]

Thanks everyone!! Great feedback. 

[PT20J]

Question: If the Flowscan transducer (supplied with JPI products -- JPI now owns Flowscan) is accurate, why do we need to adjust the K-factor in the engine monitor?

The transducer works by the fuel spinning a multi-spoked wheel. As the spokes turn, they interrupt the light path between an LED and a open collector photo transistor. The pullup resistor for the transistor is in the engine monitor. So the wheel spins and pulses are output and counted. K-factor is the number of pulses (in thousands --hence the 'K' for kilo) output per gallon of fuel. Each transducer comes from the factory with a calibration tag listing the K-factor for that specific transducer. The calibration is done at 16.0 gph. So, two points: 1) Each transducer is calibrated and supplied with a K-factor. 2) except for the spinning wheel, the system is ALL DIGITAL. 

I have spent a lot of time calibrating my transducer. For every tank of fuel I used for the past two years, I have noted the pump reading when refilling and compared it to the fuel used on my JPM 700 and paralleled Shadin Miniflo-L. The two monitors agree when set to the same K-factor. What I ended up with for a transducer with a K-factor of 28.898 is 30.50. This is reasonably accurate for my average trips, but it does vary 1-2% depending on whether I burn a tank on one trip (mostly constant flow rate) or a bunch of short trips (more variability in flow rates due to changing power settings -- more takeoffs and landings).

I have gone to great pains to make sure the installation follows Flowscan's recommendations (the Mooney M20J factory installation has the air vent on the bottom rather than the top). My conclusion is that the 201B is accurate at 16.0 gph but reads somewhat high at lower fuel flows.

The EI red cube claims to be more accurate. However, there are lots of complaints around the web of high failure rate. The Flowscans seem to run for many years before the bearings wear out and the wheel sticks.

Skip

 

[carusoam]

Let’s see...

1) The hardware comes precalibrated pretty well...

2) Use the starting K factor and see how well it does...

3) care for repeatability comes in while filling the plane... pay attention to the details...

4) three decimal places of accuracy... requires refilling the plane the same way each time very important...

To the level of marking the floor to have the wheels in the same place, in the event the ground is not level...

5) How you operate the plane, or even if the engine burned the fuel is not going to matter...

you can remove a fuel line after the sensor and run the test multiple times...

6) Flow rate will have small affect on accuracy...   So do your best to be most accurate for cruise fuel flow rates...

7) Realistically... fly your normal flights, collect your normal data...   going to extremes, probably won’t improve your data any...?

PP thoughts only, not a mechanic...

Best regards,

-a-

[Warren]

In general mechanical flow meters are sensitive to turbulence in the flow.  In the installation manuals there are specified straight sections leading into the flow meter.  The purpose of this is to stabilize the flow for consistent readings throughout the flow range.  Of course, this is almost impossible in most aircraft environments and many have a 45 or flexible hose with a bend leading to the flow meter.  Turbulence changes with flow rate adding a variable that the mechanical flow meter cannot accommodate by adjusting a fixed flow constant.

The best you can do is to arrange the input with minimal bends near the flow meter.  And, you probably cruise at close to the same settings and fuel flow for the majority of your fuel burn.  Adjust for accuracy here and it should get you close enough.

[MikeOH]

When I first bought my F I spent a bunch of time trying to calibrate the flow meter to the nth degree.  What I empirically determined is that the variability due to how level the ramp is, and temperature, pretty much has made me accept about a half to one gallon error for a 45 gallon fill.  So, maybe a 1-3% error.  Last fill was within 0.1 gallon, but a couple of fills back was 1 gallon off.

IMHO if I'm down to worrying about 1 gallon (6 minutes) accuracy I've really screwed up

[Skates97]

22 minutes ago, MikeOH said:

When I first bought my F I spent a bunch of time trying to calibrate the flow meter to the nth degree.  What I empirically determined is that the variability due to how level the ramp is, and temperature, pretty much has made me accept about a half to one gallon error for a 45 gallon fill.  So, maybe a 1-3% error.  Last fill was within 0.1 gallon, but a couple of fills back was 1 gallon off.

IMHO if I'm down to worrying about 1 gallon (6 minutes) accuracy I've really screwed up

I went through the same when I had the 830 installed and then when I replaced it with the 900. I have times where it is dead on and when it is off by up to a gallon, which isn't bad when I'm adding 40 gallons. I have found a lot of places I'm am not parked perfectly level at the pumps and it doesn't take much to change whether I get the full 26 in a tank or 25.5.

I should add that I always stick the tanks before every flight, I don't rely on what the gage is showing even though I have the CIES senders which are very accurate. 

[PT20J]

You can rationalize it any way you want, but if they were accurate over all flows, you would not be fiddling with the K-factor.

Skip

[MikeOH]

2 hours ago, PT20J said:

You can rationalize it any way you want, but if they were accurate over all flows, you would not be fiddling with the K-factor.

Skip

Sorry, but I don't think listing actual variables like how consistently level the plane is, and temperature, to be mere rationalization.  Adjusting the K-factor WITHOUT holding those constant is bound to lead to frustration.  That is NOT to say the K-factor is constant over all flows.  That's just one more variable that is hard to account for.

Trying to chase accuracy down to 0.1 gallon is a waste of effort and, unless you are down to your last 6 minutes, of dubious practicality even if that repeatable accuracy is actually achieved.

[PT20J]

Over the last year, thanks to Covid limiting my travel, most of my refueling has been at the same pump, so that eliminates one variable.

I ended up with a K-factor of 30.40 with a transducer calibrated at 28.898 which is off by 5.2% or 2.6 gallons on an average 50 gal refill. However the variation with this K-factor refill to refill is still about 1-2%. I agree that's accurate enough, but it was a hassle and took a lot of gas to get it this close.  I'm curious if anyone has a red cube and what their experience is. 

Skip

[jaylw314]

1 hour ago, PT20J said:

Over the last year, thanks to Covid limiting my travel, most of my refueling has been at the same pump, so that eliminates one variable.

I ended up with a K-factor of 30.40 with a transducer calibrated at 28.898 which is off by 5.2% or 2.6 gallons on an average 50 gal refill. However the variation with this K-factor refill to refill is still about 1-2%. I agree that's accurate enough, but it was a hassle and took a lot of gas to get it this close.  I'm curious if anyone has a red cube and what their experience is. 

Skip

It should be noted that refueling your car in the morning so the gasoline is denser is an OWT, since all gas stations have underground storage tanks.  On the other hand, every AvGas tank I've seen at a fuel pump is above ground, so that could be one more variable in the mix. 

[PT20J]

I looked up the expansion factor of gasoline and found one source listed it a 0.00095 per degree C. Large storage tanks would have a high thermal mass so the temperature would likely not change much during a single day, but would be expected to change seasonally. A 20 deg change in a 50 gal fill would result in about a 1% change in volume or half a gallon. This might account for some of the refill to refill variation, but does not explain the initial 5% average measurement error.

Anyone else care  to share their final K-factor setting after calibration?

Skip

[Tom 4536]

57 minutes ago, PT20J said:

I looked up the expansion factor of gasoline and found one source listed it a 0.00095 per degree C. Large storage tanks would have a high thermal mass so the temperature would likely not change much during a single day, but would be expected to change seasonally. A 20 deg change in a 50 gal fill would result in about a 1% change in volume or half a gallon. This might account for some of the refill to refill variation, but does not explain the initial 5% average measurement error.

Anyone else care  to share their final K-factor setting after calibration?

Skip

My new flowmeter and JPI 900 were installed last Fall. My k-factor is now 27.5 and my error is 2.3%. During Covid I have been using the same fuel pump and usually refuel at about 6am. I doubt that temperature is playing much of a factor in my calibration.

Tom

[Ragsf15e]

3 hours ago, PT20J said:

I looked up the expansion factor of gasoline and found one source listed it a 0.00095 per degree C. Large storage tanks would have a high thermal mass so the temperature would likely not change much during a single day, but would be expected to change seasonally. A 20 deg change in a 50 gal fill would result in about a 1% change in volume or half a gallon. This might account for some of the refill to refill variation, but does not explain the initial 5% average measurement error.

Anyone else care  to share their final K-factor setting after calibration?

Skip

I just computed mine and adjusted it in my jpi930.  Started with 29.88.  New setting is 30.45.  Very close to what you came up with.

[MikeOH]

3 hours ago, Tom 4536 said:

My new flowmeter and JPI 900 were installed last Fall. My k-factor is now 27.5 and my error is 2.3%. During Covid I have been using the same fuel pump and usually refuel at about 6am. I doubt that temperature is playing much of a factor in my calibration.

Tom

Just realize there is the delta T between that of the remaining fuel in YOUR tanks versus the temperature of the fuel in the storage tank; the temperatures are unlikely to be the same.

[Tom 4536]

1 hour ago, MikeOH said:

Just realize there is the delta T between that of the remaining fuel in YOUR tanks versus the temperature of the fuel in the storage tank; the temperatures are unlikely to be the same.

The airport fuel tanks are above ground. My plane is at the same airport above ground. The plane fuel tanks and the airport fuel tanks sit all night without any sun before I fuel the plane. At 6am how much difference can there be in the temperature in the airport tanks and mine. It is not like we have any control over the temperatures and in fact who cares if our readings are off 2.3%. I don't.

[carusoam]

57 minutes ago, Tom 4536 said:

who cares if our readings are off 2.3%. I don't.

The guy that has burned 97 gallons of fuel, and is only a few miles away from landing, in his long body Mooney...

Accuracy is everything...

Sort of depends on wether it is over estimating or under estimating...

Best regards,

-a-

[MikeOH]

1 hour ago, Tom 4536 said:

The airport fuel tanks are above ground. My plane is at the same airport above ground. The plane fuel tanks and the airport fuel tanks sit all night without any sun before I fuel the plane. At 6am how much difference can there be in the temperature in the airport tanks and mine. It is not like we have any control over the temperatures and in fact who cares if our readings are off 2.3%. I don't.

Exactly!  YOU have NO control over the temperatures.  Unless you refueled the last time at the same temperature that you are refueling now, there is an error.  Also, do you ALWAYS refuel at your home airport?

Finally, we are in violent agreement over your last sentence; with errors of around 1-3% it's pretty moot.  As you say, "who cares?"

That's the real point of all my comments: The idea that you can achieve sub 1% accuracy in your fuel totalizer is a pipe dream.

[MikeOH]

17 minutes ago, carusoam said:

The guy that has burned 97 gallons of fuel, and is only a few miles away from landing, in his long body Mooney...

Accuracy is everything...

Sort of depends on wether it is over estimating or under estimating...

Best regards,

-a-

At which point THAT guy has exhibited very poor judgement unrelated to the accuracy of his fuel gauges.

KNOWING the real-world accuracy of your instruments is EVERYTHING.  Expecting accuracy beyond that actual capability is dangerous.

I use a conservative fuel burn and my watch and treat the totalizer as a cross check.

[Tom 4536]

18 hours ago, carusoam said:

The guy that has burned 97 gallons of fuel, and is only a few miles away from landing, in his long body Mooney...

Accuracy is everything...

Sort of depends on wether it is over estimating or under estimating...

Best regards,

-a-

I set my k factor so I have a little more remaining in the tank than the totalizer says.

Tom

[amillet]

8 hours ago, Tom 4536 said:

I set my k factor so I have a little more remaining in the tank than the totalizer says.

Tom

That’s kinda like my wife setting her car clock 5 minutes fast.  What good does it do if you know it’s fast?

[MikeOH]

2 hours ago, amillet said:

That’s kinda like my wife setting her car clock 5 minutes fast.  What good does it do if you know it’s fast?

Well, maybe it's just me, but being a bit late to a meeting because I 'thought' my watch was a bit fast is quite different than looking at empty fuel gauges and saying, "well, I know the tanks still have a couple gallons left!"