What's new
Van's Air Force

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

Speed Numbers for my Airplane

S

Scott DellAngelo

Well Known Member
For those who are interested, without intersection fairings, no paint, rear tie down ring in, etc.... as best I can come up with are virtually identical numbers to what Van lists (perhaps 2-3 mph more). I am spinning a 3-bladed Catto 66X72 prop with O-320-E3D with the higher compression pistons and an e-mag and a p-mag. I turn a bunch of RPM's down low flat out to get the top speed numbers, but it does do it.

After some good airspeed calibration runs last night my dynon true airspeed reads about 6 knots too low at around 150 knots. I am one of the few that received a heated dynon pitot and I have the safeair1 install kit with flush static ports. After some reading of people having problems with flush static ports I might mess around with cutting the head off the pop rivet Van's uses for the static port and glueing it over top of the flush one to see if that helps at all.

At any rate, for those who spend time worrying about numbers mine is another data point saying that Van lists honest numbers.

Scott
#90598 - N598SD Flying - 31.5 hrs
 
Scott DellAngelo said:
For those who are interested, without intersection fairings, no paint, rear tie down ring in, etc.... as best I can come up with are virtually identical numbers to what Van lists (perhaps 2-3 mph more). I am spinning a 3-bladed Catto 66X72 prop with O-320-E3D with the higher compression pistons and an e-mag and a p-mag. I turn a bunch of RPM's down low flat out to get the top speed numbers, but it does do it.

After some good airspeed calibration runs last night my dynon true airspeed reads about 6 knots too low at around 150 knots. I am one of the few that received a heated dynon pitot and I have the safeair1 install kit with flush static ports. After some reading of people having problems with flush static ports I might mess around with cutting the head off the pop rivet Van's uses for the static port and glueing it over top of the flush one to see if that helps at all.

At any rate, for those who spend time worrying about numbers mine is another data point saying that Van lists honest numbers.

Scott
#90598 - N598SD Flying - 31.5 hrs
Click to expand...

Scott, my FP 9A flys right on Van's cruise speed specs, maybe a tad faster. With the Dynon AoA probe and Van's static ports my TAS reads 3%-4% low. I calibrated the Dynon and my backup ASI on the ground, and they both agree to +/- 1 Knot, and are accurate to within experimental error. In flight, they don't always agree due to different response rates and friction in the conventional ASI.

I did shim my Pitot mounting to lower the tip of the pitot a bit. In theory, this should not matter, but in practice it helped somewhat with slow flight indications.

So, the 6 knot error you are seeing with your setup is about what I ended up with as well. I would not be concerned with your static ports too much, as long as you know the airspeed error in the cruise range and have verified Vs1 and Vs0 speeds and have done your glide tests and climb tests to find Vx and Vy.

If you play with your static ports, you'll affect your altimeter readings, which can lead to problems with your mode C, so be careful. You can check this by flying side by side with a certified IFR aircraft that's had a recent pitot-static calibration. +/-100 feet is probably ok (certified accuracy is +/-50 I think). Try it at different airspeeds, although the certified bird may not be able to keep up :)

I'm a perfectionist, and I was miffed for a while due to the 'large' error in my airspeed indication. I was determined to find the fix, but never did. After a while you just accept it and enjoy flying.

Just for a hoot, take your GPS in the car. You'll find that most car speedometers over read by about 3-4%. Other than making the car 'seem' to perform better and generating more revenue for maintenance, I think this stems from potential liability concerns.

Vern





:)
 
Jim, any time but 8.5 more hours before there can be a ride included. Hopefully if we get decent weather for the next several days I can knock it out by the end of this weekend. I am travelling for work next week and if weather is good and I am done with phase 1 I will be going to the U.P. of Michigan memorial weekend. I am very excited to not have to do the 6+ hour drive anymore rather a sub 2 hour flight. :)

Vern, I agree with everything you said. I just plan to temporarily try the pop rivet "glued" on just to see what (if anything) it changes. For something to do if nothing else. Down low my airspeed seems pretty good. The dynon airspeed reads lower than the Van's round gauge below 90. (I tend to believe the dynon is close down there based on stalls and such).

Scott
#90598 - N598SD Flying - 31.5 hours
 
Sounds great. I've been talking to someone, (Dave Hu) that just bought a kit and lives in Plainfield as well. We're trying to get our schedules to work so he can see some work in progress and get some tool and work bench ideas.

Our travel schedules keep getting in the way! I hope to connect with him for a shop visit next week.

Jim White

90789 - N789JT (RES)
 
David actually came over to see my shop, now returned to a garage on Saturday. Good to hear he is going to get together with someone in process so some of his reservations can hopefully be resolved.

Scott
#90598 - N598SD Flying - 40 hours (phase 1 done)
 
Static Error

Scott,

My 9a is also equiped with the flush static ports and the Dynon pitot (not heated--yet). My TAS from the Dynon read right at 7-8% low based on many calibration runs against my Garmin gps. I have since added a 40/1000th aluminum washer (flat and 3/4" in diameter) that was made on a friends lath, bonded to the exterior and it took out about 1/3 of the error--now running about 5% low. I intend to add at least another 45/1000th and check again. I was told by an aero engineer and kit plane designer that the air next to the skin is apparently "dead" and that the static port needs to be raised into the boundary layer airflow. The way it was resulted in minor pressurization of the port and therefore a lower relative difference with the pitot and low readings. The trick is finding the minimum washer thickness to enter the boundary layer.

Cheers,

db
 
Got the second washer installed and flew the plane. I am in the area of diminishing returns. The first washer brought the error down to 5% from 7.5%, the second washer only reduced it to 4.2%. I am now wondering if the flat surface of my static port vs Vans domed static port could have an effect----any commnents??

Cheers,

db
 
Sharp edges?

db

Try your washers again, but round off the outer edges.
You could be creating your own turbulent layer with the washer sharp edges....

gil in Tucson
 
db1yg said:
Got the second washer installed and flew the plane. I am in the area of diminishing returns. The first washer brought the error down to 5% from 7.5%, the second washer only reduced it to 4.2%. I am now wondering if the flat surface of my static port vs Vans domed static port could have an effect----any commnents??
Click to expand...
  1. Have you done recent pitot and static leak checks?
  2. Is the OAT indication accurate?
  3. Have you checked the EFIS for instrument error? Mine had small errors (http://www.kilohotel.com/rv8/article.php?story=ASI_Calibration_2).
  4. What method are you using to determine TAS from GPS data? Some people use methods that are not accurate.
  5. Don't draw any conclusions from tests on one flight. There are a lot of reasons why one set of tests could be bad. Do the tests on another flight to see if the results are repeatable.
  6. Maybe the dome shape is important, as you commented. The only way to know for sure is to do testing with a different shaped washer.
 
Pitot/static system has been checked--no leaks

OAT on the Dynon D10a checked on ground for accuracy--checks good and the probe is located under the hs on the fuse side so it should be getting good data (no engine heat or sun).

The EFIS has not been calibrated although it reads within 1-1.5 kts of my analog a/s guage

TAS tests were done in smooth air using a/p on N-E-S-W headings long enough to stabilze then ave the results -- the results at the d. Alt and power settings are within 1 kt of prior speed tests in the same conditions

I will try rounding the edges of the static port and seeing if this makes a difference and post the results.

Cheers,

db
 
db1yg said:
OAT on the Dynon D10a checked on ground for accuracy--checks good and the probe is located under the hs on the fuse side so it should be getting good data (no engine heat or sun).
Click to expand...
Checked on the ground with the engine running/prop spinning? Anywhere on the fuselage there will be some error. I don't understand why everybody doesn't put their OAT sensor out under the wing.
 
As I posted way back in this thread mine is off (dynon TAS on the low side) by 6-7 knots at typical cruise speeds. I have the fancy safeair1 flush static ports. OAT is under the wing just behind the pitot mount and seems to always be within a degree of correct. Anyways since I have read many people say that the flush ports don't work too well, last night I cut off the heads of the Van's pop rivet static ports and will glue them on tonight to see if/what any differences are. If it's better i'll probably keep them on, if not i'll pull them back off. I'll post the results either way. I figure I will use the pop rivet heads instead of a washer only because they are small, slightly rounded, and I have them so why not.

Scott
#90598 - N598SD Flying - 50 hours
 
db1yg said:
The EFIS has not been calibrated although it reads within 1-1.5 kts of my analog a/s guage

TAS tests were done in smooth air using a/p on N-E-S-W headings long enough to stabilze then ave the results -- the results at the d. Alt and power settings are within 1 kt of prior speed tests in the same conditions
Click to expand...
My EFIS has an ASI error of about 3% at low speed, and over 1% at high speed. You are assuming that your EFIS has zero error. I'm betting the EFIS has some error. You should have the instrument error checked.

The average of the ground speeds on four cardinal headings is not equal to the TAS unless the wind is zero. The amount of error depends on the ratio of wind speed to TAS. If the wind is low, the error is low. The error is about 1% if the wind speed is 20% of the TAS. The error is about 4% if the wind speed is 40% of the TAS. The wind speeds at altitude can be high enough to produce a significant error.

You are chasing an error of about 4%. You want to know where it is coming from, so you can fix it. You need to know how much of that error is due to instrument error in the EFIS. You also need to use an accurate test technique. Otherwise it is like measuring with a micromenter and cutting with an axe.

I recommend you record GPS ground speed and track on four headings, with the headings separated by roughly 90 degrees. Put the data in the spreadsheet that NTPS has provided. More info at this thread, and this one.
 
I checked my TAS data against the NTPS equation by installing the data from three of the four heading runs I collected and posted prior---the result was within 1/2 kt of my original data. The NTPS spread sheet uses 3 tracks.

I have not had the calibration of the Dynon checked but will in the future.

I do plan to round the edges of the trial washer on the static port and see if it makes a difference.

Cheers,

db
 
db1yg said:
I checked my TAS data against the NTPS equation by installing the data from three of the four heading runs I collected and posted prior---the result was within 1/2 kt of my original data. The NTPS spread sheet uses 3 tracks.
Click to expand...
The NTPS spreadsheet has several sections. One part uses three tracks, and another part offers an alternative that uses four tracks. These parts are on different sheets inside the main spreadsheet. I prefer the four track method, as it uses the redundant data to evaluate the quality of the data set.

What are the tracks and ground speeds for the four runs?
 
Kevin,

Ground speeds (gps) and tracks.

N 360 149 kts
E 090 153 kts
S 180 162 kts
W 270 158 kts

AVE 155.5 kts

I used the first three (N-E-S) in the three track NT spreadsheet from this data as well as my baseline (one washer on static port) gps speed data. The average gps tracks came out to 155.5 kts. while the three track NT came out to 156 kts. The base line comparisons were also close--within .25 kts.---assuming I did this correctly.

FYI-The aircraft is a 9a with main gear fairings but no nose gear fairings as yet. The power setting was 68% (your spread sheet) and P. Alt was 6450'.

BTW, Dan--My Dynon remote mag is located aft of the baggage bulkhead. The Dynon OAT must be connected to the remote mag which processes the data and sends it to the EFIS. The early style OAT is cable length sensitive--you can shorten it but not lengthen it. The cable will not reach the wing.



Cheers,

db
 
db1yg said:
Ground speeds (gps) and tracks.

N 360 149 kts
E 090 153 kts
S 180 162 kts
W 270 158 kts

AVE 155.5 kts
Click to expand...
That is a very good looking set of data. Using all four runs, the NTPS spreadsheet comes up with 155.6 kt, and a standard deviation of 0.1 kt. I can't come anywhere close to producing data of that quality, but I've never done this testing on an aircraft with an autopilot.

You originally said the runs were on headings of N-E-S-W headings, but the data above shows N-E-S-W tracks. What gives here?

I wonder whether the Dynon EFIS corrects the indicated OAT for ram rise. The ram rise could be worth up to about 3 deg C, which would make about a 1% difference in the calculated CAS given a TAS. The sense of this error (if present) would be to make the indicated OAT be too high, which would make the TAS read on the high side for a given CAS, or the calculated CAS be on the low side for a given TAS. I think this hypothetical error is in the right direction to possibly explain part of the speed error you are seeing (but I'm tired, and I'm drinking a beer, so I may have puzzled this out in the wrong sense).

It might be interesting to do a test. Accelerate to max speed in level flight, at some altitude. Note the OAT after it stabilzes. Now, keeping altitude constant, do a slow deceleration to the stall. If the OAT indication remains constant, then Dynon has done a good job at correcting for the ram rise. If the OAT indication decreases as you slow down, then they are not properly accounting for ram rise. In this case, stabilize at low speed (same altitude as before), and note the OAT after it stabilizes. Report the altitude, and OAT vs speed for the high and low speed points, and we can figure out how to correct the indicated OAT values.
 
Last edited:
Thanks for the info Kevin!

The test was in very smooth air and I was using my TruTrak Pic Pilot and AlTrak alt hold.

BTW, my mistake--these were headings as dialed in to my Pic Pilot digital DG--not gps ground tracks. The dg has been calibrated.

Drink a beer for me.

Cheers,

db
 
Kevin,

If no gps signal is present the Pic Pilot goes into "heading hold" mode, not track mode!! First paragraph of Basic Operation.

I'm done.

db
 
db1yg said:
Kevin,

If no gps signal is present the Pic Pilot goes into "heading hold" mode, not track mode!! First paragraph of Basic Operation.
Click to expand...
OK. For future airspeed cal work, I recommend you use the Pic Pilot to fly a desired track, then record the tracks and ground speeds from the GPS. Do a four-sided box pattern, and use the NTPS spreadsheet.

Let us know the results if you do the OAT test I mentioned above. I'm also interested to learn the results of any more testing with different variations on your static ports.
 
Ok, I pro-sealed the heads from the Van's static ports (pop rivets) over my flush safeair1 static ports to move away from the skin. I did two airspeed runs and seem to have moved my 6-7 knot low TAS reading on the Dynon down to more like 3 knots low. I'm happy, they will stay.

I did my runs 4 way with GPS tracks using the pictoral pilot and altitude hold. Inserted into a couple different formulas on the internet and the data agreed pretty well.

Scott - #90598 - N598SD Flying - 52 hours
 
Scott,

Thanks for the update. My Dynon is still 4-5% below TAS with about .082 washers bonded to the flush port. My next move will be to round the edges of the "square edged" washers to see if I get some correction--like you did.
If I can get to within 2% I will be happy.

Cheers,

db
 
Static problems can be hard to troubleshoot. Keep in mind that any extrusion/imperfection forward of the static port can cause turbulent air to hit the static port. Look for rivets, etc. that might be causing you a problem.

The F4-D used a "Burble Amplifier" on the front of its radome to create a boundary layer for its static port. This was no more that a round aluminum "cap" slightly larger in diameter than the radome, so as to overhang the radome by 1/8" or or so all the way around. You could tell when there was a problem because altitude hold would porpoise. Here's a picture of the burble amplifier (shiny thing at the front of the radome):

http://aeroweb.brooklyn.cuny.edu/database/museums/getimage.htm?id=4126
 
I now have the a/s error on my Dynon EFIS down to 1.5-2.0%. I am going to live with this. I adjusted it down to this level by moving the port out into active air with one .040 machined washer (rounded edges) as described in prior postings and cutting most (I left about 1/64th protruding) of the shaft off Vans dome rivet and bonding it on the washer. Actually I enlarged the port on the washer by the diameter of the rivet shart and used this to insure correct alignment between the dome rivet port and the washer port.

Now I simply add 3 knots to the Dynon TAS if above 160 kts and 2 if below.

Cheers,

db
 
db1yg said:
I now have the a/s error on my Dynon EFIS down to 1.5-2.0%.
Click to expand...
It is good to read that you have been able to greatly improve the accuracy of your airspeed indications.

I wonder how much of the remaining error is due to the static source, and how much is instrument error in the EFIS. I checked the instrument error in my D-10A EFIS against a water manometer, and found that it read about one knot too low at low speed, had negligble error from 130 to 150 kt, then started to read too high, with the error increasing linearly to 3 kt too high at 240 kt.
 
db1yg said:
Now I simply add 3 knots to the Dynon TAS if above 160 kts and 2 if below.

Cheers,

db
Click to expand...

Almost exactly my difference too, except best I can tell i'm more like add 3 above 145 and 2 below.

Scott
#90598 - N598SD Flying - 72 hours
 
static ports on my cherokee

My 64 cherokee 235 has flush static ports but piper puts a 1/8 inch thick plate (for the lack of a better word) just in front, and I mean just!! the plate is butted up against the forword edge port hole. The plate is about 3/4x1/4 +/- Maybe they had the same issues with that airframe.

Just my 2 cents
 
Bill Ervin said:
My 64 cherokee 235 has flush static ports but piper puts a 1/8 inch thick plate (for the lack of a better word) just in front, and I mean just!! the plate is butted up against the forword edge port hole. The plate is about 3/4x1/4 +/- Maybe they had the same issues with that airframe.
Click to expand...
That is one of the standard remedies for static system errors, if the manufacturer wants to avoid looking for a better location for the static port. A dam ahead of the static port reduces the pressure seen at the port, which can help if the indicated airspeed was too low. A dam just behind the static port increases the pressure seen at the port, which can help if the IAS was too high.
 
Scott DellAngelo said:
I will be going to the U.P. of Michigan memorial weekend. I am very excited to not have to do the 6+ hour drive anymore rather a sub 2 hour flight. :)

Scott
#90598 - N598SD Flying - 31.5 hours
Click to expand...

Scott,

I was curious where you go in the U.P. We have a place near Curtis and I have been trying to convince my wife to leave a car up at Newberry so we could fly back and forth. Cuts a 4.5 hour drive down to about 3.0 hours even in a Super Cruiser. I am in the process of buying an RV8 which should bring the trip down to 1.5 hours max!

Greg
 
You must log in or register to reply here.
Back
Top