Rotax oil pressure fluctuations

[kluge]

Hello! We are desperately looking for some good idea ...

Situation: Rotax 912 ULS from 2007, 600h, with the old VDO oil pressure sensor. Dynon D180 as EFIS/EMS instrument
-Shortly after takeoff oil pressure fluctuations from normal 3.x to 1.x. Not reproducible on the ground or at the next test flight.
-Shortly after takeoff again the same oil pressure fluctuation.

Changed the Sensor (as the most probable and simple cause), damaged the oil pump housing, therefore
-new-type Rotax sensor
-new oil pump housing
-Dynon factory firmware update, configuration for new type sensor.
-"Similar" oil pressure fluctuations when oil is hotter than 60 deg C/140 deg F.
-cleaning of the oil pressure valve ball: "little better"
-replacing the oil pressure valve ball with the mushroom type: "again a little better"
-Using a bluetooth oil pressure sensor parallel to the installment above:
Fluctuations in the Dynon-Rotax-sensor installment, no fluctuations with the Bluetooth instrument.

=> ???

You may see this in the film:

Setup for Rotax-Sensor, 18 Mb


Wiring

So
a) both the old VDO and the new Rotax-Sensor could be defective.
b) the pressure valve could be defective (but the outer part, the oil pump housing is new and the ball is replaced with a mushroom type. For this hypothesis, the bluetooth instrument has to be heavily damped/averaged. Values don't look like that.
c) the Dynon instrument is crazy (but as far as I understood this with the VDO sensor voltage is measured, with the new sensor current is measured. Different setup though). All other values are displayed without fluctuation.
d) something with ground/plus wiring is wrong. But the wiring has been checked/redone
e) the new sensor transmits some spikes/dips and the old sensor was defective?

If we (intentionally) wrongly configure the Dynon instrument for a Dynon-supplied Kavlico-Sensor, the values are way off (6.something) but stable and proportional to the bluetooth-instrument.

Does all this this makes any sense to you?

 

[jakej]

Not sure what is going on there however if you did actually use the Kavlico unit then the red wire (Kavlico) needs to go to pin 18 D37 (5volts only). Did you replace & solder in a new 200 oHm resistor, just in case it got damaged/broken ?

I'd try connecting up a mechanical gauge to the same point as the sensor & do some ground running - that should isolate when the problem is exactly, either the electrics/electronics or the engine systems.

 

[kluge]

The Kavlico is the Dynon-provider sensor, right? Don‘t know the make of the Rotax-provided. The wiring diagram I uploaded appears to reflect the one at page 3-6 of the Dynon  D180 installation manual.

As I understand it, it is a voltage divider. Measurement is done at pin 6. The old Rotax/VDO sensor had a terminal with mass connector and one for pin 6. This should be a voltage divider too, measured against an internal reference/resistor. Thus pin 6 is involved with both sensors. Any crude signal processing might explain (part of) the fluctuations before and after sensor change.

What sou see in the video is basically what you are asking for: the Dynon shows the value of the Rotax sensor and the Phone shows the value of the Bluetooth-Sensor (should be similar, but more precise than an analog instrument), both connected to the engine running at the ground.

 

[jakej]

The power wire into the Kavlico sensor needs to be the 5volt wire (paired with the Map sensor if fitted) white/red stripe & remove the resistor as it’s not used with the Kavlico - Dynon May chime in here

 

[Dynon]

Our hunch is that it's a real fluctuation, especially if other channels aren't exhibiting any issues. But first, definitely get it wired correctly... what Jake said: no 200 ohm resistor with the newer style (Kavlico) sensor. And pin 18, not 15, for power to the sensor. It needs 5V, not 12V. See http://www.dynonavionics.com/includes/guides/Addendum_Kavlico_Sensors-Rev_C.pdf for the correct diagram.

 

[kluge]

But first, definitely get it wired correctly... what Jake said: no 200 ohm resistor with the newer style (Kavlico) sensor. And pin 18, not 15, for power to the sensor. It needs 5V, not 12V.

That addendum from 2012 supplements the installation manual H from 2010. It refers to the Dynon-supplied Kavlico sensor, but not to the Rotax-sensor.

Or is the Dynon-supplied Kavlico-sensor identical to the Rotax-sensor and are you saying that the new installation scheme diagram with 5V refers to both the new Rotax and the new Dynon-sensor?

The actual Dynon firmware has a setup 4 for the new Rotax sensor and 5 for the Kavlico?

Sorry for being that slow.

 

[jakej]

The Rotax sensor is wired as per the manual. It’s not identical.
Kavlico is wired as per instructions & no resistor 
When you are changing anything, regardless of manufacturer, you should always get & use the latest available data.

 

[Dynon]

Ah. I misunderstood part of your post, and thought you had fitted the Kavlico as your replacement. If you've confirmed that the wiring is going back to the pins you have identified, and that you've added the appropriate resistor as indicated, your wiring should be correct based on your diagram. The same conclusion applies so far though - almost definitely not the EMS based on the observations so far. Is it stable when the engine is off?

 

[jakej]

If you are unable to fault the wiring then I'd suggest as per my reply post#1 use a mechanical gauge to see if the fluctuations still exist - if they do then you have a problem somewhere else & not with the electric/electronics IMO.
Any pressure gauge will do (provided it can handle the normal running pressure) doesn't matter if its in PSI or metric, you are just looking to see if the same fluctuations are there or not.

 

[kluge]

That's what you see in the linked video: simultaneous reading of the Rotax-Dynon installation and an independent (battery powered) Sensor-Bluetooth-phone display. The independent Sensor doesn't show the fluctuations.

 

[kluge]

The same conclusion applies so far though - almost definitely not the EMS based on the observations so far. Is it stable when the engine is off?

Every value (RPM, MP, Oil temp, CHT, IAS, Alt ...) is normal under all conditions except the oil pressure when oil is hotter than ca 60 deg C/140 deg F. Thus if the engine is off, value dispayed is 0,something bar oil pressure as in the books for the new rotax sensor.

We did not monitor 12V-output of the dynon E37 and ground and oil pressure simultaneously. This will be the next but one task.

Once, several years ago, the flap motor stuck. The circuit breaker did not go off, thus the system voltage went low. Under this condition, the displayed value for (IIRC) oil pressure and /or oil temp went to the low red. Everything else displayed continued to be normal. Thus these virtual instruments appered to be susceptible for voltage drop of parts of the system. Don't know if somebody misused the Dynon 12V output for something else...

EDIT: sorry, i probably broke the video links, here they are again:
https://cloudstorage.uni-oldenburg.de/s/OsS987hCQ5dq7fE
shows simultaneously
-the rotax-sensor displayed at the Dynon, Setup 4 correct for Rotax sensor
-an independent battery powered oil pressure sensor with Display via Bluetooth and mobile phone
You may see the fluctuations in question only at the Dynon.

https://cloudstorage.uni-oldenburg.de/s/kzop3IsLK3shzrJ
shows simultaneously
-the rotax-sensor displayed at the Dynon, Setup 5 (incorrect for Rotax sensor, correct for the Kavlico-sensor)
-an independent battery powered oil pressure sensor with Display via Bluetooth and mobile phone
You may see no fluctuations, but the values are way off (what is to be expected)

 

[kluge]

An update:

Kavlico sensor, Dynon-Display
https://cloudstorage.uni-oldenburg.de/s/9SnWCPDjxSSuCDr with fluctuations

Kavlico sensor, analog instrument
https://cloudstorage.uni-oldenburg.de/s/yK1wZiaqBsZh8op no fluctuations

This appears to me to point to the Dynon EMS.

What we checked:
Rotax sensor, at the engine, voltage Pin 6-Ground: no fluctuations
Rotax sensor, external air pressure, heated to 70°, Voltage Pin6-ground: no fluctuations
Rotax sensor, external air pressure, heated to 70°, Voltage Pin6-external 12V: no fluctuations
Kavlico sensor, external air pressure, heated to 70°, Voltage Pin6-ground: no fluctuations
Kavlico sensor, external air pressure, heated to 70°, Voltage Pin6-external 12V: no fluctuations
VDO sensor, at the engine, analog instrument: no fluctuations
VDO sensor, at the engine, Dynon: values off (it's a 0-5bar type instead of the required 0-10 bar), little fluctuations
Rotax-sensor, at the engine, setup for kavlico: values off, little fluctuations
Bluetooth sensor, at the engine, mobile phone display: no fluctuations

Fluctuations only appear when values are displayed at the Dynon. Fluctuations appear to be less when the displayed values are in the upper 20% of the range. Maybe someone from Dynon may relate this to the internal measurement setup/schematics of pin 6 input.

The Dynon just came back from firmware update and Horizon calibration.

 

[Bomar]

You need to verify that you really have the correct Kavlico sensor (they all look identical).  Use a flashlight and magnifying glass to read the part number that's stamped on the sensor. 

Then use the part number cross-reference chart on page 27-1 of the SkyView Installation Guide to figure out which Kavlico sensor you have.

http://dynonavionics.com/public_html/yabbfiles/Attachments/SkyView_System_Installation_Guide-Rev_AD_v15_3.pdf

I have a strong suspicion that you don't have the 150psi sensor.

 

[kluge]

What leads you to this assumption?
Both the new-type Rotax-supplied and the Kavlico sensor show fluctuations when connected to the Dynon while the values themselves are correct and the engine appears not to have any real pressure fluctuations.

I think we will switch to the VDO-sensor rotax supplied before 2008, but with the new metric thread.

 

[n456ts]

If you think it's a Dynon issue or sensor issue you need further testing. Remove the oil pressure sensor. Connect it to an air pressure regulator with an analog gauge. Begin applying shop air to the sensor by controlling the pressure with the regulator. The readings should match with the analog gauge and the Dynon reading.

"fluctuations" can often be due to the oil system sucking in air somewhere between the tank and the oil pump.

 

[kluge]

If you think it's a Dynon issue or sensor issue you need further testing.  Remove the oil pressure sensor.  Connect it to an air pressure regulator with an analog gauge.  Begin applying shop air to the sensor by controlling the pressure with the regulator.  The readings should match with the analog gauge and the Dynon reading. 

We did that: heated sensor externally with air pressure: no fluctuations of voltages or dynon instrument.

"fluctuations" can often be due to the oil system sucking in air somewhere between the tank and the oil pump.

This sounds too tempting to be true.
With an mechanical gauge, we see the fluctuations.
We saw them with old oil pump housing, old pressure valve spring, old ball, old rotor as well as with new housing, new spring, new mushroom but old rotor (looks well usable). So we are running out of hypotheses.

 

[jakej]

My reply #8 suggested the mechanical gauge check & finally you state there are fluctuations still using that method - although many parts have been replaced the ‘fault’ still exists.
I would not fly with this engine until the cause is found.

 

[n456ts]

Based on your reply, you have an oil system problem, not an instrument issue. As jakej stated, don't fly until this is resolved.

 

[kluge]

To sum this up: we systematically examined most of the factors that could affect oil pressure indication in the cockpit.

These factors are:

0. div
Mass/sensor cable connections
Vibrations, heat affecting the sensor

1. Oil pump
4- and five-star rotors of the oil pump
Oil pump housing
oil pump front cover

2. Oil pressure regulation
valve ball
valve mushroom
valve spring
valve bottom screw (low/high sleeve)

3. Pressure sensing
a. analog purely mechanical non damped
b. Old pre 12/07 VDO Rotax supplied sensor (still available from VDO)
c. New Kavlico Rotax-supplied sensor
d. Bluetooth-pressure sensor

4. Pressure indication
a. mechanical gauge for 3a
b. analog electrical instruments for 3b and c
c. Dynon D180 Display for 3 b and c
d. mobile phone for 3d

Results:
I. Cabling and heat, vibrations could be excluded by external voltage, external cabling, synchronous readings of built in and externalized heated sensors.
Pump rotors did not affect the readings
II. Every other item listed above influences the pressure read in the cockpit.
III. the Rotax engine has indeed marked oil pressure fluctuations, amplitude 0.3 to 0.6 bar. This is factory normal. Cause is most probably the hydrodynamics of the oil pressure valve, hence the influence of every part of the valve.
IV. This can only be seen with a sensor-instrument combination fast enough to display the fluctuations.
Fastest sensor is the Rotax/Kavlico, Rotax VDO is way slower, mechanical non damped gauge is fast, but gets in resonance quickly-useless.
Fastest display is the Dynon, analog electrical instrument is way slower, Bluetooth was the slowest.

Conclusion
Thus, the disturbing values displayed with the Dynon/Kavlico-new Rotax combination show the reality not seen before as the old sensor was to slow to show the fluctuations. Semi-official statement from Rotor confirmed these fluctuations to be seen on the bench regularly.

Solutions
Our several-thousand-dollar solution :'( : old rotor sensor type installed at the firewall for protection against vibration, Rock stable values with the Dynon-Display. I dislike the high pressure oil lines outside of the engine, but the alternative with new sensors would be looking at constantly changing oil pressure values while inflight.

What Dynon might contemplate: add some averaging/damping at the display setup for the new sensors.

Redesign of the pressure valve regulation. The low/high sleeve spring screw and the mushroom-like ball replacement show that Rotor tried to adress this in the past (to no avail).
There are some promising ideas third party ideas which appear to work, but their user base is limited

 

[jakej]

old rotor sensor type installed at the firewall for protection against vibration, Rock stable values with the Dynon-Display.

That is what all (should do IMO) lycoming/Continental owners do, engine vibrations are quite often stated by many as the cause of sensor failures.