Oil Pressure Sender

[DCASonex]

I and some others, have examined the effects of the Jabiru PRV (Pressure Relief valve) and ways to mitigate it to some detail. Simply put the stock Jabiru PRV turns into a hydraulic jackhammer, much of the time when it is operating, slamming its piston from one end of its bore to the other. This produces high pressure spikes, and fatigues the metal in things like oil coolers and wears out the wipers in the VDO sensors. The Jabiru's 4 lobe oil pump also produces pressure pulses, but not a bad as the PRV. Due to that, would want a very high burst pressure rating on anything in the oil circuit of these engines.

David A.

[lpaaruule]

This is a drag, but better to find out on the ground, than in the air.

I did some research, and ended up buying the Sensata - Kavlico - P255-150G-E4A from Mouser.com (it's is the one Dynon sells, but Dynon is out of stock at the moment). I'll keep the Swift as a short term backup - besides, I doubt Aircraft Spruce would like a used one with a broken tab.

Ugh...

[Rick524]

https://dynonavionics.com/bulletins/sup ... 50620E.php

[WaiexN143NM]

hi paul,
lots of research from you and others. how much is the kavlico you bought? Dynon store $129. also paul is the part # correct? i went to mouser, used their search, and looked thru the senders but cant find that part #.
and you are currently flying with the swift? so i take it with the swift that the burst pressure is 3x the pressure rating.
keep us informed on the kavlico and install
david how can we mitigate the effects of the PRV?
thanks,


michael.

[builderflyer]

Jabiru has done their own testing to determine the cause (i.e. cavitation) of the high oil pressure spikes and found that the spikes were due principally to not allowing the oil to warm sufficiently before applying full power. Another contributing factor was not using multi-grade oil. For those of us who are operating the older engines, Jabiru also recommended modifying the oil pump port plate to reduce the potential for cavitation. This modification was done to my engine in yr2004 before the first flight occured in yr2005. Newer engines came from the factory with this modification already accomplished. See the latest version of this Service Bulletin JSB004-4 dated 12/19/2017.

Art,,,,,,,,,,,,,,,,Sonex taildragger #95,,,,,,,,,,,,,,,Jabiru 3300 #261

[DCASonex]

The modified back plate appears to be a bit of a fudge. The original back plate looks to have been properly designed IF the gear-rotor pump was properly positioned so that the inlet port closed and outlet port opened at point of maximum chamber capacity. However, that is easier said than done, as the chamber is moving, and is determined the point where the lobe of the inner gear reaches maximum engagement with the outer gear, and that places it close to center so any error there is magnified at the point of maximum capacity. The bolts of the gear-rotor pump are no where near an adequate means of setting this location, and Jabiru's instruction to turn the engine through two full turns insures that nothing binds, but does absolutely nothing to insure proper positioning of point of maximum capacity relative to pump ports. Switching ports too soon or to late causes problems, including broken pumps and pump keys. What the modified back plate does is increase the overlap where both inlet and outlet ports are connected to the chamber which lets high pressure oil from the outlet side flow back into the inlet producing even more pulsation. Fortunately, The momentum of the moving oil minimizes this effect.

I built a test rig that mounted A Jabiru pump from my retired engine along with filter and PRV to a milled block of aluminum that duplicated essential elements of the Jabiru crank case, and drove it with a variable speed motor and monitors output of solid state sensors with a recording scope. Various photos, short videos, and results of that are on the Jabiru/ CAMit groups forum >>> https://jabcamit.groups.io/g/main <<< I machined a back-plate that looks to have duplicated the Jabiru original and tested with that after devising a way to properly position it, and the change in pulsations was not that significant so keeping the modified back-plate seems sensible, and the pulsations from the pump are far less damaging than those from the PRV.

Suspect the requirement to pre-warm oil to a higher temperature before applying maximum power is simply to reduce the available pressure and need for that PRV to open. The problem with the PRV itself is that it has a small exposed piston area that oil pressure works against, but when opened, oil then contacts a much larger surface slamming the piston to the rear of its chamber, which quickly lowers pressure letting the piston return to closed Repeat - Repeat. When this happens in the test rig the result is dramatic, and i have a well oiled test area to confirm it, as the resulting severe pulsations dislodge any hoses not securely anchored.

The upshot of my and others experimenting points to replacing the Jabiru PRV with one where the piston slides in a sleeve with ports cut into the side of the sleeve, works best, but not to give the store away, for those who are continuing to develop this, will cation that there are some tricks to keeping the pumps own pulses from upsetting this type of PRV. Due to wife's health, and that a solution seems to have been found , i discontinued work on this, but such PRVs are being tested by others. If any further interest, suggest joining the above mentioned Jabiru/ CAMit engines forum. Lot of knowledgeable folks on that.

David A.

[builderflyer]

The upshot of my and others experimenting points to replacing the Jabiru PRV with one where the piston slides in a sleeve with ports cut into the side of the sleeve, works best, but not to give the store away, for those who are continuing to develop this, will cation that there are some tricks to keeping the pumps own pulses from upsetting this type of PRV. Due to wife's health, and that a solution seems to have been found , i discontinued work on this, but such PRVs are being tested by others. If any further interest, suggest joining the above mentioned Jabiru/ CAMit engines forum. Lot of knowledgeable folks on that.

David A.

David A, I haven't followed any of the discussion on this topic on the Jabiru/Camit engines group ever since my email account was involuntarily changed and I stopped receiving message from that group. So please excuse my ignorance. With your and the others extensive amount of work on this topic, my question for you is: isn't it possible for the Jabiru factory determination that the root cause of the high pressure spikes is cavitation in the oil pump environment is correct.......and then it's the cavitation's effect on the PRV to worsen the spikes or add to the spikes significantly as shown in your testing? So even if the PRV's design can be improved, in your opinion would that be necessary, not just desirable, if operating the engine in conformance with Jabiru's recommendations (warm the oil sufficiently before applying full power and use a multi-grade oil) contained in their Service Bulletin were followed and those recommendations eliminated the cavitation?

We really appreciate all the work you have done..........it is a clear demonstration as to how talented some of the people are in the world of experimental aircraft.

Art,,,,,,,,,,,,,,,,,,,,Sonex taildragger #95,,,,,,,,,,,,,,,,,,Jabiru 3300 #261

[DCASonex]

Art,

If you have a Jabiru engine, it would be worthwhile reconnecting to the Jabiru / CAMit forum there are some interesting discussions on other matters no ongoing there.

In answer to your question, When I made a back plate for the Jabiru pump in my test rig, that as closely as possible duplicated the old original Jabiru design and insured that the pump housing was properly aligned with it, I did not encounter any problems. However, I did not see any significant reduction in pulsations from the pump that was my reason for making that back plate, so concluded that the new design did no real harm and enlarges the window for positioning the pump housing enough to accommodate the errors so went back to testing with the later modified back plate. I also did not find that it made any difference in hammering of the PRV. The pump’s pulsations are at a much higher frequency than the hammering of the PRV, so should be much less destructive.

In my test setup with VFD motor driven Jabiru pump, with an oil filter horizontally mounted so as to duplicate conditions where air may be trapped in the engines filter, I used a variety of oils, standard aviation oil at room temperature and oil thinned with kerosene to approximate the viscosity of aviation oil at normal temperatures. The PRV would hammer in all cases. Did not spend much time with the stock PRV as I was primarily concerned with making and testing alternative PRVs.

Am quite sure the stock PRV in my CAMit 3300 hammers. When RPMs increase, pressure goes up to a point, then suddenly drops off a bit and the digital display becomes erratic, same as happened in my test rig when the PRV started hammering. In engines with looser tolerances the oil pressure when warm may not reach pressures high enough to open the PRV, same may be true of older 2200s with the smaller oil pump. Jabiru no longer offers the 4 lobe pump, and now has a new 6 lobe. Given the back plate’s port overlap, I suspect this will result in pumping less oil, and thus lowering pressure, and in one case where very low oil pressure was seen, and the new pump installed, pressure seemed to be even lower.

Hope to get around to installing one or more of my PRV this year. Not much flying last year, and it is now Winter, temp about 12°F here this morning, and due to get colder through the next week, hanger is not insulated, so hibernating.

David A.

[lpaaruule]

I got my Kavlico - P255-150G-E4A today. The orifice is actually a tunnel that is at least 1 inch deep. It looks like it would be really easy to tap the orifice with a #8, and install a screw with a small hole drilled down the middle. I could install the screw with both loctite, and #8 Nordlock washers.

Having said that, I don't know if it's a good idea to mess around with it. I'll probably install it tonight as/is.

[WaiexN143NM]

Hi all, paul,

Thanks. for the update. the part number now works for me on mouser site, $137.84.
let us know how install and performance works out.
thanks for sharing your research.

michael.