Failing Oil Pressure Sensor or Something Bad?

[Ragsf15e]

So Ive noticed that my oil pressure fluctuates.  Generally from 71-73psi about once a second.  The JPI 930 is constantly moving.  Then there’s drops very occasionally to say 60psi for half a second and back up.  Is that even possible in the engine?  I have two pictures below.  My flight today on 2” data capture.  Biggest fluctuations are in a pretty high speed descent near the end, but you can see the constant fluctuations.  Then below is a flight from 2 years ago... solid line, but it is on 6” data there.

 I called JPI and got right to tech support.  They say sensors don’t fail like that, but it could be a loose connection.  And the big jumps were at high airspeed (vibrations?).  On the ground it appears steadier.

Thoughts on if this is actually happening in the engine?

thanks!

[RobertGary1]

What does the mechanical gauge show at that time. 

[Ragsf15e]

4 minutes ago, RobertGary1 said:

What does the mechanical gauge show at that time. 

Don’t have any other gauges.  Maybe those guys with the 830 and the original gauges are onto something?

[A64Pilot]

Your looking at what is sometimes called analysis paralysis, it often comes from hooking a digital device that will display tiny increments, whether accurately or not is arguable, but the faithful believe something will measure to .001 because it displays that, a issue is that the analog device can’t measure within 5 PSI even when it was new, decades ago, and it hasn’t gained in accuracy since then.

‘There is a reason why the stock gauges have only a couple of numbers on them, with big green arcs

[A64Pilot]

They way some pressure transducers work is by the caveman method of compressing carbon powder, the tighter it’s compressed the lower the resistance. That’s how old microphones work too, so it’s more sensitive than you might think.

‘I woud bet that what your seeing is very likely the varying voltage of your electrical system, and not pressure fluctuations, maybe.

Try pulling your alternator field CB and see if the numbers change, if they don’t, that in my opinion eliminates voltage fluctuating as an issue.

‘I’m not familiar with your aircraft, if you have a alternator switch use it of course.

Edited March 26, 2021 by A64Pilot

[Ragsf15e]

Thanks, ill try that.  Seems weird that it changed from a couple years ago to now...

[MikeOH]

3 hours ago, Ragsf15e said:

Thanks, ill try that.  Seems weird that it changed from a couple years ago to now...

Didn't you state that the data capture rate is faster than the rate 2 years ago?

 

[Ragsf15e]

2 hours ago, MikeOH said:

Didn't you state that the data capture rate is faster than the rate 2 years ago?

 

Yes it is 4” vs 2”.  That could make it a little fuzzy, but it’s like night and day. 
I’m guessing it’s a loose connection on the oil pressure sensor, but I was hoping someone here had see something like this before or could tell me that it’s basically impossible for the engine to cycle pressure like that (especially down to 60 for like half a second).

[carusoam]

Depending on where the OilP is being read... it could oscillate in a similar fashion...

Make sure the JPI sensor is reading from the proper location...

 

Compare to older data, see if anything has changed since the sensor’s first flights...

If you look closely at the data being graphed... it’s not continuous, two decimal points of accuracy...

It looks like noise that is averaging +/- 1 unit.... and the error drifts a little up and down...

 

There is a service called Savvy...   send the data over... and ask... ‘have you seen this on any other IO360s?’

OilP sensors are pretty complex micro computer devices.... that can be capable of all kinds of wacky behavior...

If the wackiness is smoothing out during different portions of flight... it’s hinting at a sensor and cable challenge more than an OilP challenge...   See what is running next to the OilP wires... is it anywhere near the ignition wires or a magneto?

Lets invite Paul @kortopates to the conversation... he may be familiar with this. (Noisy OilP data on a JPI)

PP thoughts only, not a mechanic...

Best regards,

-a-

[Will.iam]

35 minutes ago, Ragsf15e said:

Yes it is 4” vs 2”.  That could make it a little fuzzy, but it’s like night and day. 
I’m guessing it’s a loose connection on the oil pressure sensor, but I was hoping someone here had see something like this before or could tell me that it’s basically impossible for the engine to cycle pressure like that (especially down to 60 for like half a second).

Why not change your data capture rate back to the same sampling rate as what you where sampling at 2 years ago and compare apples to apples to see if there is a change or if your are just reading the finer detail from a higher sampling rate. 

[carusoam]

Set your data to the smallest number possible... because you can.

It probably won’t change the data being captured any... as the noise is real...

The noise is being captured and recorded...

Bumping the data capture rate closer to one point of data every second will help for the non-steady flight regimes... like the run-up...   

This event is all through the steady cruise portion of the flight... the real numbers would be really close as it drifted around over time...  

There is one point where the OilP appears to drop... if that were real data... you would want to be on the look out for corroborating evidence... what else is misbehaving? RPM, oilT?

Best regards,

-a-

[Skates97]

It's really easy to check the connections between the sensor and harness. Snip the zip ties and slide the shield back. When I upgraded my 830 to the 900 I found that some of them did not have the star washer between the terminals as the installation manual calls for. I would check that and make sure that they are tight. 

[MikeOH]

1 hour ago, Ragsf15e said:

Yes it is 4” vs 2”.  That could make it a little fuzzy, but it’s like night and day. 
I’m guessing it’s a loose connection on the oil pressure sensor, but I was hoping someone here had see something like this before or could tell me that it’s basically impossible for the engine to cycle pressure like that (especially down to 60 for like half a second).

I know mine bounces around; it bugs me but it's always been that way and it hasn't changed over three years.  I'd try what Will.i am suggests and see if it returns to the old 'normal.'

[MatthiasArnold]

I had a similar observation. Servicing the connectors solved the issue. 
Regards,

Matthias

[Ragsf15e]

11 hours ago, Will.iam said:

Why not change your data capture rate back to the same sampling rate as what you where sampling at 2 years ago and compare apples to apples to see if there is a change or if your are just reading the finer detail from a higher sampling rate. 

Thanks! Not sure why I didn’t think of that, but I’ll give it a try.

[Ragsf15e]

10 hours ago, Skates97 said:

It's really easy to check the connections between the sensor and harness. Snip the zip ties and slide the shield back. When I upgraded my 830 to the 900 I found that some of them did not have the star washer between the terminals as the installation manual calls for. I would check that and make sure that they are tight. 

I think this is probably gonna be what I need to do.  Were you looking at the wires where they connect into the harness at the back of the actual jpi unit in the panel?

[Ragsf15e]

4 hours ago, MatthiasArnold said:

I had a similar observation. Servicing the connectors solved the issue. 
Regards,

Matthias

Thanks very much!

[Skates97]

12 hours ago, Ragsf15e said:

I think this is probably gonna be what I need to do.  Were you looking at the wires where they connect into the harness at the back of the actual jpi unit in the panel?

No, there is a lead that comes off the sensor. At the end of the lead are two ring terminals where that connects to the harness. Check those connections. 

[Ragsf15e]

9 hours ago, Skates97 said:

No, there is a lead that comes off the sensor. At the end of the lead are two ring terminals where that connects to the harness. Check those connections. 

Ahh, thanks.  I’ll have a look at that.

[Yetti]

Start with the easy things.  Check the oil level.

[Ragsf15e]

36 minutes ago, Yetti said:

Start with the easy things.  Check the oil level.

Well yes, it’s good.  The old jpi data shows it’s been doing it pretty consistently for about a year, so I think it’s not falling out of the sky.  It’s been the same from 5-7qts, so I guess I’m at sensor or wires.  I’m gonna start poking around tomorrow.