P

prkaye

Well Known Member
I've done a search and read several threads about the Dynon oil temp probe, and the common problems with grounding of this probe. If I understand correctly, the probe is grounded through the threaded fitting (which then finds ground through the firewall, or through the hose and then throught the engine?).
I've had slightly unstable readings (rapid oscillations by a few degrees). One post suggested using a hose clamp and a dedicated ground wire. So my plan is to get a large diamer hose clamp, clamped right around the cylindrical body of the probe, with an attached wire leading directly to my ground bus-bar (on my firewall). Will it work to directly ground the outer casing of the probe like this? Is the outer casing of the probe electrically connected to what the probe expects to be ground?
 
Phil
I have grounded a sensor, from another company, by using a hose clamp on the case. Make sure to clean up a spot on the unit with sandpaper where the wire will contact the unit. You can easily check the ground path from the case body to the threads on the unit using an ohm meter.
 
One ground only

You should use only one ground between the engine and the battery. If you run a seperate ground for your sensor (or PMAG or whatever) in an attempt the fix a problem caused by a bad engine ground, other devices will also use it, think path of least resistance, that includes the starter with hundreds of amps. The second ground wire you just installed may go up in smoke. This may not happen until later when / if the main ground becomes weakened.
So, investigate the braided wire engine ground you have for clean and tight connections, make it good and then look elsewhere for the problem.
Tim
 
Last edited:
except I doubt the problem is with a bad engine ground (the braided strap from the crankase to the ground bus/battery terminal). I suspect the problem is the connection through the threads on the sensor itself, given that I used thread sealant.
 
Oil temp sensor

Phil doesn't your oil temp sensor use a crush washer? The temp probe normally has machine threads and uses a crush washer to seal the flange. The other senders are cantelevered from a 1/8" NPT and remote mounted, so thread sealant on those, but not on the temp probe.
In any case, adding a second ground wire, unless it's big enough to handle the starter/alternator loads, is not good practice and could lead to smoke in bad places.
FWIW, Tim
 
Consider a local ground lead

In the instrumentation world we sometimes need to reduce the ground resistance locally while maintaining strict adherence to a single grounding path or alternately a dedicated signal return grounding topology.

In the fist case we would install a ground lead locally using a clamp or washer with a lug we can connect a ground wire to. We then connect the other end of that wire to a point as electrically close to the engine end of the braided ground strap as possible. The only current through this wire will be shared between the path through the threaded connection on the engine and the dedicated wire. Very low. We would also look at the possibility of electrically isolating the sensor from the engine and only using the dedicated ground jumper on the sensor. This approach is often difficult in a retrofit application.

The most robust approach is a grounding/signal return topology that isolates all of the signal return from any path that carries ground return current. This is usually very difficult as a retrofit and is typically addressed as a systems requirement early in the design process.

Cheers,
larosta
 
Ground Cable

I had similar problems with the oil temp. reading on my Skyview display. I also noticed a subtle change in cranking speed when starting, lower voltage when my PC 680 battery was fully charged- 13.6 as opposed to 14.2- and my Plane Power alternator began to go offline for a few seconds periodically.

I checked all connections at the recommendation of the people at Plane Power and found a slight amount of movement between the ground cable and terminal at the engine side of the ground cable. I removed the cable from the engine and was able to pull the terminal off by hand. This cable was part of Van's wiring kit, and I had checked the security of the terminal before installing it. There was carbon build-up in the cable to terminal connection, so I cleaned the terminal and cable well and soldered the terminal to the cable, as it was impossible to crimp it securely.

All the previously reported problems are now resolved. Voltage is 14.2, oil temp is steady, and starter cranks well.
I suspect that engine shake loosened the terminal from the ground cable.
Even though I was very careful with all my grounding, this was not something I would have expected. I almost didn't check it. I hope this might help those having mysterious electrical problems.
 
Last edited:
The chances of it being the ground of the sensor to the block are almost zero- but if you want to check it with an ohmmeter, you can. Just hook the ohmmeter to the brass on the sensor and then the aluminum it screws into. You'd need more than about 4 ohms to make a difference. There's no current in this connection (well, maybe 1mA), so all you need is a thousandth of an inch of brass touching aluminum with no sealant to make a good connection.

The issue is always the ground between the EMS and the ground of the engine. We've seen painted blocks, painted frames on composite planes, loose ground bolts, corroded braids, and more that have led to this issue.

As mentioned, you can't run a wire to fix this. This is like putting a garden hose in parallel with the water main out in your street that has an obstruction in it. You'll do bad things to the garden hose ;)

Is this a D10/D120/D180 or a SkyView?
 
It's a D-180.

The issue is always the ground between the EMS and the ground of the engine
Click to expand...

Hmmm, it's been a few years since I installed all of this, but I'll check the connection on the ground wire coming out of the D-180. If I recall my setup correctly, I have a braided strap running from the negative battery terminal to a brass ground block on the firewall, and another braided strap running fom th engine to this same ground block. The ground wire from the Dynon woud be connected with a spade terminal direclty to this firewall ground block. So the only weak point here could be that spade terminal connection.

The oil temp reading doesn't fluctuate at all unless the engine is running, which suggests to me that vibration is a factor.

I just read on the D-180 installation manual they recommend grounding the instrument panel or D-180 mounting tray to engine ground with a 14AWG cable. I don't think I have done this (can't remember), but something else to check and do if needed.
 
Last edited:
FWIW, I once connected the same sensor type (single wire, case ground) to the accessory case port with a steel-braided line, the line braid providing an acceptable ground to the engine block. It did of course, but in the midst of a Grand Brain Fart, I used a AN816 adapter in the hardware connecting the sensor to the line. At 20~30 hours the oil pressure gauge started making slow oscillations from low to normal, which will get your attention when flying over a pulpwood forest. Yep, it was the anodized aluminum AN816. Not sure if the root cause was non-conductive anodizing or a dissimilar metals issue, but I am sure I won't be using one again.

Along the same lines, your firewall block is aluminum? Is it anodized? If not, or if you've retapped the threads to eliminate the anodizing, I'd be inclined to think that pipe sealant would be a good thing, as it might tend to keep all forms of electrolyte out of the threads. Pull out the ohmmeter!
 
Along the same lines, your firewall block is aluminum?
Click to expand...
Do you mean my ground block? My Firewall ground block is the brass BNC one - forests of tabs on both sides of firewall with a big brass bolt through the firewall between them. The battery and engine are connected to the bolt by big braided cables.

If you mean the sensor mount, my sensor is plugged into a vans VA-168 Transducer mounting block which is bolted to the firewall. The VA-168 appears to be non-anodized bare aluminum. I'm not at the hangar now, but I'm pretty sure I used stainless steel fittings in the VA-168. The oil line from the engine to the VA-168 is a braided steel line.
 
Well I'm glad I dug into this problem. While working on it this past weekend, I discovered my main engine/battery ground connection was loose! My Firewall ground block is the brass BNC one - forests of tabs on both sides of firewall with a big brass bolt through the firewall between them. The battery and engine are connected to the bolt by big braided cables. The nuts on the big brass bolt were finger-loose! I had always visually inspected the ground connections during my Annuals, but never put my fingers on the nuts to ensure tightness. I wonder whether vibration was causing intermittent contact through this loose connection, causing hte fluctuating oil temp readings? Anyway, I snugged the nuts down good and gouped some torque seal over them. I also added a dedicated ground wire from my Dynon mounting tray (to the BNC forest-of-tabs ground block).
I'll test fly this coming weekend and see if the readings settle down. Regardless, it was fortuitious that this investigation led me to discover the loose main ground connection!
 
If you look at old (late 70's, early 80's) cars, you will find TONS of 1-wire sensors that were grounded through the engine block. Look at any modern car today and you will NOT find any 1-wire sensors. They all have at LEAST 2 wires (signal and ground. A lot of modern sensors also require a regulated 5V as well).

The reason is that the car companies all found that single wire sensors just are not reliable in the long run. They could get away with it (1 wire is cheaper than 2...) when it was just going to a gauge and a human was generally ignoring it anyway, but the switch to EFI where a computer was monitoring it (and reacting to the data) several times per second forced them to fix it.

If you have a ground block, run a ground wire from there to the sensor and see if it stabilizes it. If it does, you can probably find a 2 wire sensor that has the exact same dimensions and specs to replace your current one.

If you don't have a ground block (using he airframe as a common ground), then you can still improve the ground of the sensor by grounding it to the engine block with a clean, no sealant connection. That will at least address the poor ground caused by the thread sealant. As someone has already mentioned, you do not want to ground it straight back to the battery.

Clint
 
And of course, if I had finished reading the full thread BEFORE posting, I would KNOW that DID have a ground block and that you found a possible cause of the problem.... !
 
So after properly grounding the dynon tray to my BNC ground block, tightening up the loose nut on the main ground block post), and adding a ground wire clamped to the sensor running to the groundblock, my oil temp readings no longer fluctuate by more than 1 degree. Nice to have that problem solved.
 
Last edited:
I just started having same problem on D180 oil temp. If I turn off the alternator in flight, will this fix the problem if it is a grounding problem? It started after I changed the oil filter and slid a cardboard box through wiring to capture oil from filter.
 
Last edited:
prkaye said:
except I doubt the problem is with a bad engine ground (the braided strap from the crankase to the ground bus/battery terminal). I suspect the problem is the connection through the threads on the sensor itself, given that I used thread sealant.
Click to expand...

I found something interesting. My oil filter assembly was painted before assembly. Luckily I thought about grounding of the oil pressure sensor. The gasket prevents contact of base material and grounding is through the four bolts. In my case, I had very poor conduction due to the paint on the housing. I cleaned off the paint under one of the bolt heads and now have 0 resistance.

You have to do the same thing with an accy case bolt.
Worth looking at.

Larry
 
lr172 said:
I found something interesting. My oil filter assembly was painted before assembly. Luckily I thought about grounding of the oil pressure sensor. The gasket prevents contact of base material and grounding is through the four bolts. In my case, I had very poor conduction due to the paint on the housing. I cleaned off the paint under one of the bolt heads and now have 0 resistance.

You have to do the same thing with an accy case bolt.
Worth looking at.

Larry
Click to expand...

Good tip, but if an engine is painted properly it shouldn't be an issue.
The pro engine shops use self adhesive dots to mask any bolt or nut locations before painting.
It is undesirable to have paint under fasteners. Paint is not very hard. It will compress over time. Even a very thin paint finish will get compressed and relieve the stretch load on fasteners.
 
You must log in or register to reply here.