Need engineering point of view on oil cooler and air pressure

I moved the oil cooler from the baffle to the firewall and not pleased with the increase in oil temp.

I have a theory and need an engineering point of view.

When a cooler is mounted on the baffle, the front surface area equals the back surface area. Since the air pressure is higher in the area where the air has not cooled the engine yet (baffle box), the higher pressure gradient forces the air through the cooler from the baffle box to the remainder of the cowl.

For the sake of argument, let's assume the baffle box has a pressure 3 times what the remaining cowl has.

When you mount an oil cooler on the firewall and feed it with a 3" scat tube, the surface area coming into the oil cooler is almost 1/3 that of the back side where the air exits. If that is the case, then the pressure gradient is reduced by 3 times and it becomes difficult for air to exit the cooler since the exit pressure is close to the pressure inside the cowl.

Based on this assumption, if one was to make a reducer for the back side of the cooler with the same surface area as the front, then the pressure would theoretically drop from 3 in the scat tube to a lower pressure in the cooler equal to the cowl pressure and then would be then increase again upon leaving the cooler through the surface area reducer. This increase in pressure would allow the air exiting the cooler to return to baffle box pressure, therefore overcome the lower pressure in the cowl.

My gut is telling me that airflow has been reduced by the inability of air exiting the cooler to overcome the pressure gradient. Application of a venturi effect on the exiting air will take care of this issue.

that long run of 3" SCAT is killing you, up it to 4".

I'm no engineer, but adding another restriction to the system will not help your cooling issue. you can help a cooler by either changing the cooling media Delta T or changing the mass flowrate through them. The only one you have control of is the mass flowrate of the air. Supply as much plenum pressure to the cooler inlet and/or lower the pressure on the outlet side.

Webb, I'm running 3" diameter scat to my cooler on an O360, no issues. An important question is transitions - what are the shapes of the transitions at both ends of the scat? Ideally, they would be smooth transitions with generous radii on both ends. These transitions will matter a lot more with respect to flow than the relatively short piece of 3" scat will.

3" scat to OC mounted low on engine mount. OC is SW, engine is O-360. Cannot get OT above 195 even when 1/2 of the 3" inlet is blocked - which is how I ran all summer in Southern Florida. Temp probe calibrated in boiling water.

Webb, There are a lot of aerodynamics and thermodynamics involved here. But I think it is a safe assumption that the cowling pressure is a lot lower than the plenum pressure. The cowling is lowered by the free stream pressure of the airflow passing by the exhaust opening while the plenum is seeing the near full stagnant pressure from the air being rammed in throught the front openings. Your restrictor will only make your problem worse. My 0-320 equipped -4 hass a baffle mounted cooler with the opening to it only about 2.38 in. in diameter. The hole size was restricted to ensure that none of the air entering the cooler had a chance to pass through the cylinder fins and pre-heat the air. I kept the opening about 0.25 in. above the fins and have no problem with temps on hot days. On cooler days I have to close off the opening just to get the oil to warm.
Alan RV-4

hydroguy2 said:

that long run of 3" SCAT is killing you, up it to 4".

I'm no engineer, but adding another restriction to the system will not help your cooling issue. you can help a cooler by either changing the cooling media Delta T or changing the mass flowrate through them. The only one you have control of is the mass flowrate of the air. Supply as much plenum pressure to the cooler inlet and/or lower the pressure on the outlet side.

Click to expand...

My premise is the plenum pressure is getting decreased by a venturi effect as it exits the cooler into the cowl aera. If you use the smaller (scat) cross sectional area, it is 7.07 sq inches. The cross sectional area of the cooler opening is 18.13 sq inches.

Using a reverse restriction should pick the exit pressure back up closer to plenum pressure and ovecome cowl pressure. Using the diagram below, my thought is the scat is the smaller diameter portion of the flow and the cooler is the larger. Problem is the pressure exiting the cooler is having trouble exceeding the cowl pressure and needs to be increased so the air will exit. A pyramide shaped exit on the backside of the cooler would increase the exit pressure. Think of the way household HVAC does the ductwork on long runs, the duct is reduced in cross sectional area as it gets farther away from the air handler. That increased pressure is needed to keep the pressure up at the end so the air will blow out of the ducts. I'm thinking along those lines. That's my theory.



If I increase the scat to 4 inches as you suggest, then the cross sectional area of the scat is 78% larger and that means the pressure exiting the cooler based on a venturi effect would be increased over a 3" scat and do a better job overcoming the cowl pressure.

3" scat has often been used with good results as exhibited by some post. I'm thinking its an exit pressure issue on mine, not the available amount of air. Position of the cooler and exit side may be part of the problem. Based on the venturi theory, if the exit side is closer to the cowl exit, the pressure is higher there than up higher in the cowl. I'm guessing but I may have "shot myself in the foot" and mounted it too low.

To overcome the pressure, my thought is to put a pyramide shaped exit on the backside with a opening of 7 sq inches (equal to that of the scat area) to ramp up the pressure exiting into the cowl so it will increase the flow of air.

Still gotta put the old sensor back just in case the Dynon replacement one is bogus. That's the first step.

I'm just guessing and I'm hoping it's a good guess. Note that a exit reducer is the easiest metal work to do. Can be farbricated in a few minutes and is a heck of a lot less work than moving the cooler, back to the baffle, or a different place on firewall, or installing a 4" flanges.

We have a lot of very intelligent engineering type folks on here. I'm hoping they will chime in too.

Webb,
Did you change the oil cooler location and the cowl at the same time? Or did you go to the ducted cooler and lose oil cooling capacity while still flying the original Vans cowl?

Both at same time. As I work through this, I think the exit side of the cooler is too close to cowl exit and the pressure is higher there and it's having difficulty overcoming the cowl pressure. ie: back pressure is killing airlfow.

DanH said:

Webb,
Did you change the oil cooler location and the cowl at the same time? Or did you go to the ducted cooler and lose oil cooling capacity while still flying the original Vans cowl?

Click to expand...

Webb,

For sure there is a lot engineering science in achieving good oil cooling but for this non-engineer builder/aviator it all boils down to mass air flow across the cooler and a pressure differential to cause it.

My 4" ducted system to a RV-10 style mount with the double pass SW cooler works very well. Typical OT is less than 170 and seldom over 200. A complete shut off of air to the cooler will get the temperature up to 210.

I believe the key is enough exit area to cause a low pressure in the lower engine compartment to assist that air flow I think is so necessary to remove heat. Also, the amount of air through a 4" vrs 3" duct is almost double.

Good luck. There's nothing more irritating than marginal cooling on any mission with these airplanes.

Webb said:

Both at same time. As I work through this, I think the exit side of the cooler is too close to cowl exit and the pressure is higher there and it's having difficulty overcoming the cowl pressure. ie: back pressure is killing airlfow.

Click to expand...

No. Pressure just outside the exit is at static, less than lower cowl pressure and certainly less than upper cowl pressure. As you move the cooler exit closer to the cowl exit, local total pressure will decrease and cooler flow will increase.

There are three possibilitites. One is that the SJ cowl does not recover dynamic pressure as well as the original Vans cowl. A major point of interest between the various inlet/exit schemes is pressure recovery; remember NASA CR3405? I do hope you made upper plenum pressure measurements before you removed the Vans cowl. If you have pressure data for the Vans cowl, a quick flight with the same speed and altitude conditions should tell the tale for the SJ cowl.

Second, as Alan mentioned it is possible to feed preheated air into the cooler duct. My own measurements with a ducted cooler show in increase of 15-18F at the cooler face as compared to OAT. It gets heated in the upper plenum before it ever goes down the duct to the cooler.

Third is simple flow restriction. Remember a detail about SCAT; the rated ID is the diameter of the wire winding. When we bend the tube around a corner the fabric folds push inward and reduce effective diameter....sometimes a lot.

Dan - first of all, thanks for earlier conversation today.

Putting it all together, I think think it's a combination of issues as we discussed.

1 - 3" scat has a cross section area of just over 7 sq inches. It used to be full frontal with just over 18 sq inches. Reduced air mass entering cooler.

2 - Van's entry box into the cooler is a poor diffuser for firewall mounting. Air is poorly distributed over the cooler entrance.

3 - 90 degree scat turn into cooler equals 10 feet of straight run. Resistance of a 10 foot run is high.

4 - Venturi effect by a 3" transition to a 18" transition makes the lower exit pressure gradient harder to overcome when air exits the cooler.

5 - oil temp sensor was replaced. Temp numbers may be suspect.

Bottom line is air just isn't getting through the darn thing.

Old sensor installed. Cooler pull from firewall. Will be installed on #4 baffle tomorrow evening. Will post again if case closed after flight.

Webb,
The small diameter SCAT ain't helping, and the squared-off oil cooler inlet plenum isn't good. You can get an incremental improvement with larger, straighter SCAT, and with a nice long transition from the round SCAT entry to the shape of the cooler face. Or you can, as you say, just put the cooler back on the baffle wall

I tend to discount #4.

Sure wish you had measured upper and lower plenum pressures with the old Vans cowl. The pressure difference drives air mass through the cooler. Less upper or more lower pressure with the new cowl would result in less mass flow.

DanH said:

I tend to discount #4.

Click to expand...

Wow, now wasn't that nice....

point 4 in post #11, not post #4...

Ken, it looks like Dan is referring to point 4 in post #11, not post #4...

DanH said:

Webb,
The small diameter SCAT ain't helping, and the squared-off oil cooler inlet plenum isn't good. You can get an incremental improvement with larger, straighter SCAT, and with a nice long transition from the round SCAT entry to the shape of the cooler face. Or you can, as you say, just put the cooler back on the baffle wall

I tend to discount #4.

Sure wish you had measured upper and lower plenum pressures with the old Vans cowl. The pressure difference drives air mass through the cooler. Less upper or more lower pressure with the new cowl would result in less mass flow.

Click to expand...

Going back to the baffle wall is the easiest and fastest. I've got the angled mount I built and will use it. I guess it's the equivalent of trying to drive a 10 penny nail with a tack hammer with the attempted current firewall set up.

Totally agree on scat size killing me.

Disagree with you on #4 but without measurements, it's a SWAG (scientific wild a$$ guess). Heck, it might just be a WAG.

Nothing like hind sight on the pressure measurement but I can truly say the new plenum is so well sealed compared to the old baffle box, I suspect it is better.

Tell you what we could to. Get a measurement on mine and have some measure another 7 for comparison. I know it won't be the same box but it would be based on the same design that was the standard Van's.

Ooops

oh ooops... must have eaten a bad bowl of wheaties this AM. My apologies Dan!!

Well, it motivated me to go fly about an hour doing nothing but taking screen shots of my OT at various flight regimes.... so it was a good thing

for a cooler mounted low on the firewall-

Might there not be the possibility of loading the cooler outlet with some dynamic pressure from the outward flow of cooling air that is occurring at the exit (for a low firewall mounted cooler)???

Certainly a simple outlet plenum could be made to "shield" the cooler outlet from any dynamic pressure to ensure that it is seeing low pressure to avoid this.

IIRC, did'nt you do something like this Dan??

Thanks

hydroguy2 said:

that long run of 3" SCAT is killing you, up it to 4".

Click to expand...

I agree with the general sentiment, but the difference in flexibility and ease of routing between 3" and 4" is huge. On Steve Smith's RV-8 he used 3.5" SCAT and made it route between the cooler plenum on the firewall and the rear baffle. It would have been a lot tighter with 4" SCAT:



It seems to cool very well.

Thanks, Bob K.

4" is tight... just fits with a -7

I used sceet tube rather than scat for the smoother inside surface...

Here is a shot of 4", it is tight, but has worked out great.

Stephen and Bob - pics you showed are using a 13 row and it won't mount to the baffle. I am dealing with a seven row.

I did remove the 3" flange and cut a 16 sq inch opening tonight on the baffle which is just 2 inches shy of the 18 sq inch opening on the cooler. Also the mount for it is tilted back 20 degrees which exposes the whole face to air.

My guess is oil cooling issues will disappear and return to normal tomorrow evening when it gets tested.

BTW - This is the only squawk encountered on the conversion to the Sam James cowl and plenum. Cylinder one is a bit cool and that requires some tweaking but is not a problem. Just some aluminum tape to find out how big to make the entrance deflector.

Where's Webb?

Webb, they've been looking for you at the Hospital.

MississippiJim said:

Webb, they've been looking for you at the Hospital.

Click to expand...

Correct.....at the psych ward.

Just remember to wave at me when you see an RV flying over the cuckoo's nest.

Psycho back again. Yes I've escaped....

Test flight tonight with some record oil temps.........

The oil cooler was moved back to #4 baffle prior to flight due to earlier high temps.

The CHT's were perfect with the exception that #1 still needs a bit more shielding as it was still a bit low.

However, the oil temp kept bouncing around and then started climbing and I don't mean a little bit. Upon time to land it was reading a bold 351. No way when the CHTs were lower, the oil pressure was perfect, etc... Also we know that the oil cooler was sufficient to cooler before I put the plenum on the engine and it should work now.

Upon landing, opened the oil filler door and the temp of the air coming out was normal as in really normal. No burnt smells. No nothing.

Pulled the top cowl and used a digital thermometer and everything normal. Engine block was 187 at the highest measurement, area around the sensor was 185, oil cooler was 154, etc... Couldn't find anything that was out of range.

Hmmmm. Pulled the sensor lead off and the Dynon read 1 degree which tells me that a high reading is occurring because it is shorting to ground. Put lead back on an the sensor read 154 and the digital thermometer read 155.

SO, I've have a sensor that is shorting to ground. Instead of trying to trace it back, at the recommendation of a friend, I'm going to just run a new wire back to the Dynon instead of driving myself nuts trying to found out where.

Like I said, this is the only major squawk I've run into.

I'm beginning to believe the firewall mount may have been sufficient and it's a sensor issue.

More to come......

Firewall mount

Just as another reference point, I have a firewall mounted standard cooler on my superior IO 360 mounted with the vans kit and a 3 inch tube and have a butterfly valve to shut off the flow. Most of the time it's 1/4 to 1/2 closed even in the summer. On a 25 degree C day it'll go to 190 F max.

No worries about baffle and cooler cracks!

Peter

Webb said:

Hmmmm. Pulled the sensor lead off and the Dynon read 1 degree which tells me that a high reading is occurring because it is shorting to ground. Put lead back on an the sensor read 154 and the digital thermometer read 155.

SO, I've have a sensor that is shorting to ground. Instead of trying to trace it back, at the recommendation of a friend, I'm going to just run a new wire back to the Dynon instead of driving myself nuts trying to found out where.

Like I said, this is the only major squawk I've run into.

I'm beginning to believe the firewall mount may have been sufficient and it's a sensor issue.

Click to expand...

I just had to replace a Dynon oil temp sensor in the RV-8 about a month ago. It did the opposite of what yours is reading... mine would jump wildly and erratically between normal (180-190) and 133. After landing, with the Dynon indicating 133 degrees, I used my handheld digital laser pointer infrared thermometer on the oil filter thru the oil door and it showed 179 degrees... normal.

After exhaustively verifying that all the electrical connections were good, and even adding a dedicated ground wire clamped to the sensor to eliminate that possibility, I still had erratic oil temp readings, so we ordered a new sensor from Dynon ($20) and after installation, oil temp readings were normal and stable again. The first sensor lasted only for about 175 hours of flight time since the airplane was built.

Moral of story: these sensors certainly can, and do, fail. In my case it was an intermittant "open" or "partial-open" happening inside the sensor itself. So it's perfectly plausible that an internal intermittent short is also a failure mode of these cheap sensors.

oil cooler configuration

see the following post.
http://www.vansairforce.com/community/showpost.php?p=561361&postcount=18

It seems that upper cowl pressure has a fairly large impact on the oil cooling capacity.

I agree with the comments that the scat airflow is probably not efficient due to all of the wrinkles, however you make it big enough and it works.

In my situation a smooth 3" tube would probably have worked as well as the 4" scat.

It would be interesting to put a pitot tube in the oil cooler duct to see what velocity the air is moving. In any case, a square plenum right off of the 3" duct seems like it would have a lot of losses as the air passes around the square corner of the plenum. That is why I made my plenum with as shallow angles as was practical to evenly distribute the air across all of the OC.

Talked to Dynon. The high the reading can be is 352 and that is totally grounded. Mine was 351. Lowest is 0 and mine was 1 with no lead and that is for the totally open circuit.

I didn't find a nick in the wire from the sensor, however it runs much closer to the spark plug wires after re-routing those for the new plenum. I'm thinking that some EMF is getting into the sensor line and causing a false connection to ground. Mainly because the temp read accurately without the engine running.

Will put a new lead to the sensor tomorrow morning with a different route to it.

In the meantime, I'm thinking about getting a GRT FR-LC-01 sensor. It works out of the box with the Dynon units and is a 2 lead sensor with it's own ground wire and doesn't ground through the case. The Dynon forum had discussion on this and after exchange, sensors were rock solid instead of oscillating.

Well, I did the re-wire from the sensor back to the firewall and went around the mags. No joy.

Only did a ground run this evening, but ran the CHT's up to 330 and the oil temp was oscillating between 175 and 185 degrees. Killed the motor and the Dynon dropped down to 165 and stopped. Checked the temp on the sensor in a few minutes later and it was 154 with a IR thermometer and Dynon showed 155.

Called tech support from Dynon and they are stumped on this one.

Here is what we know.

The higher the temp on the sensor, the lower the resistance which indicates somehow current is getting in or in someway the thing is shorting to ground. This means that a loss of ground isn't happening, it is just the opposite. If I had a loose wire the temp would read lower all the way down to zero degrees which is a totally open circuit. I also verified this when I disconnected the lead from the sensor and the Dynon read 1 degree.

New wire back to firewall from the sensor. Didn't go from firewall to unit since I didn't touch a thing under the panel. Didn't find a loss of insulation on the wire I replaced.

All temps verified by a laser IR thermometer and are normal. When engine is not running but master is on, the Dynon reads pefect and does not have a fluctuating reading.

When running, the temp shown starts going out of sight after the engine heats up. Yesterday, it went to 351 which is a totally closed circuit. After checking the temp with the thermometer, it was 1 degree different on the Dynon and sensor without the engine running and was in the 150's. Ground run today was 20 degrees hotter on the EMS with engine running than what it showed when the engine was not running.

This has done this with 2 sensors. One old and one new. No changes on sensor type in the EMS.

Results are repeatable that after killing the engine but leaving the monitor on, the EMS shows a temp that agrees with the thermometer.

All other sensors work perfectly.

What has been changed on the plane.

New cowl and plenum.
Spark plug wires routed in a different entrance in the baffle.
New spark plugs installed (shorter for plenum).
Baffles lowered for new plenum.
Mags overhauled due to 500 time (Unison).
New fuel pressure sensor (works perfectly).
Prop extension.
Prop re-balanced.
New wire for sensor back to firewall.
Oil cooler moved to firewall and then back to baffle #4.

Plane had 498 hours and this was never a problem until the plane got a nose job.

Read the Dynon forum and several folks switched to the GRT sensor with the dual lead so that ground wouldn't run through the engine case bacause they had problems. I've placed an order for one (not expensive - $19). Also, one of the maybe's that came from Dynon tech support was the new sensor may be bad and engine vibration could be entering an AC current into the wire.

Fortunately, I'm not worried about the oil temp since the cooler is on baffle and was sufficient to cool the engine before the nose job. This has also been verified by a laser IR thermometer.

Somehow, voltage is getting into circuit, question is how.

It's now just a head scratcher and I want to get it fixed.

Any suggestions/solutions?

Webb said:

Well, I did the re-wire from the sensor back to the firewall and went around the mags. No joy.

---------------------------------------------------------------------

This has done this with 2 sensors. One old and one new. No changes on sensor type in the EMS.

---------------------------------------------------------------------

Any suggestions/solutions?

Click to expand...

That pretty much eliminates everything but the Dynon itself.

Erratic oil temp

Mike S said:

That pretty much eliminates everything but the Dynon itself.

Click to expand...

Could be, however when the engine is off, the Dynon is reading the resistance values from a totally closed and open circuit and inbetween. I've checked the resistance on the sensor against the reference chart and the unit is doing it's job.

Spent an hour reading the Dynon forum and I think I may of have found the answer. I think that oil has gotten on my ground connections from the engine to the airframe. During pulling a sensor, moving an oil cooler, changing oil filters, etc.., my best belief is that oil has gotten on the ground connections from the engine to airframe. I'm going to pull all engine gound connections, all alternator connections, and scuff and clean the contact points.

In one of the post, the Dynon tech made the statment that erratic reporting is 100% due to grounding issues. Here is what I found digging in the the Dynon forum archives that best describes why:

Oil temp is the sensor that is most sensitive to voltage changes. The 180-190 change you are seeing is only about 20mV of change. The jump to higher temps represents about 100mv.

The voltage change is coming from your ground between your engine and your battery/master bus. The alternator on the engine uses this ground path as part of the circuit, so all the current the aircraft is drawing goes across this connection. This connection must have a small amount of resistance, and the current going through it causes a small voltage drop. As the current changes, this voltage changes, and you see it on the oil temp sensor. The current in an airplane is changing all the time due to transponders, radios, autopilots, and all sorts of other things.

The solution to this is to get a bigger or better ground between the engine and your systems. First, check that the ground you have is free of paint, oil, powdercoat, corrosion, and any other insulators. We've seen painted engines, painted airframes, tiny wires, and all sorts of other stuff that causes this problem. If you're 100% sure that the ground between the engine and the aircraft is as solid as it can be and you're still seeing this, then you'll need to increase the size of this wire.

We've had a lot of people say they'll just run a ground wire from the oil temp sender to the EMS. Unfortunately this doesn't work, since this just becomes a very small increase in wire size versus the overall ground connection. You need to work on the main ground.

kiwipete said:

. On a 25 degree C day it'll go to 190 max.

Click to expand...

190 C oil temp? I hope that's a misprint. That's 374 F. Hope you meant 190 F (and mixed units)

Removed and cleaned ground from engine to frame. Much better but still not where it should be. Ground is #2 cable. Still has an oscillation of 2 degrees.

Steadied out around 217 even though cooler on baffle. Landing lights cause 10 more and shutting down engine dropped it by 15.

Lasered temps and couldn't anything close. Bottom of oil filter was 185 and dipstick oil was 178. Suspect a bad sensor as one problem. Do have a new one coming.

Will call Dynon and rattle their cages in AM.

Guess I could run a second ground since the Dynon folks say it's a ground problem.

Mixed units

That'll teach me for mixing my units 25 degree C day 190 F oil.


Peter

Webb said:

Guess I could run a second ground since the Dynon folks say it's a ground problem.

Click to expand...

I used a #11 Adel clamp (MS21919-WDG11) with the rubber boot removed from it to go around the body of the oil temp sensor to attach a ground wire. It made a really tight fit and solid electrical connection once the screw and nut were tightened all the way down. The clamp also provided me a good place to attach a safety-wire to the sensor.

Neal@F14 said:

I used a #11 Adel clamp (MS21919-WDG11) with the rubber boot removed from it to go around the body of the oil temp sensor to attach a ground wire. It made a really tight fit and solid electrical connection once the screw and nut were tightened all the way down. The clamp also provided me a good place to attach a safety-wire to the sensor.

Click to expand...

Did that with a hose clamp. Didn't make a single difference. Dynon's comment was it wasn't a sufficient enough ground. Only difference made so far was cleaning the ground strap from the engine the ground which reduced the wild erratic readings.

Now I only get 2 degrees fluctuation, 10 degrees more on a load such as landing lights, and 10 or more on shut down. I think that the sensor is also providing bad resistance numbers to the Dynon.

91 degrees ambient temp is 519 ohms, 190 degrees is 79.7 ohms, and 217 is 48.8 ohms. As the temp goes up, the difference in resistance becomes less and less and the chance for an bogus reading is much more likely.

Alternator current is getting into the sensor line according to them. Since it's a single wire and uses the engine case for the ground, this makes sense.

Having a GRT 2 wire sensor that doesn't use the case will probably make the last bit of difference by not allowing some voltage to sneak into that circuit.

Grounding Thots

Webb, I don't understand why you didn't redo the wire from the firewall to the Dynon.

The fact that your reading is more accurate with the engine off leads me to believe that vibration could be a factor - I recommend carefully checking all crimps, especially any made by Vans.

Having chased a lot of ground fault issues in racecars, I have three grounds on my airplane engine - the standard braid and two thick insulated wires - all attaching to different firewall locations - I also use anticorrosion conductive paste.

Hans

If you're getting big oil temp reading variations just by turning on and off other electrical stuff, that sure sounds like either a +12V supply and/or grounding problem that's affecting the power supplied to the Dynon EMS itself... or maybe there's a problem with the Dynon's own internal voltage regulation circuitry. You might want to try watching the input voltage that's feeding the Dynon itself, with a handheld DMM, while turning on and off the landing lights, starting/stopping the engine or alternator field breaker, etc, to see if you've got any unusual voltage swings at the EMS's power supply inputs.

ETA: On the RV-8 in my avatar, I wired in a dedicated ground bus with a copper strip behind the instrument panel that all avionics/electronic stuff is grounded to, and that's connected with a #8 wire to where the negative battery cable connects to the firewall with a 5/16" stainless bolt with pair of big SS fender washers sandwiching the firewall, and on the engine side, a #2 cable under that bolt goes to the engine's case to ground it. I've also got another #8 wire from that big thru-firewall bolt, going to a ground bus bar under the floor which provides ground to all the ship's lights. I believe in overkill on the ground bus and wiring on all the RVs I've wired