More airflow to the RV-12 Voltage Regulator - Proven!

I just finished my "hood scoop" mod to substantially increase airflow to the RV12 voltage regulator and NOT moving it inside the cabin. The concern is that overheating of the VR is most likely to be occurring during ground operations, when the engine rpm is low, airflow to the VR is low, and VR output may be high.

This mod is done as follows. I did NOT want to relocate the VR inside the cabin. Note that I also installed the Bender Baffle cabin heating modification so this had to be compatible.

This is the basic design of the scoop, which goes in the radiator duct.



You need both a scoop and a piece of 5/8" ID aluminum tubing. I chose that because I wanted to reuse the existing corrugated tubing supplying air to the VR.



The location of the scoop has to be far enough forward to not interfere with the Bender Baffle. In this pic, the lower cowl is UPSIDE DOWN and you are looking in from the square radiator opening.



With the scoop glassed in, I cut a hole in the top of the duct that matched the oval opening of the cut tube, then glassed in that tube. I had more than enough leftover corrugated tubing from the original installation to do all this. This picture is looking DOWN on the cowl, rightside up. To the left is FORWARD. on the right you can just see the corner of the square gasketing where the duct opening mates to the radiator. That oval hole is adjacent to the front of the forward pilot-side cylinder on the engine.



So here is the installation. The aluminum tube glassed to the duct is wrapped in electrical tape to seal the corrugated hose. Note the quick-remove coupler in the middle of the corrugated hose going to the VR - for removing the bottom cowl. The diagram shows how the corrugated tube is held in place by either safety wire or narrow tie wraps through holes cut in the aluminum outer tube. (I did not think of this - I saw it done by Joe Gores).



So, it "should" be better. After all, the existing setup has a narrow round opening that is at a right angle to the airflow it is trying to "catch." On the ground, that airflow itself comes from the SCATT tube opening which is at the very root of the prop - an area that does not produce much thrust. I had previously done a test with a dowel and yarn to prove to myself that there was a LOT of airflow at idle RPM going through the radiator part of the duct. But, how to prove that the scoop works?

Here is the setup.


I tested both the original source and the scoop source by connecting the tubing from each to a ~70 inch piece of clear plastic tubing run into the cockpit. I tested on the ground with the engine at 2300 RPM (typical warmup idle) and at 3500 rpm. For each case I let the tubing blow on my face and also into the water container, where I could see bubbles if I submerged the tube. (I tried to take pictures but they did not come out well. I was alone.) Now yes, the 70" of tubing does increase the resistance and the flow at my face will be less than would be at the VR. But that restriction is the same for both cases.

The RESULTS!
A> From the standard location/source off of the SCATT tube-supplied shroud, at 2300 RPM, only the barest wisp of air flow could be felt holding the tube to my face! No flow would occur if the tube was submerged more than about 1/8 inch into the water, and only a few slow bubbles would appear if the tube was held touching the water's surface. There was almost no change change at 3500 RPM.

This seems to verify that cooling of the VR during ground ops with the standard setup is marginal.

B> From the new location of the scoop in the duct. At 2300 RPM, much more flow was felt when holding the tube up to my face – easily 3 to 5 times as much as “A”. Vigorous bubbling occurred when the tube was submerged up to about ½ inch. There was also substantially more flow at 3000+ RPM.

So I am calling this enough evidence for proof.

I like that manometer you got there

Thanks

Your work and posting of results is appreciated Bill!

I have been interested in any progress in this area of VR cooling as well as VR options when and if I need to replace mine. I have seen several good ideas and will continue to evaluate anyone's results who take the time to post.

One difficulty is it is hard to see where your scoop and tube are installed so please consider some additional photos or drawing. Thanks

Rotax regulator frying

I had a similar problem on my wife's Rans S7. The Duccatti regulator/rectifier fried at about 370 hours. I replaced it with a John Deer "clone" regulator rectifier PN AM101406 that I bought on Ebay for $60 bucks.

I made a "crude" cooling tube ( made out of a piece of 3/4" hot water rated PVC pipe and formed the curves with a heat gun) which I stuck out through the right front nostril/air cooling intake and routed to blast air over the rectifier/regulator.

I do not know if it is the cooling air or the regulator I used but it is working perfectly 700 plus hours later.

The John Deer clone is 1/3rd the cost, can be purchased locally if your in a jamb while away from home. In our case it seems to be more stable on the voltage gauge.

Bill, good for you doing some actual testing to prove whether it is doing what you want, but I think there is a possibility that it still is not.

Your simple manometer was referenced to cabin pressure. It is possible that the pressure in the engine compartment in the zone where the regulator is mounted, is enough different from cabin pressure that you still wont have much flow.

rvbuilder2002 said:

Bill, good for you doing some actual testing to prove whether it is doing what you want, but I think there is a possibility that it still is not.

Your simple manometer was referenced to cabin pressure. It is possible that the pressure in the engine compartment in the zone where the regulator is mounted, is enough different from cabin pressure that you still wont have much flow.

Click to expand...

Scott...you know how to rain on Bill' parade��. Any suggestion on how to measure? This was looking very promising. What about encapsulating with a down tub cut like the fuel tank vent line? Venting to the outside beneath the lower engine cowling.

Here is why it is valid.

1. During the test, the top cowl was off and the canopy was open four inches "sitting on the handle."

2. I was not using the water as a manometer. I was using it as a visualization of "is air really flowing through the blast tube and if so, comparatively how much." Looking at how vigorous is the "flow of bubbles." Just like holding the end of the tube next to my face provides a "feel" of the airflow.

3. The face-test and the bubble test (for BOTH blast tube source locations) were done under identical conditions, a few minutes apart. Under those same conditions, RELATIVE TO THE ORIGINAL LOCATION off of the SCATT tube shroud, the scoop is producing MUCH more flow. Which is what I wanted to know.

4. The fact that the top cowl was off during the test has no effect because the SCATT tube and duct both get their air through the lower cowl. And if the back pressure in the overall fully cowled engine compartment is different with the top cowl on, then that would affect both test locations equally because they both "dump" into the same place.

5. I cannot come up with a scenario by which the test is in error and the original location could possibly give MORE flow than the scoop in fully cowled flight conditions. Yes, the original location will get ram air - but so will the duct. And I think the real issue is NOT overheating of the VR in flight conditions, but in ground conditions, particularly that long warmup of the oil. That "flow" from the original location against my face was very, very low. Much higher with the scoop.

In fact, I think that during ground conditions, the original location, with that metal "cap" on the VR, is worse than having no corrugated tube and cap at all. Proving that would require a temp sensor on one of the fins and testing those conditions.

So I still am of the opinion that the test was valid. And thatvthe scoop is much better for those ground operations were cooling is the worst. And logically, much better in-flight as well.

The question is whether the regulator overheats in ground operations. No proof that it does. This laboratory study indicates that it takes the regulator 45 minutes to reach operating temperature. Ground ops will not last that long in most cases.

After a typical start-up the under cowl temps will be low during ground ops and unless all systems and lights are powered up the regulator should not be working hard. An exception to this would be charging a low battery, which could pull a lot of elect load after startup. Keeping the battery on a maintainer makes a lot of sense for this reason and because it is good for the battery.

So far, I have not seen any definitive temperature values recorded for the regulator. In another thread there was discussion of placing a sensor on the regulator and recording it's temperature with the aircraft EFIS data. Joe sent me a link for a sensor, but I didn't follow up on it. I think there is one available unused input left on the D-180 and there are probably several on SV. Point is, if temperature is the issue, there is not much use doing a lot of regulator relocation work or ducting work unless you put together a way to record the pre-existing conditions and the results of changes.

And, from previous posts on the failures, temperature may not be the main issue at all with this regulator, it might be quality control.

Bill_H said:

Here is why it is valid.

1. During the test, the top cowl was off and the canopy was open four inches "sitting on the handle."

2. I was not using the water as a manometer. I was using it as a visualization of "is air really flowing through the blast tube and if so, comparatively how much." Looking at how vigorous is the "flow of bubbles." Just like holding the end of the tube next to my face provides a "feel" of the airflow.

3. The face-test and the bubble test (for BOTH blast tube source locations) were done under identical conditions, a few minutes apart. Under those same conditions, RELATIVE TO THE ORIGINAL LOCATION off of the SCATT tube shroud, the scoop is producing MUCH more flow. Which is what I wanted to know.

4. The fact that the top cowl was off during the test has no effect because the SCATT tube and duct both get their air through the lower cowl. And if the back pressure in the overall fully cowled engine compartment is different with the top cowl on, then that would affect both test locations equally because they both "dump" into the same place.

5. I cannot come up with a scenario by which the test is in error and the original location could possibly give MORE flow than the scoop in fully cowled flight conditions. Yes, the original location will get ram air - but so will the duct. And I think the real issue is NOT overheating of the VR in flight conditions, but in ground conditions, particularly that long warmup of the oil. That "flow" from the original location against my face was very, very low. Much higher with the scoop.

In fact, I think that during ground conditions, the original location, with that metal "cap" on the VR, is worse than having no corrugated tube and cap at all. Proving that would require a temp sensor on one of the fins and testing those conditions.

So I still am of the opinion that the test was valid. And thatvthe scoop is much better for those ground operations were cooling is the worst. And logically, much better in-flight as well.

Click to expand...

Well, just one persons opinion.... We never run out of those here do we...
(please keep in mind I wrote possibility. I am not saying you are wrong. I am simply saying that I don't think your test has unequivocally proven anything..... just my opinion)

The best way to verify would be using a true manometer with one port connected to the air supply hose and the other adjacent to where the regulator is mounted.
If there is a pressure differential between the two points then that will translate into a flow from the lower pressure to the higher pressure.
The higher the differential, the higher the flow.

Well, Vans clearly believes the problem to be temperature related, else why change the plans to relocate the VR to the inside of the cabin? I presume that belief is backed by evidence?

I had to replace my VR at ~120 hours. (Lots of other threads on that.) So I wanted to increase cooling without relocating it to the interior. This was a different approach than some others that have been posted. There's no question that there is more flow with the scoop - just point it at your own face!

Tony, I question that 45 minute rise to full temp as being applicable. That was on a lab bench, with fins uncovered - not next to an exhaust pipe and covered with a metal cap with just a whisper of air through it.

rvbuilder2002 said:

Well, just one persons opinion.... We never run out of those here do we...
(please keep in mind I wrote possibility. I am not saying you are wrong. I am simply saying that I don't think your test has unequivocally proven anything..... just my opinion)

The best way to verify would be using a true manometer with one port connected to the air supply hose and the other adjacent to where the regulator is mounted.
If there is a pressure differential between the two points then that will translate into a flow from the lower pressure to the higher pressure.
The higher the differential, the higher the flow.

Click to expand...

You could "borrow" a digital manometer from your local "heating & airconditioning company and run test in actual ground and flight conditions I.e. Cowling closed. Test results would be interesting..

It is a bit more complicated than measuring a "static pressure differential" at the start of the blast tube and at the exit of the cooling fins on the VR. (See Bernoulli...) What you are really wanting to measure is the FLOW or VELOCITY of air in the blast tube itself. A small inline rotameter or something like that would be better for exact readings.

Boyd Birchler said:

I had a similar problem on my wife's Rans S7. The Duccatti regulator/rectifier fried at about 370 hours. I replaced it with a John Deer "clone" regulator rectifier PN AM101406 that I bought on Ebay for $60 bucks.

I made a "crude" cooling tube ( made out of a piece of 3/4" hot water rated PVC pipe and formed the curves with a heat gun) which I stuck out through the right front nostril/air cooling intake and routed to blast air over the rectifier/regulator.

I do not know if it is the cooling air or the regulator I used but it is working perfectly 700 plus hours later.

The John Deer clone is 1/3rd the cost, can be purchased locally if your in a jamb while away from home. In our case it seems to be more stable on the voltage gauge.

Click to expand...

Please check your PM's.

MMiller said:

I examined six failed regulators, four were from various RV-12's. None of the RV-12 failures indicate ambient overheating was the cause of the failure. Three failures were the new version regulator, Ducati # 362001. All three indicate vibration was the cause of the failure.
.....
In all six regulators, there were no failed electrical components, only failed connections. With that said, this regulator can generate a lot of heat at full output, so cooling is a good thing. Van?s likely moved the regulator into the cabin to comply with Rotax temperature requirements for the regulator.

Mike Miller

Click to expand...

I love the "E" in Experimental! The mandatory "Experimental" marking in our RV-12s should be taken as a badge of honor!
Great job Bill for one more alternative in our collective effort to "save the regulator".
I believe that Mike zeroed in on the fundamental issue with objective observation after a remarkable investigation. As a result I decided to go a slightly different way, discussed in a previous post:
http://vieilleburette.blogspot.com/2015/03/ducatti-voltage-regulator-relocation.html
As it is not clear that any solution is superior so far but as it is also clear that the voltage regulator is poorly cooled at the initial Firewall location, my solution documented above aims at minimizing the burden of replacing the VR in case of failure during a long trip away from my home field. In my case, if the VR fails, the only tool I need to continue my trip is a Phillips screwdriver (to open the upper cowl). Of course, this assumes that you have a backup VR on board which is part of my solution. This extra cost is small compared to getting stuck somewhere waiting for a replacement.

Jean-Pierre - I would have done your exact mod except that I also wanted the Bender Baffle and there didn't seem to be room. I made that decision before I removed the cowl and made a bunch of measurements and actually found out the size and rotation of the baffle. I now think that your mod would fit forward of the baffle. But I wanted to test my scoop mod anyway!

My old regulator had just "started" to exhibit the impending fail behavior (not kicking in until the first time ~3500 was reached). So I carry it as a backup. I did open it up and looked at the CB and did not see such clear indications of a bad connection.

With the VR located forward of the baffle, and the baffle closed, there would still be a lot of VERY cold air in the duct surrounding the regulator (you don't close the baffle unless ambient is very cold!)

If I get a failure with the scoop I will move to your mod (revised). The fiberglass duct top is not very thick forward of the baffle, so I would also reinforce it with some layers around the square hole.

Your mod is also better because post-shutdown the regulator is not exposed to the hot under-the-cowl temps, with its tight cap installed.

Bill, regarding the failed connections, they are sometimes difficult to find. I found a bad one but there were 2 more that were less visible that Mike found. I think he is right about the vibrations being the cause but I believe that the high temperatures are an aggravating factor. At some point the contact becomes intermittent before failing completely. Your regulator had not failed completely therefore it is not surprising that the failed solders are not visible. In any case I would not trust it as a back-up. Check the documentation provided by Mike in his previous posts, you should be able to locate the critical solders which correspond to the SCR leads. I would just redo these solders and check the VR again.
My point was not to claim my cooling solution is better than any other proposed including yours. I believe that with the improvements in cooling these regulators will be good for a few hundred hours. My point is whether your VR fails at 100 hours or 400, when you get there it would be nice not to end up stuck somewhere.

Another tested solution

After blowing the third VR in about 500 hours I think I found a solution that works and is simple.
Ducati claims the max temperature is 195 F
I attached temperature measuring strips to every VR I installed and found they all indicate 180 as the max, so it blows earlier than specified but there may also be hot spots.

On the last one I riveted two additional heatsink plates on both sides combined with the blast cap rivets. Just aluminium scrap as long as the VR and 3 inch wide.
Also fitted a large aluminium screening plate about 8 inches wide and 4 inches high between the VR and close exhaust pipe, using the VR mounting bolts and some conductive paste.
After about 50 hours the temperature is max 160 F which looks like a much better margin to the 195.
Keep my fingers crossed but bought a Silent Hektik for just in case

I like the idea of the shield. A lot of heat must radiate off the exhaust, probably even more so after shut down when airflow ceases. Has anybody tried wrapping that bend with exhaust header tape as a means of insulation? I've never had any experience with it and am aware of conflicting views about whether it damages pipes, although main issue seems to be moisture retention with pipes exposed to rain (motor bikes).

Jack Moore

I think we've discussed tape wrapping the exhaust when trying to fix the oil temperature problems some of us living in hotter areas were seeing. This was also caused from radiated heat from the exhaust affecting the cooling capabilities of the oil cooler sitting way too close to the muffler.
I believe the conclusion was that moving the oil cooler away was the solution as wrapping the exhaust would just kill the stainless steel because of not allowing the exhaust to get rid of the heat.