Learn something most people do not know: Anodizing, the surface finish seen as the Gold color on our parts, is a very effective electrical insulator. You can put a multi-meter on it and test this if you doubt it. It is really magic to me how something as conductive as aluminum, can be fully electrically isolated my a microscopically thin chemical conversion layer, but it is.
All it takes it bolting the part down, and generally the fasteners, particularly lock washers and pipe threads will cut through the outer layer and make the part conductive. However, on our 2400-L Starters, we have been extensively anodizing the brackets, and on some of them powder coating the main plate and the nose. This has lead to some of them having difficulty getting a solid ground path, even though they are bolted directly to the motor.
The solution is this little grounding link made from #4 wire and two copper fittings. It connects the 3/8″ starter stud directly to one of the bolts mounting the 5th bearing housing to the engine case. The standard oil cap shows how small the part is.
Above, the link installed on a 3.0 liter Corvair. This bypasses all the anodized brackets. and lets the current flow into the case, and back to the Engine Ground Cable at the rear of the engine. Why this is important: Although the starter is tiny, it is very powerful, and it can use more than 300 amps at full load (the Corvair doesn’t typically demand this of the starter) If the motor does not find a solid ground by the time it engages, it will actually try to ground through the crankshaft, making impressive sparking where the flywheel teeth meet the starter. Heavy metal pyrotechnics aside, it is terrible for the engines bearings to do this even for a moment. The perfect small solution is the little grounding link. We are including them with all of our 2400-L starter kits now, but if you have a 2400-L series starter, we encourage you to install one.
9 Replies to “Starter grounding link, an important improvement.”
WW, I have this as a topic to discuss with you at CC. I recently discovered how well anodized finishes resist electrical current. Fairly timely as it was testing the conductivity of my engine oil temp sender which was 1/8″ NPT needing to be inserted into a 1/2″ adapter (anodized), then inserted into the 1/2″ port on the oil housing (anodized). BTW I did read the directions and have my temp sender in the correct location. I’m wondering how to get this sender to ground properly, for without it the gauge will not read reliably or at all.
Mr. Dewenter’s comments about grounding the oil temp sender are dead on, and since I can’t be at CC#39 to sit in on your conversation with him, I decided to look how I can solve it for my particular sender while I was also deciding how best to reconfigure the mounting of my nason switch and oil pressure sender. I want to lay out my thought process and decisions and get your comments, if any, and if and where I may have gone astray.
First, the sender is mostly brass and about 1 ¼” long. It has a ½” hex head just below the connection pin for the wire to the gauge, then the sensor body is about 7/8” long, the top half of which is 1/8” NPT threads and the bottom half is a bare 21/64” dia cylinder. It is currently screwed into a brass adaptor that I bought at the hardware store. Checking the continuity between the brass hex and the rear oil case shows no measureable resistance. I have to assume that that will not be the case when I replace the brass fitting with an AN 912-6D. There is 1/8” or more of the sender’s pipe threads showing between the top of the brass fitting and the shoulder of the sender’s hex head. A ring terminal, beefed up with 1 or 2 brass washers would fill this gap and provide good connectivity with the brass body of the sender. 2” of wire from that terminal to another ring terminal will give a direct ground path to the oil filter housing mounting bolt that is directly below the oil temp sender well. Problem solved… until a few issues arise:
1. A 1/8” pipe is slightly larger than 3/8”. In order to have an easy fit, 3/8” ring terminals and washers are slightly than 3/8”. Nevertheless, it may be necessary to wallow out the terminal and washers slightly. If so, I need to take care to remove any burrs, gouges, etc. I’ll also need to keep an eye on them for crack developments during condition inspections. Issue resolved.
2. The AN fitting is 1” long whereas the brass fitting is 3/16” shorter. Even if the sender were to be screwed all the way into the AN fitting, the end of the sender will be buried well inside the fitting. Examination of the brass fitting shows that the case is the same for it. Examination of the oil temp sender well in the oil filter housing shows that the space where any sender will sit is an alcove off to the side of the main oil flow to the cooler. Therefore, the sender will be reading the temp of the fitting directly and indirectly, the temp of the oil. This may introduce an error of a few degrees, but it will not be significant. Issue resolved.
3. The weight of the sender is 12 grams. The weight of the 1 quarter in my pocket is 5.7 grams, therefore, William’s 3 quarter weight limit for a 1/8” NPT fitting is about 17.1 grams. Even if the sender weighed significantly more than that limit, it would not be an issue in this case because the center of gravity of the sender sits well inside the fitting when it is installed and not cantilevered out beyond the end of the fitting where it can flex back and forth like a tuning fork. Issue resolved.
4. The sender will not fit into the AN fitting. About 1/3 of the length of the hole thru the fitting is threaded. The remainder of the hole is not. A 5/16” drill bit Is a loose fit for the hole. A 21/64” bit will not fit in the hole at all. The unthreaded end of the sender bottoms out against this undersized hole before the threads come close to engaging with the fitting threads. By definition, every other AN 912-6D fitting will be the same. The choice is to either stay with the brass fitting with its extra weight and potential hidden flaws, or drill out the AN fitting and potentially introduce flaws by doing so. While I have no training as an A&P, I have been told that the standard to be maintained is that if a piece of hardware or a fitting is no longer in pristine shape, that it should be discarded and replaced with a new piece. A fitting that has been drilled out and possibly re-tapped is the very definition of “no longer in pristine shape”. Nevertheless, replacing it with a new fitting is not an option. I had decided to drill out the AN fitting. I will also have to run a 1/8” NPT tap into the fitting afterwards to at least chase the threads, and perhaps cut them deeper to ensure that I have a snug fit for the ring terminal and washer. Since this fitting is in direct contact with the oil flow, it is vital that I remove ALL chips, burrs ,flakes, shavings, dust, etc. from the fitting before installing it since the next stop for any such particle is a bearing surface. Issue decided, risks accepted.
5. Since I am drilling out part of the AN fitting, should I drill it out bigger than necessary and create a pocket around the unthreaded portion of the sender to allow more contact with the oil? I have decided not to do this. Drilling a slightly larger hole would probably result in an air pocket around the sender which would introduce significant errors in the temp reading. Drilling a hole big enough to eliminate this possibility runs a significant risk of compromising the integrity of the fitting and probably would not improve the quality of the temp reading anyway. Issue decided, risks minimized.
P.S. I thought about building a Rube Goldberg arrangement of pipes, tees, and senders, and sending a picture of it balanced on a ruler with a quarter, a nickel, two dimes, and a roll of pennies with a caption to the effect that “I didn’t have 3 quarters, but it still adds up to 75 cents”. But I figured that you weren’t in the mood to have your chain yanked, even if it was in jest.
Dave, this may not be helpful in your situation but some Oil Temperature senders are two wire types so that a ground is supplied directly. A case in point is the GRT FT-LC-01 and while you may not be using a GRT system to monitor the oil temperature it is listed as compatible with the Dynon engine monitoring systems so you might consider that as a positively grounded option.
Sarah, No, the temp sender very clearly uses a single wire from the gauge with grounding thru the engine case. To be honest, I am quite comfortable with the solution that I have described except that it feels sacrilegious to be doing surgery on a brand new AN fitting. However, experience has taught me that sometimes I develop tunnel view on a particular solution and when somebody else looks at it, they ask ‘Why didn’t you do a, b, & c?’ to which I reply ‘Oh…I didn’t think of that’. So, I’ve posted this verbose description and I figure that William, you, or somebody else will point it out if I am doing something that will make the engine un-airworthy or just exceedingly stupid.
These kind of constant updates are the ONLY reason I have confidence to fly an auto conversion motor. It is humbling / heart warming to know someone with a brain and heart has my back. I am working hard toward having a mindset open to receiving and implementing the great wisdom you freely share.
I have the old style starter but I painted most of my parts so I plan on adding this to my engine, no harm in having an assured high quality ground for the starter. Just because the engine is assembled and test run does not mean there is not room for improvement.
I would go a little crazy troubleshoting if i couldnt get a ground thru the starter body. 🙁
WW – Is the bolt on the 5th bearing housing still long enough to get the proper torque with the addition of the copper fitting, or is a slightly longer bolt required?
The 1/16″ lug on the link doesn’t interfere with getting the full strength of the bolt.