CPI Installation Show and Tell

[Toobuilder]

While I opened the door to the performance of Ross' new CPI ignition product in this thread, I figured I'd show you the nuts and bolts here.

First off, Ross has a bunch of parts to make a complete installation kit for the 4 banger Lycomings, but he's still ramping up the 540 stuff. I've been following development of CPI for a while and just couldn't stand it anymore- so after begging him for the basic components, I had to develop my own "installation kit".

First up is the hall sensor bracket. I came up with the .250 thick mount plate, a machined .750 dia x 1.9 spacer, and a 1.9 inch long, 5/8 hex "nut", which I machined to length, drilled and tapped for the case stud.
The extra long NAS bolt shown replaces the lower case through bolt, and the two 10-32 lock nuts complement the capscrews holding the hall sensor to the bracket. Also, the spacer and nut are machined to the exact length needed to provide the correct air gap for the hall sensor. If anyone wants to try this installation before Ross gets the 540 parts out, shoot me a PM, I can provide more specific dimensions/drawings.

Here is the whole thing installed and ready for the ring carrier (flywheel)

Here's the ring carier installed. Ignore the hieroglyphics on the back side. That was me trying to get my head around magnet positioning. (Quite easy, actually)

Though the coil can be mounted anywhere, I decided to keep the wires short and mount it on the backbone of the case. Simple aluminum extruded angle stock and a couple more spacers.

Here is the coil ignitor module, mounted just behind the coil on the backside of the baffle. The overly thick mount plate shown is my attempt at a "cold plate" heat sink. That, coupled with the large surface area of the rear baffle should ensure the ignitor never generates more heat than it can handle.

And finally, the whole package. (And yes, the rear baffle seal is flipped out of the way for the picture).

You will have to excuse the seemingly hasty installation shown on this picture. I'm out of room on this panel and this is one of the few places left that I can get at it easily. Eventually, the brain will be hidden. Once the curve is set, I'll lock down the settings and forget about them.

 

[Mike S]

Thanks Mike, pretty sure I am not the only person waiting/watching for this thread.

 

[M McGraw]

Thanks

These pictures just saved Ross a waisted telephone call. The CPi manual states it would be advisable to place a blast tube on the coil pack for longevity. I was wondering why not just place the coil pack on the front side of the baffles and avoid the known "leak". Do you have any temperature data on the coils after shutdown? Do you anticipate a moisture issue?

 

[Mike S]

Mike, did you flip your rear baffle seal over for the photo, or is that how it is normally??

 

[Toobuilder]

No temp data, but Ross says the thermal demands on the components at 2700 RPM is almost nothing. These are designed for much higher RPM/duty cycle.

That said, on my lasy flight I immediately took the cowl off and put hands on everything - nothing alarming to me based on touch. Certainly no warmer (and likely cooler) than the underhood temps of a typical car.

As for water intrusion, all the connectors are the current standard automotive, o ring sealed units. I'm sensitive to mounting the components under a cowl or panel seam where water could drip when parked, but I think I'm pretty well protected here. Time will tell.

 

[Toobuilder]

Mike, did you flip your rear baffle seal over for the photo, or is that how it is normally??

Flipped for the picture.

 

[Toobuilder]

Flew it again this morning hoping to get a whole bunch of data but ran low on gas before I could get the baseline speed performance to settle in (nothing to do with the new ignition). Anyway, I landed, shoved it in the hangar and pulled the cowl. Using my trusty optical pyrometer I measured the coil (185f at the base, much cooler on the top) and the ignitor (155f). Ross says these are rated for 125c, so it would appear they are plenty happy in this location. I'm guessing the highest temps they will see is post shutdown heatsoak in the summer. As long as the fan is turning, there should be plenty of cooling.

 

[Toobuilder]

I have more than 3 hours on the system now and most of that has been taking data to figure out the best curve, and also to run some "what if" scenarios. That data is for another thread, but I do have some bottom line data relevant to this thread: What did this ignition do for me WRT fuel flow and performance?

As a baseline, my old cruise numbers used to be 191 KTAS @ 11.5 GPH. (16.6 MPG)

Economy: So for a test this morning I leaned the engine until I found my "old" 191 KTAS. Resulting FF: 10.9. (17.5 MPG) So it looks like I picked up over half a gallon of fuel per hour for the same performance.

Engine output: so let's turn up the heat a bit and use my old fuel flow of 11.5 and see what that does for speed. Resulting speed: 198 KTAS. Same fuel flow at higher speed means better MPG, or in this case, 17.2 MPG.

Not a bad improvement!

Just as another data point I also richened it up to peak EGT (12.0 GPH) and that gave me 202 KTAS (still better economy than the baseline- 16.8 vs. 16.6) and going to 100 ROP gave me 205 KTAS @ 16.0. (12.8 MPG)

 

[Toobuilder]

Just as an update, Ross has developed an additional advance feature that is enabled with the flip of a switch. This optional feature is adjustable from 0 to 20 degrees and is added on top of whatever curve you have set. In my case it will be used as a LOP switch so that I can takeoff, climb, fly acro, formation, etc. with the standard curve optimized for ROP ops, then when in LOP cruise, flip the switch and bump the advance to the optimum LOP setting.

I flew it today briefly and it works like a charm. I was unable to take any hard data, but based upon prior testing, the 4 degree bump I have set should give me a knot or two.

 

[Toobuilder]

We took the airplane on the first real cross country and I've had the opportunity to test the new LOP feature. Bottom line: when in LOP cruise the speed difference between the optimal ROP curve and the slightly more advanced setting is 3 knots.

Testing has led me to believe that my particular engine "likes" about 31 degrees advance at best power mixtures, and about 35 degrees LOP at my normal cruise altitude. The picture below shows the screen used to set the LOP advance. Up to 20 degrees is available, but you see mine is set at 4.

Below is the switch I use to add the 4 degree advance. In practice I use a conservative advance curve optimized for ROP ops on the CPI ("norm" position), then when it needs more advance for the slower burning LOP condition I throw this switch and within 30 seconds 3 more knots show up on the EFIS.

 

[rvbuilder2002]

Thanks for the write-up on this system Michael.
I look fwd to following as you do more testing.

Also glad to see the hard line protection of the CPS wires (very important in my opinion). Is there room for a hard shield on the last little bit of exposure of adjacent to the ring gear support pulley? Seems like that would then give you fully protected sensors.

 

[Toobuilder]

Thanks for sharing your interest. I think the addition of another player in the EI game is big news - Especially one as versatile as this one is proving to be. I'll keep posting updates as long as there is a legitimate need.

The wire armor addresses possible damage from routine maintenance as well as the unlikely event of impact damage from a large bird or other random high velocity FOD. Yes, I worried about that little bit of exposed wire too. Primarily from a thrown belt. I was cooking up a belt guard, but as I started reassembling the baffling I realized that the front baffle serves as an excellent barrier to that particular hazard. There is a fairly easy way to armor that last little bit, but I dont think it's required.

 

[erich weaver]

Sorry for the thread drift, but where does one find nice rounded blast tube inlets like the one in your picture here?

thanks
erich

 

[Toobuilder]

Standard SCAT tube flanges from Spruce and other suppliers.

The one shown is 2.5 inch and feeds the eyeball vents in the cockpit - it was not installed for cooling the coil pack.

 

[rvbuilder2002]

Yes, I worried about that little bit of exposed wire too. Primarily from a thrown belt.

That was the reason for my comment.
I realize it is likely rare, but I have read a report of a forced landing caused be a shredded belt destroying the ign sensor wires.

 

[Toobuilder]

I read of that (or a similar) incident too. Id like to see how the front baffle on the incident airplane is set up and how the wires were routed. In my case, the tube armor extends past the Adel clamp on the sensor bracket (not present in picture #2, first post). The inch or so of "exposed" wire is entirely behind the front baffle AND a secondary baffle. I dont believe a thrown belt has enough energy to penetrate or remove the front baffle and take out the short bit of wire. It's possible, but not probable - IMHO.

OTOH, if someone can show me a picture of a mangled front baffle, I would certainly reconsider my position.

 

[Tom Martin]

Great work Michael!
Getting some speed back while running LOP is a very positive result. I gave up on LOP operations a few years ago due to the speed losses. I look forward to more reports.
Are you still running one mag on the bottom plugs?

 

[Toobuilder]

Tom, Yes I'm still running the Slick on the bottom plugs. I'm getting comfortable with how this system works so I expect to remove the mag and be full electronic in a couple of weeks. At that point I can really start playing with the advance curves/mixtures and see what the engine likes.

We also have a second CPI system going on another Rocket at the moment. His engine is a botuque build with big compression and other hot rod stuff, so it will be interesting to see how his ignition curves compare to my little factory stock 260.

 

[Slowpoke]

Hello all,
Mike (Toobuilder) help me complete my install on my Rocket last weekend. I had enough time to get a quick test flight in before dark, she ran smooth. I was having issues getting her to run L of P when running mags, well during my short test flight I was able to run L of P very easily running one mag and one of Ross CPI units. Programming of this unit is not hard and allows you a lot of possibilities if you want to use them.

Now the install is more labor intensive than the Emag but the versatility of programming is worth it to me. Not to mention the Emag is late to the game.

I installed both controllers as I am planning on going full CPI after some testing. The controllers are mounted to the back of my iPad cradle. The iPad is easily removed if for some reason I wanted to make adjustments.

We trimmed the edges of the hall sensor to get clearance on the edges but keep the tolerances tight for the sensors them selves.

 

[Toobuilder]

That CPI brain mount idea is pretty slick. Great way to use otherwise "wasted" real estate. I'm stealing that idea for myself!

 

[DanH]

I installed both controllers as I am planning on going full CPI after some testing. The controllers are mounted to the back of my iPad cradle. The iPad is easily removed if for some reason I wanted to make adjustments.

Very clever! I love stuff like that.

 

[Toobuilder]

I have some more good news for those of you considering this system:

If you have gone online and read the installation manual, you might have noted that the published power requirements were pretty high. When I installed the system I followed the instructions and wired it up for the 10 amp load listed. It was not until yesterday however that I actually took measurements in an operating environment. It would appear that this manual is anticipating use in motorsports applications well above 10,000 RPM. Our slow turning aircraft engines are a whole lot less demanding.

Cut to the chase, I hooked up my trusty multimeter in line and found that at idle RPM the entire system (coil and brain) pulls just 0.7 amp. Dropping the coil during runup (brain only) brings it down to 0.1 amp, and finally, max RPM during climbout the whole system pulls a modest 1.56 amps.

This means that wiring and any backup battery (if used) can be very light. It also means the components are generating significantly lower heat than they are designed for.

 

[rv6ejguy]

Updates

Right, you can fly for many hours with these units if your alternator takes a dump.

Based on customer feedback, we've made some recent changes to software to better suit aviation applications:

100 rpm programming increments replace 250 rpm ones for direct drive engines

As Mike showed back a page or so, the knock input can be re-purposed for adding a user programmable amount of ignition advance to the programmed curves with the flick of the switch. Useful when running LOP where flame speed is much reduced. This allows you to get peak cylinder pressure optimized for best torque.

User adjustable MAP offset to better synch the CPI display with other cockpit instrumentation.

Some new CNC'd rear coil mounts for the 540 should be ready by the end of next week along with crank sensor mounts for the 540.

 

[rv6ejguy]

Looking through some comments here, none of our components require blast tubes or additional cooling air. They run very cool. Note the total current draw that Mike measured. The 6 cylinder module heat sink only ran at 54C in still 14C air during testing after 90 minutes at 4500rpm. It runs much cooler at 2500 rpm.

The Hall Effect sensor is mounted out in ambient air, not inside the hot engine. It's fully encased in epoxy so there are no moisture or vibration concerns.

The coil packs are designed to be engine mounted so the components are rated for the temperatures and vibration. Again, epoxy filled and all FF connections are waterproof.

The six cylinder coil module is heat sunk to the aluminum billet coil mount but this has very lower power dissipation anyway.

The MAP sensor in integral to the cockpit mounted module so does not see any heat or high vibration.

 

[Toobuilder]

Just flew with the updates this morning and the 100 RPM steps are smoother during taxi and follow the RPM on the EMS much better. Though the MP pressure display is user adjustable, I found it spot on right out of the box and like the RPM, it tracks the EMS throughout the whole range.

Great work Ross!

 

[rv6ejguy]

Just flew with the updates this morning and the 100 RPM steps are smoother during taxi and follow the RPM on the EMS much better. Though the MP pressure display is user adjustable, I found it spot on right out of the box and like the RPM, it tracks the EMS throughout the whole range.

Great work Ross!

Cool. Glad to hear it's all good news. I hope to have time tomorrow to final test fit the 540 crank sensor mount on the scrap case/ crank and make sure all that is good to go. Just too busy today to get to it.

Look forward to getting your feedback on Adam's Rocket soon.

 

[Toobuilder]

Now flying the dual CPI system. Had to wait a few days for the 50 MPH winds to die down but got a quick flight in this evening. Mag drop is minimal (~30RPM) on either system compared to the 125 on the mags. Climb was strong all the way to altitude (it is fun being able to maintain 3,000 FPM climb as you pass through 9,000 feet!) and going way LOP (9.7 GPH) still delivered 185 KTAS. BTW, that was my "old" cruise numbers at 11.5GPH So it looks like I picked up my expected 1.5 GPH with the switch to full electronic ignition. And yes, the "LOP" switch gave a noticable boost when this lean. Didn't check my max cruise numbers as I was running out of daylight, but a low level (3,000 MSL) max effort speed run delivered 222 KTAS in level flight. I don't do that very often, but it appears to be a few knots faster than I've seen before.

Much more testing to come, but this system is great so far!

 

[rv6ejguy]

540 Parts

Thanks for the report Mike.

Here are a couple photos of the 540 Hall sensor mounts for the CPI/ EM-5. Dual pickup shown. We've added some 10-32 threaded holes to bolt cable protection to if desired.

 

[rv6ejguy]

540 Rear Coil Pack Mounts

Just back from the anodizer. These mount the 6 cylinder coil packs off the magneto studs. These are last bits to complete the 540 CPI kit which should be available Feb. 15th.

 

[rv6ejguy]

New Vids

I did a couple of videos about the CPI for those who might be interested:

Basic features and programming:
https://www.youtube.com/watch?v=lBEOQdZBOS8

Some of the hardware:
https://www.youtube.com/watch?v=olQwqfxm2BE

 

[Mike S]

Ross------great stuff, thanks for sharing the links.

Question-----for a full dual setup, do I just buy two complete packages or are you making a setup that combines some of the items, brain box comes to mind?

Thanks, pretty sure you will be selling a lot of these.

 

[Toobuilder]

Mike-

These are completely independent systems. The only place they come together is the mechanical pairing of the dual hall sensor. But even this is independent because they do not talk to each other - they just share the same block of aluminum.

So even if you intend to start out with a single system, you will typically install the dual hall mount and just stow the wires from the unused hall pickup until you need it.

 

[rv6ejguy]

Ross------great stuff, thanks for sharing the links.

Question-----for a full dual setup, do I just buy two complete packages or are you making a setup that combines some of the items, brain box comes to mind?

Thanks, pretty sure you will be selling a lot of these.

For a dual system it's not quite double the components, but close. Only one crank sensor mount, magnet set and drilling block. Pretty well everything else is duplicated. The dual Hall sensor is slightly less expensive than two of the single ones also.

We're finishing off the documentation right now for the six cylinder setups. Should be shipping production models by the end of the month.

 

[Mike S]

Do you pre load basic timing map in the CPU??

Or, at least provide a guide as a place to start?

 

[Toobuilder]

Regarding the ignition curve development, the advantage of this system is that you can easily copy any effective curve you have seen on an aircraft engine - or even just lock it down like a magneto until you get comfortable. I suspect as more people start working with them we will start to see some fairly clear discriminators between the more stock engines like mine and the big compression hot rod types - like Adam's.

The other thing to consider is the RPM curve really does not need to be all that sophisticated. I have my full advance in by 1800 RPM and zero MP advance at anything above 24 inches. Thats pretty conservative. Its essentially emulating fixed magneto timing at high power.

The real trick is just figuring out what the engine likes at various cruise power settings. And even this is not likely to hurt anything with experimentation unless you go WAY off the reservation. I've run anywhere from 25 to 40 degrees in cruise and the only thing I've seen is some speed loss, and EGT/CHT changes.

It is my intent to nail down Idle, 100%, max cruise, economy cruise, and LOP (handled as a separate advance value via the switch). Once I have the hard points perfected, then I believe it will be an easy thing to just interpolate all the points in between and just "fill in the blanks" so to speak.

There has been much discussion on the merits of advanced timing on this and other forums over the years and one thing certainly rings true: when it comes to advanced timing, "more" is not better. Variable timing is a great thing for an aircraft engine, but there is no one size fits all solution.

This system allows someone to come up with the exact timing curve that suits their engine, airframe, exhaust, and flying style.

 

[Mike S]

Mike, thanks for the reply, pretty much how I think the development of a curve should be handled.

Dont want to be disrespectful or flippant, but all I am looking for is a simple yes/no for the question I am looking to get answered; is there a basic timing map already in the CPU so there is a known and safe place to start from??

 

[rv6ejguy]

Mike, thanks for the reply, pretty much how I think the development of a curve should be handled.

Dont want to be disrespectful or flippant, but all I am looking for is a simple yes/no for the question I am looking to get answered; is there a basic timing map already in the CPU so there is a known and safe place to start from??

We'd enter a very conservative base timing map pretty much like Mike said so it's super safe for starting and running a 9 to 1 CR engine on 91 Mogas. Something like 24 degrees rpm timing max with a few degrees of MAP retard thrown in above 26 inches. The user can then adjust to his engine specs, needs and fuel type.

Our philosophy is a bit different from some of the other EI makers who don't want to tell you what their box is doing behind the scenes. You can see the whole picture on the screen and do anything you want with it.

That control comes with responsibility of course. You don't want to try to run a 10 to 1 engine on 87 octane pump gas with 40 degrees of timing WOT at sea level...

We are gathering timing info from several flying aircraft now and will consolidate that info to better advise people in the future what has proven to work with different CRs and fuel types.

 

[Mike S]

One more question.

Can this be installed without having to remove the prop?

 

[rv6ejguy]

One more question.

Can this be installed without having to remove the prop?

Usually we have the flywheel off, but I suppose you could actually drill and tap the flywheel in place. Never thought of that actually until now... I haven't actually tried it but no show stoppers come to mind at the moment.

Toobuilder could maybe comment on that possibility, although he had his prop off to do his 540.

 

[Toobuilder]

...Dont want to be disrespectful or flippant, but all I am looking for is a simple yes/no...

Yes Mike, I went a long way to give you a non answer. Sorry about that.

Let me try and do better on the next question:

Concerning installation without removing the prop, I believe if you can get the front baffle off then you can do it without too much trouble. The magnets can be definitely be installed with the prop on, and that is probably the most challenging part of the whole thing. You will need to trim the front baffle to clear the hall mount, as well.

 

[Toobuilder]

I hung the second system on the baffle.

Of note is the fact that I'm using the large surface area of the baffle as a heat sink for the igniter module.

 

[Slowpoke]

I found some time to do some testing with one of Ross's CPI units installed.

A couple of issues I was having prior to installing this ignition was, starting kick back, running lean of peak and the dreaded detonation on takeoff. My engine is running 10:1 pistons and was timed originally at 25 deg.

It took a couple of quick flights to figure out how I wanted the CPI programmed for a starting point. It also helped that Mike and I are doing this at the same time so we could compare notes and bounce ideas back and forth. The programing is easy and totally flexible, basically it looks as follows:
1. Starting at 0 deg
2. Idling at 33 deg
3. Take off power 20 deg
4. 7500' cruise 31 deg to 35 deg

I am see the same cruise numbers as Mike was seeing within 1 or 2 KTS when he had one unit installed. But I have not had any kick back on starting, she will run LOP when I ask now (50 deg LOP smooth) and it has not detonated on takeoff with full power. These are all big pluses for me so far. This unit's flexibility is great and allows the set up to match the engine it is being installed on.

Not to mention Ross's customer service is outstanding - Thanks Ross.

I guess you can say I am happy so far

 

[rv6ejguy]

540 CPI KIt Now Available

Today was the official release of the production Lycoming 540 CPI kit which follows the 4 cylinder 320/360 kit released in January. As Mike and Adam have generously showcased some of the features here, I won't repeat too much more.

We've decided to make the "LOP switch" standard at no extra charge.

The panel mount kit is optional as are things like plugs, wires, boots if you want to get those locally.

You must have a stock Lycoming 8 7/16 ID flywheel to use mount the magnets. Some aftermarket flywheels like those from Sky Dynamics won't work without modification.

This unit allows you to program any rpm advance curve and any MAP advance/ retard curve you want, in flight, if you wish.

http://sdsefi.com/cpi.htm

 

[rv6ejguy]

More Lycoming Parts for CPI/ EM-5

We finally got around to adding long reach spark plug adapters to our inventory for the IO-390/580 engines and any others using long reach plugs.

We also now have 3 types of coil pack mounts for both 4 and 6 cylinder engines- firewall mount, magneto cover mount and top rear crankcase mount.

The CNC shop has been very busy for months now making many new components. We should have a complete inventory before the end of August so buyers will have more choices to do things the way that suits them best.

You'll see some new info on the website starting in late July.

 

[Chkaharyer99]

Michael, Adam and/or Ross

I recently visited a friend nearby and saw his SDS CPI system install. Very clean. My complements to Ross in his design and packaging.

I'm really interested in Ross' offerings but must confess I have much to learn on this subject.

My RV-8 is equipped with Slick mags, single buss (bus), one battery, one alternator.

In upgrading to the CPI system, does one need a more complex dual bus system and/or a bus manager / two batteries / second alt / other?

Would a Vertical Power system be compatible with the CPI?

I'd like to hear what Michael and Adam and any other CPI customers are doing in this regard. Any explanation of why is also appreciated.

I do not intend to thread drift. If my questions, comments are not appropriate for this thread, my apologies and please disregard.

 

[rv6ejguy]

Michael, Adam and/or Ross

I recently visited a friend nearby and saw his SDS CPI system install. Very clean. My complements to Ross in his design and packaging.

I'm really interested in Ross' offerings but must confess I have much to learn on this subject.

My RV-8 is equipped with Slick mags, single buss (bus), one battery, one alternator.

In upgrading to the CPI system, does one need a more complex dual bus system and/or a bus manager / two batteries / second alt / other?

Would a Vertical Power system be compatible with the CPI?

I'd like to hear what Michael and Adam and any other CPI customers are doing in this regard. Any explanation of why is also appreciated.

I do not intend to thread drift. If my questions, comments are not appropriate for this thread, my apologies and please disregard.

Not sure what electrical mods Michael and Adam did but a big consideration is that these devices only draw a couple amps in cruise so assuming your battery got the engine started and you have some sort of low voltage warning device (important with two EIs) to tell you when your alternator takes a dump, you'd have a couple hours to find a place to land if you shed load.

Yes the primary battery could go U/S so a small 7 amp/ hour backup battery could run the CPI for a couple hours as well. 2 hours in either direction should cover you on a 4 hour flight.

I'd say use whatever makes you comfortable for the usual missions and terrain you fly over.

I'm no fan of VP. Don't like so many eggs in one basket, especially if your engine life force (ignition) is running through them. A couple good toggle switches would seem to be more reliable and a lot less expensive.

I'm interested to hear what some of our customers have done in this regard too. Dialogue is good, it often brings some new/ better ideas into our heads.

 

[Toobuilder]

Full disclosure: I am not equipped for IFR flight, and truth be told, I not that interested in night flight anymore. That said, I am single alternator and single battery, but consider my system somewhat redundant in the fact that primary power is off the alternator, with backup being switched to the battery in front of the main contactor. In other words,if I see an alternator failure I'll shed unnecessary equipment and run off the ship's battery buss. If I have a hard fail of the ship's buss I can dump it all with the master, yet still have the battery reserve through this one (switched) dedicated circuit. If the alternator throws a belt, the ship's buss fails AND the battery has an internal failure, well, it's just not my day.

The dual CPI system draws very little power, so as long as I catch a failure early enough, the main battery is going to get me on the ground with plenty of reserve. I would expect a fully charged battery to far exceed the fuel supply, in fact.

When my mission gets a little more demanding I will be looking at a more robust power supply, but I'm comfortable with the configuration for the current mission profile.

 

[erich weaver]

Bob Nuckols follower here. I think he would mostly agree with your philosophy with the exception of suggesting an annual battery capacity test and knowing exactly what your electrical load is after shedding all but essential items. The fact that your battery is fully charged and started the plane isn't enough and certainly doesn't guarantee you will only be limited by fuel supply.

Erich

 

[Toobuilder]

I've measured the load at maximum power and as a result I'd expect a dual system to require 3 amps or less at cruise. While a battery capacity check is a good thing to do for any of our increasingly electrically dependant airplanes, I'd like to point out that I am not likely to take off with a failed alternator. My envisioned failure mode is going to occur enroute, after the battery has recovered from the engine start sequence. Also, I have the capability to shed EVERYTHING (to include the master contactor) and still have ignition power.

I have yet to decide if the next iteration of my electrical system is going to include a second alternator or not, but another battery is a near certainty.

 

[Toobuilder]

I hung the second system on the baffle...

To those contemplating following my lead and hanging the coil on the baffle, you should reconsider. I'm seeing cracking in the corner after only a few hours. I didnt reinforce anything here and was just hoping for the best. It bit me.

Its a fine location if you are willing to treat it like an oil cooler installation with plenty of beef up, but I'd suggest looking at the billet mounts that Ross provides, adel clamped to the engine mount, or hang it on the firewall like Adam did.